Post-Election Audits | CTVotersCount.org

Audit Report: Flawed by lack of transparency, incomplete data, and assumed accuracy

Lack of transparency in the process, provides no basis for public confidence in the process, in the audit, and ultimately in our election system. The purpose of the audit is to determine the accuracy of the optical scanners, that purpose is negated when the accuracy is assumed. A statistical calculation based on randomly selected data, omitting some of that data not randomly chosen for omission, is invalid.

Update 8/12: We have received a clarification of the official report from the Deputy Secretary of the State, which modifies our opinion <read>

******************

Last week, the University of Connecticut released its official post-election audit report on the November 2010 election, just short of seven months after the election: Statistical Analysis of the Post-Election Audit Data, 2010 November Election <read>

Like previous reports, this official report fails to provide confidence in the post-election audit process and in the accuracy of the election itself.

The audit data received by the VoTeR Center contains 867 records. Among the 867 records received by the Center, 20 records (2.3%) were incomplete. This report deals with 847 records (97.7%) among which 799 records (94.3%) are from the original data and 48 records (5.7%) were revised based on the follow up conducted by the SOTS [Secretary of the State’s] office.

As demonstrated by the Coalition audit reports there are major shortcomings in the post-election audit process where official hand counts do not match optical scanner counts and incomplete reports are submitted. Based on the UConn report, these differences are addressed in three inadequate ways:

1) Some results are recounted by state officials outside of public view:

The VoTeR Center’s initial review of audit reports prepared by the towns revealed a number of returns with unexplained differences between hand and machine counts. The vast majority of records with high discrepancies were concentrated in the following three districts: East Haven (Deer Run School) with the highest reported discrepancy of 180, Hartford (Burns School) with the highest reported discrepancy of 170, and Preston (Town Hall) with the highest reported discrepancy of 55. Additionally, one or more discrepancies were reported in all but one district for the town of Orange; here the highest reported discrepancy was 14, however this could not be explained as no questionable ballots were reported. Following this initial review the SOTS Office performed additional information gathering and investigation and, in some cases, conducted independent hand-counting of ballots in the four districts mentioned above. The final information was conveyed to the VoTeR Center on June 17th of 2011 for the 48 records pertaining to those districts. The rest of the records (799 out of 847) discussed in this audit report are the original records reported by the towns. [Emphasis ours in all quotes]

We interpret this to mean that some of the counts with differences were recounted by state officials behind closed doors. And based on those counts the Secretary of the State’s Office assumed that all the counts in those districts with differences were hand counting errors. Further, assuming that when election officials make an error in counting in one case in a district and that the machine was accurate in that case, then in all cases where there were differences in those districts that the machine did not make errors.

For the last couple of years we have repeatedly, to no avail, requested that recounts of ballots for the audit be announced and open to public observation or at least the Coalition be notified and given the opportunity to observe. Lack of transparency in the process, provides no basis for public confidence in the process, in the audit, and ultimately in our election system.

As a service to the public, the Coalition provides transparent access to the official audit count reports. You can see scanned copies of the original reports from the towns mentioned <here>

2) Other differences were not recounted but “affirmed” to be hand count errors:

187 records (22.1%) showing discrepancy of 2 to 5 votes, 42 records (4.9%) showing discrepancy of 6 to 13 votes (for this group, although no manual review of the discrepancies was conducted, the SOTS Office affirmed that the discrepancies were due to hand counting errors),

We interpret this to mean that the Secretary of the State’s Office has such faith in the accuracy of our optical scanners and the election process that they assume that any differences must be caused by hand counting errors. The purpose of the audit is to determine the accuracy of the optical scanners, that purpose is negated when the accuracy is assumed.

3) Some audit reports contained incomplete data and such data was not included in the report:

The audit data received by the VoTeR Center contains 867 records, where each record represents information about a given candidate: date, district, machine seal number, office, candidate, machine counted total, hand counted total of the votes considered unquestionable by the auditors, hand counted total of the votes considered questionable by the auditors, and the hand counted total, that is, the sum of undisputed and questionable ballots. This report contains several statistical analyses of the audit returns and recommendations. The statistical analysis in this report deals with the 847 records that are sufficiently complete to perform the analysis.

We interpret this to mean an assumption that the legally mandated audit of three races in a randomly selected 10% of districts is valid even if some of the results are not reported. A statistical calculation based on randomly selected data, omitting some of that data not randomly chosen for omission, is invalid.

Spelling out our concerns, as we said last year, after the November 2009 official UConn report:

All counting and review of ballots should be transparent and open to public observation. Both this year and last year we have asked that such counting be open and publicly announced in advance. [And again in 2011 to the new administration]
Simply accepting the word of election officials that they counted inaccurately is hardly reliable, scientific, or likely to instill trust in the integrity of elections. How do we know how accurate the machines are without a complete audit, any error or fraud would likely result in a count difference, and would be [or could have been] very likely dismissed.
Even if, in every cases officials are correct that they did not count accurately, it cannot be assumed that the associated machines counted accurately.
Simply ignoring the initial results in the analysis of the data provides a simple formula to cover-up, or not recognize error and fraud in the future.

As we have said before we do not question the integrity of any individual, yet closed counting of ballots leaves an opening for fraud and error to go undetected and defeats the purpose and integrity of the audit.

We also note that in several cases officials continued to fail perform the audit as required by law or to provide incomplete reports.

There are other flaws in the audit law.

For instance, there is no legally mandated deadline for the towns to submit audit reports to the Secretary of the State’s Office or for UConn to provide the analysis. We believe seven months after an election is a long time for the public and candidates to wait.
As the Coalition covered in our August 2010 Post-Election Audit Report and the November 2010 Post-Election Audit Report, the list of districts used in the random district drawing is inaccurate and challenging to verify. This also negates reliability, accuracy, and confidence in the random audit.
The November 2010 election, most glaringly pointed out the need for the audit to select from all ballots in the election, not just those counted by optical scanners in the polling place. Among the omitted are centrally counted optically scanned ballots and all originally hand counted ballots. The Coalition Bridgeport Recount Report demonstrated to the public that hand counted ballots can be counted inaccurately on election night. The official election system was not able to audit or recanvass those ballots and has never officially, as far as we know, recognized even the possibility the original hand counted, hand transcribed, and hand totaled numbers may be inaccurate.

As we have pointed out, over and over:

We have no reason to question the integrity of any official. We have no evidence that our optical scanners have failed to count accurately. However, if every difference between a hand count and a scanner count is dismissed as a human counting error then if a machine has or were ever, by error or fraud, to count inaccurately, it would be unlikely to be recognized by the current system.

People and scanners have and will make counting errors. The solution is transparent counting, multiple times to insure accuracy, along with credibly ballot security.

Three districts selected for May post-election audit

This morning, Secretary of the State, Denise Merrill randomly selected three districts for post-election audit from the the districts using optical scanners in the May municipal elections. Exempted from the audits were districts from two towns with post-election close vote racanvasses and two who used ballots, but not optical scanners.

This morning, Secretary of the State, Denise Merrill randomly selected three districts for post-election audit from the twenty-four districts using optical scanners in the May municipal elections. Two alternates were also selected. Exempted from the audits were districts from two towns with post-election close-vote racanvasses and two that used ballots, but not optical scanners.

This is a good time to recall our stand that the audit law should be amended to included all ballots and districts in the audit, with no exemptions. The three major exemptions are districts with recanvasses, centrally counted absentee ballots, and all hand counted ballots. We know from experience that manual counting can be performed with great accuracy, yet also with significant errors. Any exemption from the post-election audit is an opening for errors to be hidden and a formula for skulduggery.

For example, the November Bridgeport election and citizen recount demonstrated about 25% of originally hand counted ballots being incorrectly counted or reported in the (still) official results. Since audits do not include hand counted ballots we do not have a way of judging how accurately other towns in Connecticut count ballots on election night. We speculate that sometimes some are quite accurate, but that sometimes some are also significantly inaccurate.

The districts selected were:

Groton – West Side Middle School Dist. 2
Groton – William Seely School Dist. 4
Stonington – Stonington Fire House

Alternates selected:

Griswold – Griswold Town Hall
Woodbridge -Center School

Public hearings for 15 election related bills – Update: Our Testimony

Today we provided testimony on ten bills. We talked six times and complemented the Committee on their new format of handling bills one at a time, allowing each person who wanted to the opportunity to testify on each bill separately. It worked very well and did not take as much time as one would expect over the old format of one opportunity per person for the day. Most of the testimony today was agreement or friendly disagreement between registrars, town clerks, state officials, and advocates. In the end we expect that better laws will result.

Note: The General Administration and Elections Committee has taken up several election bills and concepts for this session. We are optimistic that some of the concepts will be developed and passed to provide increased election integrity. Many of the bills taken up, often well intended, have unintended negative consequences. We are highlighting several of them to point out highlighting several of them to point out the good, the bad, and the unbelievable.

Update: 1/14/2011:

Related: Secretary of the State Merrill’s press conference CTNewsJunkiee report and CTMirror.

***********Original Post:

On Monday the General Administration and Elections Committee will hold hearings of 15 elections related bills <agenda>

Most of these bills became publicly available on the Legislature’s website on Wednesday morning and late Wednesday they appeared on the agenda for Monday’s public hearing. A lot to absorb in a few days, yet we can say at this point that there are some good, not so good, and some highly questionable proposals in these bills.

Several of the bills are consolidations and committee rewrites of other bills, specifically for no-excuse absentee voting and an associated Constitutional amendment. One focuses on “Election Integrity” dealing with some of the issues raised by ballot shortages in Bridgeport, moderator training, registrars identifying polling places to the Secretary of the State etc. Another focuses on “Post-Election Audits”, authorizing local officials to audit via “independent machine rather than a tabulator” as an alternative to the current manual count.

Many of the bills are “Technical Bills” primarily intended to make small “technical” changes to the statutes to adjust to the move from lever machines to optical scanner. By our count these bill total 256 pages of details and redundancy since many address the same existing statutes. It seems they must have been written by different groups but perhaps based on previous bills which failed to pass the legislature over the last three years. We are in the process of reading through the bills to prepare detailed testimony for Monday, with suggestions for revisions, deletions and improvements. After reading through all the bills and discussing some portions with other advocates we can summarize a few items at this point:

It is all well and good to replace “machine” with “tabulator”, replace “registrars” with “registrars of voters”, and remove “he”s throughout, we doubt these changes will make any difference in the interpretation of the law by registrars of voters or the courts.

There are many changes that seem necessary to the conduct of elections by optical scanner that are not included. For instance, several sections do not seem to recognize that in addition to optical scanners and absentee ballots we now have polling place paper ballots, both scanned and hand counted to deal with secure and count.

The election laws, I suppose like many others, remain highly convoluted from amendments over the years and redundant. For example there are extensive, almost completely redundant, separate sections for primaries and elections. The problem is that they often differ in critical, substantial ways that make little sense. One bill makes a positive change that would allow officials from other towns to serve in primaries as they the existing law allows for elections. Asked to serve in another town for a primary, I declined to avoid breaking the law.

Within the bills are many good changes as well as some needing improvement. For example, the bill for “Integrity of Elections” calls for registrars to file plans for ballot printing with the Secretary of the State, locations of each polling place, and the names of moderators for each polling place to the Secretary of the State in advance of the election. Our reading indicates that moderators can be rejected by the Secretary but with no specified deadline for such rejections. We have called for the Secretary of the State to have an accurate list of polling places to restore the integrity of the post-election audits. While we applaud that change to enable an accurate list,we will also suggest a possibly more efficient method to accomplish that same goal.

But also within these 256 pages are several significant changes that may or may not be advisable. For instance, one calls for a demonstration “device” for voting in each polling place instead of a demonstration “machine” – it is hard to tell what would satisfy the requirement or if such would also be required for the IVS machines intended for voters with disabilities; our reading of another clause intended to codify the current recanvass procedures would eliminate a critical step in the process; another in our reading would significantly change the counting of cross-endorsed candidate votes – to the likely detriment of candidates and voters.

I am a strong supporter of new techniques and technologies that support independent machine auditing, but we will oppose change authorizing local officials to audit via “independent machine rather than a tabulator” as an alternative to the current manual count. While well intended, the proposed law provides no restrictions on such a machine, no requirements for the process, no standards, no guarantee the process would be anything like the successful example in Humboldt County, CA, and no budget for implementation. Voting integrity and confidence require that any independent machine audit be required to meet requirements that provide for public transparency and validation. In other words, vendors need to dot the “i”s and cross the ‘t”s in software and hardware products, while the law must require election officials to also dot the “i”s and cross the ‘t”s in implementing such audits. Municipalities that balk at spending a few hundred dollars on an audit when they are randomly selected are hardly in a position to acquire such equipment, let alone evaluate the equipment, and develop an effective, satisfactory process.

Nov 2010 Election Audit Observation Report

Coalition calls again for legislature to act.
Citizen observation and analysis show little, if any, improvement in
November post-election audits.

The Coalition noted significant differences between results reported by optical scanners and the hand count of ballots by election officials across Connecticut. Compared to previous audits, the Coalition noted little, if any, improvement in the attention to detail and in following procedures in the November 2010 audits.

For the full report visit http://www.CTElectionAudit.org

Coalition calls again for legislature to act.
Citizen observation and analysis show little, if any, improvement in
November post-election audits.

Why We Need Audits and Recounts: AccuVote Missed 0.4% of Ballots in Aspen Elections

How do we know that our Dieblod/Premier/Dominion AccuVote-OS voting machines count ballots and votes accurately in each election, in each polling place? Maybe sometimes they do and sometimes they don’t.

The reason we need paper ballots, audits and recounts is to verify that citizens’ votes are counted accurately. How do we know that our Dieblod/Premier/Dominion AccuVote-OS voting machines count ballots and votes accurately in each election, in each polling place? Maybe sometimes they do and sometimes they don’t.

In Aspen, Colorado they did not, missing 11 ballots out of 2544. Premier AccuVote Machines Missed 0.4% of Ballots in Aspen Elections <read>

On May 5th 2009, Aspen (CO) held municipal elections for mayor, two city council seats and a ballot measure. Pitkin County’s Premier (formerly Diebold) AccuVote optical scan voting machines failed to register 11 (0.4%) of 2,544 ballots, which was discovered due to the ballots also being counted on Election Day at a central location with a separate system. Premier is one of the three largest providers of voting equipment in the United State…

To underscore the importance of the missing 11 ballots, it is not uncommon for manual recounts of optical scan elections to find new valid votes that were discounted by the optical scan voting machine, either because the machine detected a stray pen mark as an over-vote (voting for more candidates than allowed), or because a voter marked a choice too lightly or outside the designated spot on the ballot, such that the machine detected an under-vote, or skipped race. Such “found” votes are common in manual recounts where humans can recognize a voter’s intent that the optical scan machine could not. In the recent Aspen election, the independent scanning of ballots did indeed allow election officials to find at least one such valid vote missed by the AccuVote voting machines.

However, this problem is unrelated to the discovery that nearly a half percent of ballots – 11 ballot cards in all – went entirely unrecorded by the AccuVote machines. According to Aspen City Clerk Kathryn Koch, both the poll book record of the number of voters who voted and the TrueBallot record of ballots processed agree that there were 2,544 ballots. The AccuVote machines, however only recorded 2,533 ballots

Unfortunately, differences in counts between humans and AccuVote-OS optical scanners in Connecticut audit reports are routinely dismissed as human error rather than potential scanner errors to be investigated. See the Coalition Audit Reports.

This is also why CTVotersCount stands for stronger audits and actual recounts in very close elections in Connecticut. We also are strongly support auditing by machine in Connecticut, with systems like the one used in Aspen by TrueBallot, and others by ClearBallot, and TEVSystems – provided Connecticut implements such systems in ways that are transparent, provide for public verification, and confidence.

Sadly these same ballots are being withheld from public scrutiny to verify this result.

74 Districts in 55 towns chosen for audit. And the surprises just keep coming.

At the drawing the Secretary announced an agreement with the Bridgeport Registrars that the 12 districts with the copied, hand counted, ballots would be voluntarily audited with all the ballots for the race for Governor counted. Surprise Surprise! It seems that not everyone in Bridgeport government agreed to the agreement

Yesterday, the Secretary of the State with the assistance of members of the Connecticut Citizen Audit Coalition chose 74 districts. This is the 1st drawing I have not attended since they started in August 2007, but the Coalition including CTVotersCount was well represented.

We had understood from Deputy Secretary Mara that the registrars would have until 12/1 to complete the audits, so we were surprised when the press release said they had to be complete by next Monday 11/22. The law says they must be complete two days before certification which is next Wednesday, although we recall past audits that went beyond. In any case, we and the registrars are all scrambling, especially since the drawing was four or five days later than usual. Hard for us to fault the registrars for not complying with the Secretary of the State’s procedures which require a three day public notice before the audits occur.

The surprises just keep coming!

Bridgeport rejects ballot recount

HARTFORD — Bridgeport officials, citing a lack of authority for Secretary of the State Susan Bysiewicz, have rejected a request to recount ballots in the 12 city voting precincts that were kept open an extra two hours on Election Day.

Bysiewicz’s office said Arthur C. Laske III, deputy city attorney, called Monday afternoon to announce that the city would not be redoing the count that they originally performed in the days after the Nov. 2 election.

Laske’s call occurred hours after a Bysiewicz news conference announcing that Bridgeport’s voting registrars agreed to perform an additional review of the ballots as part of a routine statewide recount of totals in 74 polling precincts.

We would like to say more, but we have audit observers to schedule.

Update from CTNewsJunkie <read>

In a joint statement Bridgeport Attorney Mark Anastasi and Deputy City Attorney Arthur Laske said Anastasi spoke with an attorney in the elections division of Bysiewicz’s office Monday afternoon to tell them “no one in the city had agreed to such a recount.“

Bysiewicz’s office has no legal authority to force them to do a recount, but those polling places were left out of the random drawing for the statewide audit of 74 polling places.

“It seems the Secretary was mistaken both as to whom her office spoke with from the City of Bridgeport and what was said on Monday,” Anastasi and Laske said. “The Bridgeport Registrars of Voters had preliminary discussions with Deputy Secretary of the State Leslie Mara about this subject, but neither Registrar agreed to the proposal or received a promised follow-up call about the proposal.”

While a voluntary audit may ensure confidence in the results of the governor’s race, Anatasi and Laske said “The City can find no legal authority which either requires or even allows the State or the City to conduct such a recount.”

“We believe that the Secretary of the State is equally aware of this absence of legal authority,” the two attorney’s said.

Post-Election Audit Report: Incremental Improvement – New Integrity Concern

Citizen observation and analysis shows some improvements along with a newly uncovered problem with the random selection process…We conclude that August post-election audits still do not inspire confidence because of:

failure in the integrity of the random district selection process,
lack of standards for determining need for further investigation of discrepancies,
weaknesses in the ballot chain of custody, and
lack of, consistency, reliability, and transparency in the conduct of the audit.

the list of polling districts for the random audit drawing was missing some districts and is otherwise inaccurate and ambiguous. The integrity of the audit requires an accurate list of districts that is verifiable by the public. We have extended our recommendations to the Legislature to include an efficient fix to this problem.

Full Report, Press Release etc.<Audit Coalition Post>

Summary, from the Press Release and Report:

Coalition Finds Small Improvements and New Problem in
Connecticut Post-Election Audits

Citizen observation and analysis shows some improvements along with a newly uncovered problem with the random selection process

This is the sixth major post-election audit observation report by the Coalition since the adoption of optical scanners and paper ballots statewide.

Coalition Executive Director, Luther Weeks noted, “Unfortunately, we discovered that the list of polling districts for the random audit drawing was missing some districts and is otherwise inaccurate and ambiguous. The integrity of the audit requires an accurate list of districts that is verifiable by the public. We have extended our recommendations to the Legislature to include an efficient fix to this problem.”

League of Women Voters of Connecticut President, Cheryl Dunson said: “Compared to previous audits, the Coalition noted continuing incremental improvements in the attention to detail, following procedures, and in the chain-of-custody by election officials. We caution that the primary audit is simpler and shorter than those for November elections which may account much more accurate counting this time.”

We conclude that August post-election audits still do not inspire confidence because of:

failure in the integrity of the random district selection process,
lack of standards for determining need for further investigation of discrepancies,
weaknesses in the ballot chain of custody, and
lack of, consistency, reliability, and transparency in the conduct of the audit.

Each of these items individually could impact the integrity of the statewide post-election audit and calls into question the credibility of the entire post-election audit.

Although most of our general observations and concerns remain, we observed improvements in following audit procedures, in the accuracy of the counting, and in the completion of forms.

Connecticut Citizen Action Group Executive Director, Tom Swan said, “The integrity of the entire audit is dependent the ballot chain-of-custody and on every step of the audit being accurately accomplished in a consistent, transparent, and professional manner. We continue to support our past recommendations to the Secretary of the State and the Legislature for improvement in the post-election audit laws, counting procedures, and chain-of-custody.”

Weeks added, “We look forward to the post-election audit of the November election. We hope to see significant improvement in following procedures along with more accurate counting, demonstrated in the November post-election audit which will involve more extensive, complex counting. ”

Observers came from the membership ranks of the coalition partners — The League of Women Voters of Connecticut, The Connecticut Citizen Action Group, Common Cause Connecticut, and Connecticut Voters Count. Without volunteer observers, nobody but a small number of local election officials would know what happens in post-election audits.

New Overall Audit Integrity Concern

A new concern surfacing this year is the inaccurate list of districts used in the random selection process which is required by law to be based on all of the districts used in the election. This directly impacts the integrity and credibility of the entire post-election audit.

Issues In Three Towns

Several districts in one town were selected, but in one case in the municipality, the ballot bag contained only blank ballots. In subsequent discussions with the registrar, she reported that a novice moderator in a multiple district polling place had sealed all voted ballots in one bag and all unused ballots in another bag. [As far as we know, this district was never counted as was not included in Audit Reports from the Secretary of the State.]

In the one district: The official Audit Report indicates 1703 machine counted ballots but only 688 manually counted ballots counted in the audit. In that same audit report 188 ballots are listed for one party with a total of 254 votes in the race audited for that party. The huge difference may represent poor counting procedures and lack of understanding of the audit procedures, however, we have no way of determining the accuracy of the audit nor of the official reporting of results. Our observer’s comments:

They never counted the ballots first…One team referenced the Tally Sheet from Election night. They recounted their votes until the figures agreed… Checking was done to the Tally Sheet off the Moderator’s report not the machine tape… I did not observe a machine tape, only the Moderator’s return with the Tally Sheet. When I asked if they had a machine tape, I was told no by one of the Registrars. When I asked if there was a tape in the ballot bag, I was told no… I have concern about the number entered on the “LT Gov”

These results may represent incompetence. However, incompetence uninvestigated transparently leaves an opening to cover-up fraud and error. The Secretary of the State’s Office had reviewed district results, yet apparently did not notice these large differences until it was pointed out by the coalition.

In one of the last district reports provided to the Secretary of the State’s Office: In one district in one municipality which audited three districts there was a significant difference between the machine counts in one race and the hand count reported. For two candidates the machine counted 262 and 154 votes while the hand counts were 132 and 78 votes for those same candidates.

Full Report, Press Release etc.<Audit Coalition Post>

Update from CTNewsJunkie: Questions Raised As State Finishes Post-Primary Audit <read>

Update: Greenwich Audit: Is it worth $1200 $480?

Greenwich Time article: Greenwich picked – yet again – to audit primary results. Greenwich registrars and Secretary of the State debate value of audit and random selection of Greenwich.

Plus – Registrar and we agree: Audit Absentee Ballots Too!

8/20/2010: Greenwich Time article Greenwich picked – yet again – to audit primary results <read>. Greenwich registrars and Secretary of the State debate value of audit and random selection of Greenwich.

We note that Greenwich has about 24 voting districts. Everyone should expect that on a statewide 10% (or 5%) random selection audit, any city with that many districts will almost always have at least one district selected for audit.
In other states with audits, each county must audit a minimum number of districts and many of those counties are smaller than Greenwich in districts, population, and resources.
Contrary to the article, no candidate loses or picks up votes in Connecticut post-election audits. No matter how inaccurate the machine or human count, candidates would have to bring the matter to court and presumably ask for a recount to change the results in any way.
Looking at the web site for Greenwich we find that “They are also responsible for hiring and training over 200 official poll workers as well as maintaining all voting equipment used for the election.” Which at $140 per poll worker would come out to $28,000. Actually we expect they pay lots more than that for the election, since the audit will certainly involve much less than a half a days work, compared to the 17 hour election day and training. Running an election, maintaining the equipment, printing ballots and the ongoing costs of voter registration and the registrars office would seem to dwarf the $1,200.
Even so, we are sympathetic to Greenwich and other municipalities. We are in favor of the state paying the cost of audits as the towns selected perform the work for every voter’s benefit.
Finally, those familiar with our Ten Myths In the Nutmeg State would understand that we do not find the Connecticut post-election audits all that stringent. those familiar with statistics would understand also that it is not the percentage of the districts selected, but the total number of districts counted and number and size of loopholes that determine the power of an audit.

Greenwich picked – yet again – to audit primary results
Neil Vigdor, Staff Writer
Published: 09:53 p.m., Thursday, August 19, 2010

Greenwich’s election officials, have once again received marching orders from the state that they must perform an audit of last week’s primary results from Riverside School.

Six of the seven times since the state started mandating audits three years ago, the town has been picked for a recount, irking its two registrars of voters.

“We’ve been chosen every time,” said Sharon Vecchiolla, the town’s Democratic registrar. “We could do without it.”

Out of a total of 722 voting precincts throughout Connecticut, 73 were randomly selected by the secretary of the state’s office for an audit, which must be conducted between Aug. 25 and Sept. 15. “I was stunned Stamford didn’t get an audit,” Fred DeCaro III, the town’s Republican registrar.

The audit will take place at 9:30 a.m. Aug. 25 in the Town Hall Meeting Room, with eight paid poll workers hand-counting the ballots cast for lieutenant governor in last week’s Republican primary won by Mark Boughton over Lisa Wilson-Foley and the Democratic primary won by Nancy Wyman over Mary Glassman. Greenwich had 12 poll locations for the primary.

Under Connecticut law, election results from 10 percent of all voting precincts in the state must be audited after elections, a mandate that was put in place in three years ago when mechanical lever machines were replaced with electronic scanners. The fax-like machines read blackened ovals on paper ballots that resemble standardized test answer sheets.

“I think we wouldn’t mind if it was dropped to 5 percent of polling places instead of 10,” DeCaro said.

Secretary of the State Susan Bysiewicz defended the audit mechanism, saying that the threshold of 10 percent is the most stringent in the nation.

“I think it is a good check and important for voter confidence in our election system,” Bysiewicz said. “I think a high threshold of 10 percent is appropriate because that’s what the University of Connecticut’s computer science department recommended as a threshold that would ferret out fraud or a computer breach.”

The state not only relies on UConn for guidance on audit sample sizes, it sends the results there for a post-mortem and uses the university to test the memory cards in voting machines before elections.

Bysiewicz said that the electronic voting machines have a near 100 percent accuracy rate since they were introduced, except for isolated instances of human error by those casting ballots.

“It means not only that our votes are secure, but that our votes are counted properly,” Bysiewicz said.

A bingo barrel was filled with the names of all 722 polling places, with a representative from the League of Women Voters of Connecticut randomly drawing individual precincts one-by-one to be audited, according to Bysiewicz.

“There are many cities of good size that had more than one precinct and there are towns that had six,” Bysiewicz said. “Sometimes towns manage not to get chosen. The idea is over the course of time, every town has one or more precincts chosen.”

Bysiewicz noted that the city of New Britain once had nine polling places chosen all at once.

A total of 319 ballots were cast at Riverside School in the GOP primary for lieutenant governor, compared to 193 in the Democratic primary.

The cost of the audit to taxpayers was estimated at $1,200 by DeCaro, who said each person doing the hand counts will be paid $140.

After the November 2008 election was audited, John McCain picked up eight votes on Barack Obama in town, while Christopher Shays picked up 14 votes on Jim Himes for Congress. It was a moot point for McCain and Shays, however, who both lost their races.

No irregularities were discovered during the other previous audits done by the town, which occurred after the November 2007 municipal election, the February 2008 presidential primaries, the August 2008 Democratic congressional primary and the September 2009 Democratic tax collector primary.

Greenwich was not selected for an audit after last November’s municipal elections, while two Stamford precincts were.

Despite the hassle of conducting frequent audits, DeCaro said it was much better than the alternative of no checks and balances.

“I wouldn’t be in favor of eliminating the audits entirely,” DeCaro sad[sic].

Update: 9/26/2010 Audit actually took an hour and counters paid $60.00 each, not $140.00; Greenwich Time article <read>

Eight paid poll workers spent less than an hour hand-counting the ballots…The eight poll workers were paid $60 each to do the hand count, [$480.00 assuming the registrars are salaried]

And another Greenwich Time article where we completely agree the registrar: Greenwich registrar recommends audit of absentee ballots <read>

If the state is going to make cities and towns go to the trouble of auditing election results, Republican Registrar of Voters Fred DeCaro III said it should require absentee ballots to be included in hand counts.

Out of the 24,996 votes cast in the Aug. 10 primary in Greenwich, 2,330, or just under 10 percent, were done by absentee ballot, according to the registrars of voters.

All absentee ballots are counted by poll workers at Town Hall, not in the individual precincts where the voters live that are subject to state-mandated audits.

“Why not include absentee ballots?” DeCaro said Wednesday following an audit of the lieutenant governor primary results from Riverside School.

See my comments on both articles at the Greenwich Time site.

Post-Election Audit Drawing: 73 Districts 45 Municipalities Selected

Today the Secretary of the State’s Office conducted the random drawing. Members of CTVotersCount and the League of Women Voters performed the selections.

The Official Press Release: Secretary of the State’s Office Draws Precincts to Have Statewide Primary Results Audited: <read>

UCONN: Failed memory cards caused by weak batteries, inadequate design

This week at the 2010 Electronic Voting Technology Workshop on Trustworthy Elections in Washington, D.C., Dr. Alex Shvartsman and his team from the Uconn VoTeR Center delivered a significant paper. It covered research into the cause of the complete failure of the AccuVote-OS memory cards, at an unacceptable rate — We suggest the costs of mitigating the problems should be born by the manufacturer and/or distributor since the ultimate cause is the inadequate design of the memory cards for their intended purpose.

This week I attended the 2010 Electronic Voting Technology Workshop on Trustworthy Elections in Washington, D.C., Dr. Alex Shvartsman and his team from the Uconn VoTeR Center delivered a significant paper. It covered research into the cause of the complete failure of the AccuVote-OS memory cards, at an unacceptable rate. <See our earlier coverage>. <The Research Report>

[W]e determined the time interval from the instant when a battery warning is issued by the AccuVote to the point when the battery does not have enough voltage to retain data on the memory card.We show that such interval is about 2 weeks. Thus timely warnings cannot be provided to protect against battery discharge and loss of data during the election process…

Recommendations

we determined the time interval from the instant when a battery warning is issued by the AccuVote to the point when the battery does not have enough voltage to retain data on the memory card. We show that such interval is about 2 weeks. Thus timely warnings cannot be provided to protect against battery discharge and loss of data during the election process…

The lifetime of the Energizer battery, when its voltage remains above the 2V needed for data retention in standby mode, at that current load, according to its datasheet [9] is 9,000 hours or approximately one year.

Given that it is possible that a memory card is used for elections once a year, it leads us to the same conclusion: For each election, a decision would be made, whether or not to replace the batteries for this election. The decision would be based on the amount of time since the batteries were last replaced and on the estimate of the service life of the battery (e.g., using the procedure at the end of the previous section).

Discussing the challenge with Dr. Shvartsman at the workshop, it seems that replacing the batteries is more complicated than might be assumed. The battery is under the memory card label, so replacement includes completely removing all remnants of the old label then preparing and placing a new label on the memory card. Shvartsman estimated the replacement cost, including labor, may be on the order of $10 per memory card.

We suggest that $10 per year per card is well worth avoiding most of the problems associated with the current huge, unacceptable failure rate. The total cost would be about $40,000 per year, somewhere in the range of $0.025 per ballot cast. To put this in context, ballot printing is about $0.45 per ballot and election costs average in the range of $5.00 to $8.00 per ballot cast. We also suggest the costs of mitigating the problems should be born by the manufacturer and/or distributor since the ultimate cause is the inadequate design of the memory cards for their intended purpose.

PS: Dr. Shrvartsman is mentioned prominently in an article posted at Verified Voting: Voting Technology Research Gets In-Depth <read>