GEORGIA — (Analysis) The recent special election to fill the Georgia Sixth Congressional District seat formerly held by Republican Tom Price, who was appointed to Donald Trump’s cabinet as Secretary of Health and Human Services, was the focus of extraordinary attention and expenditure.
More than any of the other 2017 special elections, the GA-6 election was seen as a proxy for approval or disapproval of the Trump presidency and as a clue to the Democratic prospects for retaking the House in 2018. Held in a district that had long been solidly Republican but that had given Trump the barest 1.5-percent plurality in 2016, this election was also the subject of intense media focus.
The Democratic candidate, 30-year-old Jon Ossoff, a former Congressional staffer and first-time office-seeker, faced a crowded field of 17 other candidates in a preliminary contest held in April. Among them was Karen Handel, former Republican Secretary of State of Georgia, along with a host of less serious challengers. If no candidate polled at 50 percent of the total vote, the two top finishers would meet in a June runoff.
With the wave of Trump disapproval continuing to grow, Ossoff was closing in on the 50-percent mark going into the April 18 election. On election night, as the returns were coming in, Ossoff held over 50 percent until a supposed “glitch” in Fulton County (the three counties in the Atlanta suburbs that comprise GA-6 are Fulton, Cobb, and DeKalb) paused the returns for several hours. When reporting resumed, Ossoff’s total had dropped below 50 percent, where it remained through the final count. Ossoff finished at 48.12 percent; Handel finished second with just over 19 percent of the vote and went through to the June runoff against Ossoff.
A “glitch;” a sudden shift from “win seat” to “runoff;” the fact that all but the mailed-in and “provisional” ballots were cast and counted on electronic voting machines with no paper record and no capacity for recounting, auditing or verification; and the extraordinary security breaches uncovered at the Kennesaw State University Election Center, the outfit entrusted with the programming of the computers and the management of voter databases – each of these factors raised red flags about what was reported as a “disappointing” Democratic result, as Ossoff fell 1.9-percent short of the magic 50-percent number.
The stage was then set for the June 20 runoff. The perceived significance of this election was mirrored in the funds that poured in for both sides – more than $50 million, an all-time record for a congressional seat. The tracking polls predicted a close election but averaged to an advantage for Ossoff, in spite of the fact that all were conducted using the Likely Voter Cutoff Model for sampling, a methodology that is generally recognized to disproportionately eliminate Democratic-leaning constituencies such as renters, students and poorer voters from the sample, thereby advantaging Republican candidates in the poll results.
With the exception of a single poll conducted by a polling firm identified with an “R” (that is, as working for Republican clients), Handel never held a lead in the polls going into the runoff election. Nonetheless, on the basis of my own experience observing and handicapping U.S. elections in the computerized voting era, I publicly predicted (speaking at a conference on June 2) with complete confidence that Ossoff would lose to Handel. Indeed, I promised that it was such a lock that, should Ossoff win, I would cease all election integrity activities and concede that I was nothing more than the wild-eyed, tinfoil hat “conspiracy theorist” that we are all so often accused of being. Fortunately for my career, Handel came through with flying colors and won by 3.8 percent, finishing with 52.87 percent to Ossoff’s 48.13 percent.
Kellyanne Conway summed up the reaction among leading Republicans when she tweeted “Laughing my #Ossoff.”
Laughing my #Ossoff
— Kellyanne Conway (@KellyannePolls) June 21, 2017
The Democrats, who have gone zero for five in special elections and are seemingly unable to win anything in spite of Trump’s lead-balloon popularity, started wailing about new strategies and new leadership. The bounteous and bitter fruits of apparent victory and apparent defeat, having had a profound effect upon political expectations and strategies, and indeed upon all aspects of political behavior, going forward.
Prior to the election, legal action to compel that votes be cast on paper (and counted by optical scanner) to provide a durable record for verification purposes failed when a judge ruled that it would be too burdensome on the state to print ballots for GA-6 and to use its existing optical scanners (“OpScans,” which were being used to count mail-in ballots) to count election day ballots. As a result, only mail-in ballots and provisional ballots – approximately 10 percent of the total votes cast – were cast on paper and in any way verifiable. The remaining 90 percent? For that, we’d just have to trust the Kennesaw State Election Center, its director Merle King, and their already-breached security protocols.
It is worthy of note that this was a single-contest election that could have easily been counted observably, in public, by hand, within two hours of the polls closing at minimal expense (though plenty of volunteers would have likely poured in). The Dutch, having taken one whiff of our 2016 elections and being aware of the security risks in computerized counting, changed their protocol after two days of consideration and counted their 2017 election by hand, joining a growing list of other advanced democracies in doing so.
Verifiable vs. unverifiable counting: an enormous disparity
The Georgia Secretary of State elections website helpfully breaks down vote totals by type of ballot cast. There are four types of voting: election day in-person voting, early in-person voting, vote-by-mail and provisional ballots. The first two are cast and counted on computerized voting machines, which permit no meaningful verification, either by audit or recount. Mail-in and provisional ballots, on the other hand, are cast on paper and counted on OpScans, the paper then being retained, by federal law, for 22 months – which would, at least in theory, permit verification processes to be undertaken. The results for each type of voting are shown in the table below:
We can see that, after winning the verifiable vote-by-mail voting by a stunning 28.4-percent margin (and the provisional voting by an even more lopsided margin of 46 percent), Ossoff also polled a narrow win in the unverifiable advance-in-person voting, only to be blown out by 16.4 percent in an unverifiable election day landslide. With the election already under a cloud of known security breaches at Kennesaw State, the larger cloud of known vulnerability to hacking and rigging of unverifiable electronic voting machine records and the still larger cloud of more than 15 years of virtually unidirectional vote-counting anomalies and red flags in the computerized voting era, this particularly glaring disparity warrants deeper investigation.
To be clear, Handel’s landslide victory in Election Day voting was absolutely shocking. It was not remotely predicted by a single poll, not even the Republican-identified by Trafalgar referred to above, which was the only poll to show Handel ahead (by 2 percent). Even this outlier poll showed Handel with a 1.6-percent lead among likely voters who had yet to cast their votes a week before election day. This leads to two obvious questions: What happened to swing voters so strongly for Handel or against Ossoff? And did the big vote-by-mail Ossoff margin simply reflect that Democratic voters in GA-6 are more likely to cast mail-in ballots?
The answer to the first question is fairly clear: there was no gaffe or scandal in the week before election day. Ossoff did not get crushed in a debate, urinate in public on a statue of George Washington or get caught in bed with a farm animal. Handel did not give a speech for the ages. The money and endorsements pouring in from both sides had already done their work – hardly anyone (3.88 percent, according to the Trafalgar poll) remained “undecided.”
The Ossoff get-out-the-vote operation, which helped Ossoff to his 6.6-percent overall lead in early voting, did not run out of money, and there was no shortage of volunteers. The only incident of note was the “baseball practice” shooting in Virginia, in which a Republican Congressman, as well as several others, were seriously wounded a week before the GA-6 election.
Although it was Republicans who were targeted, the response to the incident was universal bipartisan condemnation – and there was a clear sense that the perpetrator was a disturbed individual and a symptom of a hyperpolarization and breakdown in norms of civility and decency in which Trump’s inflammatory rhetoric is regarded by many, including Republicans, to have played a key role. There was a strong sense, too, that in this environment, either side might be the target of such violence. There was no sign that election day voting in GA-6 was swung so sharply by either this incident or any other late-breaking news event.
The second question requires a bit more digging. If it turned out that GA-6 Democrats had displayed a historical tendency to mail in their ballots, that would suffice to establish a benign explanation for the verifiable/unverifiable disparity – though it would hardly serve to remove the clouds of a priori suspicion. It was a simple enough exercise to download and organize the archived data for the past several GA-6 elections from the Georgia elections website. As shown in the chart below, it is not Democrats but Republicans who consistently prefer to vote by mail in GA-6.
That is, until 2017. Suddenly – in the two Ossoff elections, preliminary and runoff – that pattern spins on its heels and it is Democrats (or, more precisely, voters who selected the Democratic candidate; analysts are attempting to gather the data necessary to determine how much of the Ossoff mail-in vote was cross-over by Republicans, a determination of great forensic significance) who flocked to the mailbox to vote.
Or did they? What if the Ossoff mail-in vote advantage reflected not simply a flood of Democratic voters suddenly deciding to vote by mail, but instead the verifiability of those paper ballots and their consequent relative immunity to risk-free manipulation? What if the mail-in votes as cast were not so wildly divergent from the in-person votes as cast? What if the unverifiable in-person votes were manipulated, when they needed to be, with a big Ossoff lead to overcome on election day? What if one of the numerous known security breaches was exploited to alter the result of the election?
Real possibility of fraud raises real concerns for democracy
If these questions seem far-fetched, you owe it to yourself (and to democracy) to ask the “opposite” question: What proof exists that the 90 percent of the vote count conducted on unverifiable and manifestly vulnerable electronic voting machines was not hacked or maliciously programmed, altered in the pitch-dark of cyberspace? Carefully consider.
You might ask Kennesaw State Election Center director Merle King or Georgia Secretary of State Brian Kemp, or any of the good stewards who conduct and control elections in GA-6 and throughout the state for that proof. The voters are owed either a cogent answer or a straightforward admission that such elections are strictly faith-based.
Some were led to speculate that the surge of Democratic mail-ins in the June runoff could have been prompted in part by the Ossoff campaign’s encouraging of vote-by-mail as a protection against electronic voting machine-based fraud. Having combed the campaign literature for that specific message, what we have found so far is that the Handel campaign inserted vote-by-mail ballot applications into at least one of its mailings, while the Ossoff campaign did not.
Such “benign” explanations, where remotely plausible, certainly warrant serious investigation. But so do the hardware and software that recorded and counted 90 percent of the GA-6 vote in endless and invisible strings of 1s and 0s. Right now, pending further investigation (which, in spite of a multi-partisan election contest recently filed in Georgia state court, is as unlikely as it has been in all of the suspect elections of the computerized voting era), we are basing national leadership, policy, and direction – all of it – on those 1s and 0s and on blind trust in the Merle Kings, Kennesaw States, Command Centrals, Triads, Diebolds, ES&Ss, Hart Intercivics, and Dominions of the world.
In what way, we must ask ourselves as the true stewards of our democracy, have they earned that trust? As it stands, our planet is riding on something none of us is permitted to see, and therefore on the voluntary ethical self-restraint of actors with every motivation in the world not to exercise it.
The solution is a very basic one: restoration of public, observable vote-counting to American elections. There is no sign that this is going to be given to us by our elected office-holders – those for whom the electoral process has, in one way or another, “worked.” We are going to have to insist, and apply the full leverage of our economic muscle (beginning with mass refusal to purchase unnecessary items) to that insistence.
We as a public are free to make bad electoral choices – that’s democracy – but the minute those choices are made for us, in our name, and falsely – that is something else. It could well be fascism via fraud. It is, at the very least, a lie. And it is up to us to reclaim our true and undubbed voice.
Jonathan Simon is Executive Director of Election Defense Alliance, a nonprofit organization dedicated to restoring observable vote counting and electoral integrity. He’s also the author of “CODE RED: Computerized Election Theft and The New American Century.” His related blog can be found at www.CodeRed2016.com/blog.