January 09, 2024

In favor of hand-counted audits

January 8, 2024

Honorable members of the [New Hampshire] Senate Election Law and Municipal Affairs Committee:

I write in opposition to SB 489-FN, An act relative to election audits. [https://www.gencourt.state.nh.us/bill_status/pdf.aspx?id=22927&q=billVersion]

As I have written to Senator Gray in the past, I am opposed to the lack of transparency and public oversight and increased reliance on hackable, riggable computerized election machinery that using scanners and ballot images for audits embodies.

No true election security expert endorses using a scanner to perform an audit, even if it's a new make or model. Manual/hand-counted post-election audits have been recommended since 2009 by both the Secretary of State's own Electronic Ballot Counting Device Advisory Committee [formerly at http://sos.nh.gov/ballotcountdev.aspx; can supply] and the national League of Women Voters [https://www.lwv.org/sites/default/files/2018-07/report-electionaudits.pdf], as well as the experts who testified before the Kobach commission: Andrew Appel, Harri Hursti, and Ron Rivest. [https://youtu.be/XkPJbbKPJ_w?si=Lr3F0xQ1Az2gxpi5&t=7391]

Verified Voting, the premier body of election experts, is unequivocal: "Audits require human examination of voter-marked paper ballots... Audits cannot rely on scanned images or machine interpretations of the ballots to accurately reflect voter intent" (p. 7, Principles and Best Practices for… Audits). [P. 7, https://verifiedvoting.org/wp-content/uploads/2020/05/Principles-and-Best-Practices-For-Post-Eleciton-Tabulation-Audits.pdf]

Let us stop to ask, Why do audits in the first place? It's because election security experts know there are too many undetectable vulnerabilities with computerized voting, and manual audits of hand-marked paper ballots are our best defense. So why are we turning to computerized scanners and digital images when the experts say we should be doing hand-counted audits of the paper ballots?

The convenience of letting a machine do the work of auditing with digital images is very seductive. Proponents will even argue that people are more error-prone than machines. The problem is that the public can’t oversee the software in a scanner as it computes votes.

New Hampshire has a long, proud tradition of meticulous hand counts and recounts. The safeguards built into those can be utilized in audits and even strengthened to allow for maximum transparency and public oversight.

Here are concerns about the bill as written:

“I. The secretary of state shall randomly select not less than 8 ballot counting devices used by towns or city wards to be audited…. The selection of towns or city wards to be audited shall be non-public and made after any ballot counting device memory cards have been programmed.

“II. The secretary of state shall not announce which devices in the towns or city wards have been selected to be audited until after the polls are open on election day.”

The whole point of random audits is to make it impossible for *anyone* to know ahead of time which machines will be selected. Allowing the secretary of state to choose the machines in advance negates the protective effect of random selection.

“VI.(a) A random sample of the ballots counted and/or scanned by the device being audited, shall be selected and shall be compared with the voters' choices recorded for that ballot.”

The previous version of this bill, SB 366-FN [https://legiscan.com/NH/text/SB366/2022], was worded slightly differently:

"VI. A random sample of not less than 5 percent of the ballots scanned shall be selected and the images of the ballots [emphasis added] selected shall be compared with the voters' choices recorded for the ballot."

Given that the audits will be conducted using a scanner, I’m guessing that ballot images will again be used for comparison. I used to think this meant that the ballot images would be compared with the paper ballots to ensure that the images are accurate copies but instead am told that the ballot images will be compared with the cast vote record (CVR), a computer-generated record based on the ballot images. If true, this proposed audit never actually verifies that the ballot images – which the scanner uses to tabulate the votes – are accurate copies of the paper ballots.

Here's what Professors Andrew Appel and Philip Stark have to say about this:

"Audit the Digital Images?

"Some vendors are promoting systems that create digital images of ballots. These vendors claim that the images make RLAs [risk-limiting audits] easier to perform because fewer (or no) paper ballots need to be inspected. That is incorrect: if a risk-limiting audit relies on images of ballots, it must check that the error in making the images from the voter-verified paper ballots plus the error the system made interpreting those images to make cast-vote records is not large enough to cause the electoral outcome to be wrong. It is a mathematical fact that this requires examining at least as many physical ballots as an audit that compares CVRs to a human reading of the paper ballots, without relying on the digital images." [P. 537, https://georgetownlawtechreview.org/wp-content/uploads/2020/07/4.2-p523-541-Appel-Stark.pdf]

In other words, as I understand it, an audit that relies on digital images of ballots requires examining at least as many corresponding paper ballots as a hand-counted audit of the paper ballots alone. Which begs the question, wouldn’t it be easier to just audit the paper ballots?

Professor J. Alex Halderman's research also demonstrates the risks of auditing digital images rather than paper ballots:

"Abstract. As paper ballots and post-election audits gain increased adoption in the United States, election technology vendors are offering products that allow jurisdictions to review ballot images—digital scans produced by optical-scan voting machines—in their post-election audit procedures. Jurisdictions including the state of Maryland rely on such image audits as an alternative to inspecting the physical paper ballots. We show that image audits can be reliably defeated by an attacker who can run malicious code on the voting machines or election management system...." [https://jhalderm.com/pub/papers/unclear-evoteid19.pdf]

Please reject this cave-in to convenience and send a message that an audit requires the same high standard of proof that New Hampshire provides for recounts – nothing less. It must be truly random – not pre-selected – and hand-counted, with ample public oversight.

Respectfully,
Barbara Glassman
Nashua, NH

May 22, 2023

Overselling digital ballot images

A response to John Brakey’s “Digital Ballot Images: Key to Trustworthy Elections & Bridging Our Great Divide”:

The cross-checking that is cited is between the digital image and the cast vote record (CVR). But the cast vote record is based on the digital image, so they may align, but what if the digital image was corrupted prior to the creation of the CVR? See J. Alex Halderman, “Unclear Ballot: Automated Ballot Image Manipulation,” https://jhalderm.com/pub/papers/unclear-evoteid19.pdf.

John Brakey holds that “even in the very unlikely event that a hacker managed to break through and alter ballot images — a far more difficult exploit than altering vote totals on a voting machine or central tabulator — this manipulation would be exposed immediately simply by spot-checking the ballot images against the originals.” Immediately? What would a statistics expert like Philip Stark recommend in the way of spot-checking? Is spot-checking mandated in the bill?

Voters are again being told to trust technology, that it is “hash-check protected,” which brings to mind Andrew Appel and Susan Greenhalgh’s reporting on “Voting Machine Hashcode Testing: Unsurprisingly insecure, and surprisingly insecure,” https://freedom-to-tinker.com/2021/03/05/voting-machine-hashcode-testing-unsurprisingly-insecure-and-surprisingly-insecure/.

I applaud John Brakey’s fight for transparency. My fear is that overselling what we can learn from ballot images will further diminish the willingness to work for the gold standard of public oversight and verification: hand counting or hand-counted audits of hand-marked paper ballots. Worst case: A hack or rig could go undetected.

John Brakey replied on Facebook: “This system is about redundancy. It is important that the original ballot is married to the ballot image and some form of a risk-limiting audit is done to verify that the ballot images have not been altered.” 

February 12, 2023

Professor Stark re GA audits

Marilyn Marks posted at Post.News. Professor Stark's testimony has relevance nationwide:

"Why GA election outcomes cannot be verified

"Now that GA's elections are over and it's pretty quiet, it's time to understand why top experts in the nation agree that GA's voting system must be overhauled to protect the vote for the future. Put aside the rancor and Trump cult disinformation campaigns and focus on the science. It's a little bit complicated, but Dr. Philip Stark lays it out in simple terms.

"Start with Dr. Philip Stark's report on Raffensperger's completely flawed audits. No race can be audited to test the outcome. That is not sustainable in turning purple Georgia!"
https://www.stat.berkeley.edu/~stark/Preprints/cgg-rept-10.pdf

January 12, 2023

In opposition to ranked-choice voting

In opposition to NH HB 345, enabling ranked-choice voting (RCV) for state party primary and municipal elections, and HB 350, establishing procedures for RCV for federal and state offices:

What the well-intentioned proponents of ranked-choice voting overlook is that it conflicts with a much more urgent and critical need for manual auditing to counter the threats of hacking and rigging and the lack of confidence in our voting machines. The complexity of ranked-choice voting would only add to the resistance to hand-counted audits.

Little-understood downside to ranked-choice voting: The RCV component of state and federal races and state party primaries will always have to be tabulated by the Secretary of State or some other state authority on their RCV-programmed central scanner. In NH, that would hold for RCV municipal elections also until local jurisdictions purchase new, RCV-capable machines. Even then, how would the ranked-choice votes for, say, Nashua Aldermen-at-Large be tabulated? By the SoS? Or maybe the city clerk?

How would we audit, let alone recount, the SoS's results? Every legislator who has a vote on RCV should watch at least a portion of the video on RCV auditing [https://youtu.be/OzF-KEMJsHc] by the RCV Resource Center to see what the challenges are. Daunting, to say the least.

RCV advocates contend that it reduces polarization. Traditional runoff elections to reach a majority of votes for the winning candidates are held in a number of states, including Georgia, Alabama, Mississippi, Louisiana, and Texas. Does anyone see a reduction in divisiveness in those states?

RCV's Achilles’ heel is the requirement for *instant* runoffs, which necessitates central, state-level, computerized tabulation to reallocate losing candidates’ votes. Traditional runoffs, on the other hand, are tabulated just like any other election. Candidates with the lowest vote totals are simply eliminated after a round; their votes are not redirected to surviving candidates. RCV advocates see its strength in exactly this capability to rank candidates. I do understand the appeal, but it comes at the unacceptable cost of local oversight, transparency, auditability, and therefore security.

See also Brad Friedman: "Why Ranked Choice Voting Is a Terrible Idea" at https://bradblog.com/?p=12608: "it's very difficult to count, virtually impossible for the public to oversee, requires central tabulation and computers to pull off, and candidates and voters in many places where it's been tried in the past have found that it's impossible to understand why some candidates won and others lost." Brad notes that Approval Voting is an audit-friendly alternative to reach winners that can attract a majority of votes.

Barbara Glassman
Nashua, NH

PS This article should be required reading: https://www.ktvu.com/news/election-error-in-alameda-county-names-new-winner-in-district-4-race

PPS Brad Friedman wrote of his 2/20/23 Bradcast, beginning at 42 minutes:

“[W]e close with details on voting system failures in both New Jersey and California in November of 2022. In both cases, the systems ended up naming candidates who actually lost to be the winners and vice versa. In Monmouth County, NJ, the ES&S voting tabulators allowed duplicate votes to be uploaded to it twice. It wasn't discovered until last month. In Alameda County, CA, the tabulators were programmed incorrectly to do the impossible math of Ranked Choice Voting (RCV) elections in the Bay area. It was discovered in late December last year by a pro-RCV organization who happened to run the math on their own computers and discovered the error.

“In both cases, the errors were discovered after losing candidates were certified as winners. How many places did that happen but wasn't ever discovered, because human beings rarely bother to count ballots to make sure the computers did so accurately. This is what comes of outsourcing public elections to private companies and running elections on computer systems that are virtually impossible to oversee by the American public.”

July 18, 2022

Re the selection of new voting machines for NH

 Honorable Members of the Ballot Law Commission and Special Committee on Voter Confidence:

It’s disappointing that only two voting machines [Dominion’s ImageCast Precinct and VotingWorks’ VxScan Precinct Scanner] were exhibited on the 12th. I hope Clear Ballot’s ClearCast Go is not out of the running. It’s the current version of the one chosen by the Bucks County, PA, commissioners when they replaced their ancient machines in 2019. 


I assume you are only considering optical/digital scanners, not universal-use ballot-marking devices (BMDs) for all voters. 

 

The Bucks County, PA, Commissioners considered both optical scanners and BMDs, even though BMDs would cost roughly twice as much, and there was no reason for non-disabled voters to use machines designed for the handicapped. Such is the power of the voting machine vendor lobby in PA.

 

A couple of election integrity advocates I had worked with formed a new group, trying again to influence the selection, and this time succeeding. SAVEBucksVotes.org compared Clear Ballot, Dominion, ES&S, Hart Intercivic and Unisyn, concluding in favor of Clear Ballot. 

 

They had so many qualms about ES&S that they devoted a separate page to it: https://www.savebucksvotes.org/expressvote-xl-problems, mostly addressing the high-priced, problematic universal-use BMD that Philly chose after intense lobbying by ES&S; but the section lower down on that page, "Ethical concerns about ES&S as a company," is well worth a look.

 

Re Dominion, in Georgia, after an examination for a lawsuit there, "a University of Michigan computer science professor who has testified numerous times on Capitol Hill about U.S. election security, J. Alex Halderman—claims that Georgia’s ballot marking devices (BMDs) 'suffer from specific, highly exploitable vulnerabilities that allow attackers to change votes despite the state’s purported defenses.'" The U.S. Cybersecurity and Infrastructure Agency, or CISA, agrees; Dominion’s touchscreen BMDs “have software vulnerabilities that leave them susceptible to hacking if unaddressed.”

 

New Hampshire would be using a different Dominion model, but the fact that the vulnerabilities Halderman discovered are so damaging that the judge sealed his report raises questions as to whether their optical scanners may suffer from some of the same weaknesses.

 

Dominion has also been damaged by Colorado clerk Tina Peters’ theft of voting machine hard drives and their release at Mike Lindell’s “Cyber Symposium.”

 

I regret being unable to attend the demonstration. Wishing you all well with a difficult decision.

 

Regards,

Barbara Glassman

Nashua, NH


May 24, 2022

Testimony to the Special Committee on Voter Confidence, Derry, NH

Honorable Members of the Special Committee on Voter Confidence:

My name is Barbara Glassman. I moved to Nashua from Pennsylvania in 2012. In 2005 I became the webmaster for a nonpartisan election integrity group trying to convince county commissioners to purchase optical scanners rather than paperless voting machines. We failed, but I’ve kept posting about election integrity, now on Facebook at Save Our Vote. [https://www.facebook.com/SaveOurVote


New Hampshire was one of the first states to adopt computerized voting with its purchase of the AccuVote in the early 1990s. Election security experts recognized the danger. Vote rigging had become much easier. All it would take is a few lines of code, riding on a memory card.


No one wanted to hear it. Neither major political party, nor the media, nor election administrators had any interest in casting doubt on the outcomes of elections. 


Security experts understood that the best defense against rigging or hacking is hand-counted audits of hand-marked paper ballots. Verified Voting has been saying that since 2004 [https://verifiedvoting.org/issues/]. The national League of Women Voters said it in 2009 [https://sites.google.com/a/leagueofwomenvoters.org/clearinghouse/rep/election-administration/lwvus-report-on-election-auditing]. In New Hampshire in 2009, the Electronic Ballot Counting Device Advisory Committee called for hand-counted audits, saying, “[T]his deserves the highest priority.” [formerly at http://sos.nh.gov/ballotcountdev.aspx; can supply] Secretary Gardner’s hand-picked experts – Andrew Appel, Harri Hursti, and Ron Rivest – testified to the need in 2017 in Manchester [at 2’3” in the afternoon session: https://youtu.be/XkPJbbKPJ_w?t=7391].


Despite New Hampshire’s long, proud history of meticulous manual recounts, Secretary Gardner resisted hand-counted audits. If he hadn’t, if he followed his own experts’ recommendations, this committee might not have been needed. A publicly observed hand-counted audit would have answered many of New Hampshire voters’ questions about the 2020 election, at least in our state.


Instead, Secretary Gardner turned to a new, high-speed scanner marketed for audits. It will scan the paper ballots and then audit the digital images. How will we even know that the images are accurate copies of the ballots? Only 5 out of every 10,000 ballots will be compared with their images to see if they match. [I'm told that the comparison was between ballot images and the cast vote record (CVR), a computer-generated record based on the ballot images, rather than the paper ballots, degrading confidence even further.]


Here's what Professors Andrew Appel and Philip Stark have to say about this: "These vendors claim that the images make RLAs [risk-limiting audits] easier to perform because fewer (or no) paper ballots need to be inspected. That is incorrect…. [T]his requires examining at least as many physical ballots as an audit” of the paper ballots alone. [P. 537, https://georgetownlawtechreview.org/wp-content/uploads/2020/07/4.2-p523-541-Appel-Stark.pdf] Which begs the question as to why it wouldn’t make more sense to just audit the paper ballots.

The convenience of letting a machine do the work of auditing with digital images is very seductive. It allows us to rationalize away the futility of trying to solve an inscrutable software problem by throwing more inscrutable software at it.

Suppose Secretary Gardner had proposed abandoning New Hampshire’s long tradition of hand recounts in favor of simply running the ballots through a new scanner. Pushback would have been immediate. How would we know which scanner was accurate without a hand recount? We can’t know. Yet that is what is offered in Senate Bill 366: machine-counted audits with a new scanner and grossly inadequate verification of ballot image accuracy.


BTW, pre- and post-election testing cannot detect every software problem. Computers know what day and time it is. Scanners can be programmed to perform well when tested and behave differently on election day. 

The VW cheating scandal is a perfect illustration. VW's software allowed their cars to pass emissions tests, only to pollute when not in testing mode. [See "What if Volkswagen made Voting Machines?" https://thevotingnews.com/what-if-vw-made-voting-machines/]

The GOP is the first major political party to recognize the fallibility of software and the need for audits. Unfortunately, a highly publicized audit in Arizona relied on inexperienced, partisan third parties that didn’t even have the confidence of some fellow Republicans. Sadly, many Democrats now view audits with suspicion and resistance. 

So if confidence in New Hampshire elections is to be restored, it will require reversing the trend toward secrecy and instead welcoming public oversight, starting with repeal of the ballot and ballot image exemptions from public records and passage of legislation that would allow moderators to do a verification count at the polls on election night, as in Senate Bill 79 [https://legiscan.com/NH/bill/SB79/2021], a measure that used to go unchallenged. Routine, mandatory audits would need to be hand-counted and publicly observed, delivering the same high standard of proof that New Hampshire recounts provide.

Designing an effective audit is a formidable task. It deserves more time and study than legislators alone can spare. I know a few New Hampshire voters who are interested in drafting a plan. I hope their voices will be heard.

Thank you for the opportunity to offer suggestions. It’s much appreciated.

Respectfully,

Barbara Glassman   

Nashua, NH 


Addendum

Verified Voting, the premier body of election experts, is unequivocal: "Audits require human examination of voter-marked paper ballots... Audits cannot rely on scanned images or machine interpretations of the ballots to accurately reflect voter intent.” [P. 7, https://verifiedvoting.org/publication/principles-and-best-practices-for-post-election-tabulation-audits/].

May 03, 2022

In New Hampshire: A continuing plea for hand-counted audits

May 3, 2022

Dear Representatives:

I write in opposition to SB 366-FN, An act requiring an audit of ballots cast in the 2022 primary and general election.

As I have written to Senator Gray in the past, I am opposed to the lack of transparency and public oversight and increased reliance on hackable, riggable computerized election machinery that using scanners and ballot images for audits embodies.

No true election security expert endorses using a scanner to perform an audit, even if it's a new make or model. Manual/hand-counted post-election audits have been recommended since 2009 by both the Secretary of State's own Electronic Ballot Counting Device Advisory Committee [formerly at http://sos.nh.gov/ballotcountdev.aspx] and the national League of Women Voters [https://sites.google.com/a/leagueofwomenvoters.org/clearinghouse/rep/election-administration/lwvus-report-on-election-auditing], as well as the experts who testified before the Kobach commission: Andrew Appel, Harri Hursti, and Ron Rivest. [https://youtu.be/XkPJbbKPJ_w?t=7391]

Verified Voting, the premier body of election experts, is unequivocal: "Audits require human examination of voter-marked paper ballots... Audits cannot rely on scanned images or machine interpretations of the ballots to accurately reflect voter intent" (p. 7, Principles and Best Practices for… Audits). [P. 7, https://verifiedvoting.org/wp-content/uploads/2020/05/Principles-and-Best-Practices-For-Post-Eleciton-Tabulation-Audits.pdf

Let us stop to ask, Why do audits in the first place? It's because election security experts know there are too many undetectable vulnerabilities with computerized voting, and manual audits of hand-marked paper ballots are our best defense. So why are we turning to computerized scanners and digital images when the experts say we should be doing hand-counted audits of the paper ballots?

The convenience of letting a machine do the work of auditing with digital images is very seductive. Proponents will even argue that people are more error-prone than machines. The problem is that the public can’t oversee the software in a scanner as it computes votes. 

New Hampshire has a long, proud tradition of meticulous hand counts and recounts. The safeguards built into those can be utilized in audits and even strengthened to allow for maximum transparency and public oversight.

Here are concerns about the bill as written:

4. “I. The secretary of state shall randomly select one to 3 percent of AccuVote devices, but not less than 4 such devices, to be used at the 2022 state general election to be audited…. The selection of devices to be audited shall be non-public and made at least 2 weeks prior to the general election, after the AccuVote memory cards have been programed [sic]."  

 

The whole point of random audits is to make it impossible for *anyone* to know ahead of time which machines will be selected. Allowing the secretary of state to choose the machines in advance negates the protective effect of random selection.

4. "VI.  A random sample of not less than 5 percent of the ballots scanned shall be selected and the images of the ballots selected shall be compared with the voters' choices recorded for the ballot."

So this bill allows as little as 1 percent of the AccuVote machines, but not less than four, to be audited, and then as little as 5 percent of the ballots from those machines to have their images compared with "the voters' choices recorded for the ballot," presumably, the voters’ hand-marked paper ballots. If my husband, an engineer, is not mistaken, that works out to a rate of 0.05 percent of the AccuVote ballots, or 5 in 10,000 paper ballots that will be compared with the corresponding ballot images to try to establish that the digital images are accurate copies of the hand-marked ballots. [Even worse, I'm told that the comparison was between ballot images and the cast vote record (CVR), a computer-generated record based on the ballot images, rather than the paper ballots.] Does anyone see that as sufficient to inspire confidence? In contrast, Massachusetts conducts hand-counted audits of the paper ballots in 3 percent of all precincts when the president is on the ballot.

Here's what Professors Andrew Appel and Philip Stark have to say about this:

"Audit the Digital Images?

"Some vendors are promoting systems that create digital images of ballots. These vendors claim that the images make RLAs easier to perform because fewer (or no) paper ballots need to be inspected. That is incorrect: if a risk-limiting audit relies on images of ballots, it must check that the error in making the images from the voter-verified paper ballots plus the error the system made interpreting those images to make cast-vote records is not large enough to cause the electoral outcome to be wrong. It is a mathematical fact that this requires examining at least as many physical ballots as an audit that compares CVRs to a human reading of the paper ballots, without relying on the digital images." [P. 537, https://georgetownlawtechreview.org/wp-content/uploads/2020/07/4.2-p523-541-Appel-Stark.pdf

In other words, as I understand it, an audit that relies on digital images of ballots requires examining at least as many corresponding paper ballots as a hand-counted audit of the paper ballots alone. Which begs the question, wouldn’t it be easier to just audit the paper ballots?

Professor J. Alex Halderman's research also demonstrates the risks of auditing digital images rather than paper ballots: 

"Abstract. As paper ballots and post-election audits gain increased adoption in the United States, election technology vendors are offering products that allow jurisdictions to review ballot images—digital scans produced by optical-scan voting machines—in their post-election audit procedures. Jurisdictions including the state of Maryland rely on such image audits as an alternative to inspecting the physical paper ballots. We show that image audits can be reliably defeated by an attacker who can run malicious code on the voting machines or election management system...." [https://jhalderm.com/pub/papers/unclear-evoteid19.pdf]

Please reject this cave-in to convenience and send a message that an audit requires the same high standard of proof that New Hampshire provides for recounts – nothing less. It must be hand-counted with ample public oversight.

Respectfully,

Barbara Glassman

Nashua, NH


January 25, 2022

Ranked-Choice Voting: A Conflict

What the well-intentioned proponents of ranked-choice voting overlook is that it conflicts with a much more urgent and critical need for manual auditing to counter the threats of hacking, rigging, and the perception that the 2020 election was stolen. The complexity of ranked-choice voting would add to the already formidable resistance to hand-counted audits. 

A wealth of resources has been directed at ranked-choice voting. “The head of Voter Choice Massachusetts, Mac D’Alessandro, told The Intercept that his organization is receiving more than $1 million in matching funds from Unite America” in 2019. Unite America is “a once obscure political organization that raised its profile earlier this year [2019] when Kathryn Murdoch became its co-chair and largest donor; she’s a daughter-in-law of Fox News founder Rupert Murdoch.” https://theintercept.com/2019/10/07/murdoch-pennsylvania-gerrymandering-lobbyists/?fbclid=IwAR0qCfiNuMklRlawN5uKBAsEVsQhmZSEVXhiDfoFkG1f-tfGyNCVxJaKjJA


I’m not suggesting there is anything sinister in the injection of Murdoch money, apart from the fact that megadonors have way too much influence over our elections and everything else. I don't doubt that Kathryn Murdoch genuinely wanted to do good on the election front and wonder whether Harvard law professor Lawrence Lessig helped convince her that ranked-choice voting is a top priority. 


I was a fan of Lessig for years; went to a couple of his New Hampshire rallies for campaign finance reform. He’s a brilliant man with an extraordinary passion for election reform. Not sure why he switched his focus from campaign finance reform to ranked-choice voting. Had he asked leading election integrity advocates, e.g., Verified Voting, for suggestions as to what is needed most, advocacy for mandatory manual audits would have come out ahead of ranked-choice voting. 


I’m not questioning anyone’s motives, but voters deserve to know all about the funding of ranked-choice voting campaigns.


I urge the advocates for RCV to read Jonathan Simon’s recent article, "The Real Steal: Electoral Forensics and the 2020 Election," as well as his book CODE RED: Computerized Elections and the War on American Democracy: Election 2020 Edition. Everyone should watch KILL CHAIN: The Cyber War on American’s Elections, featuring Harri Hursti. I hope RCV proponents will read my plea for hand-counted audits and consider supporting that cause.


BTW, there are at least two other alternatives to the current system of plurality voting that are audit-friendly. Subsequent runoff elections to reach a majority of votes for the winning candidates are held in a number of states and European nations, including Georgia, Alabama, Mississippi, Louisiana, Texas, and France. And there’s Approval Voting https://www.electionscience.org/library/approval-voting/.


See also Brad Friedman: "Why Ranked Choice Voting Is a Terrible Idea" at https://bradblog.com/?p=12608: "it's very difficult to count, virtually impossible for the public to oversee, requires central tabulation and computers to pull off, and candidates and voters in many places where it's been tried in the past have found that it's impossible to understand why some candidates won and others lost."


Little-known fact about ranked-choice voting: The RCV component of state and federal races and state party primaries will always have to be tabulated by the Secretary of State or some other state authority on their RCV-programmed central scanner. In NH, that would hold for RCV municipal elections also until local jurisdictions purchase new, RCV-capable machines, as I understand it. Even then, how would the votes for, say, Nashua Aldermen-at-Large be tabulated? By the SoS? Or maybe the city clerk?

How do we audit the SoS's results? While auditing a sample of local precincts would be necessary, it wouldn’t be sufficient without also auditing the tabulation of the ranked choices by the SoS’s scanner. I’ve started watching the video on RCV auditing by the RCV Resource Center. Every legislator who has a vote on RCV should watch at least a portion of the video to see what the challenges are. They look daunting, to say the least, especially from the perspective of a voter who wants to see hand-counted audits.

RCV's Achilles’ heel is the requirement for *instant* runoffs, which necessitates central tabulation and computerization to reallocate losing candidates’ votes. Traditional runoffs, on the other hand, are tabulated just like any other election. Candidates with the lowest vote totals are simply eliminated after a round; their votes are not reallocated to surviving candidates. RCV advocates see its strength in exactly this capability to rank candidates. I do understand the appeal, but it comes at the unacceptable cost of local oversight, transparency, and auditability.

Barbara Glassman

Nashua, NH

January 13, 2022

In New Hampshire: A plea for hand-counted audits

In opposition to SB 366-FN re audits 

January 11, 2022


To the Honorable Members of the Senate Election Law and Municipal Affairs Committee:


I write in opposition to SB 366-FN and ask Ms. Melillo to please add this to the legislative record. 

 

As I have written to Senator Gray in the past, I am opposed to the lack of transparency and public oversight and increased reliance on hackable, riggable computerized election machinery that using scanners and ballot images for audits embodies.


No true election security expert endorses using a scanner to perform an audit, even if it's a new make or model. Manual/hand-counted post-election audits have been recommended since 2009 by both the Secretary of State's own Electronic Ballot Counting Device Advisory Committee [formerly at http://sos.nh.gov/ballotcountdev.aspx] and the national League of Women Voters [https://sites.google.com/a/leagueofwomenvoters.org/clearinghouse/rep/election-administration/lwvus-report-on-election-auditing], as well as the experts who testified before the Kobach commission: Andrew Appel, Harri Hursti, and Ron Rivest. [https://youtu.be/XkPJbbKPJ_w]


Verified Voting, the premier body of election experts, is unequivocal: "Audits require human examination of voter-marked paper ballots... Audits cannot rely on scanned images or machine interpretations of the ballots to accurately reflect voter intent" (p. 7, Principles and Best Practices for… Audits). [P. 7, https://verifiedvoting.org/wp-content/uploads/2020/05/Principles-and-Best-Practices-For-Post-Eleciton-Tabulation-Audits.pdf


Let us stop to ask, Why do audits in the first place? It's because election security experts know there are too many undetectable vulnerabilities with computerized voting, and manual audits of hand-marked paper ballots are our best defense. So why are we turning to computerized scanners and digital images when the experts say we should be doing hand-counted audits of the paper ballots?


The convenience of letting a machine do the work of auditing with digital images is very seductive. Proponents will even say you can't trust the people doing hand counts. New Hampshire has a long, proud tradition of meticulous manual recounts. The safeguards built into those recounts can be utilized in audits and even strengthened to allow for maximum transparency and public oversight.


Secretary Gardner chose not to follow the advice of his own hand-picked experts. With his departure, we now have the freedom to reassess.


If that does not happen, then here are concerns about the bill as written:


4."I.  The secretary of state shall randomly select one to 3 percent of AccuVote devices, but not less than 4 such devices, to be used at the 2022 state general election to be audited. … The selection of devices to be audited shall be non-public and made at least 2 weeks prior to the general election, after the AccuVote memory cards have been programed [sic]."  


The whole point of random audits is to make it impossible for anyone to know ahead of time which machines will be selected. Allowing the secretary of state to choose the machines in advance negates the protective effect of random selection. It is not random selection.


4. "VI.  A random sample of not less than 5 percent of the ballots scanned shall be selected and the images of the ballots selected shall be compared with the voters' choices recorded for the ballot."


So this bill allows as little as one percent of the AccuVote machines, but not less than four, to be audited, and then as little as 5 percent of the ballots from those machines to be selected for comparison with "the voters' choices recorded for the ballot." Does this refer to the hand-marked paper ballots? If my husband, an engineer, is not mistaken, that works out to a rate of 0.05 percent of the AccuVote ballots, or 5 in 10,000 paper ballots that will be compared with the corresponding ballot images to try to establish that the ballot images are accurate copies of the hand-marked ballots. Does anyone see that as sufficient to inspire confidence? In contrast, Massachusetts conducts hand-counted audits of the paper ballots in 3 percent of all precincts when the president is on the ballot.


Here's what Professors Andrew Appel and Philip Stark have to say about this:

"Audit the Digital Images?

"Some vendors are promoting systems that create digital images of ballots. These vendors claim that the images make RLAs easier to perform because fewer (or no) paper ballots need to be inspected. That is incorrect: if a risk-limiting audit relies on images of ballots, it must check that the error in making the images from the voter-verified paper ballots plus the error the system made interpreting those images to make cast-vote records is not large enough to cause the electoral outcome to be wrong. It is a mathematical fact that this requires examining at least as many physical ballots as an audit that compares CVRs to a human reading of the paper ballots, without relying on the digital images." [P. 537, https://georgetownlawtechreview.org/wp-content/uploads/2020/07/4.2-p523-541-Appel-Stark.pdf


In other words, an audit that relies on digital images of ballots requires examining at least as many corresponding paper ballots as a hand-counted audit of the paper ballots alone.


Professor J. Alex Halderman's research demonstrates the vulnerability of auditing digital images: 

"Abstract. As paper ballots and post-election audits gain increased adoption in the United States, election technology vendors are offering products that allow jurisdictions to review ballot images—digital scans produced by optical-scan voting machines—in their post-election audit procedures. Jurisdictions including the state of Maryland rely on such image audits as an alternative to inspecting the physical paper ballots. We show that image audits can be reliably defeated by an attacker who can run malicious code on the voting machines or election management system...." [https://jhalderm.com/pub/papers/unclear-evoteid19.pdf]


I sincerely hope you will take this opportunity to reenvision the path forward.


Respectfully,

Barbara Glassman

Nashua, NH