Unintended consequences could undermine Georgia voting

One of the risks of policymaking is the pesky law of unintended consequences, which simply states that actions—legislative or otherwise—often come with unanticipated effects. International efforts to ban use of the pesticide DDT has almost certainly led to many more deaths, and U.S. government meddling to protect the sugar industry forced sugar-related companies to reduce their domestic workforce by about 10,000 people.

These are just two examples of the law in action, and recent events have proven that Georgia is not immune to this law. In 2024, legislators enacted a measure they hoped would increase voter confidence, but it was poorly conceived. If lawmakers don’t address this next session, it could change elections as we know it—making them more expensive, delaying their conclusion and eroding voter confidence in the process.

The bill in question is Senate Bill 189. It was a 24-page omnibus elections proposal, and tucked within it are four lines that should make Georgians shudder: “The official tabulation count of any ballot scanner shall be based upon the text portion or the machine mark, provided that such mark clearly denotes the elector’s selection and does not use a QR code, bar code, or similar coding, of such ballots and not any machine coding that may be printed on such ballots.”

This is just legalese for votes cannot be counted using computer codes on ballots, like the ones we have used for years; only plain readable text can be used to count them. The justification for this change is that there are some elements in Georgia who do not trust the QR codes used in Georgia’s Dominion voting system. They worry—without a shred of evidence—that the codes can be corrupted and change their selections. Without scanning plain text, they would have no idea if their votes had been stolen, or so they say.

This is just a continuation of the unfounded conspiracy theories stemming from the 2020 election, which audit-after-audit and lawsuit-after-lawsuit demonstrated to be legitimate. Never mind that for now. If lawmakers want to replace the code with plain text, that is fine by me. The technology already exists to do so and has been studied in Georgia, but the General Assembly failed to provide a legal and timely path to implementation.

As my friends at the Eternal Vigilance Institute wrote, “Georgia’s 2024 statewide optical character recognition (OCR) audit pilot demonstrated technical feasibility but also revealed that full statewide adoption requires new software certification, county training, and a funded transition plan.” The law goes into effect July 1, 2026—sandwiched between critical statewide primaries and general elections.

So what happens if legislators do not tweak the law and the feds refuse to get involved? Then there is only one legal method of counting votes: Do it by hand. For those with a baseless distrust of Dominion voting machines, this might not seem like much of an issue, but for everyone else, it is hugely problematic. Just look at what other locales have found.

Nye County, Nev., decided to give hand-counted ballots a try. It took weeks to count the ballots, and after the first day of counting, the county clerk estimated an astounding 25% error rate. Shasta County, Calif., also resolved to look into the matter. In a feasibility study, elections officials there estimated, “To perform a full manual tally in a Presidential Election with 50 contests, the Elections Department would have to add at least 1,255 additional temporary staff members to perform a full manual tally, at a cost of $1,651,209.68.” This is roughly three times the cost to simply use their Dominion voting machines.

Georgia is of course not Nye or Shasta counties, but hand-counting would certainly be much more expensive and conceivably harder to implement. Based on Shasta County’s numbers and all things being equal, very rough back-of-envelope math suggests Georgia might need to allocate an additional $110 million and 85,000 temporary staff members to hand-count ballots. If Georgia follows this path, it would still take weeks to determine election winners in close elections, and the error rate could mar the races, which would irreparably harm voter confidence.

This is not tenable for Georgia. Perhaps the easiest path forward is for the General Assembly to pass legislation delaying the implementation of the problematic provision until the software is certified, workers are trained and the endeavor is funded. Failure to act, on the other hand, would either trigger the aforementioned hand-count or the feds would take over our elections, which could be equally bad.

The federal government can’t even pass a budget. So their ability to effectively administer our elections is questionable at best. In whatever direction lawmakers go, they need to keep the law of unintended consequences top-of-mind.