Year: 2008

The Times, they are a Learning

A new editorial on the New York Times lays out the case for election integrity and credibility in: A Tale of Three (Electronic Voting) Elections <read>

Electronic voting has made great strides in reliability, but it has a long way to go. When reformers push for greater safeguards, they often argue that future elections could produce the wrong result because of a computer glitch or be stolen through malicious software. That’s being too nice.

There have already been elections in which it is impossible to be certain that the right candidate was declared the winner. Here are three such races. It is not just remarkable that these elections were run so badly, but also that the flaws are still common — and could easily create havoc in this fall’s voting.

They provide three examples and three lessons. The first and third of which we have not learned well enough in Connecticut

Lesson: Electronic voting makes large-scale vote theft easy. A patch slipped onto voting machines or centralized vote tabulators can change an election’s outcome. Every piece of software must be scrutinized by neutral experts. If there is not enough time, election officials need a backup plan, such as conducting voting entirely on paper ballots…

Lesson: Electronic voting machines must produce a voter-verifiable paper trail for each vote so voters can see that their choices register properly. In a disputed election, the paper, not the machine tallies, should decide who wins…

Lesson: Paper ballots alone are not enough. There must be strong audit laws that mandate comprehensive hand recounts when an election is close.

After the 2000 election debacle, Americans demanded a better system of voting. What we have gotten is new technology with different flaws. If the presidential race is close, this year’s “hanging chad” could be a questionable result on electronic voting machines that cannot be adequately investigated.

Partisan Consultant Behind The Congressional Firewall

Should partisan consultants be totaling our elections? Be provided access behind our firewalls? Managing data for congressional committees? Especially committees involved in voting technology?

Bob Fitrakis, The Free Press: Behind the firewall: Bush loyalist Mike Connell controls Congressional secrets as his email sites serve Karl Rove <read>

One has to wonder about the implications of the premier partisan campaign IT man, steadfastly loyal to the country’s most well-known security-industrial complex and CIA family, serving as the man behind the U.S. Congress’ firewall…

SourceWatch notes that Connell developed the websites for the House Intelligence, Judiciary, Financial Services, Ways and Means, and Administration Committees. According to SourceWatch, Connell teamed up with R. Rebecca Donatelli, Chair of the D.C.-based Campaign Solutions, to form Connell Donatelli Inc. (CD Inc.) as a specialized online advertising agency in July 2004. One of CD, Inc.’s first activities was to become the registrant, administrator and tech organizer for the anti-Kerry group Swiftboat Veterans for Truth’s website swiftboatvetsfortruth.org…

Connell also handled the IT system work for the Bush-Cheney Re-election Campaign and worked for Secretary of State J. Kenneth Blackwell in designing the system that allowed the real time outsourcing of Ohio’s presidential vote count to a Chattanooga, Tennessee server site.

Feinstein to Voters: Waste $, jeopardize Democracy.

Update: Take Action: E-Mail our Senators to oppose S. 3212 <read and send> This is especially important because Senator Dodd is on the committee and the hearing is this week!

The well intentioned but flawed 2002 Help America Vote Act launched the large scale move to electronic voting. Many states have paperless DREs (Touch Screens), some have paper trails of questionable value, and others like Connecticut have true paper ballots with inadequate chain-of-custody and inadequate post-election audits. What could possibly make things worse? The answer is the latest bill from the U.S. Senate, the folks who brought us HAVA. The bill, S. 3212, would add costs and subtract safety

As we stated when the bill was outlined by Senators Feinstein and Bennett, the primary minor benefit of the bill would be to unite voting integrity advocates in opposition. Our friends at Verified Voting have produced an analysis of the bill <read>

A number of troubling provisions require us to urge opposition to S. 3212:

1. S.3212 allows “independent” vote records that would exist only in computer memory to be used to verify electronic vote totals.

2. The non-paper verification methods allowed by S. 3212 would increase the costs and burdens of conducting elections without the benefit of increased confidence and auditability.

3. Language in the bill would exempt from any verification requirement those paperless voting systems purchased before January 1, 2009 to meet HAVA’s accessibility requirements. This would leave millions of voters (particularly those with disabilities) dependent on insecure paperless electronic machines for the foreseeable future.
Continue reading “Feinstein to Voters: Waste $, jeopardize Democracy.”

Audio: Voice Of The Voters – Connecticut Update

Update: Audio is now available of the Nutmeg State update on Voice of the Voters. The Connecticut segment is about 2/3 of the way through just after the weekly update by John Gideon <Audio>

We look forward to more state updates over the coming weeks. So far, Connecticut is the best of the litter. but sadly that only serves to show how far the whole Nation has to go and the risks to democracy we face in November.

Summary report prepared before the broadcast:
Continue reading “Audio: Voice Of The Voters – Connecticut Update”

Witness To A Crime

I have just completed reading Witness To A Crime, by Richard Hayes Phillips. The result of three years of persistent, detailed investigation of the 2004 election in Ohio. This book proves several times over that the election was stolen.

From the book:

“it is unlikely that 14 voters in one precinct in inner-city Cleveland were wavering between John Kerry and Michael Peroutka, and decided to vote for both candidates, thus spoiling their own ballots. It is also unlikely that 14 voters made identical mistakes on their on their ballots. It is more likely that these 14 ballots were punched in advanced…

“In these fourteen precincts, there were 1780 gay-friendly Bush voters and 360 straight ticket gay-friendly Republicans, or the ballots are not real.”

I choose these samples as they give a flavor of the information covered in the book. There are lot of details, there are a lot of numbers, but there are also two stranger than fiction stories: 1) The story of detailed review of reported statistics which don’t add up to anything but fiction along with details only available because a team of researchers copied and photographed ballots in order as they were stored. 2) The story of election officials giving contradictory explanations, false information, destruction of records by “mistake”, delaying records requests, and otherwise obstructing the investigation.
Continue reading “Witness To A Crime”

Debunking Pre-Election Testing Myths

We have said it here as one of our Ten Myths:

Myth #6 – Memory card errors cannot affect the outcome of our elections because election officials conduct pre-election testing of our electronic voting systems.
Reality

  • Pre-election testing of electronic voting systems will detect only basic errors such as junk memory cards, wrong candidates, and machines that simply don’t work. Pre-election testing cannot detect all errors and programming attacks.
  • Computer science tells us it is impossible to test completely. Recent academic reports outline many ways that clever programming can circumvent detection during basic pre-election testing.

    • Don’t take our word for it. An OpEdNews article covers the issue and the myth in more detail. Don’t take their word for it. They have a list of fifty studies including the one from UConn which we highlighted in testimony to the legislature.

    From OpEdNews, Debunking Pre-Election Testing Myths:

    Debunking myths can be a full time job in the election integrity world. Someone recently asserted:

    “As for vote switching, not sure how many times I have to tell you, the election goes thru a logic and accuracy test that proves the votes are counted correctly. There is no vote switching … on ES&S machines. Not sure where you get this information. You shouldn’t believe everything you hear.”

    Malware can easily defeat pre-election testing and certification processes: logic and accuracy tests cannot “prove” that software is free of malicious code. Assertions that no vote switching has ever been shown to have occurred on an ES&S system or any other computerized voting system is explained by the fact that malware (malicious software code) can be self-erasing.

    Tampering Questions: Georgia 2002, Ohio 2004

    The Raw Story: GOP cyber-security expert suggests Diebold tampered with 2002 election

    Georgia 2002     (When surprising results gave the Governorship to Sonny Purdue and Max Cleland lost his re-election to the Senate):

    Spoonamore received the Diebold patch from a whistleblower close to the office of Cathy Cox, Georgia’s then-Secretary of State. In discussions with RAW STORY, the whistleblower — who wishes to remain anonymous for fear of retaliation — said that he became suspicious of Diebold’s actions in Georgia for two reasons. The first red flag went up when the computer patch was installed in person by Diebold CEO Bob Urosevich, who flew in from Texas and applied it in just two counties, DeKalb and Fulton, both Democratic strongholds. The source states that Cox was not privy to these changes until after the election and that she became particularly concerned over the patch being installed in just those two counties.

    Ohio 2004 (The decisive electoral votes to George Bush):

    At the Ohio press conference yesterday, the former McCain adviser said Michael Connell, of the Republican Internet development firm New Media Communications, had designed a system that made possible the real-time “tuning” of election tabulators once Ohio Secretary of State Kenneth Blackwell had outsourced the hosting of vote counting on the same server which hosted GOP campaign IT systems. He said he didn’t believe Connell was behind the alleged fraud, but that he should be considered a key witness.

    BradBlog: GOP Operatives Leading Data Security Expert Joins Press Conference, Case, Notes Fraudulent Patterns That Should Have Triggered Investigation

    This case has the potential to put some of the most powerful people in the country in jail, according to Arnebeck, as he was joined by a well-respected, life-long Republican computer security expert who charged that the red flags seen during Ohio’s 2004 Presidential Election would have been cause for “a fraud investigation in a bank, but it doesn’t when it comes to our vote.”
    …
    “In the 2004 election, from my perspective, on any of the programs we run for any of my credit card clients, the results from the 14 counties, those are the sort of results that would instantaneously launch a credit card fraud investigation or a banking settlement investigation.”

    Spoonamore’s reference to the “14 counties” refers to the so-called “Connelly Anomaly” in which down-ticket candidates got more votes than John Kerry.

    The name comes from the candidacy of C. Ellen Connelly, an African-American woman who was running for the Ohio Supreme Court in 2004. She was endorsed by pro-choice and civil rights groups, and was relatively unknown to Ohio voters, in addition to being vastly outspent by her opponent in the campaign. Yet, somehow, Connelly got scores of thousands more votes than did John Kerry at the very top of the ticket.

    Fairfield Panel: Rosy Views and Thorny Issues

    While I was on vacation, the Fairfield Citizen News had a report covering the panel in Fairfield last week, with added information, apparently obtained from our web site, and an interview with the Secretary of the State. Overall it is a good article fairly covering what was said on the panel. The two hours flew by. I would have loved to have had more time to respond to questions and statements by both the other panelists and the audience. The entire article is well worth reading. Below are some of the statements in the article and my comments: <my opening statement> <Fairfield Citizen News Article>

    Secretary of the State Susan Bysiewicz is in favor of the system. She explained Wednesday, “I firmly believe in Connecticut we have one of the strongest voting security laws in the country the Audit Bill It’s the toughest audit law in the country. After every election we put every precinct on a card and into the bingo box where 10 percent are chosen for public hand count.” She explained that 10 percent equaled 77 precincts in the state. “That’s a lot of ballots,” she said. “We shouldn’t take the machine’s word for it.”

    On the measure of 10% we have the toughest audit law, but on other measures such as reliability, follow-up, transparency, and independence we have a law that is insufficient, unreliable and ineffective. See the 10 Myths.

    Continue reading “Fairfield Panel: Rosy Views and Thorny Issues”

    NY Times: “Check The Vote”

    Once again the New York Times gets it right on electronic voting coming out in favor of post-election audits and reading their local Brennan Center reports for accurate information. <read>

    Computer scientists have shown that it is easy to tamper with electronic voting machines in ways that are all but impossible to detect. The machines also make mistakes on their own. Just this month, the elections supervisor of Palm Beach County, Fla., apologized after machines there failed to count 14 percent of the votes cast in a city commission election.

    The answer is voter-verifiable paper trails: paper records of each vote cast that voters can check to ensure that their preferences were accurately recorded. When these paper records are created, malfunctioning or dishonest machines can then be detected by a careful audit that compares the electronic vote totals with the votes recorded on paper.

    Unfortunately, states don’t require such scrupulous audits. A 2007 study co-authored by the Brennan Center for Justice at New York University School of Law found that most of the 38 states with voter-verifiable paper trails did not even require audits after every election. The states that do have audits do them inadequately.

    An earlier accurate Times editorial was Broken Polls. Contrast this to the mythical world of the Hartford Courant which ignores all negative evidence including our own UConn Research.

    Testing/Inspecting For Democracy Is Too Much Work

    Why bother inspecting restaurants, bridges, trucks, and voting machines? It is just too difficult and costly.

    While Connecticut is way behind in inspecting restaurants as we have been with bridges and trucks, we fit right into the trend evidenced by two national stories – the Election Assistance Commission finds its just too hard on vendors to insist that voting machines actually be certified – and in New York we find that half of the voting machines delivered by Sequoia do not work.

    As summarized by John Gideon <read>

    The Board of Advisors is advising [PDF, pg 7] the EAC, via resolution, that they need to speed-up the certification process for voting systems. They want the system to be what it was under the old, rubber-stamp system headed by the National Association of State Elections Directors (NASED). They want the same system of testing and certification that has resulted in our voting systems failing in many elections and not even being compliant with federal standards.

    Incredibly, the Board’s recommendation to the EAC goes so far as to admit that a failed “common practice” of the past should, apparently, be re-instituted under the newer certification system. “The common practice since the introduction of electronic voting systems,” they wrote, “has been to make hardware and software upgrades based on issues found in the most recent election in sufficient time to improve the voting systems for the next general election.”

    Douglas Kellner as quoted by Kim Zetter of Wired on Sequoia in NY <read>

    Douglas Kellner, co-chair of the New York State Board of Elections, expressed frustration with the vendor, saying it appeared that Sequoia was using the state’s acceptance testing process to find problems with its machines in lieu of a sound quality-control process.

    “There’s no way the vendor could be adequately reviewing the machines and having so many problems,” he told Threat Level. “What it tells us is that the vendor just throws this stuff over the transom and does not do any alpha- or beta-testing of their own before they apply for certification testing. Then they expect that we’ll identify technical glitches and then they’ll correct those glitches. But correction of those glitches is an extraordinarily time-consuming process. And its very disappointing that this equipment is not ready for prime time.”

    But New York has nothing on the Nutmeg State where UConn tests reveal that less than half of our election officials faithfully follow pre-election testing procedures. As for the restaurants it seems Connecticut occasionally still inspects them, but nowhere near as often as required by our own laws.