Dem
51
GOP
49
In 2019, U.S. District Court Judge Amy Totenberg (kid sister of NPR reporter Nina Totenberg) ruled that the Georgia voting machines, which didn't have a paper trail, were untrustworthy and couldn't be used anymore. This forced the state to scrap them all and spend millions of dollars to buy new ones. This wasn't as lucrative for Dominion Voting Systems as suing Fox News, but they did manage to sell Georgia thousands of machines.
As usual, someone sued and the case got handed to Totenberg. The new machines are much better than the old ones because they print a piece of paper with the voter's choices on it. They are then scanned and saved for a possible hand recount if needed. This is far better than the old machines, which were hackable and didn't have a paper trail.
So what's the problem? The new machines print bar codes on them and it is the bar codes that are actually scanned. There isn't any way for the voter to know that the information encoded in the bar codes corresponds to how they voted. The manufacturer (or a hacker) could potentially take a ballot in which the voter voted a straight Democratic ticket but encode a straight Republican ticket in the bar codes. The people suing don't like the bar codes at all. They want printed ballots with no bar code so the scanning machines have to scan the human-readable votes. Even better would be paper ballots marked by hand and hand counted, with no machines involved at all. However, previous elections have shown that when voters mark their ballots by hand, they sometimes make a real mess of it.
The situation is not terrible because it can be tested. After an election, officials could randomly select, say, 200 ballots and run them through the scanner and see the results. Then they could manually count all 200 ballots. If the results agreed, then even if the voting machines are cheating (unlikely) they are probably not doing it more than 0.5% of the time. If the officials made the test with 1,000 random ballots and no errors turned up, then the machines are probably not changing even 0.1% of the ballots.
The problem is that even if Totenberg believes the people who don't like the bar codes, it would force Georgia to put out a tender for new machines, make a choice, buy them, test them, and train workers on them this close to an election. Even if the state could pull it off, it will surely undermine voter confidence in the election. Just imagine the headlines the day after a decision to dump the machines. They will probably be something like: "Judge doesn't trust the voting machines; orders them thrown out." How will that affect voter confidence in the election? What would Donald Trump say?
One compromise ruling the judge could make is to allow the machines but state that the electronic total the scanners give is just preliminary. It could be given to the media on Election Night with the footnote that the count is not official. The next day, the state would have to hand count every ballot. That would be the official result. If the difference with the mechanical count exceeded a certain threshold, there would have to be a second count, this time with observers from each party examining each ballot. If the second hand count agreed with the first hand count but disagreed with the machine count, then the state would have to instruct Dominion to change the software to stop printing bar codes altogether and requiring all future elections to hand-count the ballots.
The ironic thing is that hand counts also have errors. One way to minimize that is to have the hand counters put all the Trump ballots in one pile as they count and the Biden ballots in another pile. Then Republican observers could be allowed to inspect the Biden pile and pull out any ballots they believe are actually for Trump (or invalid) and vice versa. Then a committee of partisan observers and election officials would rule on every challenged ballot.
Of course, there are races other than for president on the ballot. If a race for a state Senate seat was close, then the above process could be repeated for that Senate district. The law could even be changed saying that for any race where the winner's margin was under 1%, this method would be applied with the state paying for it. But if any candidate who lost by more than 1% wanted the "separate pile method" used, the candidate would have to pay for it. Something like this would probably be workable. (V)