Atlanta, Georgia, September 17, 2026
Mark Lindeman, Policy & Strategy Director, Verified Voting
Full Testimony:
Good morning, Chairman Anderson, Chairman Burns, members of the committee. My name is Mark Lindeman, and I’m policy and strategy director of Verified Voting, a national nonpartisan nonprofit organization. Our mission is to strengthen democracy for all voters through the appropriate use of technology in elections. So we appreciate your work to set voting system standards on behalf of all Georgia voters, and personally, on behalf of my extended family in Augusta, I thank you for taking on this very important study.
Your basic task, as I understand it, is to scope out the criteria for a voting system that provides hand-marked paper ballots, tabulated electronically, as the primary voting method, and excellent ballot-marking devices as an alternative. To frame up my remarks, I want to briefly address the benefits of that approach. Then I will talk a bit about the voting system market nationwide, and especially how the federal Voluntary Voting System Guidelines, or VVSG, may affect your work. And finally, I’ll address a few specific issues.
I’ll start with broad strokes. We all have heard a lot about voting system security, not just today, but over the years. Let me propose a baseline fact: there’s no such thing as a voting system that’s perfectly secure and reliable. Nobody has designed election technology that can only ever do what it’s supposed to do. And, let’s face it, that’s true of people too. So, it’s important for election technology and people to work together and even to check each other’s work.
Now one of our basic frames is resilience. In a resilient election system, when things go wrong, the problems can be contained, detected, and corrected. You can strive for perfection, but if you’re designing for resilience, you always have a good Plan B. We argue that voter-verified paper ballots are crucial to resilience. Voter-verified paper ballots provide a means to check electronic tabulators, to know if they’re right and to correct them if they’re wrong. The paper ballots increase the system’s resilience against technology failures, mistakes, cyberattacks, and even fearmongering. Georgia’s routine risk-limiting audits help confer that resilience.
In that context, we support a move to hand-marked paper ballots for most voters because it tends to reduce reliance on electronic technology and put the paper ballot, which is the definitive record of voter intent, at the center of the voting experience.
That said, let me be clear: Georgia voters deserve access to ballots that they can mark by hand, but they also deserve access to excellent ballot marking devices that can help them mark their ballots, if they need or prefer that help. That isn’t just a matter of law. It’s an issue of basic fairness. Ballot marking devices can’t be an afterthought. Election workers can give priority to voters who need it, but they should not gatekeep who can use the machines. And we think that more than a few voters should use the BMDs. It doesn’t have to be hundreds in a polling place, but to protect voter privacy and dignity, it shouldn’t be just a handful. Voting equipment isn’t very accessible if almost nobody uses it.
Now, what I’ve just described may seem ambitious, but it’s broadly how most in-person voters around the country vote already. Let me show two slides, based on our online database called the Verifier, where we have tracked voting systems around the country going back to 2006. Here’s the headline: over two thirds of U.S. registered voters live in counties where both hand-marked paper ballots and ballot marking devices are available on election day. Georgia is among just four states that only provide ballot marking devices.
This second map illustrates something you’ve probably noticed: Georgia is considerably the largest state by population to have a statewide voting system. For that and other reasons, it can be confusing to relate experiences elsewhere to the Georgia case. But I think you’ll always find good examples that Georgia officials can learn from and adapt.
Now, let me talk briefly about the federal Voluntary Voting System Guidelines, or the VVSG. Currently, not by law but by rule, Georgia requires voting systems to meet the VVSG standards. Georgia’s current voting system is certified under VVSG 1.0, the first version. In early 2021, a new version was adopted, that’s VVSG 2.0.
But for various reasons, it has taken a while for manufacturers to get new systems all the way to certification. Right now, just three companies have VVSG 2.0-certified systems – one of those was certified the day before the last. Two other systems are in testing, and a sixth is awaiting approval of a test plan.
We think that overall, VVSG 2.0 is a real step forward. But we’re not convinced that lacking VVSG 2.0 certification at this time should disqualify a voting system. In particular, VVSG 2.0 requires voting systems to provide capabilities whose use is optional. So some manufacturers may have spent extra time to implement features that Georgia will never use. You should consider whether Georgia-specific standards and robust testing could stand in for VVSG 2.0 certification at the time of purchase.
In my remaining time, I want to briefly touch on a few points about hand-marking, ballot on demand, and usability testing.
First, we’re aware of concerns that some voters will mismark their ballots, possibly leading to polling place disruptions or lost votes. Improvements in hardware, software, and even ballot design over the last 20 years have helped prevent or mitigate these errors, but voters do make mistakes. To get some sense of how voter error might play out in Georgia, I examined about 25,000 hand-marked absentee ballots from 2024. I found that in about one out of every 250 ballots, voters were deemed to have overvoted, usually by voting for two candidates in a vote-for-one contest. Some overvotes were caused by stray marks, and sometimes voters just seemed to misunderstand the rules. If an in-person voter overvotes, the scanner will return their ballot with a warning. This gives voters a chance to correct their ballots and to learn from their mistake. Thoughtful voter education and patient poll workers can go far to manage the transition to hand-marking. Hand-marking as the primary system has worked well around the country, and it will work well here. You just need to prepare for it.
Next, ballot on demand printing. As you know, ballot on demand printing equips an advance voting site to print the correct ballot for a voter who lives anywhere in that county, rather than having to preprint many kinds of ballots. It’s a great technology. I have used it to vote early in several elections, always without a hitch. But like any technology, problems can arise, especially if the roll-out is rushed. Let me share a cautionary tale that ends in a major success story.
As you may know, Maricopa County, Arizona – which is the fourth largest county in the nation – ran into trouble with its ballot on demand printers at some election day vote centers in November 2022. The printers worked fine in the primary election, but in November, Maricopa used longer ballots and heavier paper, and some printers started to fail. The ballots still printed, but on some of them, parts of the timing marks that identify the ballot style flaked off, and the tabulators often rejected those defective ballots. So voters had to cast them in secure compartments to be tallied later. Of course this problem caused confusion, delays, and long lines at some voter centers.
Maricopa was able to resolve this problem by phasing out the brand of printers that had proven unreliable and buying more of another brand that performed well and at a large scale. In the November 2024 election, we estimate that about 470,000 Maricopa County voters used ballot on demand, either during in-person early voting or on election day. That would rival or surpass any Georgia county. And by the way, we believe that almost 600,000 early voters in Miami-Dade used ballot on demand in the same election.
Let me add that if Georgia counties plan to use ballot-on-demand on election day, they need a robust plan to provide emergency ballots. Paul Lux addressed that topic at the last meeting. As with any other technology, you don’t want to count on it working when you need it most.
Finally, an ideal procurement process should include extensive usability testing, including but not limited to voters with various disabilities and various language needs, and covering every part of voting. That includes verification of BMD ballots, preferably independent of the voting system. In that connection, summary ballots tallied by optical character recognition may be preferable to BMD ballots that resemble hand-marked paper ballots, sometimes called bubble ballots. That’s because some voters may have personal assistive devices that can read the text on summary ballots, but are not able to read bubbles.
I will stop there and thank you for your attention to me and all the presenters.