My comments follow – I think he’s wide of the mark.
Montana’s Paper Ballot System Ensures Election Day Accuracy
By: Brad Johnson, Montana Secretary of State
Montana has the cleanest, most secure elections anywhere in America. Our paper ballots give us the most reliable backup in a worst-case scenario. Our procedures prevent unauthorized tampering with ballots or equipment. Our personnel are dedicated and competent. And the counting machines we use have adequate security.
However, a recent study commissioned by the Ohio secretary of state’s office found security flaws in a large number of election machines, including three used by counties here in Montana. But none of those machines is used to actually record an individual’s vote. None of those machines is the only means of knowing what the vote count is. And none of those machines has ever failed a test for us.
Montana does not use machines to record votes. We use paper ballots as the gold standard of ballot security. Machines are only used to count those ballots.
What does that mean? It’s simple: No matter what goes wrong with the machine, come hell or high water, we will have a reliable record of every single vote cast.
The machines mentioned in the Ohio study used in Montana are the Automark ballot marking machine and the M-100 and M-650 ballot counting machines. The League of Women Voters has called for those machines to be decertified by the secretary of state’s office. In other words, those machines would no longer be authorized for use in elections here. However, because of the measures in place in Montana elections, we are not convinced this is necessary.
Each machine first meets or exceeds federal standards before ever being purchased. Shortly before the election, a number of each machine to be used, chosen at random, is tested with sample ballots. This process is open to observation. Any attempt to tamper with the machines would be detected before the count.
On Election Day, election judges again test a random sample of every machine to be used, employing a stack of sample ballots. Again, the process is open to public view. Once again, any tampering with a vote counting machine would be detected.
In the event of a recount, state law requires that results be tabulated by hand. So any attempt to tamper with the results of an election would be caught by a hand count.
Machine counts have never, not once, been shown to be inaccurate by a hand count. Even the League of Women Voters admits that there is no proof or even strong suspicion of machine counts tampering.
The Ohio study found, and the League of Women Voters is concerned about, ways in which these machines could be compromised if someone gained physical access to the machine. But Montana’s county election officials take steps to prevent that. At some point, we have to rely on people. Trained, competent people take care of and supervise the machines and keep them secure.
Montana’s election systems work. They deliver clean, accurate, fair results in every election, every year. That’s because the local election officials in all of our counties are dedicated, skilled people who are committed to the democratic process. It’s also because the Legislature requires paper ballots.
The League of Women Voters is to be commended for their dedication to clean, accurate elections. But our election systems have been tested by years of use. Tampering with them on the eve of a presidential election isn’t a good solution.
——–
That’s a good piece by Johnson, and Montana’s elections to this time have been secure, so far as I can tell. Exit polls of our results have been dead-on, and recounts have tended to reinforce official counts. Furthermore, Montana is an out-of-the-way place – presidential candidates hardly care about our three electoral votes, and our coming federal elections will probably be lopsided. The Gubernatorial race is another matter.
Still, Johnson’s piece is not totally reassuring. The prime faults with our system are three: 1) we have no control over the software that counts the votes; 2) we don’t do random audits of election results to assure accuracy, and 3) exit polls, now done by private concerns, are kept private. (Also of concern, Johnson’s statement that “trained, competent people take care of and supervise the machines and keep them secure.” More about that later.)
The scanning equipment we use to count the votes is a simple operation – the machines read and tally the pencil marks on ballots left by voters. I’m no programmer or computer geek, but I did do some BASIC programming in college, and swear I could devise a program to do such a simple operation. We’re basically looking at 1980’s technology here. The idea that this is complicated, secretive work and that patents are at risk and fortunes are at stake is absurd. The software code ought to be open for public inspection (which, in reality, translates into computer geeks poring over it looking for holes.)
That it is not, that vote-counting software is proprietary, is a gaping hole in our election security system. That Johnson places his faith in the pre-election tests is symptomatic of the credulity of election officials in blindly trusting the companies that make the equipment to behave well. Programmers could easily write code to override the results, to kick in at certain times and then self-eradicate. Such programming, wired into the memory cards that the machines depend on, could easily get past Johnson’s election day pre-tests.
And indeed, in New Hampshire, where a recount is underway, memory cards have mysteriously disappeared.
Second is the idea of audits. OK – that’s my background. I’m an accountant, and accounting, like dentistry, is best taken with a dose of nitrous oxide. But here’s the scoop: Don’t trust. Audit. Do so randomly. When a random audit turns up a problem, turn up the heat. Go in deeper, learn more.
Montana’s precincts all ought to know that there is a chance that their outcome, no matter how lopsided, might be recounted by hand. These “trained, competent people” are, after all, just people, susceptible to error and working low-paid, low-reward jobs. Furthermore, they are not generally sophisticated about how counting machines can be undermined, so their oversight is not necessarily reassuring. And local election supervisors are usually invested in their own work, and usually the last to suspect anything is wrong, the first to vigorously defend the apparent outcome.
Any company of any competence relies on oversight, and random auditing is a critical feature of any control system, The fact that we don’t have it in our elections is absurd.
Exit polling, so reliable until the arrival of the Bush family, is nothing more than an audit tool. It ought to be financed by taxpayers, as we are the ultimate auditors of our own results. Other countries, like Germany, do exit polling publicly and as a matter of course. In our country, it is done privately, the the results are kept secret. We know that the exit polls in New Hampshire don’t support Hillary’s surprise win. Why? Two reporters slipped up and told us so. Exit polling ought to be, like random audits, part of the election verification system. That it isn’t is, again, absurd.
And finally, I ask and don’t know the answer, who are these “trained, competent people” that Johnson talks about? Are they public employees? One would hope so – one of the problems that states like Ohio and Florida and New Hampshire have had is that supervision of elections is done by private companies. Employees of these companies have enjoyed free access to the counting machines right up until the election – they have been free to insert last-minute patches and replace memory cards. It’s a travesty.
A secure election system will rely on paper ballots as the ultimate source of the vote count. Because of sheer volume, we must use scanning equipment to count the votes. The software used on these machines and their memory cards must be open to public inspection, and the machines themselves kept under lock and key with recorded access only by publicly supervised and authorized officials until election day. Pre-election tests should be run by government employees before the public uses the machines. After the elections, the state should conduct random audits of various results as a control check. The state might as well, while they are at it, conduct exit polls – a further audit check on reliability of results.
Montana does some of these things, but not all of these things. Until Montana does all of these things, our elections are not secure. Johnson has done what would be expected of a man in his position – he attempts to assure us of the integrity of the system. He fails.
PS: It ain’t so bad here as in South Carolina, where a primary will be conducted this week and next week. There they have no paper trail to verify results. None. They are entirely dependent on the machines. It is high comedy.
Piece of Mindful 