Wired's Latest "E-Voting" Story
There's been plenty of misleading and incomplete coverage of the electronic voting issue, most of which ignores the civil rights implications of this debate, but Wired.com takes the cake. The latest article from Wired.com's Kim Zetter, entitled. "How E-Voting Threatens Democracy", abandons any pretense of engaging in balanced journalism. Much of this lengthy article is a paean to Bev Harris of Black Box voting fame, and the Avi Rubin group whose work I've criticized here.
Two problems are most glaring in Wired.com's latest effort. The first is that the fact that, in six pages of this story, it fails entirely to address the civil rights advantages of Direct Record Electronic ("DRE") voting systems, compared with paper-based systems like punch cards and optical scans. It doesn't deal with the passionate advocacy on behalf of people with disabilities, in support of conversion to electronic voting systems which have accessibility features that would allow many to vote independently for the first time in their lives. Nor does it mention that touchscreen voting allows easier multilingual voting, or that it lowers the racial disparity in uncounted votes compared with other systems.
The second problem is that Wired.com continues to use problems with paper-based voting systems to attack electronic voting, which it refers to as "e-voting." (I avoid this term, because it invites confusion between internet voting and DRE voting, which occurs on stand-alone machines that aren't wired to the internet.) As noted here, Wired.com has previously done the same thing, using problems with optical scan voting to argue that "e-voting" is insecure. Here again, Wired.com relies on incidents where paper-based optical scan systems caused problems to argue that electronic voting can't be trusted. This overlooks the fact that the kind of problems that have so often caused difficulty with paper-based systems -- including unclear marks and hanging chad -- don't exist with electronic voting. While these sorts of problems result in thousands of lost votes in each election, even the Wired.com story must admit that to this point the doomsday scenarios envisioned by Harris and others haven't materialized. As the Wired.com story begrudgingly admits on the last page, there's "no evidence" of electronic voting fraud by election officials or voting system manufacturers.
On the other hand, there's plenty of evidence of fraud and error with paper-based systems, which accounts for quantifiable numbers of lost votes. For example, in California's 2002 recall election, an estimated 176,000 votes were lost due to the use of punch cards. In Ohio's 2000 presidential election, the number of lost votes due to inferior paper-based systems was at least 70,000. One can perhaps quibble with these estimates, but there's no reasonable basis for disputing that large numbers of votes are lost by the use of good ol' paper. The ones who bear the brunt of these problems include African Americans, people with disabilities and non-English speaking voters, as well as those who are illiterate. All of these groups would benefit from conversion to touchscreen voting.
None of this is to deny that we should be concerned about the integrity of electronic touchscreen systems, as we should with all other systems. It does not, however, follow that a contemporaneous paper replica (the so-called "voter verified paper trail") is the answer. Nor does it justify the disregard for the civil rights dimension of the electronic voting debate evident in so much coverage of this issue.
Controversy Over Voting Machines in Ohio
Here in Ohio, a fierce controversy is brewing over whether or not to move to electronic voting. As noted here, Ohio is one of the last bastions of the "hanging chad" punch card, used by a substantial majority of the state's counties. (Disclosure: I'm part of the legal team that seeks to end Ohio's use of punch card voting machines.) Secretary of State Kenneth Blackwell gave counties until Tuesday, March 30 to decide which of three electronic systems to move to, according to this story from WHIOTV.com. The Cincinnati Enquirer reports here that the Hamilton County Board of Elections (where Cincinnati is located) was among those set to make a decision Tuesday evening. This story also reports that the State Controlling Board is refusing to release some $130 million in federal money, as the result of security concerns that have been raised about electronic voting systems.
HAVA Money on Its Way in May?
Electionline.org reports here that $2.3 billion in funds appropriated pursuant to the Help America Vote Act will be distributed to states in May, after more than a year of delay. Each state and territory's HAVA plan was just published in the Federal Register, as required by HAVA before funds can be distributed. Electionline.org also describes other hold-ups in getting much-needed money to the states so far.
Another Editorial Opposing Touchscreen Voting Ban
This one from the Tri-Valley Herald in Northern California, which joins other papers that have come out against Senators Perata and Johnson's proposal to prohibit touchscreen voting in California. See here for a description of their proposal and my views, and here for links to other editorials on the subject. The TVH editorial notes that touchscreens appear to have worked well in 11 of 14 California counties in the March 2004 primary.
Jim Adler's Blog
Jim Adler, CEO of VoteHere, Inc., has started a blog dealing with electronic voting security issues. VoteHere develops electronic voting software, including the VHTi, a system of electronic voter verification described here. Jim's been a thoughtful voice in the electronic voting debate. He takes the security concerns surrounding electronic voting seriously and shares some of them, but doesn't believe that the contemporaneous paper replica (the so-called "voter verified paper trail") is the magic bullet. His blog should be worth reading.
Open Voting Consortium
A group called the Open Voting Consortium is attempting to create a "voter verifiable" voting machine using "open source components." Described on their website is a project "aimed at developing a PC based voting machine they claim will be easier to use, more tamper-resistant, and cheaper than commercially available voting machines." See here. Among the advantages of computerized voting, the site notes, is greater disability access and multilanguage capacity. The OVC is planning to demonstrate a prototype that produces a paper ballot on April 1, 2004. (Unfortunately, it seems they've chosen an inauspicious date.)
My take: I've blogged previously, see here, on whether open source is a good idea. Of course, creating software for a prototype is only one part of developing a secure and accessible voting system that will actually work in a real-world election environment. We should encourage this sort of experimentation, but avoid leaping to conclusions about whether any particular prototype is best. And we should certainly avoid legislation requiring any particular fix, such as a contemporaneous paper replica, while such experimentation is taking place.
EAC to Hold Hearing on Electronic Voting
The newly constituted Election Assistance Commission held its first public meeting today. Created under the authority of HAVA, the EAC's responsibilities include the development of voting system guidelines. See here for a summary of the EAC's duties.
The commission voted to hold a hearing, within forty-five days, on the reliability and security of electronic voting. Chairman DeForest Soaries commented, "Many questions exist about the reliability and security of electronic voting devices. The Election Assistance Commission is responsible to ensure that every American who wants to vote in November can vote and that their vote is counted. We must have a thorough airing of the critical e-voting issues and American needs to know the result."
My take: This is a positive development. The EAC was created, in part, to deal with the sorts of security concerns that have been raised about electronic voting and to reconcile these concerns with the imperative of promoting voting equality, embodied in HAVA and other laws. It is far better to allow the EAC to deal with the concerns that some have raised about electronic voting, than to try to pass legislation in an overheated election year climate. As HAVA's co-authors urged in their "Dear Colleague" letter earlier this month, Congress should give HAVA a chance to work, rather than rushing to enact legislation that may ultimately prove counterproductive to that law's goals.
Update: Rick Hasen has now posted the EAC's press release.
Newsweek Story on Alternative Verification Methods
Newsweek has a story in its March 29 issue on the electronic voting controversy, which is available on MSNBC here. What's interesting about this story is that they don't just talk about the contemporaneous paper replica, or "voter verified paper trail." Instead, they discuss other means of verifying the accuracy of electronic voting that are being developed. It also notes that Rebecca Mercuri, who first came up with the idea of the CPR, is now expressing concerns that it might not be implemented correctly. She now advocates that the actual count take place not from the electronic ballots themselves, but from the paper replicas of electronic ballots. This suggestion, of course, would reduce electronic voting machines "to little more than ballot printers," as the co-authors of HAVA put it in this letter to their colleagues. It also would destroy the ability of people with visual or reading disabilities to cast their votes independently, since there would be no way for them to verify that the printed-out ballots correspond to their choices.
All of this suggests that it would be a grave mistake to lock into place a particular method of vote verification, as the Holt Bill and the Graham/Clinton Bill requiring the CPR would do. See here and here.
Just Democracy Website Up and Running
Just Democracy is a law student-led group seeking to ensure equality at polling places throughout the country in the 2004 elections. Its website www.justdemocracy.org, is now up and running. As noted here, J.D. seeks to form chapters at law schools throughout the country to monitor and prevent problems in November. They've already got chapters started at several law schools. This is a great initiative and I hope that it emboldens law students throughout the country to play an active role in making sure that we don't see a replay of 2000 in 2004.
California Releases Draft Standards for CPR
California Secretary of State Kevin Shelley has released draft standards for the contemporaneous paper replica (or "voter verified paper trail"), to be required on electronic voting machines effective July 1, 2006. If nothing else, these standards clarify that getting a workable CPR system in place will be no easy matter. It will certainly not be "simple," as some critics of electronic voting have supposed. In particular, Shelley's proposed standards would require that Direct Record Electronic ("DRE") voting machines have attached printers that would print out paper copies of the electronic ballot. These would not be receipts, but would instead have to be printed in such a way that they could be viewed but not touched by the voter. There would also have to be something to store both cast and "spoiled" paper ballot replicas, and readers that can interpret the information on the ballot replicas. In addition, the draft standards generally require that the systems ensure readability, privacy, and accessibility for people with disabilities. They'd also have to allow voters to spoil their ballots at least twice, in the event that the paper replica doesn't correspond to their intended choices.
My take: The unfortunate outcome of Shelley's decision to require a CPR will likely be to dissuade counties from moving to electronic voting systems at all. As I've mentioned previously, there's no certified system in existence that can produce the CPR. Nor is it clear that a workable system that conforms to the parameters set forth in these guidelines can be developed by 2006. As these guidelines clarify, it's going to be complicated, difficult and expensive to implement any DRE system, if a CPR is required. The ultimate effect, therefore, may well be to lead many counties to decide that the game isn't worth the candle and to stick with paper-based systems -- something that will hurt all voters, but especially those who have the most to gain from implementation of state-of-the-art technology, such as people of color and people with disabilities.
Recent Editorials on Electronic Voting
At least a few newspapers have now taken positions in opposition to the baby-with-the-bathwater approach to electronic voting, exemplifed by the proposal by some California lawsmakers to ban electronic voting entirely, discussed, here and here. The San Diego Union-Tribune has this editorial, calling this approach "misguided." It emphasizes what many seem to have forgotten: that lost votes were and are routine with the punch card machines that electronic systems are replacing. The Desert Sun offers a similarly reasonable perspective here.
For a rundown of the debate in California see electionline.org's weekly update. I'm quoted on the conflict between lawmakers seeking to ban electronic voting and disability rights activists seeking to ensure accessible voting for all. Electionline.org also notes a recent editorial from Florida's Sun Sentinel, opposing Rep. Robert Wexler's new lawsuit aimed at requiring that printers be added on to Florida's electronic voting systems.
Just Democracy Website
Just Democracy is the group of law students, discussed here, trying to establish a nationwide network to prevent polling place and registration problems in the 2004 elections. They've now got a webite, www.justdemocracy.org. Although still under construction, it contains an email address from which more information can be obtained.
Problems with Paper Used to Attack Electronic Voting!
Some folks are evidently determined to attack electronic voting, regardless of the evidence, and in spite of its advantages for traditionally disenfranchised voters like African Americans and people with disabilities.
Case in point: Wired News has this story, written by Kim Zetter, regarding problems in counting optical scan paper ballots in the March 2 primary in Napa County, California. Optical scan use paper ballots that voters mark to indicate their choices, similar to a standardized test. In Napa County, a procedural error resulted in the miscounting of paper ballots. In particular, the ink from gel pens used by some voters couldn't be read during the machine count.
What's remarkable about this story is not that there were problems with an optical scan system. To the contrary, we've long known that paper-based systems like optical scans and punch cards can result in problems, not the least of which is a racial disparity in uncounted votes -- something, incidentally, that the Wired story doesn't mention.
What is remarkable about the Wired story is its attempt to convert a problem that occured with a paper-based voting system into an attack on electronic voting! The title erroneously refers to "Lost E-Votes." Worse still, the author of the article cites this as a "string of problems with the new generation of electronic voting machines." But the reality is just the opposite. In fact, this incident -- hardly the first of its kind -- exemplifies the problems with the paper-based systems that electronic touchscreen machines are replacing.
Either the writer of this story doesn't understand the difference between optical scan and electronic voting systems, or she's affirmatively trying to mislead readers by confusing the two. Contrary to the impression that the story conveys, the Napa County problem confirms that paper ballots are susceptible to error. It also shows that, regardless of what system is used, problems will occur if proper election procedures aren't in place.
Palm Beach County Undervotes
Yesterday's South Florida Sun-Sentinel features this story (noted in Rick Hasen's blog yesterday) on blank ballots cast in Palm Beach County's March 9 primary. What's notable about this is not the percentage of non-votes -- which was actually quite low -- but the fact that there was only one item on the ballot, causing some to wonder whether the undervotes were intentional. A total of 289 electronically cast ballots in Palm Beach County didn't register a vote, for a no-vote rate of less than one percent, about half that of Brouward County. It's therefore not at all clear that any of the Palm Beach County undervotes are unintentional, given that some voters will show up at the polls out of a sense of civic duty but then choose not to cast a vote. It's also worth emphasizing that the no-vote rate in Florida's contested 2000 election was about three times higher than what Palm Beach County reports in the March 9 primary.
My take: Should we be worried about this? Probably not, based on what's come out so far. But this does point to the importance of making sure that we invest in voter education and poll worker training before elections, especially with new technology. With all the focus on the machines themselves, it's easy to forget that no system will work well if voters and poll workers don't understand what they're supposed to be doing.
San Mateo Registrar on Proposed California Legislation
Warren Slocum, the chief election official in San Mateo County, has written this post on the recently proposed California bill to decertify electronic voting machines, discussed here. Although Slocum has been vocal in expressing concerns about electronic security and supports the contemporaneous paper replica, he opposes decertification of electronic voting as a "meat cleaver" approach. As he puts it:
The problems experienced in the three touch screen counties [Alameda, San Diego and Orange] were not security violations brought on by hackers and crackers out to do election evil. The problems, as I understand them, result from poor management, poor training, poor equipment and human error. Nevertheless, these are serious issues which need to be addressed immediately. But we should not "throw the baby out with the bath water."Slocum has some interesting suggestions as to how problems can be avoided in November, including the Secretary of State convening a blue ribbon task force and examining "best practices" of jurisdictions that have successfully moved to electronic voting.
My take: Although I disagree with Slocum on whether a CPR (the so-called "voter verified paper trail") is a good idea, I agree that the bill to ban electronic voting machines is a terrible idea. As he points out, it would be far more risky to force the conversion to new voting systems in just a few months. His suggestions for what can be done between now and November also seem constructive. Improving procedures is the key to making sure that we realize the considerable advantages that voters -- especially such traditionally disenfranchised groups as racial minorities and people with disabilities -- stand to gain from implementation of new voting technology.
A group of law students have started a program called "Just Democracy" to promote equal access to the vote in the 2004 elections. The CalTech/MIT Project estimated that, in 2000, between 2.5 and 4 million voters were unable to vote as a result of polling place or registration problems. Just Democracy aims to prevent such problems from recurring, by starting chapters at law schools throughout the country to identify problems at polling places. It doesn't appear that they've got a website up yet, but a more complete description -- along with an email address to write for more information -- appears below.
'Just Democracy’ Project Seeks to Provide Clear Path to the Ballot
In an effort to prevent the confusion and mistakes that marked the 2000 election, a group of Harvard Law School students have launched a project to ensure that 2004 presidential election voters are given proper access to the ballot. The new group, Just Democracy, plans to recruit and place more than 1,000 law students with expertise in election law at what they believe could be high-risk polling places around the nation.
“Whatever one’s politics, seeing eligible, motivated voters wrongly denied their right to participate is an offense to democratic principles,” said Becca O’Brien, a second-year HLS student who will direct the project along with second-year student Micah May. “Our vision for Just Democracy is of a completely non-partisan network of law school chapters. Each will be deeply rooted in its own community, and the chapters will be connected to each other through a national network and a commitment to supporting the rights of every voter.”
Though most of the media attention in the 2000 presidential election focused on the difficulty of attaining an accurate count of the ballots, student organizers contend that a more disturbing problem arose when eligible voters were denied access to the polls. According to a recent study by the Caltech-MIT Voting Technology Project, between 2.5 and 4 million voters were wrongly excluded from the last presidential election. Problems and mistakes that led to the exclusions included voter registration errors and confusion at polling places. Other problems identified in the 2000 election were voter intimidation and improper operation of polling places.
“Just Democracy could make a real difference during the next election,” said Harvard Law School Professor Heather Gerken, an expert on election law. “These students are trying to make the promise of democracy a reality. As someone who teaches election law and cares deeply about democratic rights, I believe this is an important step in the right direction.”
Throughout the next several months, Just Democracy will establish chapters and training programs throughout American law schools. These local organizations will work with community groups to identify specific problem polling places and the appropriate volunteer deployment strategy. On election day, these volunteers will assist poll workers with any questions or problems they might encounter, will hand out state-specific voters’ rights information, and will work to ensure that people who turn out to vote are not wrongly turned away.
Law students interested in starting or joining a chapter of Just Democracy, or anyone interested in contributing or in learning more, can email email@example.com.
My take: I've written before about the danger that problems that surfaced in 2000 will be repeated in the forthcoming elections. See here and here. Among the most vulnerable are people of color, linguistic minorities, and voters with disabilities. While I've focused on voting machines, there's a lot more to the problem than that. The recent issues that have surfaced in San Diego and Orange County demonstrate that having adequate procedures in place is at least as important. The Just Democracy project looks to be a great way for students, whatever their political beliefs, to make a difference in the forthcoming elections by trying to ensure equality in the voting process.
More of the Same from the New York Times
Continuing with it's simpleminded approach to a complicated problem, the New York Times on Sunday published this editorial repeating its support for the contemporaneous paper replica. Oblivious to the fact that there's not a machine certified or successfully implemented anywhere that can produce a CPR, the NYT continues to insist that there this can be implemented in time for the November election. They've still failed to address any of the practical concerns that have been raised about this proposed requirement, much less the civil rights disadvantages addressed in the Leadership Conference on Civil Rights statement opposing the CPR. NYT readers deserve a thorough analysis of this issue, rather than a feelgood proposal that won't solve anything.
California Lawmakers Seek to Ban Touchscreen Voting
Chasing a train that has long left the station, two California lawmakers are pressing to halt plans to use electronic touchscreen voting machines in the forthcoming elections, according to this story in the Oakland Tribune. State Senators Don Perata and Ross Johnson are asking Secretary of State Shelley to decertify these machines and disallow their use in November. If he refuses, they're threatening to press for legislation to prevent the use of electronic voting in November
According to Dr. Henry Brady of U.C. Berkeley, an uncommon voice of reason in this debate: "If we go to paper ballots tomorrow, believe me, we will have problems with paper ballots. We've got to have a little patience and understanding . . . .We also have to be tough with these firms and say they just have to do a better job. We should be pretty upset with them, frankly. This is not good."
My take: Brady's got it exactly right, but is way too kind. In fact, what these senators are proposing is completely nuts. To force counties to switch away from touchscreen voting to a new system at this late date would invite disaster, and suggests that Perata and Johnson have no idea of how complicated it is to implement a new voting system. Do they genuinely expect counties to make a 180 degree turn, obtain paper- based systems, and get these systems up and running smoothly in less than eight months? Ain't gonna happen.
Their proposal would, moreover, throw out the baby with the bathwater. While there have been problems with the implementation of Diebold's Accu-Vote, Sequoia's electronic machine has performed very well in Riverside since 2000, and in other counties. To make matters worse, these senators' proposal would ban the only type of voting system that allows people with disabilities to vote independently, at a time when the California Secretary of State has just been sued for not doing enough to provide accessible voting. See here. Can you spell T.R.O.?
As I noted yesterday, the problems reported in San Diego and Orange County could have been avoided if better procedures had been in place. The contemporaneous paper replica (or "voter verifiable paper trail") that some are urging wouldn't have done anything to prevent the problems that occurred on March 2. People who really care about voting rights should instead be working to ensure that better procedures are in place by November. These might include improved poll worker training and having backup paper ballots at polling places, in case the electronic machines don't work. Yet some electronic voting critics, rather than trying to make things better, are launching into the histrionics that have sadly characterized their ill-advised campaign for the contemporaneous paper replica, deflecting attention from more practical means by which to enhance the security of electronic voting.
Nevertheless, Perata and Johnson deserve credit for honesty. Rather than proposing the false fix of a "voter verifiable paper trail" which would do little to improve security but is really a poison pill for electronic voting, they're being candid about what they really want: to stop a electronic voting, irrespective of its benefits for people with disabiltiies, linguistic minorities and people of color. This is a bad idea, in my view, but it at least suggests the possibility of a more candid debate about the pros and cons of electronic voting.
Orange County Poll Worker Problems
The Los Angeles Times on Tuesday published this report on problems that occurred in Orange County last week, in the implementation of its new electronic voting system. According to the report, poll workers gave many voters the wrong ballot. Henry Brady of U.C. Berkeley, co-author of Counting All the Votes, observes that, "This is a procedures problem more than anything else. It's not a problem with a new kind of voting system." Fortunately, it appears that the races weren't close and the error didn't make a difference.
My take: The Orange County incident highlights the necessity of having adequate procedures in place, particularly adequate poll worker training, before a new system is implemented. Some critics of electronic voting will undoubtedly cite this episode to argue that we should stick with paper-based systems, or require a contemporaneous paper replica for all electronic votes. But this is exactly the wrong lesson to draw. The so-called "voter verifiable paper trail" wouldn't do anything to fix this sort of problem. If the wrong electronic ballot is voted, the paper trail will simply reproduce the error. So too, the problems in San Diego's election, discussed here, wouldn't be solved by a paper trail. What's needed instead, as Prof. Brady notes, are better procedures.
Unfortunately, the paper trail debate has diverted attention from the absence of adequate election procedures, which leads to human errors by poll workers and others. In fact, the CalTech/MIT Report concluded that of the 4-6 million votes lost in the 2000 election, approximately 1 million were lost due to polling place operations. Rather than focusing exclusively on the equipment, we should be focused on ensuring that adequate procedures are in place to ensure that no votes are lost or mistakenly cast.
It's not the machines. It's the procedures, stupid.
Preliminary Report on San Diego Election
The County of San Diego has released this report on the use of the Diebold electronic touchscreen system during the March 2, 2004 election. The report discusses the failure of some of the precinct control modules (PCMs) to boot properly, causing delays in opening some polling places. The PCM is a machine that encodes the cards, which are then given to voters to activate their ballots on the touchscreen units. The county has formed a review team to further investigate the cause of the problem, and make sure it doesn't recur in the November election. The report also notes that the touchscreen units allowed many disabled citizens to vote independently for the first time, and that it allowed voting in three languages (English, Spanish and Tagalog).
Graham and Clinton Introduce Paper Trail Legislation
The A.P. reports that Senators Hilary Clinton and Bob Graham have introduced legislation to require a contemporaneous paper replica (CPR), the so-called "voter verified paper trail," for electronic voting systems. As I noted in a post last week, the principal co-authors of the Help America Vote Act -- two Democrats and two Republicans -- have written a "Dear Colleague" letter opposing such legislation.
According to the A.P. story, Senators Clinton and Graham both insist that this bill has nothing to do with their personal political aspirations. Don't believe it. Mandating that printers be attached to electronic voting machines by November 2004, as their bill reportedly would do, is particularly unrealistic, given that there's no machine certified (much less successfully implemented anywhere) that can do this. If that's indeed what this bill requires, it would suggest that they're not really interested in improving election security but are simply doing this to make political hay.
What's not clear from this article is whether these senators have considered the adverse civil rights consequences of the CPR, discussed here and here. More on this when I actually see their bill.
Voters with Disabilities Sue California Secretary of State
A federal disability rights lawsuit against California Secretary of State Kevin Shelley and four county registrars was filed today in Los Angeles. This lawsuit, AAPD v. Shelley, specifically takes aim at Shelley's November 2003 directive requiring that all electronic voting systems generate a contemporaneous paper replica ("CPR").
Because the complaint doesn't yet appear to be available on-line, I'll describe it at some length. The plaintiffs include individuals with visual and manual dexterity limitations, who assert the right "to cast a ballot independently, unassisted and in secret in the same manner as all other citizens." Also among the plaintiffs are the American Association of People with Disabilities, the California Council for the Blind, and the California Foundation for Independent Living Centers. In addition to Shelley, the defendants are registrars of counties that use paper-based optical scan systems as their principal voting method.
The lawsuit arises from the failure to provide an accessible voting system, that allows citizens with visual and manual limitations to vote independently. It notes that Direct Record Electronic ("DRE") voting machines that allow independent voting are now available, but are not currently provided to all individuals with disabilities who need them. In fact, DRE machines are the only ones that currently allow accessible voting.
Plaintiffs seek nullification of Secretary Shelley's November 2003 directive requiring a CPR (the so-called "voter verifiable paper trail") for all DREs. In particular, the complaint alleges that: "The Secretary of State’s restrictions would require new features for DRE systems that even he recognized in his directive did not exist yet and that will not be approved by 2005, and that would impose discriminatory burdens on disabled voters in violation of federal and state law. " It also notes California Attorney General Bill Lockyer's opinion that requiring a CPR for all electronic voting systems would violate federal and state disability rights laws.
Because Shelley's directive amounts to a prospective decertification of existing DRE systems -- which are the only accessible systems currently available -- it will have the effect of discouraging counties from adopting accessible voting technology. In fact, the complaint alleges, the impact of Shelley's action has already been to cause Los Angeles and Sacramento Counties to suspend their plans to move toward accessible DRE systems.
The lawsuit contains four claims for relief: 1) a claim under the Equal Protection Clause, alleging that the actions of Shelley and the registrars deny the fundamental right to vote of people with disabilities, 2) a claim of disability discrimination under Title II of the Americans with Disabilities Act, 3) a claim under section 504 of the Rehabilitation Act, alleging disability discrimination by federally funded entities, and 4) a claim under the California Elections Code, section 19227(a), that defendants have denied equivalent access to the vote for people with disabilities.
My take: This is a very important test case, one that can be expected to have a nationwide impact on the debate over electronic voting. The lawsuit recognizes what has long been apparent to many civil rights advocates: that calls to require that all electronic voting machines generate a contemporaneous paper replica will inevitably discourage counties from moving to accessible technology. (See my views quoted in Rick Hasen's blog on November 25, 2003, regarding the disability rights consequences of Shelley's directive.) The biggest losers will include not only individuals with disabilities, but also linguistic minorities and people of color, who stand to benefit most from the replacement of the "hanging chad" punch card with electronic systems.
Given that Secretary Shelley adopted the CPR requirement despite the Attorney General's recommendation that it would violate the ADA and state civil rights laws, it can be expected that Shelley will vigorously fight disability advocates in this lawsuit. What's less certain is how counties -- such as Sacramento and Los Angeles -- will react. These counties had planned to convert to DRE systems, but are discouraged from doing so by Shelley's November 2003 directive. Indeed, the California State Association of Counties and registrars in counties that use electronic voting have already been vocal in their opposition to Shelley's November 2003 directive.
If these counties are wise, they'll side with plaintiffs in this lawsuit, and join disabled advocates in seeking rescission of the directive to require that electronic voting systems generate a contemporaneous paper replica. So too, those counties that have already converted to DRE voting should come in on the side of plaintiffs, and oppose Shelley's ill-advised directive.
Finally, any jurisdiction throughout the country that's considering adoption of a CPR requirement should take note of the California lawsuit. Imposing a CPR requirement will likely expose election officials to similar lawsuits under the Americans with Disabilities Act and other civil rights laws.
Could Ohio Be the Next Florida?
Perhaps so, if some legislators from my new home state have their way. According to a story in Friday's Cleveland Plain Dealer, a group of legislators is now urging that the state reverse course, after three years of planning to replace the discredited "hanging chad" punch card.
In the 2000 election, punch cards were used by 70 of 88 Ohio counties. As in other states, punch cards had a no-vote rate significantly higher than any other system. But in contrast to such states as Florida and California that have now upgraded, Ohio hasn't yet replaced its punch cards, making it one of the last bastions of the "hanging chad." Yet Rep. Bill Seitz, a Cincinnati Republican, is suggesting that keeping the punch card might be cheaper than purchasing a better system.
If the Plain Dealer's story is accurate, this appears to have turned into a feud between Republican Secretary of State Kenneth Blackwell and some members of the Republican-controlled legislature. These legislators are using security concerns regarding electronic voting as an excuse to avoid replacement of the antiquated punch card system.
The real losers of this feud will be Ohio's voters, many of whom are likely to be stuck using precisely the same obsolete equipment that caused so many problems in Florida's 2000 election. The biggest losers will be African-American voters, who are disproportionately harmed by the continued use of "hanging chad" punch card systems. (Disclosure: As I've mentioned before, see here and here, I'm co-counsel for plaintiffs in a lawsuit challenging Ohio's continued use of punch card voting machines.) Whatever one thinks of electronic voting, it provides no excuse to stay with the discredited and discriminatory punch card machine. At the very least, Ohio counties should move to optical scan systems that result in significantly fewer lost votes than punch cards.
Worse still, it appears that Ohio is likely to be one of three or four swing states, if the presidential election turns out to be close. And that race is likely to be very close in Ohio, as the major candidates' visible presence here even in these early months attests. Thus, we're faced with the prospect of having another decisive election that is within the substantial margin of error resulting from punch card voting. One would think that Ohio's public officials would be doing everything in their power to avoid a repeat of the 2000 election fiasco.
Let's hope that the Florida nightmare isn't repeated in Ohio. The odds are of course against it. But if it does happen ... well, you heard it here first.
Media Alert: Disability Groups to Sue California Over CPR Requirement
On Monday, four disability rights groups will file suit against California Secretary of State Kevin Shelley and four counties, challenging the decision to require a contemporaneous paper replica ("CPR") for electronic voting systems. The lawsuit, to be filed on behalf of people with visual and manual dexterity limitations, will demand accessible electronic touchscreen machines in time for the November 2004 elections, and will challenge the legality of Shelley's November 2003 directive to require a CPR for all electronic voting machines.
According to a media advisory issued on Friday, "the suit will allege violations of the Equal Protection Clause of the Constitution, the Americans with Disabilities Act, and other state and federal laws." More details will be announced at a press conference on Monday at Loyola Law School, held by representatives of the American Association of People with Disabilities, California Council of the Blind, California Foundation of Independent Living Centers, Western Law Center for Disability Rights, Disability Rights Education and Defense Fund, and the law firm of Howrey Simon Arnold & White.
My take: Secretary of State Shelley has left himself open to precisely this sort of lawsuit by requiring the CPR. By mandating a system that has not yet been certified or successfully implemented anywhere, he provided a strong disincentive for counties to move to accessible touchscreen systems. Moreover, there's no electronic system with a CPR that allows people with disabilties to vote independently. Accordingly, the CPR requirement imposed by Shelley is counter to the accessibility goals of the Americans with Disabilities Act and the Help America Vote Act. It may also violate California state disability access laws, which in some cases go beyond the requirements of federal law. More comments when I see the complaint.
HAVA Co-Authors on DRE Paper Trail
The four principal co-authors of the Help America Vote Act have written a "Dear Colleague" letter responding to proposed legislation that would require a contemporaneous paper replica, or "voter verified paper trail," for electronic voting systems. The letter doesn't appear to be available on the web yet, so I'll quote from it at some length.
The March 3, 2004, letter from Sen. Dodd, Sen McConnell, Rep. Ney and Rep Hoyer begins by stating:
As the principal authors of the Help America Vote Act (Public Law 107-252) (HAVA), signed into law by President Bush on October 29, 2002, we feel compelled to express our concerns about recent legislative efforts that promise enhanced electronic voting system security. Various proposals have been introduced in the House and Senate, but a common feature of these bills is they would amend HAVA to require that all voting systems, including electronic and computer-based systems, produce or accommodate a "voter verified paper record." Not only are such proposals premature, but they would undermine essential HAVA provisions, such as the disability and language minority access requirements, and could result in more, rather than less, voter disenfranchisement and error.
The legislators acknowledge the concerns that have been raised regarding the security of Direct Record Electronic ("DRE") voting systems, but explain that HAVA anticipates such issues and and provides a mechanism for dealing with them. In particular, HAVA creates a Technical Guidelines Development Committee to assist the new Election Assistance Commission ("EAC") in developing standards to ensure the security of DRE voting.
The letter goes on to note that electronic voting systems have a "demonstrated track record" of improving accessibility, particularly for people with disabilities. It asks that HAVA be given a chance to work, instead of rushing to adopt something that would turn the best available systems into "little more than ballot printers" and inhibit the accessibility goals that HAVA was intended to promote. The letter concludes by stating:
There must be voter confidence in the accuracy of an electronic tally. However, the current proposals would do nothing to ensure greater trust in vote tabulations but would be guaranteed to impose steep costs on States and localities and introduce new complications into the voting
Questions regarding voting systems security, as well as many others, need to be examined by the entity responsible for doing so under existing law, the Election Assistance Commission, before Congress begins imposing new requirements, just months before the 2004 presidential and congressional elections, that have not been fully considered. The security of voting technology is a non-partisan issue. We encourage you to allow HAVA to be implemented as enacted and provide those who are charged with ensuring the security of voting systems the time and flexibility needed to get the job done effectively.
My take: This is an excellent letter, which effectively explains why the CPR would undermine the principal goals of HAVA, especially access for disabilities. Let's hope that it stops (or at least slows) the momentum that appears to have been growing for such ill-conceived proposals as the Holt Bill.
Update: The HAVA sponsors' March 3 letter is now available here.
USA Today Op-Ed on Electronic Voting
Don Campbell of USA Today has an op-ed in support of electronic voting. Campbell emphasizes a point that electronic voting's critics have glossed over: the substantial reduction in lost votes that can result from conversion from punch cards to Direct Record Electronic ("DRE") systems. In Georgia, for example, Campbell notes that the uncounted vote rate dropped from 5% in 1998 to 1% in 2002 following the conversion to electronic touchscreen machines. That's not to deny that there are security issues surrounding electronic voting. Nor is it to deny that there are things that can be done to beef up DRE security, including those that the League of Women Voters has suggested and others that I've mentioned in prior posts, see here and here. But the breathless rhetoric of electronic voting's die-hard opponents, including the insistence of some that paper is the only answer, has blinded many to the considerable advantages of paperless DRE technology, exemplified by the statistics that Campbell cites.
Early Returns on New Voting Systems
Many counties in California and other states used new voting systems in yesterday's Super Tuesday election. As is to be expected with implementation of any new voting system, a smattering of problems were reported in counties using electronic voting systems for the first time, according to this A.P. report. So far, most of these problems seem to be the result of human error, such as leaving voting machines unattended, something that presents a risk analogous to leaving ballot boxes unattended.
Yesterday's Los Angeles Times also included a story regarding the transition to electronic voting. This story notes that these systems have audio hookups allowing blind voters to vote without assistance. It also summarizes the security concerns and "conspiracy theories" that some have raised -- though the article doesn't discuss the racial disparities generated by the outgoing generation of paper-based voting systems.
It will be interesting to examine the data from those counties that are making the conversion from paper-based systems -- like the "hanging chad" punch card -- to electronic voting. If 2000 election results are any indication, we should expect to see a substantial overall decrease in unvoted ballots and a closing of the racial gap. More uncertain is what the result of Los Angeles County's use of the "Inkavote" system. This is a type of optical scan system in which voters used felt-tip pens to mark a card that looks very much like a punchcard ballot without the chad. See the L.A. County Registrar-Recorder site for pictures.
While the InkaVote won't result in hanging chads, it may result in some of the same user interface problems that plague punch card systems. These include misaligned ballots, resulting in improper marking, and voters having difficulty checking their votes since the ballots don't have the candidate names on them. This system could therefore generate some of the same disparities -- including racial disparities -- that we've consistently seen with punch card systems. For a very helpful collection of documents regarding punch card performance see this page put together by U.C. Berkeley's Henry Brady.
Unfortunately, we may not get very reliable data from the March 2004 election. Given that there was no real contested race at the top of the ticket (and no I don't count Kerry v. Edwards which has been over for weeks), it will be difficult to ascertain how many unvoted ballots are intentional. We'll have a much better idea of the impact of the shift in voting technology after the November 2004 election.
Op-Eds on Electronic Voting
There's been a lot of attention to Direct Record Electronic ("DRE") voting over the days leading up to this Super Tuesday. The New York Times has yet another op-ed raising the specter of fraud with electronic machines and urging a contemporaneous paper replica, the so-called "voter verifiable paper trial". The latest piece yet again disregards the civil rights advantages of touchscreen technology and the likely result of requiring a CPR: namely, to discourage counties from going to DRE systems that provide considerable benefits to racial minorities, linguistic minorities and people with disabilities.
In this op-ed, Adam Cohen correctly states that new voting technology is creating a "whole new generation of conspiracy theorists." He then proceeds to join in the theorizing. In particular, he asserts that with electronic voting, there's no way to be sure of who won the election.
While I suppose that's theoretically true, the same could also be said of elections conducted with paper-based systems such as punch cards and optical scans. We can't know "for sure" that the ballot boxes weren't intercepted, or that ballot boxes weren't stuffed.
Incidents of fraud and error with good ol' paper are not, by the way, merely the stuff of conjecture like so many of the anti-DRE conspiracy theories floating about, such as the ones regarding Georgia's recent election, noted here and here. These are things that have really happened, and not just in LBJ's 1948 Senatorial campaign. Ballot box lids were found floating in the San Francisco Bay as recently as 2001, while in Brouward County, Florida over 200 ballots were found in a file cabinet four months after a 2002 election. Can we really be sure that every vote was properly counted? And how do we know for sure that paper ballots haven't been lost, added, or replaced in other elections without detection?
The answer, with either a paper-based or electronic system is that we don't know for sure ... but that we can reduce the risks by ensuring that proper procedures are in place. With paper voting, this means establishing protocols to make sure that the ballots are kept secure, something that doesn't always happen but usually does. With electronic voting, that means doing things like testing the machines before during, and after the election to ensure they work as they're supposed to, avoiding hooking them up to the internet, making sure that the certified software is the same as that installed on machines, and using machines that have multiple internal backups.
Cohen blithely asserts that a CPR, which he erroneously refers to as a receipt, would solve the problem. (The reference to a receipt is wrong, since voters wouldn't keep the printed paper replica as they would an ATM receipt to which he falsely compares the CPR -- the whole point of having the CPR is that it's kept in the custody of election officials to serve as an audit trail.) He doesn't address 1) the fact that no CPR system has been successfully implemented anywhere or even certified, 2) the user interface problems that adding a printer would create, 3) the administrability difficulties of adding a printer and the not-so-great results in the two cases where such a contraption has been tested on just a limited basis, or 4) the likely result that requiring a CPR will force counties to stick with inferior paper-based systems that are particularly bad for voters of color and those with disabilities. And all of this is entirely aside from the long history of problems with keeping paper ballots secure.
For a different and more realistic view of elections, see this op-ed by Florida Secretary of State Glenda Hood. As she notes, the process of certifying and implementing technology takes time, if it's to be done right, and can't be done in time for the 2004 elections. Assertions that credible elections are impossible without a paper trail, she contends, rest on an unfamiliarity with the processes that already exist.
I agree with this to a point, though there are undoubtedly things aside from a CPR that can be done to make electronic voting more secure, including those listed above. For more information on such recommendations, see the California Ad Hoc Touchscreen Task Force report, which I discussed here
The Bottom Line: One thumb up for Hood, two thumbs down (way down) for Cohen.