June 23, 2021

New Hampshire Election Audit, part 1

Based on preliminary reports published by the team of experts that New Hampshire engaged to examine an election discrepancy, it appears that a buildup of dust in the read heads of optical-scan voting machines (possibly over several years of use) can cause paper-fold lines in absentee ballots to be interpreted as votes. In a local contest in one town, preliminary reports suggest this caused four Republican candidates for State Representative to be deprived of about 300 votes each. That didn’t change the outcome of the election–the Republicans won anyway–but it shows that New Hampshire (and other states) may need to maintain the accuracy of their optical-scan voting machines by paying attention to three issues:

  • Routine risk-limiting audits to detect inaccuracies if/when they occur.
  • Clean the dust out of optical-scan read heads regularly; pay attention to the calibration of the optical-scan machines.
  • Make sure that the machines that automatically fold absentee ballots (before mailing them to voters) don’t put the fold lines over vote-target ovals. (Same for election workers who fold ballots by hand.)

Everything I write in this series will be based on public information, from the State of New Hampshire, the Town of Windham, and the tweets of the WindhamNHAuditors.


In the November 3, 2020 general election in Windham, New Hampshire, the race for State Representative was very close–24 votes difference–so one candidate asked for a recount. The recount showed that she lost by hundreds of votes–not just 24. The hand recount of the optical-scan paper ballots disagreed with the numbers claimed by the optical-scan voting machines to an extent that was shocking. The citizens of New Hampshire demanded to know: were the op-scan machines hacked? Were the machines misconfigured and reading marks from the wrong place in the paper? Did creases in absentee ballots cause the op-scan machines to misread votes? Was the recount itself erroneous?

The Secretary of State said that there was no provision in New Hampshire law that permitted him to do a re-recount, or to examine the voting machines for hacking. So a Republican legislator introduced a bill for a forensic audit to find out what had happened. That bill passed the legislature unanimously:

… [T]his act authorizes and directs an audit of the ballot counting machines and their memory cards and the hand tabulations of ballots regarding the general election on November 3, 2020 in Windham, New Hampshire of Rockingham County district 7 house of representatives for the purpose of determining the accuracy of the ballot counting devices, the process of hand tallying, and the process of vote tabulation and certification of races.  

A forensic election audit team shall be formed to complete the audit described in section 3 and it shall be comprised of:  one person designated by town of Windham,  one person designated jointly by the … secretary of state and attorney general, [and one person to be chosen by those two auditors]. … The results of the audit shall not alter the official results of the Rockingham County district 7 house of representatives race as determined by the ruling of the ballot law commission on November 25, 2020, upholding the recount of that race.

Governor Sununu signed bill SB43 into law on April 12, 2021. On April 26 the town of Windham chose Mark Lindeman of the Verified Voting Foundation as its auditor, and the Attorney General (upon the advice of the Secretary of State) chose Harri Hursti, a well known expert on the cybersecurity of voting machines and bank ATM machines. Those two then selected Philip Stark, of the University of California at Berkeley, as the third auditor.

This is truly a “dream team” of experts. They know what they’re doing, they’re experienced with election audits and the forensic examination of voting machines, and they know how elections work. New Hampshire could not have found anyone better prepared to get to the heart of the matter.

What happened in Windham

The town of Windham has a single polling place, in which 10,006 ballots were cast: 6697 in-person and 2949 absentee scanned by machine, and 80 absentee counted by hand (because overseas “UOCAVA” ballots are in a different format that the machines didn’t accept).

In New Hampshire, absentee ballots are processed in the polling place, on election day. That is, there’s a preprinted voter list. When registered voter shows up to vote in person, their name is found on the list, and crossed out. An absentee ballot envelope is processed by checking the voter’s name (from the envelope) against the same preprinted voter list, and crossing it out; then the absentee ballot is removed from the envelope, and scanned into the same voting machines that are used for in-person ballots.

That procedure is a way of checking that the same person doesn’t vote absentee more than once, or doesn’t vote both absentee and in person: their name appears only once on the list, and can be crossed off only once.

In Windham on November 3rd, there were four optical-scan voting machines, Global Election Systems model “AccuVote OS”, purchased in 1998 and 2000. The 6697 in-person and 2949 scannable absentee ballots were fed into those four machines during the day. That’s something like one ballot every 10 or 15 seconds into each machine, all day long!

At the close of the polls, each machine printed its result out onto a “results tape”–thermal cash-register paper listing how many votes each candidate got. There were four “results tapes”, one per machine; here’s a small portion of just one of them, for the State-Rep contest:

In this race, there were 8 candidates running for 4 seats, and every voter got to vote for four out of 8.

Election administrators aggregated the four voting-machine results tapes by hand (along with the 80 hand-counted ballots) onto a paper worksheet to produce an official “Return of Votes” form, signed by the Town Clerk.

In the race for State Rep, in 5th place out of 4 seats, Kristi St. Laurent was only 24 votes behind the 4th-place candidate Julius Soti. So she asked for a recount. By state law, those are done by hand, by the office of the Secretary of State, in Concord (the state capital).

The results of the recount were:

OriginalRecountDiff
Kristi St. Laurent (D)4,4564,357-99
Henri Azibert (D)2,7872,808+21
Valerie Roman (D)3,4153,443+28
Ioana Singureanu (D)2,7642,782+18
Julius Soti (R)4,4804,777+297
Mary Griffin (R)5,2925,591+299
Bob Lynn (R)4,7865,089+303
Charles McMahon (R)5,2565,554+298
TOTAL33,23634,401+1,165

As you can see, something is grossly wrong–either with the machine counts, or the hand counts, or both. The people of New Hampshire, and their legislators, wanted to understand how this happened–and how to prevent it in the future.

The audit team looked at,

  • the voting machines (and their ballot programming, and their performance under the kind of heavy usage they saw last November);
  • the paper ballots (and associated records), as preserved by the Secretary of State from the November recount
  • and the context, learning from election administrators in Windham and other New Hampshire towns about polling-place procedures.

What caused the discrepancy

The audit team has not yet written their official report, so everything I’ll describe here is only their preliminary findings as described in their tweets.

The auditors supervised a careful recount by teams of 5 people (caller, checker, 2 tallyers, flagger). Members of each team included Democrats, Republicans, and independents. The hand count was within 0.05% of the official state recount, for every State Rep candidate (the State did not recount the races for President, Senator, etc.). So, basically, the official state recount was right, the voting machines were wrong; why?

There is zero evidence that the voting machines were hacked. Forensic exams show that the software in these machine matches the reference machine provided by the Secretary of State’s office (the audit team will continue to examine that software to make sure the SoS’s copy is right), and there was nothing unexpected in the memory cards.

Folds in ballot papers. Ballot papers marked by voters in the polling place were (generally) not folded. Absentee ballots have identical printing; they were folded in thirds before they were mailed to the voters; marked by the voters at home, refolded and mailed back to the Town Clerk, and then (on election day) scanned through the same voting machines as the in-the-polling-place ballots.

What happens if a fold goes through one of the “vote targets” (the ovals that voters are supposed to fill in with a pen to indicate their choice)? The voting machine is not supposed to interpret a fold as a mark. But the audit team found that in many cases, a fold would be read as a mark:

Microscopic photo of fold through a vote target. From @WindhamNHAuditors

The fold lines in absentee ballots weren’t always in exactly the same place, but in hundreds of ballots they crossed through the target for Democratic candidate Kristi St. Laurent. Now, consider what happens if an voter marks all 4 Republican candidates (and none of the 4 Democratic candidates), by blackening their ovals with a pen. If the optical-scanner also interprets the fold line through one Democratic candidate as a mark, then the machine “thinks” there are 5 votes in a vote-for-4 contest; and that’s an overvote, so all of those votes won’t be counted.

Many voters voted straight-party: all R, or all D. The effect of this was that hundreds of votes for the Republican candidates were not counted. An all-Democratic vote, with a fold line through a blackened target, would not be affected.

In the recount, with humans looking at the votes, the votes were counted accurately. Humans are not likely to interpret a fold line as a vote! So the official Windham results (after the recount) are trustworthy.

But many towns in New Hampshire use the same AccuVote optical-scan voting machines. We can legitimately wonder whether some elections in other towns (that were not recounted by hand) got the wrong result. I’ll discuss that question in my next post.

In my next article I’ll examine:

  • What do the national voting-machine standards say about how voting machines should distinguish fold-lines from intentional vote-marks?
  • Could a build-up of dust make the voting machine more likely to misinterpret folds as marks?
  • What do election administrators in other states do to avoid this problem?
  • Should New Hampshire throw away its voting machines and buy new ones, or throw away its voting machines and count votes by hand? Or are there measures they could take to use these same machines in a trustworthy way?
  • What series of circumstances led to this problem in Windham 2020, and how could those corrective measures prevent anything like this from happening again?
  • Could improved technology in optical-scan voting machines be less susceptible to this problem?

Comments

  1. Douglas W. Jones says:

    I recall a conversation (by phone) with a local election administrator in another state who used the AccuVote OS in her jurisdiction.

    I asked her about preventative maintenance. She said “what do you mean?” I explained about how paper sheds fibers, dust, and how fingerprints accumulate on things people touch, making a layer of crud. Cleaning this off matters. She seemed quite surprised. I asked if she had a service contract for the machines. She said it had been dropped years ago because it was too expensive.

    In conversations with other election officials, not the same jurisdiction, when asking about mark sensing, I have frequently heard the line: “These machines are certified to rigorous federal standards, so we don’t worry about that.” Unfortunately, this misrepresents the standards. What the standards guarantee is that, in principal, if adjusted and maintained appropriately, the machines can distinguish votes from non-votes (including folds). It is up to the administrator to do that adjustment and maintenance. Without it, the standards are irrelevant.

  2. Harvie Branscomb says:

    There are variables related to the fold that aren’t yet mentioned. Auditor Hursti pointed out the distinct effect that the direction the fold takes ( up or down). He also noted the extent to which the ballot sheet is flattened upon insertion matters a lot. I didn’t hear mention of the extent to which the paper itself collects dust and dirt at the place where the fold is made.

    My concern goes beyond the most obvious discrepancies that seems about to be explained. I worry that there will be an assumption that all the sources of error inherent in the NH use of the machine have been located.

    I believe further research is needed to learn about other potential errors. I understand that this device was put into service before certification tests were popular and largely has escaped serious testing for accuracy and reliability in spite of it’s use in many elections for many years.

    In particular the overvote detection performance is worthy of study. In NH the option to have the voter or an election judge confirm the intention to overvote is disabled. I myself have studied the machine long enough to know that the machine should not be allowed to interpret overvotes without human interaction.

  3. Dwight Shellman says:

    Thank you for this explanation Dr. Appel. I completely agree that the Windham auditors are the varsity squad of post-election audits. Their efforts have been very impressive from the start.

    I conducted 8 separate elections with AcccuVote optical scanners of the same vintage in a county with about 10k voters from 2010-2013. Harvie Branscomb told me back then he encountered a similar situation in another county – that it was a known risk. I never doubted him, but could only state that we never encountered this issue in any election in my county. We maintained a robust routine maintenance regimen, tested all different types of ballot stock in the pre-election logic & accuracy test (folded and flat ballots printed commercially and in-house with Runbeck’s Sentio ballot-on-demand system), and then performed a rigorous post-election audit. Our commercial ballot printer precisely folded absentee ballots using the fold mark overlay the ballot layout software generates so folds do not bisect target areas (ovals), and we were careful to fold ballots in the same way for the few hundred we prepared manually and mailed in-house. Based on my personal experience, AccuVote optical scanners are old & clunky & slow & decidedly not sexy, but they are reliable workhorses and very accurate.

    This system has since been retired in my state, but it’s important for people to understand that this type of machine tabulation error for any voting system is quite rare and isolated, and never occurs in the overwhelming majority of other jurisdictions that use the same or similar systems. Also, I learned a lot of things I know about elections the hard way, by doing them the wrong way before I figuring out the better or correct way, or hearing about similar experiences from others. If these initial indications are confirmed by the auditors’ final report, the election officials in Windham obviously should reassess some of their business processes. But the fact that this happened to them does not mean they’re incompetent or inattentive or acted with malicious intent. It means they’re fallible human beings who probably never encountered or heard of this vulnerability before, and now will improve procedures so they don’t make the same mistake twice. That’s how every election official I know learned their craft. By all accounts, the Windham officials conducted the election very well, and preserved all of the original artifacts necessary to conduct a forensic audit at all. I hope the good people of Windham bear those points in mind.

    Thanks again.

    • Harvie Branscomb says:

      Yes Dwight Shellman became an Election Director in a County adjacent to mine after my county abandoned use of the device thanks to my demonstrations of inaccuracy of the same device. He did not encounter a problem in his elections but at that time Colorado did not have sophisticated tabulation audits and LATs were not sufficient to show such problems. I expect he never inserted the same intentionally marked test ballots ten times as a test of reliability of interpretation as I did. I had conducted extensive tests using test ballots intentionally marked to be easy for a human to interpret but difficult for this machine interpret. I could regularly get 8 votes where ten were expected when the same ballot was inserted 10 times.

      The lack of interest in my results or disbelief they were real is why my work could not have protected NH from suffering from a very similar error in 2020 to one my county encountered in 2005. My testing followed an problem election during which I was a candidate. Stray marks on a “vote for 3” contest caused by loose toner, but only after folding, and compressing led to ghost votes and overvotes. Only after too many of these invisible overvotes were detected was a problem suspected. Suspicion led to a hand count and that full hand count confirmed the problem existed in the form of a 5 percent discrepancy. It was very similar to the NH situation.

      A serious lack of effective means to share information. about inherent errors in voting systems is a reason why jurisdictions still find problems that have already been solved elsewhere.

      Dwight says Accuvote devices are very accurate and he has told me that before but an adequate test for accuracy would have to be at least as much as recently occured in NH – a hand count or perhaps two to compare with the machine count. Even then not all defects may have been revealed.

      Many officials developed confidence in the accuracy of the Accuvote but that confidence was based on limited test experiences allowed during elections. The post election audit in CO at the time wasn’t strong enough to substantiate a narrow margin election and a hand count in parallel with the full official machine count such as Windham benefitted from was virtually impossible to obtain. We must develop methods to achieve confidence from public access to an evidence base and move beyond expressions of confidence.

  4. Matt A Stuart says:

    Shouldn’t a ballot with over-votes be rejected by the machine and treated as a spoiled or challenge ballot? Just curious.

    • Andrew Appel says:

      On this voting machine, as in most precinct-count optical-scan voting machines, election administrators can configure that machine to either (1) notify the voter that there are overvotes, and let the voter decide whether to take the ballot back and correct it or to cast it anyway; or (2) don’t notify the voter, and accept the ballot anyway.

      Then either way, if a ballot with overvotes is cast, the overvote may be in only one contest out of several on the ballot. In any contest without an overvote, the voter’s vote will be counted. An any contest with an overvote, the vote will not be counted.

    • Douglas W. Jones says:

      The Help America Vote Act of 2002 requires that optical-scan voting systems have the ability to return overvoted ballots to the voter and to sort out such ballots for human attention when the voter isn’t there. Federal rules explicitly talk about the ability to do this and do not require that jurisdictions enable this ability.

      This is all a matter of states’ rights. The Federal government can ask that voting equipment have certain features and abilities, but they avoid dictating to the states anything about the use of these features and abilities. The only real exception is when it can be shown that state procedures are discriminatory against voters on the basis of race or other protected category — in such cases, civil rights law and disability law apply, allowing the Federal government to force the states to do things.

      The net result is that mediocre election administration is perfectly legal under Federal law so long as it is not mediocre in a discriminatory way. So long as votes are miscounted at random, without regard to race and disability, all is well, as things now stand.

      • Deborah Sumner says:

        Dr. Jones, I disagree with your interpretation of HAVA. Maybe you can convince me I’m wrong.

        This overvote notification requirement for optical scans isn’t part of the “voluntary guidelines.” It was required in order to receive federal money and NO waiver was allowed.

        See Section 301 (a) (1) (A))( i-iii)

        https://www.doj.nh.gov/election-law/documents/title-iii-hava.pdf

        (d) EFFECTIVE DATE.–Each State and jurisdiction shall be required to comply with the requirements of this section on and after January 1, 2006.

        Ironically, I used your information and Dr. Stark’s court testimony when three of us notified the NH AG in Sept. 2017 that NH was NOT in compliance with the federal requirement. The response was they weren’t going to do anything about it.

        At the time I knew of two fraud possibilities with overvotes in the AccuVote system. I now know of four and have informed state government officials, the town and city clerks association and legislators so they ALSO know what they are.

        September 9, 2017 (sent by regular mail)

        Gordon J. MacDonald, Attorney General
        NH Department of Justice
        33 Capitol Street
        Concord, NH 03301

        Dear Mr. MacDonald,

        We recently learned that even though NH has chosen not to comply with federal standards for voting equipment, it is still supposed to be following minimum standards specified in Section 301(a) of HAVA, which includes 3) notify the voter when more selections are made than permitted (over votes).

        See links to Press Release and 2015 standards below.
        https://www.eac.gov/assets/1/28/EAC%20Updates%20Federal%20Voting%20System%20Guidelines-News-Release-FINAL-3-31-15-website.pdf

        https://eac926.ae-admin.com/assets/1/Documents/VVSG.1.1.VOL.1.FINAL.pdf

        NH is the only state we know of that does NOT program its computers to return over voted ballots to election day voters for possible correction.

        Known reasons for over votes:

        1. If a voter changes his/her mind, crosses out one vote and fills in another oval, voter intent could be determined by visual inspection, but not by the computer)
        2. Creases on absentee ballots counted as “votes”
        3. Voter intent not recognized by computer
        4. Stray marks/dust counted as “votes”
        5. Computer or programming error
        6. Specks on the paper due to the printing process
        7. Two KNOWN possibilities for fraud (let me know if you want to know what they are).

        Philip Stark, statistics professor and associate dean of mathematical and physical sciences at UC Berkeley, has testified in court that “There are documented instances where scanners have high rates of erroneously inferring that valid votes are over votes.”

        See link to Doug Jones article:http://homepage.cs.uiowa.edu/~jones/voting/optical/

        Over vote rates for Jaffrey appear to be extremely high, but since the 2010 law requiring reporting over votes to Concord was changed, there is no way to know if this is typical of the rest of the state.

        In Nov. 2012, 71 of 2,843 Jaffrey ballots tabulated by the AccuVote were recorded as over votes, 2.5% of the total. Nov. 2014, 21 of 1,945 ballots, 1.1%. Nov. 2016, 49 of 2,867 ballots, 1.7%.

        It appears likely that hundreds, even thousands of voters are being disenfranchised every election, in violation of RSA 659:64 Determining Intention of Voter and HAVA. That can make a difference in close races such as we had in November (information previously shared with your office).

        We respectfully request:
        1. In consultation with Anthony Stevens, Assistant Secretary of State, issue a directive to LHS Associates to begin programming all AccuVotes to return over voted ballots to election day voters for possible correction, Make sure local officials know how to preserve voter privacy when that occurs and how to use the over ride function if needed for absentee ballots or voters who don’t have time to vote a new ballot. (See RSA 656:42)

        2. Require reporting of over votes to Concord and that the Secretary of State prepare proper reporting forms for each election that include a candidate for federal office. (Amend RSA 659:73).

        3. If you determine legislative action is needed for either/both of the above, to work with a legislator to draft the legislation and speak for it in committee.

        4. If you decide to take no action on this request, please let us know ASAP and no later than Sept. 18.

        Copy: Anne Edwards, Associate Attorney General (sent 9/11/17 via email)
        Brian Buonamano, Election Law Attorney (sent 9/11/17 via email)

        • Douglas Jones says:

          I am not a lawyer, but Section 301 (a) (1) (B) appears to modify Section 301 (a) (1) (A) by permitting educational and instructional material to be substituted. This is crucial for vote-by-mail voters, since without (B), section (A) would require somehow notifying the voter of an overvote, after the ballot was stripped of any connection to that voter. Section B does not limit itself to postal voters, so the net result has always appeared to me (as long ago as 2001 when I first read a draft) as permitting educational materials and instructions, all subject to no quality standards at all, to be substituted for the mechanism required by part (A).

  5. Marvin SAMMONS says:

    Bunk! Machines are bad. No ID is BAD*. No Excuse mail in BAD*. Dropboxes BAD*. Fake pandemic BAD.

    * unless legislation specifies.

Speak Your Mind