Which system of voting will return faith in election outcomes?
That was the question at hand for Montgomery County’s Election Board in February and March.
County Clerk Sondra Sixberry wants a return to fully hand-marked, hand-counted ballots. In this system, cameras and eyeballs would verify accuracy and accountability. To support her proposal, Sixberry offered evidence from Co-Founder of the Chester County Coalition for Voting Integrity, Professor Jana Nestlerode.
While humans rely on machines to reduce human counting errors across banking, budgeting, payroll, and other industries, machines’ trustworthiness have become a point of controversy in elections. Nestlerode says it’s not about small acts of fraud–those are already illegal–but system-wide exploitation.
Nestlerode, who teaches and researches at West Chester University in Pennsylvania, argues that the core problem with voting machines is not “vote fraud” by individuals but “election fraud”: large‑scale, electronic manipulation of computer‑counted vote totals that local poll workers would never see.
Such concerns stem from assertions that machines can be infected by viruses. They can be altered or mis-programmed by partisan vendors. And because proprietary software is secret and often shielded from forensic review, Nestlerode, Sixberry and others contend that citizens and officials have no real way to know whether machine tallies match voters’ choices, even when systems are not connected to the internet.
Nestlerode cites patterns such as consistent “red shifts” between unadjusted exit polls and reported results, as well as past vendor conflicts of interest, as circumstantial evidence that computerized tabulation is a “hacker’s paradise” and incompatible with transparent democracy.
Sixberry looks to Nestlerode’s case for returning to precinct‑level, hand‑counted paper ballots, where ballot counting occurs openly, under the eyes of candidates, voters, cameras, and community observers. The argument is that hand counting builds civic participation, reduces dependence on private technology vendors, avoids machine purchase and maintenance costs, and—citing the Caltech/MIT Voting Technology Project—can yield fewer spoiled and uncounted ballots than touchscreen systems when done properly.
At the February 10 election board meeting, which was well-attended, Sixberry advocated expanding the use of hand‑marked paper ballots and exploring hand counts at the precinct level. This was framed as a “Trust but Verify” compromise to rebuild voter confidence. The county currently uses touchscreen tablets that print a paper ballot with a barcode that must be run through a separate scanner/tabulator. While none of these machines are connected to each other or to the internet, Sixberry and some members of the public argued that hand‑marked, hand‑counted ballots would better show “voter intent,” avoid dependence on software and foreign‑owned vendors, and reduce long‑term equipment costs as older machines age out.
Note: while the voting machines are not connected to anything, the e-pollbook machines used to scan and verify voter IDs are connected to a system to prevent double-voting.
What research says about hand counts vs. machines
Every day, humans rely on machines to count and perform calculations for them. Their cars tabulate miles driven. Gasoline pumps have safety measures to ensure the cost per gallon and the charge are accurately metered. Same with household energy. Calculators help balance budgets. While some systems are hackable, election machines are among those designed with safeguards.
The Bipartisan Policy Center (BPC) concludes that machine tabulation of paper ballots is more accurate, faster, and cheaper than counting all ballots by hand, especially at scale. BPC’s elections research endorses both hand‑marked and machine‑marked paper ballots as secure, provided they produce a voter‑verifiable paper record that can be audited and used in recounts, explicitly distinguishing that from “paperless” options. BPC notes that machine tabulators are “the linchpin of accurate and timely election results” and warns that efforts to replace them with full hand counts would in fact “undermine election integrity and sow distrust” because of delays and higher error rates.
Nearly all jurisdictions in the US have turned to voting machines because they are more accurate and far faster than hand counts. Hand counts can have error rates up to 50 times higher than machine counts when large numbers of ballots are involved. Hand counts end up being more costly when it comes to time, trust, and dollars. The Bipartisan Policy Center cites one study in Michigan where shifting toward hand counts required more precincts and staff. It took longer and yielded lower accuracy.
Cost, maintenance, and reliability
BPC notes that almost every state now has formal processes to test and certify voting equipment. Most rely on a federal testing program that has been in place for nearly 20 years, and machines must meet error‑rate standards and produce a paper record. Election officials in every state test equipment before each election, and 49 states conduct post‑election audits that typically involve manually checking a sample of paper ballots against machine tallies, precisely the “trust but verify” approach discussed in Montgomery County. BPC’s security explainer also underscores that 96 percent of voters in 2026 will cast ballots with a voter‑verifiable paper trail, which is already the model used in our county.
On cost and maintenance, Sixberry correctly notes that many of the county’s ballot‑marking devices are aging and will eventually require replacement at a per‑unit cost in the $3,500–$3,800 range, and that e‑pollbooks (the machines that verify voter IDs) will also need to be replaced when their state certification expires. However, BPC analysis shows that trying to substitute human labor for tabulators is typically more expensive than replacing equipment, once you account for poll‑worker wages, overtime, recruitment and training, and the need to double‑staff counting teams to catch human error.
As one business owner at the February 10 meeting noted: his highest cost is labor, not machines, and he doubts that “manpower will be cheaper than machines.”
For the system Montgomery County actually uses—a voter‑verifiable printed ballot plus a separate scanner—the best‑practice model is to keep using machine tabulation, invest in normal replacement of tabulators and e‑pollbooks, and layer on statistically sound audits.
Objections to machine voting
From the February 10 discussion, the main objections to voting machines were:
- ●“Any machine can be hacked,” even if not connected to the internet.
- ●Concerns about overseas vendor ownership and distrust of QR‑coded ballots.
- ●Confusion between the county’s current system, which retains a paper record of each vote, and direct‑recording machines that store votes only electronically.
- ●Fears that software or engineers could invisibly change results at the state or server level.
- ●General preference for talking to a person rather than “trusting a machine.”
Several participants who’ve worked in some IT positions stressed that “anything with electronics” is theoretically vulnerable if someone gains access, but others clarified that Montgomery County’s voting tablets and scanners are in “airplane mode”–not networked and not connected to any server. Only the e‑pollbooks (which manage the voter list, not votes) network, and they are locked down by state rules.
BPC’s nationwide review explains the layered safeguards around such systems: strict physical and access controls, pre‑election testing, certification, non-internet-linked tabulators, and post‑election audits in nearly all states. It also emphasizes that tabulators always work from a paper record that can be recounted by hand if needed, which the county already does in recounts.
Most of the concerns centered around other counties’ systems—especially paperless and centralized systems—which Montgomery County does not use. BPC explicitly distinguishes those older systems from the modern system our county uses. Again, MoCo systems are all hand‑ or machine‑marked paper ballots, voter‑verifiable, tabulated by certified equipment, and backed by audits.
The other members of the Election Board, former state legislator and Republican Party Chair Phil Boots and Democratic Party Chair Virginia Servies, discussed a constructive path forward, consistent with Sixberry’s updated “Trust but Verify” statement.
- ●Keep the current touch‑screen ballot‑marking plus scanner system as the primary method.
- ●Offer hand‑marked paper ballots on request for voters who prefer them, as Indiana law already allows.
- ●Use the IC 3‑12‑14 procedural audit, or a routine post‑election hand audit of selected precincts, to compare hand counts with machine results and publish the (likely very small) differences.
That approach directly addresses hacking fears, preserves accessibility and speed, respects state law, and follows research showing that machine counting of paper ballots—with meaningful audits—is both more accurate and more cost‑effective than full human hand counts.
At the end of the February 10 meeting, the board postponed a decision.
When they reconvened on March 10, a small crowd heard board members Virginia Servies and Phil Boots report that state election officials would not allow a post‑primary hand recount of ballots to “verify” machine accuracy. Servies noted that the board must follow state law and the state constitution.
Sixberry continued to press her “Trust but Verify” idea, citing Indiana code provisions she believes would allow comparing past hand counts to machine tallies, suggesting dual locks on voting machines and overnight camera/security monitoring, and arguing that voters should have a choice between machine‑marked and hand-marked ballots. Both are paper ballots.
While some attendees voiced strong distrust of machines and proposed a “We the People” referendum to decide the issue, Boots and Servies said this is not permitted under Indiana law.
Ultimately, MoCo is following state law and using machines in the most accurate, efficient and legal means possible, and no member of the election board or community has found evidence to cast doubt on the accuracy and veracity of local election results.
“I’ve served on the election board for ten years, and I am very confident in our elections,” said Servies. “We have always had honest, accurate reports.”
Note: We reached out to Sondra Sixberry for further comment. We didn’t hear back yet.