Kathy Dopp's Blog

As of about 8:10 p.m. E.T. Clinton was slightly leading Obama in the exit polls 51.6% to 47.8%. This was too close to call according to news media. We will be releasing a more detailed commentary of the PA exit polls soon. Current adjusted exit poll results are available here:
http://www.cnn.com/ELECTION/2008/primaries/results/epolls/#PADEM

Regardless of whether Obama or Clinton "wins" the PA Democratic Primary according to PA official machine counts, what we will never know is whether or not PA voters' ballots were accurately cast or counted because 90% of PA voters voted using invisible e-ballots without any voter-verifiable paper ballot record that could be used to verify the counts and for independent auditing. And we will never know what trade secret instructions were compiled into humanly unreadable machine language and were used to secretly cast and count invisible e-votes in PA.

We know that there were many problems with PA's voting systems today. Some of PA's election problems are listed here by BradBlog.com:

http://www.bradblog.com/?p=5918

Some PA election officials are reporting "few problems. However, mainstream news reports show otherwise.

Problems (from most recent to earliest) Include:

4:12pm PT: Wrong machine deliver, machine failure, court-appointed machine inspector denied entrance to polls. http://www.electionjournal.org/

3:54pm PT: Voter-service phone lines jammed across state.
http://www.nytimes.com/2008/04/23/us/politics/22cnd-campaign.html?_r=3&hp&oref=slogin&oref=slogin&oref=slogin

3:45pm ET: Machines down, get jammed, in Bucks County. http://blogs.phillyburbs.com/blog.php/?p=28259&cat=0

3:09pm ET: More machine breakdowns in Philly, voters only allowed provisions if both machines malfunction. http://cbs3.com/topstories/polling.places.pennsylvania.2.705732.html

3:01pm ET: Obama camp reporting problems, difficulty getting machines started, voting probs, Philly official says "all is well". From http://rawstory.com/news/2008/Obama_camp_charging_voting_problems_in_0422.html

12:28am ET: Machines down in African American area of Philly, voters leave without voting, registration probs. Via Email from computer security, e-voting expert, Dr. Rebecca Mercuri...

11:00am ET: Registration roll probs in Philly, provisionals only handed out if both machines malfunction. From Philly Inquirer/Daily News PA Primary Blog... http://www.philly.com/philly/blogs/politics/Vote_early_to_avoid_problems.html

10:50am ET: Both machines break down, 100's leave without voting, Obama's name not on provisional ballot! From Philly Inquirer/Daily News PA Primary Blog...
http://www.philly.com/philly/blogs/politics/More_voting_machine_malfunctions.html

10:45am ET: Electrical probs lead to ES&S iVotronic machine malfunctions, use of paper ballots in Green Tree. From CBS affiliate KDKA 2...
http://kdka.com/local/Poll.Green.Tree.2.705572.html

10:44am ET: Machine breakdowns in Delaware County and South Philly. From Philly Inquirer/Daily News PA Primary Blog...
http://www.philly.com/philly/blogs/politics/18002279.html

9:45am ET: Ballot design probs in Palmer Township (Northhampton County). Delegates not listed under candidates. From Lehigh Valley Live...
http://blog.pennlive.com/lvbreakingnews/2008/04/palmer_township_voter_reports.html

9:43am ET: Fewer machines in Northhampton County this year. From Lehigh Valley Live...
http://blog.pennlive.com/lvbreakingnews/2008/04/voters_might_need_to_wait_in_l.html

9:14am ET: Voting machines malfunctioned, votes lost in Bethlehem. From Morning Call...
http://www.mcall.com/news/local/all-n-l-voters-042208-cn,0,4187550,full.story

6:15am PT: Machines "on fritz" as Philly polls opened, voters given provisionals. From Philly Inquirer/Daily News PA Primary Blog...
http://www.philly.com/philly/blogs/politics/Grumbling_at_the_provisional_ballots.html

To read the Brad Blog article in its entirety see:
http://www.bradblog.com/?p=5918

More reported problems w/ PA election today

http://hotlineblog.nationaljournal.com/archives/2008/04/philly_voting_t.html

http://www.philly.com/philly/blogs/politics/Extended_Hours_Paper_Ballots.html

Voter Action, asked a Philadelphia judge to extend poll hours until 10 p.m., is making its case with voter affidavits. One Richard Brown, of the 5700 block of N. 12th Street in Philadelphia, swore an affidavit that both voting machines at his polling place were broken when he arrived there at 6:55 a.m. It took nearly two hours for the machines to be either repaired or replaced, according to Brown's affidavit, and more than an hour for election officials to produce paper provisional ballots. Roughly 75-100 would-be voters left before that, according to Brown's affidavit. However, the judge denied the request.

---------------------------------------------
BACKGROUND INFO

Blackbox Voting (http://blackboxvoting.org) named PA the worst state to vote in. Other articles re. the flawed PA voting systems include:

Major flaw in State of Pennsylvania online voter registration puts user data at risk
Posted by Nathan McFeters March 18th, 2008
http://blogs.zdnet.com/security/?p=965

Election Fraud in Pennsylvania?
Michael Collins, Sunday, 20 April 2008
http://www.scoop.co.nz/stories/HL0804/S00288.htm

Pennsylvania - State of Denial
Marybeth Kuznik
http://votepa.us/

The Pennsylvania Primary: Democracy of the Gods
Tuesday's Election Will be 'Unrecountable, Unverifiable, and Unauditable'...
http://www.bradblog.com/?p=5916

Pennsylvania's Primary: Paperless and Unverifiable
By Verified Voting Foundation April 15, 2008
http://verifiedvoting.org/

Video - Voting Trouble in PA? (Re. Clueless PA Officials who claim that PA's machines are "HAVA-approved" and that "I got elected on them, so there are no problems.")
http://www.youtube.com/watch?v=lHfnLXAOjDA

Long Lines expected:
Pennsylvania Voter Registration Total Hits Historic High for a Primary Election
http://www.foxbusiness.com/article/pennsylvania-voter-registration-total-hits-historic-high-primary-election_569836_1.html

--
Without any routine post-election independent check of the machine counts, we can assume that PA's election results are inaccurate, even if there were no attempt to manipulate votes, because we've seen that often central tabulators fail to count memory cards or incorrectly transfer counts between internal databases, and most cannot print a report of the individual voting machine counts, so that means that the only way to detect these errors is to manually audit voter verifiable paper ballot records for entire precincts - which is impossible, given the lack of any voter-verifiable paper ballot records on the vast majority of PA's voting machines.

Kathy Dopp

The material expressed herein is the informed product of the author Kathy Dopp's fact-finding and investigative efforts. Dopp is a Mathematician, Expert in election audit mathematics and procedures; in exit poll discrepancy analysis.

http://utahcountvotes.org
http://electionmathematics.org
http://electionarchive.org

History of Confidence Election Auditing Development & Overview of Election Auditing Fundamentals
http://electionarchive.org/ucvAnalysis/US/paper-audits/History-of-Election-Auditing-Development.pdf

The National Election Data Archive (NEDA) urges the Colorado Legislature to pass legislation to favor voter marked paper ballots and to promote ballot marking devices for voters who can benefit from a computer enhanced voting interface; and to require valid independent election audits in Colorado.

According to NEDA, DRE voting machine-printed paper ballot records do not accurately record voter choices because:

DRE voting machine ballot printers jam, run out of paper, may be incorrectly loaded with paper, or otherwise fail to print

The “three strikes you’re out” rule for DRE voting machine printed records, which give the voter only two chances to refuse to approve the machine’s paper ballot, can prevent the most diligent voter from having a valid printed paper ballot record

Studies show that only approximately 30% of voters take the time to verify DRE printed ballot records and that about 30% of these voters are able to accurately verify all their ballot choices correctly. This means that perhaps 10% of voters accurately verifying DRE printed ballots.

There are several ways that programmers can fraudulently switch the votes of the up to 90% of voters who do not accurately verify their DRE printed ballot choices.

* There are several ways for voters or other attackers to undetectably damage DRE printed ballot records and electronic ballot records that would be unobservable until after elections.

Full Letter to the Colorado Legislature:

http://electionarchive.org/ucvAnalysis/CO/Ltr2ColoradoLegislature.pdf

---

NEDA also pointed out flaws of the Colorado auditing procedures and explained some fundamental concepts of election auditing to the Colorado Secretary of State's Office:

http://electionarchive.org/ucvAnalysis/US/paper-audits/CO/NEDA-Response2CO-AuditingRules.pdf

Discrepancies between exit polls and vote counts in the February 5th, 2008 “Super Tuesday” U.S. presidential primary strongly suggest that a vote miscount may have occurred, favoring Hillary Clinton at the expense of Barack Obama, and on the Republican side favoring John McCain at the expense of Mitt Romney.

States with the least likely (most suspicious) election results given their exit polls are Massachusetts and New Jersey in both the Democratic primary election in the Republican primary elections. Massachusetts, New Jersey and New York are just a few states that did not conduct any post-election manual audits to check the accuracy of machine vote counts, leaving their primary elections susceptible to undetected manipulation consistent with their exit poll discrepancies. Not one U.S. State currently uses all the fundamental measures that are required to ensure accurate elections – paper ballots, scientific post-election manual audits, verifiable ballot security, ballot and voter reconciliation, and public access to election records. The Super Tuesday discrepancies between exit polls and the actual vote counts are startling.

Read the FULL REPORT here:
http://electionarchive.org/ucvAnalysis/US/exit-polls/SuperTuesday2008/SuperTuesday2008DemPrimary.pdf

NEW HAMPSHIRE’S DEMOCRATIC PRIMARY ELECTION RESULTS ARE SUSPICIOUS

Pre-election polls projected that Barrack Obama would win the New Hampshire Democratic primary election. An average of seven opinion polls predicted that 38.8 percent were going to vote for Obama, while 30 percent would vote for Clinton. The opinion polls came close to predicting the final results for New Hampshire’s hand-counted votes - 39.2% for Obama and 34.9% for Clinton - but New Hampshire’s Diebold/Premier machine-counted votes reversed the outcome.

The reversal of the machine and hand counts is consistent with programming errors counting votes cast for Obama, for Clinton and votes cast for Clinton, for Obama.

To see this consistency of New Hampshire’s election results with programming error, analysts examined Clinton and Obama vote shares out of votes cast only for Obama and Clinton. Overall, Clinton’s hand count share of such votes is 47.07% to Obama’s 52.93% share and a virtually exact reverse pattern occurs with machine counts where Clinton’s share is 52.95% to Obama’s 47.05%.

A statistical analysis of New Hampshire’s Democratic primary by the National Election Data Archive rules out precinct-size and seems to rule out demographic factors as possible causes for the reversal of Obama and Clinton’s machine and hand-counted results; and shows that the pattern is consistent with vote miscount favoring Clinton.

The National Election Data Archive’s New Hampshire analysis and raw data is posted on the Internet at ElectionArchive.org

http://electionarchive.org/ucvData/NH/DemPrimary2008-PairedPrecinctStudy.pdf

and

http://electionarchive.org/ucvData/NH/

About 80% of New Hampshire ballots were counted by Diebold/Premier optical scanning machines without any post-election manual audits to verify the machine count accuracy.

Press reports hypothesized theories for why Clinton beat Obama in New Hampshire including:

1. the "Bradley effect" (closet racism) that white voters lie to pollsters and "say" they'll vote for a Black, but given a secret ballot don't,

2. the “damsel in distress” theory that Clinton's tears brought women voters out for her,

3. the “good weather” theory, and

4. the “economy was key” theory.

It would be interesting to know why these effects would only occur when ballots are counted by Diebold/Premier voting machines but not when ballots are counted in public view by hand.

The “electronic miscount” theory could be a more plausible explanation for the discrepancies between the opinion polls and the machine-counted results.

Could someone have mis-programmed – by accident or on purpose – the optical scan machines such that Hillary’s votes went to Obama and Obama’s votes went to Hillary?

UPCOMING RECOUNT OF THE TUESDAY, JANUARY 8, 2008 NEW HAMPSHIRE PRIMARY ELECTION

The Secretary of State (SOS) of New Hampshire announced that there will be a statewide recount of the paper ballots beginning on Wednesday, January 16, because presidential candidates Democrat Dennis Kucinich and Republican Albert Howard requested it.

http://www.sos.nh.gov/recount%20press%20release.pdf

It is imperative that, not only ballots, but poll books and absentee and provisional voter records be inspected for a representative sample of recounted precincts. Post 2004 election investigations of "recounted" (and non-recounted) precincts in Ohio discovered large numbers of "phantom" ballots for which no voting records could be found, and disenfranshised voters for whom no ballot could be found - see http://baiman.blogspot.com/

U.S. Representative Kucinich asks for a Recount of NH machine counts in the interest of election integrity

http://www.dennis4president.com/go/homepage-items/kucinich-asks-for-new-hampshire-recount-in-the-interest-of-election-integrity/

Kucinich is asking for donations to pay for the Democratic Primary recount

http://www.dennis4president.com/go/homepage-items/help-defend-the-integrity-of-our-voting-system/

CONCLUSION:

Is the suspicious pattern in New Hampshire’s Democratic primary results caused by voting machine counting? We have no clear idea, because we have no confidence in the unaudited machine vote counting process. Knowing how easy it is to corrupt machine-counted election results, it is appalling that New Hampshire and other states do not routinely conduct post-election manual checks of the accuracy of machine vote counts. Human mistakes and worse are inevitable, and without routine post-election measures to detect and correct mistakes, and without public oversight over security and chain of custody of ballots, inaccurate vote counts and incorrect election results are inevitable.

The full analysis of the National Election Data Archive is available at ElectionArchive.org

http://electionarchive.org/ucvData/NH/DemPrimary2008-PairedPrecinctStudy.pdf

QUESTIONS THE PRESS COULD ASK DURING EACH ELECTION ARE:

Does this state conduct any publicly observable post-election independent manual counts to check the accuracy of machine counts? (This would require voter-created, or at least voter-checked, paper ballots.)

Were the detailed vote count data and raw polling data made publicly available immediately after the election for analysts to detect any possible suspicious patterns?

Were the invisibly-machine-counted vote counts checked, audited, provably correct?

Is there public oversight over chain of custody procedures for election records and ballots?

What are the state’s procedures for securing paper ballot and election records; and how may the public participate?

DATA AVAILABLE FOR ANALYZING THE NEW HAMPSHIRE JANUARY 8, 2008 PRIMARY ELECTION RESULTS

Another Statistical analysis which reaches the same conclusions as the National Election Data Archive, with links to vote count data

http://call-with-current-continuation.blogspot.com/2008/01/statistical-exploration-of-new.html

New Hampshire Secretary of State Web Site

http://www.sos.nh.gov/presprim2008/index.htm

2008 New Hampshire Republican & Democratic Primary Results

http://ronrox.com/paulstats.php?party=REPUBLICANS

http://ronrox.com/paulstats.php?party=DEMOCRATS

The type of voting method each town uses:

http://www.bbvdocs.org/NH/state/Jan-08-votingsystems-NH.txt

http://www.sos.nh.gov/voting%20machines2006.htm

National Election Data Archive
http://electionarchive.org/ucvData/NH/

Pre-election opinion polls:

http://www.realclearpolitics.com/epolls/2008/president/nh/new_hampshire_democratic_primary-194.html

Note: No exit poll data was publicly released after the election unless it had been adjusted to match the final unofficial vote counts. I.e. No public exit poll data is available to use to judge the accuracy of the election results as occurred after the 2004 presidential election.

CNN Exit poll info:

http://www.cnn.com/ELECTION/2008/primaries/results/epolls/index.html#val=NHDEM

FOX News Exit Polls: Women and Seniors Like Clinton

http://youdecide08.foxnews.com/2008/01/08/fox-news-exit-polls-independents-like-obama-mccain/

SOME OTHER ARTICLES ON THE NEW HAMPSHIRE DEMOCRATIC PRIMARY ELECTION

NY Times on the planned NH recount on

http://thecaucus.blogs.nytimes.com/2008/01/11/candidates-push-for-a-nh-recount/

Keith Olberman Covers the Story

http://www.youtube.com/watch?v=rMJB4OCE1fo

Diebold Again: Did Hillary Really Win New Hampshire?
by Dave Lindorff

http://baltimorechronicle.com/2008/011108Lindorff.shtml

New Hampshire to Recount Ballots in Light of Controversy
by Kim Zetter January 11, 2008

http://blog.wired.com/27bstroke6/2008/01/new-hampshire-t.html

Do NH Primary Statistics Show Election Fraud?

http://salonesoterica.wordpress.com/2008/01/09/do-nh-primary-statistics-show-election-fraud/

Informative article w/ good links to exit poll info

http://www.margieburns.com/blog/_archives/2008/1/9/3455407.html

Diebold favors Hillary, hand count for Obama

http://presscue.com/node/38034

Election Integrity Questioned in New Hampshire - Dennis Kucinich formally requests recount
By Michelle Wolski, Epoch Times Florida Staff

http://en.epochtimes.com/news/8-1-11/63935.html

Electronic voting machine results questioned in New Hampshire primary
Dan Kaplan

http://www.scmagazineus.com/Electronic-voting-machine-results-questioned-in-New-Hampshire-primary/article/104145/

Alternet: http://www.alternet.org/story/73551/

Primary Concerns - Hoisting a few red flags about the elections
by Robert C. Koehler, Tribune Media Services, January 10, 2008

http://www.commonwonders.com/archives/col429.htm

Ben Mosley’s Blog

http://benmoseley.blogspot.com/2008/01/do-nh-primary-statistics-show-election.html

Analysis of Hand-counted versus Diebold-counted Precincts

http://i4.tinypic.com/823g1mt.gif

http://my.barackobama.com/page/community/post/rahmainepugh/CWg

Where Paper Prevailed, Different Results, By Lori Price

http://www.legitgov.org/nh_machine_vs_paper.html

European press: It wasn't a miracle - Hillary won via a rigged vote by Michael Carmichael, Global Research, January 14, 2008
http://www.globalresearch.ca/index.php?context=va&aid=7794

PLEASE DONATE

US Count Votes, DBA National Election Data Archive urgently needs your donations if it is to continue its work to try to obtain public oversight over the integrity of election results via mandatory routine vote count audits and public access to election records and election data, and public oversight over ballot security procedures. Donating money to get out the vote efforts, to political candidates, or to efforts to educate voters on issues, makes little difference if votes are not accurately counted. NEDA can not continue its efforts without funding.

http://electionarchive.org/fairelection/donate.html

THANK YOU. PLEASE PASS ON THIS ANNOUNCEMENT.

CONTACT: Kathy Dopp kathy@electionarchive.org 435-658-4657

A new analysis using Utah's 2004 General Election results, finds that obtaining 95% confidence-level in federal and state election outcomes only requires auditing a reasonable 5.5% of all precinct races overall.

The National Election Data Archive has created a Powerpoint presentation showing simply "How to Determine Initial Sample Size for 95% Confidence-Level Vote Count Audits"

Mandatory Post-Election Vote Count Audits

and a smaller pdf version of the powerpoint presentation:

Counting overall 5.5% of Utah's precinct-race counts using this method achieves a higher confidence-level (95%) at lower cost than any flat manual election audit such as the 10% audit Connecticut uses.

Utah's overall audit rates to achieve 95% confidence-level for the 2004 general election are:

For all federal and state races audited - 5.5% of precincts for all federal and state-level elections.

Federal & Statewide Races total audit required - average 3% of precincts

State Senate Races total audit - average 11.2% of precincts

* State House Races total audit - average 30.9% of precincts

The minimum required audit rate for a race is 1.5% (due to Utah's having total 29 counties and 1958 precincts) and maximum audit rate is 100% for one close Utah State House race. Audit amounts will vary state by state depending on the number of close races and how many vote counts there are in each race.

Typically a larger overall percentage of precincts must be audited, but not a larger "amount" of precincts, for races with fewer total number of precincts. The last slide in NEDA's powerpoint presentation explains why a flat percentage audit of precincts (or other vote counts) - the approach adopted by most states currently - is mathematically incorrect.

A general rule of thumb that can be applied without using any fancy mathematics was suggested by M.I.T. Professor Ronald Rivest: Estimate the audit sample size for any race by dividing one (1) by the margin (in percentage of ballots) between the winning candidate with the least number of votes and the runnerup (the losing candidate with the most votes). This easy calculation (1/margin%) is not precisely accurate but provides a confidence-level of roughly between 70% to 100%. Eg. if the margin is 1%, then manually audit 1/0.01 = 100 vote counts; and if the margin is 50%, then manually audit 1/0.50 = 2 vote counts. More precise methods to determine vote count audit sample sizes are described in the National Election Data Archive's powerpoint presentation.

The powerpoint presentation also includes a description of a new method by Aslam, Popa, and Rivest for weighting random selection of precincts by the amount of possible error that the precinct could contribute to the margin between the winning and losing candidates.

NEDA is also releasing a spreadsheet to make it easy to see how to calculate vote count audit sample sizes using unofficial election results:

MODEL LEGISLATIVE PROPOSAL FOR 95% CONFIDENCE-LEVEL VOTE COUNT AUDITS:

After developing legislative language for election auditing since January 2008 in consultation with various experts, the National Election Data Archive finalized a legislative model for election auditing:

Legislative Text Requiring 95% Confidence-Level Election Audits.

The National Election Data Archive recommends combining its bill text with a
Legislative Request for PUBLIC ACCESS TO ELECTION RECORDS:

http://utahcountvotes.org/legislature/PublicAccess2ElectionRecords4Utah.pdf

Currently no state yet subjects its election results to scientific independent election audits.

The 95% confidence-level vote count audit legislation proposed:

1. uses a flexible definition for "vote counts" so that precincts, machine counts, batch counts or even individual ballots can be audited - depending on the capabilities of the voting systems (most systems are not auditable on the ballot level currently).

2. uses scientific definitions for "random selection" and "confidence-level" to flexibly permit any valid auditing procedure that will assure that at most 5% of incorrect election outcomes would be certified (since we expect far fewer than 100% of outcomes to be incorrect, this assures that well more than 95% of election outcomes are correct)

3. specifies procedures ensuring that audits cannot be manipulated or overcome

4. specifies responsibilities for auditors and election officials, and establishes an expert State "Election Audit and Recount Committee" that is appointed by expert department chairs of math, computer science, and political science departments

5. specifies publicly observable manual counts and random selections

Colorado SOS Announcement
http://www.bradblog.com/Docs/CO_CertificationResults_121707_Final.pdf

Overview - Premier/Diebold
http://www.elections.colorado.gov/WWW/default/Voting%20Systems/certifications/2007/Premier%20Election%20Solutions.pdf

Detailed Report - Premier/Diebold
http://www.elections.colorado.gov/WWW/default/Voting%20Systems/certifications/2007/01-A2_project_overview_binder_premier.pdf

This is an initial assessment of the CO SOS study of Diebold/Premier voting systems by Computer Scientist David Webber.

From: David RR Webber (XML)
Date: Dec 19, 2007 1:35 PM
Subject: RE: [vote-technology] Question re. CO Voting Study & Diebold

Kathy,

OK - the detailed report reveals that Diebold was given a conditional "Yes" pass on accuracy and security - but did in fact fail or partially fail many tests in those categories. Overall it appears they were not approved for use because of failing in other areas.

So it seems inaccurate to say they passed all the tests.

Eg:

"At some point, the Testing Board had to come to a final conclusion with respect to GEMS. They chose at the end to change the "suspended" items to a "fail" status. That approach is perfectly valid in a test methodology, under the condition that it is applied consistently. In reality, if Diebold does not rectify the problems associated with its GEMS election management software, it will not be able to provide any of the voting devices.

Those devices require GEMS for use in an election".

For me - the tests also did not prove the Diebolds DREs are safe. The Smartcards remain a huge area of concern. Just because the ones used in testing passed - you have no idea what is on the encrypted smart cards used for a live election - nor how they effect balloting. The smartcard provides an almost perfect means of introducing changed behaviour on election day - that is practically undetectable.

The VVPAT printer records offer some recourse - but would the following tests detect descrepencies between the VVPAT records and the digital counting records that could prove tampering / malfunctions?

BOTTOM LINE - if Diebold was an airplane - it would not be getting FAA approval for passenger use - although you could probably taxi it around the runway OK!!!

DW

Why Did the Hotel Hosting the “2007 Post Election Audit Summit” Bar Election Audit Expert Kathy Dopp (myself) from Talking to Conference Attendees?

Since June 2005, I have developed new election auditing mathematics and procedures in collaboration with statisticians and other mathematicians. My research work has tried to advance the state of election auditing by developing effective auditing methods , by identifying the critical elements that distinguish effective from ineffective methods, and by bringing attention to certain deficiencies and ethical conflicts in other organizations’ election auditing work. I founded and direct the National Election Data Archive, a 501(c)(3) group of volunteer statisticians and mathematicians devoted to ensuring accurate election outcomes.

I registered online to attend the 2007 Post Election Audit conference in Minneapolis, MN, booked air travel and made hotel reservations; but the next day the conference organizers informed me that that I was not in fact registered saying that there was no room left at the conference. I resolved to go anyway, to continue my work with others to ensure that election outcomes are an accurate reflection of voters’ choices, even if I was not permitted to participate in the conference workshops. The 2007 Post Election Audit conference was sponsored by The American Statistical Association, Verified Voting Foundation, The Brennan Center for Justice at NYU School of Law, Common Cause, Citizens for Election Integrity Minnesota, and The Florida Voters Coalition.

On the first day of the conference, October 25th, as I entered my Embassy Suites hotel room, hotel staffers called to tell me that I was not permitted to attend the conference. Before I had spoken to one conference attendee, hotel staffers approached me and told me that I was not permitted to speak to any conference attendees in hotel common areas, that I was not permitted to be in some common areas of the hotel near to the conference area, and that I was not allowed to provide any conference attendees with any information that I had prepared on election auditing. I knew several conference speakers and attendees and planned to confer with them. However, when I tried to speak to anyone I knew who was walking through a hotel common area, the hotel manager followed me and stood between me and that person, preventing any conversation by blocking visual contact and talking, stating to all present that I was not allowed to give them the information on election auditing.

While at the hotel, the hotel did not receive any complaint against me, nor did I violate any ordinance. Yet the conference hotel manager dogged my footsteps and while I waited in line at the front desk to sign up for Internet access, evicted me from the hotel when he saw a conference attendee ask for, take, and try to put away the one page election auditing information.
http://electionarchive.org/ucvAnalysis/US/paper-audits/CriticalAuditMethodology.pdf
Then the hotel manager ordered me to sit in a chair and not move, called for police to escort me to my hotel room to pack my belongings, and made me sign a “trespass” form (I needed a ride to another hotel).

Here is my written report of events: http://electionarchive.net/docs_pdf/info/US/AuditingSummitPoliceComplaint.pdf The hotel told police that I was “expected to disrupt the conference”. Why? I have promoted election auditing at election industry conferences since 2005 without causing any disturbance, and had no intention to disrupt this one.

Why Did Conference Organizers Seek to Exclude Me, and What Possible Reasons Would Any of The Organizers Have to Treat Me In This Manner?

Prior To the 2007 Post-Election Audit Summit Some of Its Sponsors & Speakers

Recommended Election Audit Procedures and Quantities Insufficient to Detect Vote Fraud at a Given Confidence Level

Promoted the Use of Un-Auditable Voting Systems

Replicated Others’ Election Auditing Work without Attribution

Claimed that 2004 Presidential Election Exit Polls Are Not Consistent with Vote Fraud

....

Conclusion:

I have consistently tried to work with the 2007 Post Election Audit conference sponsors to achieve a common goal and agrees with them on most election auditing fundamentals. There was basis for several conference organizers, whose analyses and recommendations I had critiqued, to seek to avoid debating me in a public forum, though the interests of scientific inquiry and policy development are best served by debate which is open to those with appropriate qualifications. But was there any basis for the a priori assumption that I would ‘disrupt’ the conference, rather than simply sharing my information and recommendations with willing listeners? Certainly I had not ‘disrupted’ any of the many previous election industry conferences and hearings I had attended. The field of election auditing is under development and much work remains to be done before transparent effective election auditing methods can be implemented. Work in the fledgling field of election auditing must be done with competence, with attention to detail, and with integrity.

The conference organizers have many questions to answer. Which organizations(s) arranged for, authorized, sanctioned, or knew in advance of what was told to hotel staffers to cause them to treat me in such a manner? Will persons viewed as potential critics attending future events sponsored by these organizations be treated in the same manner? If these organizations do not approve of my treatment, then what steps have they taken to ensure that such treatment will never occur again at events that their organizations sponsor?

It appears that rather than welcoming potentially critical views, particularly those which tried to improve upon their methods or which suggest that the election fraud phenomenon is more ‘serious’ and ‘actual’ than the conference organizers are willing to admit, they chose to preempt and stifle any debate by excluding and removing from the conference environs one who espoused such views. This is a great and chilling length to go to in order to ensure conformity of belief. The trampling of rights of peaceful expression is a dangerous road to travel, particularly in the America of 2007, where civil liberties have been undergoing erosion.

To read the entire briefing see:

http://electionarchive.net/docs_pdf/info/US/MathematicianOustedFromAuditConferenceHotel.pdf

----

Contact the hotel or one of the sponsoring organizations to complain:

Embassy Suites Hotel, MN Airport – Manager William Gheen - phone 952-854-1000

Hilton Hotels Corporation – phone 800-445-8667 or 310-278-4321 or email april.wilson@hilton.com

Verified Voting http://www.verifiedvoting.org/contactus.php

The Brennan Center http://www.brennancenter.org/article.asp?key=30

The American Statistical Association http://www.amstat.org/index.cfm?fuseaction=main

Common Cause phone 202-833-1200

School of Law – Boalt Hall (conf. speakers) http://www.law.berkeley.edu/clinics/samuelson/faculty.html

NYU Law School (The Brennan Center) Compliance Officer http://www.nyu.edu/compliance/officer/

Stanford University (Verified Voting) http://www.stanford.edu/dept/ombuds/how_we_help.html

Citizens for Election Integrity Minnesota http://www.ceimn.org/contact_us

The Florida Voters Coalition http://www.goallthewayflorida.com/html/contact_us.html

National Science Foundation (NSF funds some of sponsors’ work) http://www.nsf.gov/oig/hotline.jsp

Office of Management and Budget (NSF funded projects) http://www.whitehouse.gov/omb/contact.html

Expert on Election Auditing Booted Out of Conference Hotel at Election Auditing Summit
October 27, 2007

Friends

I am writing this from the Holiday Inn Hotel in Bloomington, MN.

As you may know I am the Executive Director of the National Election Data Archive, a group of volunteer statisticians and mathematicians devoted to the integrity and accuracy of election results. I have a masters of science degree in mathematics from the University of Utah.

Unfortunately, on October 25th, 2007 I was booted out of the Embassy Suites hotel where the 2007 Post-Election Auditing Summit was held - although I had not violated any ordinance. I had not even tried to walk into the conference area: and had been flawlessly polite, considerate, and friendly to everyone. The hotel management had not received the tiniest complaint about me. Yet I was literally hounded and harassed from almost the moment I entered the conference hotel. Obviously the conference organizers had planned for my disrespectful treatment prior to my arrival.

Here is my report of events that I filed with the Bloomington, MN police:
http://electionarchive.net/docs_pdf/info/US/AuditingSummitPoliceComplaint.pdf

This mistreatment reflects a lack of respect for civil rights and an intolerance for openness and dissent on the part of the organizers of the 2007 Post-Election Auditing Summit.

No one can say that I did not give it the old college try to work WITH the organizers of the conference on election auditing.

Only an expensive civil legal action would be able to uncover which conference organizers arranged for my mistreatment at the election auditing conference and what fabrications they told the hotel management in order to convince them to harass me from the minute I walked in the door of the hotel.

On the plane on the way to the Summit, I had time to finish reading one of the election auditing papers of one of the presenters at the Post-Election Auditing Summit, and have finished writing one correction to his work in a concise 3-page paper, something which I had hoped instead to discuss with him personally at the conference. Berkeley's Philip Stark successfully re-derived a formula that I had previously derived in July 2006, but he used a bound for his calculations which was too high for the purpose of ensuring correct election outcomes. Stark's methods for analyzing exit poll discrepancies may prove very useful to election auditing in the long run.

http://electionarchive.org/ucvAnalysis/US/paper-audits/Corrections2PhillipStark.pdf

-----------

What is it that the organizers of the 2007 Post-Election Auditing Summit (http://electionaudits.org) do not want you to know?

There will be a follow-up paper to briefly describe some prior election work by some of the conference organizers and speakers that may possibly explain some of the reasons why its conference organizers, prior to my arrival, planned measures to prevent me from speaking with any conference attendees, including with state and county election officials and legislators.

As you may be aware, I am an accomplished expert on election auditing methods and should rightly have been a speaker at any conference on the election auditing topics and I certainly should have been permitted to attend the conference. As a human being, even without any expertise, I should have been permitted to freely speak with conference attendees in the public areas of the conference hotel.

Here is a chronology of virtually all work to develop new methods to audit elections that has been done since 1975:
http://electionarchive.org/ucvAnalysis/US/paper-audits/History-of-Election-Auditing-Development.pdf

Here is my informational handout on election auditing that conference organizers apparently did not want the conference attendees to receive:
http://electionarchive.org/ucvAnalysis/US/paper-audits/CriticalAuditMethodology.pdf

but perhaps this information is what conference organizers did not want attendees to learn about:
History of Confidence Election Auditing Development & Overview of Election Auditing Fundamentals
http://electionarchive.org/ucvAnalysis/US/paper-audits/History-of-Election-Auditing-Development.pdf

and also this information about two of the conference speakers prior incorrect or unethical works
http://electionarchive.org/ucvAnalysis/US/IncorrectElectionDataAnalysis-06.pdf

The National Election Data Archive (NEDA) published a history of the development of election auditing methods from 1975 to present including an overview of election auditing fundamentals.

According to Kathy Dopp, NEDA’s Executive Director, election auditing experts are asking two fundamental questions:

“What size manual count is needed to detect machine vote count error which could produce an incorrect election outcome?”

and

“Given any discrepancies between machine and manual counts found in an audit, what confidence is there that the election outcome is correct, and should we certify the election or expand the audit?”

However, according to Dopp, there is disagreement among election auditing advocates over which methods to use to calculate the audit sample sizes. Some groups have widely disseminated recommendations on election auditing which would result in insufficient audit samples whenever there are misrecorded undervotes or overvotes and such samples would be insufficient to determine, for instance, if the contested 2006 Sarasota County Florida Congressional District 13 race was correctly decided where there were 18,000 undervotes (no votes recorded in the race).

NEDA’s chronology "History of Confidence Election Auditing Development (1975 to 2007) & Overview of Election Auditing Fundamentals" is available online at its web site http://electionarchive.org

Dopp says that an exhaustive review of election auditing literature shows that election auditing experts agree on many issues such as

election audits should be designed to provide 99% probability that voters’ intended choices correctly determine election outcomes;

election records necessary to verify the integrity of the audit must be publicly available

manual audits and random selections must be publicly observable, and

the public must be able to verify the secure handling of paper ballots and election records.

NEDA's report, which took several months to create, describes prior election auditing work and the work that remains to be completed before confidence election audits can be implemented, including the creation of an illustrated manual & a toolkit with step-by-step instructions for conducting election audits, including training materials, and describing possible methods for auditing unofficial election outcomes and securing ballots and election records for different types of voting and election systems.

Some states will need to pass legislation before meaningful election audits can be implemented. For instance Florida needs to increase the time period between Election Day and the date when election results are certified to allow adequate time for audits; Ohio needs to allow independent auditors to handle ballots during the canvass period; and Utah needs to allow public access to election records and detailed vote counts necessary for valid audits.

NEDA urges everyone to read its report and to work together to implement confidence election audits. The full report is posted online at:

History of Confidence Election Auditing Development (1975 to 2007) & Overview of Election Auditing Fundamentals

http://electionarchive.org/ucvAnalysis/US/paper-audits/History-of-Election-Auditing-Development.pdf

One page (2-sided) hand out for election officials on election auditing:

http://electionarchive.org/ucvAnalysis/US/paper-audits/CriticalAuditMethodology.pdf

--------------
PLEASE TAKE A FEW MINUTES TO TAKE THESE ACTIONS NOW:

Please pass along copies of NEDA’s “History of Confidence Election Auditing Development (1975 to 2007) & Overview of Election Auditing Fundamentals”
http://electionarchive.org/ucvAnalysis/US/paper-audits/History-of-Election-Auditing-Development.pdf

to your own county and state election officials and to your state legislators and ask them to implement election audits in your own county and state.

Roster of State Election Offices
http://electionarchive.org/ucvInfo/US/StateElectionDirectors.pdf

and to your US Representatives and US Senators
http://www.house.gov and http://senate.gov

and please ask them to implement legislation requiring election audits of federal elections such as H.R.811 or S.559.

Twenty eight election reform groups from all over the U.S. publicly support "The Voter Confidence and Increased Accessibility Act of 2007" (House Resolution 811) and have responded to the National Association of Counties' (NACo) concerns about H.R. 811.

Read their response to NACo here:
http://electionmathematics.org/VoteYesHR811.pdf

----

NACo Statement
“Requires every county in the nation to replace current equipment”

Fact
Most counties would not have to replace current equipment. Optical scanners meet the bill's requirements – as is. Some existing ballot marking devices can be easily upgraded before 2012. The 2012 deadline gives ample time to develop, test, certify and implement improvements for voters with disabilities.

NACo Statement
“Multibillion-dollar unfunded mandate“

Fact
Current authorizations are sufficient to cover the costs, including the 2008 requirements for voter verifiable paper ballots and audits. HR811’s CBO score says authorizations are sufficient and the bill authorizes appropriation of such sums as may be required.

NACo Statement
“Preempts state laws allowing a reel-to-reel paper trail“

Fact
The bill allows reel-to-reel paper through 2011. Reel-to-reel paper roll ballots are proven to be prone to errors and paper jams, violate ballot secrecy, are not durable, and permit hackers to manipulate manual audits.

NACo Statement
“Requires many counties to replace their equipment twice – once in 2008 and again in 2012“

Fact
Counties required to replace paperless voting systems by November 2008 may purchase equipment today which will meet the 2012 requirements, such as optical-scan paper ballot systems and upgradeable ballot marking devices or ballot printers.

NACo Statement
“Requires parallel voting systems – new electronic systems that don’t yet exist along with sufficient pre-printed paper ballots for all voters“

Fact
No parallel systems are required. Existing optical scan paper ballot voting systems are auditable, less costly, simpler to operate, can be simply upgraded to meet the bill’s 2012 requirements for voters with disabilities, and are less vulnerable to Denial of Service attacks, power outages, electronic failures, and hacking.

NACo Statement
“Imposes cumbersome and untested audit process which postpones the certification of election results“

Fact
Manual independent auditing is a routine procedure in virtually all fields, can be conducted in a few days, and will improve the accuracy of election outcomes and increase voter confidence.

NACo Statement
“Prohibits recounts using any automated equipment“

Fact
The bill does not prohibit machine counts, but it does require that between 3% and 10% of precincts be manually counted to ensure the accuracy of electronic tallies.

NACo Statement
“Prevents the use of vote centers and early voting unless county officials print thousands of different ballot styles in multiple languages for each location“

Fact
Many States currently use optical-scan paper ballots for early voting, accommodating multiple languages in multiple precincts. New Mexico uses a "ballot on demand" system with great success to eliminate the need to print ballots ahead of time.

NACo Statement
“Ignores EAC standards and voting system certification“

Fact
HR811 does not change Federal certification or voting system standards. The Election Technology Council says that it takes 4 to 9 months to Federally certify new products. HR811 provides four years, plenty of time until 2012, to certify and implement accessibility improvements. Federally testing and certifying software prior to an election does not ensure its proper functioning during an election; and States do not use certified software to count votes because ballot definitions which control vote counting are never Federally tested or certified. Hence the focus is on audits.

---

SOME GROUPS SUPPORTING "The Voter Confidence and Increased Accessibility Act" (HR811) INCLUDE

The National Election Data Archive
Verified Voting
VoteTrustUSA
The Electronic Frontier Foundation
Brennan Center for Justice at NYU School of Law
Southern Coalition for Secured Voting
People for the American Way
Common Cause
Move On
Americans United for Democracy Integrity, & Transparency in Elections, AZ (AUDITAZ)
Voting Integrity Alliance of Tampa Bay (VIA Tampa Bay)
Sarasota Alliance for Fair Elections (SAFE)
Florida Voters Coalition
Count the Vote (GA)
Defenders of Democracy (GA)
Georgians for Verified Voting
Iowans for Voting Integrity
Secure, Accessible, Verifiable Elections - SAVE Our Votes (MD)
Citizens for Election Integrity Minnesota
Coalition for Peace Action (NJ)
North Carolina Coalition for Verified Voting
Voting Matters/Oregon
VotePA
VoteAllegheny (PA)
Election Reform Network (PA)
Concerned Voters of Centre County (PA)
Gathering To Save Our Democracy (TN)
Utah Count Votes
Doug Jones, Computer Scientist, Iowa University
Aviel D. Rubin (Avi), Computer Scientist, John Hopkins University
Douglas A. Kellner, Co-Chair, New York State Board of Elections
Barbara Simons, Former President of the Association for Computing Machinery

---
for complete doc with endnotes:
http://electionmathematics.org/VoteYesHR811.pdf

Please contact your US US House Representative in support of HR811.
http://www.house.gov

Ask your Representative to Vote "YES" on HR811 and "NO" on any unfunded mandate amendment that would seek to gut its 2008 deadlines for paper ballots and election audits.

Support Clean Elections
http://electionarchive.org/ucvAnalysis/US/SupportCleanElectionsIn2008.pdf

Important Facts About the Voter Confidence and Increased Accessibility Act
http://electionarchive.org/ucvInfo/US/FactsAboutHR811.pdf

"First U.S. Scientific Election Audit Reveals Voting System Flaws, But
Questions Remain Unanswered"

Rather than republish my article here, please read it on one of these two international web sites, including Scoop in New Zealand, the same web site that got me started working on election integrity back in March 2003.

http://www.scoop.co.nz/stories/WO0705/S00162.htm#a

or here:

http://peacejournalism.com/ReadArticle.asp?ArticleID=20699

or here:

http://electionarchive.org/ucvAnalysis/OH/CuyahogaElectionAudit.pdf

We recommend splitting current election reform legislation into several separate bills which each address a few topics . In the interest of brevity, a list of detailed comments follows this one-page list of recommendations.

Immediate Measures for True and Accurate Elections

1. Manual Audits: Require sufficient verifiable manual audits of election results to ensure that outcomes are correct for all federal races. Provide funds for Audits and Auditable Voting Systems. Require replacement of paperless un-auditable voting systems. Require State Election Audit and Recount Committees, and create a U.S. Election Audit and Recount Committee whose functions would include approving election audit and recount procedures for federal elections; and setting standards and reasonable time frames for state auditable, audit, and voter service reports (see item #5 and definitions).

2. Security Precautions: Outlaw Wide Area Network connections to, and wireless capability in, voting equipment; prohibit voting or transferring any voted ballots through any electronic network; and require states to make their security procedures available for public review and input.

3. Prohibit Practices that Disenfranchise Voters: See a specific list in “Detailed comments” section.

Sunshine Provisions and States’ Rights

4. Public Oversight Of Elections: The public can help to ensure the integrity of elections and audits and prevent voter disenfranchisement only if the public has Access to Election Data and Records: In reasonable time frames, require election officials to make publicly available in original paper and electronic form all election data and election records that could reveal fraud or errors in elections or are necessary to verify voter service reports and manual audits, prior to certification of results; and Public Right to Observe: Require jurisdictions to allow citizens to observe all aspects of elections. Election Monitoring Website: Create a GAO website for publicly displaying the auditable, audit, and voter service reports from the states.

5. Voter Service Reports: Require states to submit timely reports of detailed election data that can be used to measure voter disenfranchisement and voter service levels.

6. Reallocate the Functions of the US Election Assistance Commission (EAC): Preserve states’ rights and do not reauthorize the US EAC.

Long-Term Improvement Of Voting Systems

7. Require Voting Technology with Disclosed Software, Security, Audit-ability, Privacy, and Independent Ballot Verification for Voters with Disabilities: Allow ample time for standards-setting including public input and prioritization of possibly conflicting requirements; development of enforcement, testing, and monitoring systems; and for development, purchase, and training cycles; and for development and adoption of State Implementation Plans. To improve existing voting systems, the entire sequential process of setting standards, product development and implementation could take five to ten years, and federal requirements should enable jurisdictions to budget for voting equipment life-spans of at least 10 to 20 years.

Enforcement

After reasonable time frames, provide swift and certain penalties when an election jurisdiction fails in a transparency, auditing, or reporting obligation; or disenfranchises its voters. The goal is for election records and auditable, audit, and voter service reports to be available for public review and oversight prior to certification of election results and prior to swearing in of federal office-holders.

Read the full proposal for federal election integrity legislation here:
http://electionarchive.org/ucvInfo/US/EI-FedLegProposal-v2.pdf

Some election integrity groups are recommending that commercial off-the-shelf (COTS) software on voting systems be exempt from public disclosure requirements of current federal election reform proposals. Are they advocating for voters or vendors?

FACT 1: “Definition: A component of a computer voting system is considered COTS (Commercial-Off-The-Shelf) if it is (1) general purpose (e.g., operating system software, hardware driver, or database system), (2) not modified, configured, or customized in any manner for voting use, (3) available for sale to the general public”

FACT 2: COTS software could be trade secret, fully publicly disclosed, or open source (free publicly disclosed software).
(Ed: For simplicity, we ignore the variety in types of licensing agreements for publicly disclosed software. OpenVotingCosortium.org has publicly disclosed voting software.)

FACT 3: There is insufficient legal means to require public disclosure of all trade secret COTS software on voting systems.

FACT 4: There is no technical means to ensure that undisclosed trade secret COTS software does not contain malicious or buggy programs that could deny voting rights or manipulate vote counts. Since COTS software specifications are written by external sources, future changes to the product will not be within control.

FACT 5: Today, there is no practical method, even when the software is publicly disclosed, given current voting system design and development methods and the variety of components within one model, to ensure that a particular voting system software actually ran on a voting machine during an election.
(Ed: Some method of public oversight must be developed, for the government serves at the people’s expense and are our employees, and not vice versa.)

FACT 6: There is no infrastructure (equipment, systems, and technologists) available to verify that any voting system software actually ran on a voting machine during an election.
(Ed: An effective infrastructure could possibly be developed over time if secure voting machines were developed with public disclosure and accountability in mind.)

FACT 7: Virtually any voting system software could be manipulated to rig elections or to disrupt service to voters. Trade secret COTS software is manufactured in China, Canada, India, and many other countries from which hacking attacks against US military and government computers have already occurred. Trusting trade secret COTS software is trusting tens of thousands of employees in thousands of companies all over America and the world.

FACT 8: Most software on current electronic voting systems (DREs, optical scanners, ballot marking devices, and central tabulating devices) is trade secret COTS software, along with custom software which is integrated from possibly many different COTS components.

FACT 9: Voting systems with substantially less trade secret COTS software could be developed and would be more verifiable and allow for better post-election forensics.
(Ed: This would just take time - probably over several election cycles.)

FACT 10: Government could provide incentives to develop open source voting systems, and could stop purchasing voting systems with unnecessary trade secret COTS software on it.
(Ed: The Open Voting Consortium is developing one.)

Some election integrity groups and their followers are demanding changes to proposed federal legislation in order to continue the use of trade secret COTS voting system software.

These election integrity groups signed formal statements requesting that HR811 be amended to exempt COTS voting system software from public disclosure:

Alliance for Democracy
Alliance for Democracy - Portland Chapter
Audit AZ
Black Box Voting
Broward Election Reform Coalition
Coalition for Voting Integrity (PA)
Democracy for New Hampshire
Las Vegas (NM) Peace & Justice Center
Main Street Moms (theMMOB.org)
MidHudson Verified Voting
Missourians for Honest Elections
New Yorkers for Verified Voting (NYVV)
North Jersey Impeach Group
Progressive Democrats of America (PDA)
Reach Out America, Nassau Co. NY
Velvet Revolution.us
Voter Action
VotersUnite.org
Voting Integrity Alliance of Tampa Bay (VIA Tampa Bay)


The position of these election reform groups is that:

“There must be no undisclosed voting system software. However, HR811's proposed HAVA Section 301(a)(9), as currently written, fails to exempt software that is truly Commercial Off the Shelf (COTS) from public disclosure. The bill must be amended to require true COTS software, such as the Windows operating system and standard printer drivers, to be escrowed and available to officials under confidentiality, but not publicly disclosed.”

Nancy Tobi of Democracy for New Hampshire states:

“The prohibition of undisclosed software does not provide any exemption for COTS software (commercial off the shelf). No existing voting equipment meets this requirement because they all use COTS, and many use Microsoft software. Microsoft will never share its code, and this requirement would make every piece of voting equipment in use today illegal, requiring jurisdictions to replace equipment at a high cost, unfunded by HR 811.”

Note: Neither of the above position statements recommends prohibiting or minimizing the use of federal dollars on trade secret COTS software that is on voting systems. Nor do they recommend increasing the time frames to develop publicly disclosed and open source voting systems. Black Box Voting, Voters Unite, Democracy for New Hampshire, Verified Voting of New York, Voter Action, Alliance for Democracy and others, recommend exempting COTS software from full disclosure or requiring possibly legally unenforceable requirements for partial COTS software disclosure.

In 2002 Common Cause and PFAW ignored the advice of technologists when they pushed for implementation of in-auditable electronic-ballot voting machines. Today, some election integrity groups are ignoring technologists in proposing a blanket exemption for trade secret COTS software which could make voting systems more costly, less secure and reliable, and certainly less transparent and verifiable.

Nancy Tobi, of Democracy for New Hampshire also, surprisingly, objects to current federal legislative proposals which mandate manual counts of paper ballots in the 2008 election. In her words:

“10. Section 328. Effective Date for Audits. Impossible effective date [2008] for implementation among states for whom no such audit function currently exists.”

Why are some election integrity groups saying that voting systems should be comprised of trade secret COTS software when much of the trade secret COTS software on voting systems is unnecessary and makes voting systems less transparent, and often less secure and more expensive than necessary?

Conclusion and Recommendations

Paperless in-auditable digital electronic recording (DRE) voting machines should be replaced immediately with currently available optical scan paper ballot systems before the 2008 election. The 2008 election should be manually audited in sufficient amounts to ensure accurate election outcomes. After that, federal funds should not be spent on voting systems which use more than the most minimal trade secret COTS software ; and all voting systems, including DREs, optical scanners, and central election management systems should be replaced within reasonable time frames with fully independently auditable, fully publicly disclosed voting systems. Expert technologists need to plan reasonable time frames to require reductions in the amount of COTS software on voting systems and ways to encourage the use of open source voting systems.

The only trade secret COTS software which should be permitted on any voting systems (machines or software whose purpose is to create ballot definition files, help voters cast voted ballots, or whose purpose is to tally or report votes or vote counts, including central election management systems ) purchased with federal funds after 2008 shall:

(a) allow for reverse engineering for voting system evaluation and testing as well as publication of evaluation and tests, including but not limited to usability, performance, errors and bugs; and

(b) be firmware used to run hardware ; and

(c) disclose any information which is necessary to meet the above requirements.

All other software on voting systems, including all operating system software; and all software for the purpose of casting, counting, reporting, or tallying votes or vote counts; and all customized software on voting systems must be fully publicly disclosed. Plus only fully publicly disclosed compilers or program-handling code shall be used to convert source code into machine executable instructions for all software using on voting systems.

Some election reform groups are asking for much more trade secret software on voting systems than is necessary. Why?

Acknowledgements

I received significant help understanding the issue of voting system software from many technologists and computer scientists since 2003. For this particular paper, I received comments and assistance from computer scientist Rebecca Mercuri and from computer scientist Arthur Keller of the Open Voting Consortium, as well as from Dan Lyons and Mark Rothstein. This paper does not necessarily reflect their positions.


--------------- END --------------

References:

[1] This story was removed from OpEdNews.com, a progressive web site, after Bev Harris of Black Box Voting called the article “false” ignoring requests by Dopp to identify any false statement in the article. Dopp has been receiving emails commending her article from technologists, but has also received emails from Nancy Tobi of Democracy for New Hampshire who says that Dopp’s purpose is to “viciously attack others in the movement and to spread disinformation, lies, and paranoid accusations of plagiarism” Here is a saved copy of the disputed article, along with Bev Harris’ request to have it removed:
http://electionarchive.net/docs_other/PositionStatements/BevTries2ShutDownOpEdArticle.htm
[2] This COTS definition is from the Open Voting Consortium’s “Proposed OVC Listed Policies: Second Draft”
http://gnosis.python-hosting.com/voting-project/January.2007/0044.html
[3] Other resources that describe the complexities of verifying software running on voting systems include:
November 2006 “COTS and Other Electronic Voting Backdoors” by Rebecca Mercuri, Vincent J. Lipsio, and Beth Feehan http://www.csl.sri.com/users/neumann/insiderisks06.html#197
“Comment on the IEEE P1583 Standard” by Charles E. Corry, Ph.D. December, 2004
http://www.vote.nist.gov/ecposstatements/Corry-EAC_comments.pdf
“Comment on the IEEE P1583 Draft Standard” by Stanley A. Klein December, 2004
http://www.vote.nist.gov/ecposstatements/comment-memo-5_3_2.pdf
“Comment on the IEEE P1583 Draft Standard – Part 1” by Vincent J. Lipsio December, 2004 http://www.vote.nist.gov/ecposstatements/2004_1221_COTS_STG_EAC.pdf
“Comment on the IEEE P1583 Draft Standard” by Rebecca Mercuri December, 2004
http://www.vote.nist.gov/ecposstatements/MercuriEACmemo.pdf
and
http://en.wikipedia.org/wiki/Custom_software
http://en.wikipedia.org/wiki/Commercial_off-the-shelf
http://en.wikipedia.org/wiki/Open_source_software
“Avoid Another HAVA Train Wreck: Software Disclosure Requirements are a Good Long Term Goal but Need to Be Redrafted in Current Federal Election Integrity Legislation.” By Kathy Dopp and dozens of technologists and computer scientists
http://electionarchive.net/docs_other/dopp/VotingSystemSoftwareDisclosure.pdf
David Wagner, computer scientist in his testimony before the House Admin committee
http://electionarchive.net/docs_other/HearingTestimony/wagner.pdf
Holt's HR 811, A Deceptive Boondoggle -- 10 Blunders to Fix by Bruce O’Dell
http://www.opednews.com/articles/opedne_bruce_o__070221_holt_s_hr_811_a_dece.htm
[3] Rivest, or others, could probably devise a method to ensure the code components running on a system. (Could be impractical). They would need to hash the code, and encrypt the hash result.
[4] According to Dr. Charles Corry, Colorado Springs, CO, former IEEE (the Institute of Electrical and Electronics Engineers) member of the voting system guidelines committee for 4 years (& former Marine corporal) October, 2006: “Many of the hard drives and apparently all of the motherboards of the voting machines are Made in China. China is known to be attacking the Dept of Defense, Commerce Dept and other government computers. The motherboard controls the computer and hiding a malicious program in the boot sector of a hard drive isn’t much of a trick, one has to assume that some or all of the Diebold voting machines are potentially, even probably controlled by China (Security 101).” And “Diebold is based on Microsoft Windows. No other operating system in the world is as subject to so many viruses, Trojan horses, hack tools, worms, or other attacks.”
[5] Their formal letters objecting to fully publicly disclosed software as proposed by Representatives Holt and Tubb-Jones, and Senators Clinton and Nelson:
“Thirteen Issues with the Holt Bill (HR 811)” by Nancy Tobi of Democracy for New Hampshire also takes current election reform bills to task for requiring independent manual counts of ballots in the upcoming 2008, claiming that any requirement for manual audits should be postponed until 2010. http://www.electiondefensealliance.org/13_problems_Holt811_Tobi
“Essential Revisions to HR 811” by John Gideon of Voters Unite is signed by Black Box Voting, by Verified Voting of New York, by Voter Action and many other major election integrity groups.
http://www.votersunite.org/info/HR811EssentialRevisions.htm
Because I have noticed that both Voters Unite and Democracy for New Hampshire have made edits to their written statements without acknowledgement in the past, I have made copies of the April 7, 2007 versions of their public position statements here:
http://electionarchive.net/docs_other/PositionStatements/13_problems_Holt811_Tobi.htm
and
http://electionarchive.net/docs_other/PositionStatements/HR811EssentialRevisions.htm
[6] On April 9, 2007 Bev Harris of Black Box Voting pointed out that in a February 13, 2007 post to a discussion forum, she wrote what she characterizes as a “dissenting opinion” that:
“a) The COTS provision, which allows Commercial Off the Shelf components, is still not entirely satisfactory. These COTS components can be used to destroy election integrity, and many obstacles remain to [a] getting the COTS source code on ALL components into escrow in the first place -- actually, this document says "software", not "source code", which is a potential concern; and [b] making sure it is actually the version used in the election, which is an exceptionally challenging task and current remedies lack both the requirement and the means to do this.”
Note that in her post, Harris correctly points out a few of the obstacles of getting the code into escrow and verifying it, but she does not dissent from the basic recommendations; and further seems to want to change “software” to “source code” when in truth, much more than source code is needed to have any hope for verifying software which runs on voting systems.
[7] Ibid endnote v.
[8] John Gideon’s letter is dated February 13, 2007.
[9] Here is a March 12, 2007 article on the topic of voting system software disclosure written with the help of numerous technologists, “Avoid Another HAVA Train Wreck: Software Disclosure Requirements are a Good Long Term Goal but Need to Be Redrafted in Current Federal Election Integrity Legislation.” http://electionarchive.net/docs_other/dopp/VotingSystemSoftwareDisclosure.pdf
[10] See “Kathy Dopp’s Election Audit Papers”
http://electionarchive.org/ucvAnalysis/US/paper-audits/KathyDoppAuditMathBibliography.pdf
[11] There should be no blanket exemption for all COTS software. Voting systems should contain the most minimal amount of COTS software that is necessary given available products and resources.
[12] There is no intent to prohibit federal funding from being spent on:
Publicly disclosed or open source COTS software, or
proprietary trade secret COTS spreadsheet tools and PCs that could be used to tabulate vote counts on condition that election officials publish all tables of data in those spreadsheets to delimited text cvs files on public web sites as soon as they are finished and before the manual audits and prior to when vote counts are made official, or
* normal trade secret COTS printers used to print out reports that are created on central election management systems.
[13] After all paperless DRE voting machines are replaced with currently available paper ballot optical scan voting systems. See http://electionarchive.org/ucvInfo/US/EI-FedLegProposal-v2.pdf
[14] Trade secret COTS components for I/O devices like disks, CD reader, DVD readers and writers, etc. See also “Proprietary Subsystems in the Diebold TSx Voting Machine – A Chart by Black Box Voting © 2007”
http://www.bbvforums.org/forums/messages/46591/ProprietaryTSx-sm-46690.pdf

Action Alert - Please Act Now to Protect U.S. Democracy

Please support this easy-to-use audit plan to ensure the integrity of US election outcomes. Contact your US House Representative and Representative Rush Holt D-NJ.
http://house.gov (put in your zip code)

To ensure the integrity of U.S. democracy which so many Americans have fought and died for, and to ensure the future well-being of humanity, we must demand a tiered election audit with sufficient manual audits for any range of margins between candidates.

http://electionarchive.org/ucvAnalysis/US/paper-audits/FourTierAudit/TieredElectionAudits.pdf
(To understand the proposal at a glance, see Table 2 on page 2.)

I heard from a reliable source that Congressman Rush Holt D-NJ may not be considering any tiered election audit plan for his new election integrity legislation. (Nor, I presume, is Holt considering a requirement for minimum audit amounts.) Holt's new election integrity proposals are due to be publicly released this week!

A tiered election audit specifies what percentage and what minimum number of vote counts must be manually counted for specified ranges of margins between candidates in the initial election results. The smaller the margin between candidates, the larger the manual audit must be to find a small amount of miscount that could wrongly alter the outcome.

A tiered election audit is a good compromise between audits which require that “large enough size samples of vote counts are manually counted to ensure 99% scientific certainty that the election outcomes are correct” which must be calculated individually for each race; and audits which require that “2% of all vote counts are manually counted” which are insufficient to ensure the integrity of either close election outcomes or outcomes of races that involve a small number of total vote counts.

Tiered election audits are a do-able way to calculate an amount of vote counts to manually audit that would be sufficient to ensure the integrity of U.S. elections.

A requirement for a minimum "amount" of vote counts to be manually audited for particular margins between candidates, would result in 100% hand counts of 100% of vote counts whenever that is necessary to ensure election outcome integrity, and a minimum audit amount would prevent election officials from subverting audits by aggregating ballots into a smaller number of larger-size vote counts which would cause percentage-based audits to be ineffective to ensure the integrity of election outcomes.

Please show this newly revised version of "Tiered Election Audits" to our US legislators:

http://electionarchive.org/ucvAnalysis/US/paper-audits/FourTierAudit/TieredElectionAudits.pdf

The US House must require tiered election audits and a minimum audit amount for all federal elections, in order to secure the integrity of US elections and ensure the future of our way of life.

http://www.house.gov/

Thank you for your dedication to the integrity of our election outcomes.

We do not want to risk the possible dire consequences of leaving unaudited elections wide-open to vote fraud and innocent miscount in the upcoming 2008 elections.

Please help us ensure the future by taking action now.

--
----
Kathy Dopp
http://electionarchive.org
National Election Data Archive
Dedicated to Accurately Counting Elections
Subscribe to announcements by emailing election-subscribe@uscountvotes.org
Please donate or volunteer.

"Enlighten the people generally, and tyranny and oppressions of body and mind will vanish like evil spirits at the dawn of day," wrote Thomas Jefferson in 1816