Play Live Radio
Next Up:
0:00 0:00
Available On Air Stations
Support NHPR and you could win a trip to Paris!

‘A Conspiracy Of Coincidences’: Ballot Folds Were To Blame In Windham Election Discrepancies

Allegra Boverman for NHPR

An anomaly in Windham's 2020 election results attracted national attention and plenty of conspiracy theories, but auditors have found that the actual explanation was much simpler: The problem stemmed from misplaced folds in its absentee ballots.

Auditors investigating discrepancies in a local legislative race released a report today concluding that fold lines on ballots were the “primary root cause” of about a 400-point gap in vote totals.

One of the auditors was Harri Hursti, an internationally known election security expert and ethical hacker. He joined All Things Considered host Peter Biello to share his findings.

Hursti’s Takeaways:

  • The machine Windham used to prepare its absentee ballots last fall wasn’t properly calibrated and folded the ballots in the wrong place. The ballots were designed to be folded with creases between vote bubbles, but in Windham, the crease overlapped with one of the bubbles for a state representative candidate, which messed with the vote totals when those ballots were fed through scanners to be tallied on Election Night. 
  • There was no evidence of voter fraud or other election tampering. Hursti says New Hampshire’s use of paper ballots makes its elections more secure in general.
  • While New Hampshire doesn’t routinely audit its election results, it is well-positioned to build on what it learned from the Windham investigation to implement a solid risk-limiting auditing system in the future. Hursti thinks the state should be ready to audit its elections by 2022.

Get NHPR's reporting about politics, the pandemic, and other top stories in your inbox — sign up for our newsletter (it's free!) today.


Peter Biello: This is All Things Considered on NHPR, I'm Peter Biello. Last year's election went smoothly in New Hampshire, except for one lingering question: Why did the vote totals in Windham's state representative race change so dramatically between election night and a state recount a week later? We now have an answer to that question. A team of auditors recently wrapped up an investigation. One of those investigators is here now to share what he learned. Harri Hursti is an ethical hacker and an internationally known election security expert. Harri, thank you very much for speaking with me.

Harri Hursti: Thank you for having me.

Peter Biello: Your audit found that the "primary root cause" of the discrepancies in Windham had to do with how the absentee ballots were folded. Can you tell us a little more about that and why it made such a big difference in the vote totals?

Harri Hursti: So, the folding itself is a function here where it's always a question of what kind of paper you are using, how it folds and, in this case, it's also the folding machine. We found out that the folding machine, which was leased by the town of Windham for other purposes, not for elections, when it was used, they did not adjust the machine to fold it where the proper place to fold it is. And the proper place is a scoring line made by the printer. The other part [that] was a contributing factor is also that the folding machine was probably slightly broken, so it didn't make the folding horizontal, it may be slightly diagonal, and that might be a contributing factor… Why was this not noticed? Because if the line would have been completely horizontal, it probably would have damaged timing marks in a certain percentage of the paper, causing the voting machines to reject the ballots. This was a conspiracy of coincidences. So many things needed to go wrong in order to have this anomaly happen.

Peter Biello: OK, and just to be clear, your team found no evidence of fraud, tampering, or any malicious activity in Windham, correct?

Harri Hursti: We did not find any kind of malicious activity. We did a thorough investigation. We looked [at] every single possible allegation, every single possible theory out there. We couldn't find anything [that] would be malicious or which would have a political bias. And I want to stress here that New Hampshire is using a principle called ballot rotation. So, not any place on a ballot has a political bias because the political parties between the jurisdictions are randomly organized. Also, if the same candidates are in multiple jurisdiction ballots, then the candidate orders are randomly organized. So, even if this would be somewhere else, it would have no political bias because everything is randomized between the jurisdictions.

Peter Biello: In other words, the candidates for any particular office are not listed always alphabetically by last name, they're switched up, and ballot machines, when they count, if they are calibrated correctly, can figure out where the candidate is and calculate those votes appropriately. That's what you're saying?

Harri Hursti: Correct.

Peter Biello: With respect to the folding problem in Windham, do you have any reason to believe that this issue was widespread?

Harri Hursti: We tried to find out if there are other places using voting machines and the volunteers, which we had to do the hand count on and assist this audit - they were all sworn-in election officials in nonpartisan offices - and we were immediately asking everybody, "Do you know any other place using the voting machine?" And we couldn't [find] any place where they would use a folding machine.

Peter Biello: For all this mystery around Windham's vote totals, the actual cause of this problem seems surprisingly straightforward. It's all about how they were folded in these machines. In your view, was this preventable?

Harri Hursti: This was something which can be prevented easily in the future. If not anything else, sending the paper in a flat envelope, making certain that people now know that folding is more dangerous than it is. Again, the underlying fact here is that at no point of time was this affecting the outcome of the election. Even with the original machine count and the hand count, which revealed the discrepancy we were investigating, never any of this changed the outcome of the election.

Peter Biello: A key goal of this process was to identify ways to prevent this from happening again. What were the most important things New Hampshire could do to shore up its elections against similar problems?

Harri Hursti: First of all, we have a list of recommendations at the end of the report. One thing which is for public confidence and making certain that the elections are always going right, having a mandatory risk-limiting audit for every race, every single time, so nobody has to request that. The Secretary of State of New Hampshire has already piloted this last year, so this process has already started. My belief is that they will continue doing this. After all, even with this folding problem, what we found is that New Hampshire has an excellent control over paper ballots and excellent control over the process. So, nobody would say that 2020 was an easy year for anyone because of the COVID and everything, so under the circumstances, this was still a well-run, good election, even when there was this discrepancy and the discrepancy was caught, which means that the system worked.

Peter Biello: In your view, how urgently should New Hampshire adopt the auditing system you described?

Harri Hursti: I mean, the risk-limiting audit is something where the first states which implemented it, it took them six to eight years to get it right. So right now, New Hampshire has the benefit of already having access to the best practices. I mean, my recommendation would be implementing that for 2022.

Peter Biello: Is there anything else about this audit that you think people should really keep in mind going forward?

Harri Hursti: Well, I think the key point here is that the reason why New Hampshire's elections are extremely well-run and trustworthy and [enabled] this audit is [because] New Hampshire is paper ballot based, and you have an excellent control over the forensic material. Without a paper ballot, without the control, you would have never been able to do a complete forensic audit and find the root causes. But that is possible only because the foundation is solid.

Peter Biello: Secretary of State Bill Gardner often likes to say about New Hampshire elections, "You can't hack a pencil." He really does like the paper system that we have. Seems like what you're saying validates his statement.

Harri Hursti: Correct. It's kind of interesting that the computer security expert, the election security experts, hackers like myself, we almost sound like Luddites because we understand how vulnerable the technology is and how important it is to have hand-marked paper ballots so that the voter's intent is recorded in a permanent media and you have an ability to go back and verify the results, because every voting machine we have today and every voting machine we have for the foreseeable future always can be hacked. Using voting machines is a must across the U.S. and in bigger jurisdictions in New Hampshire, however, always, always verify.

Peter Biello: Harri Hursti is an ethical hacker and an election security expert. Thank you again for speaking with me. I really appreciate it.

Harri Hursti: Thank you for having me.

Peter Biello is the host of All Things Considered and Writers on a New England Stage at New Hampshire Public Radio. He has served as a producer/announcer/host of Weekend Edition Saturday at Vermont Public Radio and as a reporter/host of Morning Edition at WHQR in Wilmington, North Carolina.
Julia Furukawa is the host of All Things Considered at NHPR. She joined the NHPR team in 2021 as a fellow producing ATC after working as a reporter and editor for The Paris News in Texas and a freelancer for KNKX Public Radio in Seattle.

You make NHPR possible.

NHPR is nonprofit and independent. We rely on readers like you to support the local, national, and international coverage on this website. Your support makes this news available to everyone.

Give today. A monthly donation of $5 makes a real difference.