By late Tuesday night we might find that we voted for a shift in the Washington power base, with some congressional races decided by margins of fewer than 5,000 votes.
In a congressional race decided by so few votes, a bad, erased, or corrupted memory card in one electronic voting machine can change the result of an election.
And that's what's behind a dangerous trend in the American electorate: Voters are losing faith that their ballots will actually count.
When so many elections are decided by so few votes, we should feel more empowered as voters, not less.
So why, as the Pew Research Center for the People and the Press reported in October, do 12 percent of all voters and 29 percent of black voters not believe that their votes will be tallied accurately?
Think about this: more than one in 10 voters think they have thrown away their votes! That’s more than 8 million voters who don’t believe in what they are doing. Will they continue to vote as election machine debacles continue to proliferate?
The problem with our pallid national debate about voting technology has been driven by conspiracy theorists and framed as an excuse for Democrats losing elections. As a result, Republicans and independents who care a great deal about the integrity of American democracy have been driven away from the debate.
This should not be a matter of Democrats versus Republicans. The problem is that election bureaucrats — both Democrats and Republicans — put too much faith in the vendors who build and maintain the machines and too much faith in the technology itself.
So what’s actually happening here is a clash between science and bureaucracy. State and county officials want to get their contracts out and their vote tallies in as fast and as cleanly as possible. They favor the illusion of precision over the necessity of accuracy and fidelity. They don't want to hear or consider bad news.
It’s also a struggle between private contractors and the public interest. Electronic voting vendors capitalize on American faith and fascination with all things flashy and new. As a result, scientific critics of electronic voting and fans of good-old-fashioned paper get dismissed as Luddites or fear mongers.
The erosion of faith in the electoral system started in the 2000 election with the aborted Florida recount and the realization that thousands of voters had mistakenly voted for Patrick Buchanan rather than their first choice, Al Gore. That technological glitch alone cost Gore the White House.
While that was a rather low-tech problem, it unfortunately yielded an expensive and foolish high-tech solution. Congress passed a law and funded transition to electronic voting around the country. It did so without mandating simple and obvious standards to ensure that votes would be secure. The simplest of these would have been to mandate a printed paper record of each vote so that controversial or wacky results could be verified later.
In the interest of expediency, and at the expense of democracy, Congress, led as usual by the voting machine lobbyists, rushed headlong into a techno-fundamentalist fog. States and counties soon followed.
Then, in 2002, Georgia reported surprising upsets in both its gubernatorial and Senate races, in which Democratic incumbents who had led in the polls found themselves out of jobs. Coincidentally, in 2002 Georgia (Democrats, it so happens) introduced new and untested electronic voting machines that prevented manual recounts or independent oversight of the software.
In 2004, more voters found their faith in the system shaken by reports that electronic voting machine screens were not calibrated properly. Voters who thought they were voting for one candidate found the machine had recorded a vote for another.
And in one Ohio voting precinct that had just registered 800 voters, electronic machines reported an extra 3,893 votes for President George W. Bush. Officials caught that error before it made it into the final statewide tally. Bush won Ohio — and the presidency as a result — by a mere 137,000 votes.
In a more alarming, likely, and less overtly partisan event, a county in North Carolina saw so many voters on election day 2004 that the memory cards in the machines filled up and overwrote the earlier votes. The county lost more than 4,500 votes that way. With so many stories like these, it’s surprising that more Americans don’t think their votes will not count.
Despite documented problems with the machines in many places in the country, counties and states have continued to contract out to private companies what used to be the most public of functions: the counting of votes.
Meanwhile, independent computer scientists and security experts have time and time again demonstrated that hackers could easily tamper with electronic voting machines.
Last month Princeton computer science professor Ed Felten and his research team opened a machine made by the notorious Diebold company using a key so common that it is used to open file cabinets and hotel minibars. Once open, anyone could mess with the insecure memory card that records and stores the vote records. A vandal might not have a partisan agenda. But the damage to democracy would be grave nonetheless.
And time and time again, teams of researchers led by such experts as Professor Avi Rubin at Johns Hopkins University and David Dill at Stanford University have shown that many electronic machines are inherently insecure and that paper records are essential to maintaining a trustworthy system.
Rubin, in his brilliant and entertaining new book, "Brave New Ballot: The Battle to Safeguard Democracy in the Age of Electronic Voting," tells stories of his efforts to get authorities to take security and design flaws seriously.
The standard model goes like this: researchers test a system for security flaws and find them present and alarming. Then the company that makes the machine responds via press release that the researcher did not have the most up-to-date system to test. Then the researchers request the most up-to-date-system. Then the company refuses to release it for testing. The company says “trust us.” The state and local officials say “trust us.” The scientists, of course, believe in evidence and proof — not faith-based systems. But no one in power seems to. So nothing changes.
The mainstream press, with its obsession for balancing “sides” in an argument, grants equal weight and credibility to scientists who have no financial stake in the outcome of a policy decision and vendors who clearly do.
Once you clear out the bureaucratic, interested, and anti-scientific smoke from the debate, it’s actually really simple to restore faith in our voting systems.
- First, states and Congress must mandate that all voting technologies be redundant — they must have paper records that can be recounted in the event of a breakdown, a malfunction, or a very close result. Either major party can and has stolen elections through fraud. And more commonly, computers break down, electricity fails, and memory cards get corrupted. Only with auditable paper records can we determine that fraud or error messed up an election.
- Second, election hardware and software code must be made available so that security experts may test it for vulnerabilities and errors. Secret code is almost always bad code. It’s a basic programmer’s rule: you need a large and diverse set of eyeballs to catch all the glitches in a pile of code. The fewer people involved, the more errors you will protect.
- Third, we should trust paper in general. Optical-scan paper ballots are by far the best medium for signaling one’s voting preferences. They are sturdy, securable, local, distributed and easy to use.
- Fourth, we should make steps to nationalize and standardize voting methods. A national non-partisan commission should issue rules that govern voting, counting, data storage, and recounts. This patchwork system in which methods vary county-by-county and state-by-state is irrational and dangerous.
- And finally, public officials who make technology decision must have no interest in or involvement with the companies that provide voting methods, machines, and services. There are already too many questions about the propriety of partisan officials counting their own votes and ruling on important election laws (see Ohio this year). Those in charge of our elections must be above reproach and beyond conflicts of interest.
“America deserves a foolproof voting system,” Rubin writes in his new book. “Whoever designs that system must be able to prove that the system cannot be cheated and be able to explain to the average eighth grader.” The current, secret, proprietary model is not going to meet that standard.
If we do not have a rich, informed, non-partisan, publicly minded nationwide examination about how we vote and how we count our votes, we are in danger of becoming like just another shaky, storefront democracy.
So if Tuesday evening we learn that a handful of seats Republicans thought they had won go the other way amid reports of malfunctioning electronic voting machines — and there is no way to run a recount — perhaps Republicans will join the critics of electronic voting and we can move to restore trust in our balloting system.
We should not wish for chaos and controversy. But without a reason to wonder and worry we are in danger of being too complacent and letting a handful of untrustworthy companies rule our most basic public ritual: the vote.
Siva Vaidhyanathan is an associate professor of Culture and Communication at New York University. His latest book is The Anarchist in the Library: How the Clash Between Freedom and Control is Hacking the Real World and Crashing the System (Basic Books, 2004). He blogs at Sivacracy.net.