The case for lever voting machines


In 2002, a measure called the Help America Vote Act mandated that voting across the country needs to be easier and more handicap friendly, leaving the clunky lever voting machines out. However, the venerable LVMs are still used in New York, especially in New York City, and there are a crop of studies popping up lately advocating their reliability.

We spoke to Dr. Bryan Pfaffenberger, a professor of Science and Technology at the University of Virginia, and the only certified scholar ever to take such a hard look at lever voting machines. He's studied the LVM industry from its birth in 1889, to its bankruptcy in 1983. Now that New York — the last bastion of lever voting machines — is phasing the units out, we asked Dr. Pfaffenberger to weigh in with his insights on the aging machines.

1. What's so great about a century old machine?
Imagine, if you can, a modern appliance with a solid 50 year service life. There are "a lot of things that are right about this technology," Pfaffenberger told DVICE, "[lever voting machines] underscore the truly astonishing technical achievement that these machines represented in their day. The list of potential exploits is finite and known, unlike electronic machines. During the entire century, there were relatively few elections where controversy arose over disputes or subverted elections."

"A little bit of dust can muss up an optical scanner," Pfaffenberger said, "they could take 50 ballots and run them through the scanners and get a different number every single time."

Lever voting machines are all hardware. "The really remarkable thing about the lever voting machines is that they rank highly on studies — which include the most modern machines — either at the top or 2nd in voting performance measures." With that kind of track record, Pfaffenberger told us, it'd be like a steamboat operating as a viable alternative for modern boats, or the Wrights brothers' airplane trumping jets.

2. Any drawbacks to using lever voting machines?
Lever voting machines are old, and require diligent maintenance. The most talked about problem plaguing the machines is a failure of the vote-counting odometers to advance, and there was a "statistically significant prevalence of candidates receiving 99 votes" instead of 100. Still, Pfaffenberger maintains that this problem is "a bit separate from the [LVM] technology as if you have a poorly run district there's no technology you can actually stick in there that'll run correctly." In the right hands, there's very little reason why a lever voting machine should fail. Nowadays, it's "astonishingly rare" for attendants to open up a unit and find a 99 count.

There has also been a big push lately for a paper trail to accompany all voting methods — something which lever voting machines historically lacked. "From a contemporary standpoint, the fundamental technological flaw is the lack of an independent audit trail." The AVM Printomatic features a paper spool, though the "point of the Printomatic is not to just print out the sheet at the end — but to print out an authoritative document to prove that all of the counters were set to 0 before the election." If these counters aren't zeroed out, a candidate could start with more or less votes from the get-go.

Despite this, the key is the "finite number of failure points." Electronic voting machines are still evolving, and so are the ways to exploit the technology. Worse than that, DREs are susceptible to mass tampering, while each lever voting machine would have to be tinkered with individually.

3. In an ideal world, would we still be using lever voting machines?
The short answer? No. "We talk about it and we daydream about it," Pfaffenberger mused, but he knows it'd be a lost cause to champion machines already on the way out. "One of the biggest mistakes we've made with our technology today is we're doing it on the cheap." And LVMs aren't cheap, or portable, or light — but they were proven to be reliable, and buyers could keep them for half a century. Counties, cities and states still buy voting machines infrequently, but modern machines are often valued for portability and affordability, not reliability.

There is a way for us to use computerized systems and make them secure. The government, the military and even corporations rely on solid machinery, but there's something all three have in common: money. If we want voting technology to be as solid as it was in the heydays of the lever voting machine, then it needs to be made the same priority, and approached with the same ingenuity.

For Pfaffenberger, lever voting machines mark "an important chapter in American history and from one which I think we can learn."