A pair of articles in today’s New York Times (“Vote Fraud Theories, Spread by Blogs, Are Quickly Buried,” by Tom Zeller Jr., and “Mostly Good Reviews for Electronic Voting,” by John Schwartz) do a good job of addressing some of the increasingly widespread concerns about apparent inconsistencies in last week’s election, but the conclusion that they come to is altogether wrong.
Based on the coverage that this issue has received in the past 48 hours — notably on NBC Nightly News and NPR’s All Things Considered — one would think that all is plainly well in the world of electronic voting, and that the election results are known to be valid.
If I may set aside the recent presidential race and my vote preference, I believe it important to emphasize that we have no way of knowing if the results were valid. It is impossible to verify.
It used to be — just a few years ago — that it was easy to verify the validity of an election. Every precinct is staffed by local folks who, in exchange for $100 or so, spend the whole of the day running the election. In the morning, before the voting begins, they open up the various voting devices, such ensuring that pulling the “Smith” lever increases the “Smith” counter. Representatives from each party are able to perform the same inspection, with even the mechanics of the more complex devices being sufficiently clear that a malfunction would be immediately obvious. Precincts that conduct their votes on paper — such as punch ballots or a simple “X” in a box — require even less inspection. Throughout the day, election monitors needed only remain in the room for it to be obvious that no changes were being made to the machines. At the end, the staff and the monitors could monitor the vote count by simply looking at the ballots as they’re being counted, or looking at the tally on the machine, having done a post-election re-inspection to ensure that the mechanics were still operating properly. In case of any controversy, a recount could be done, which meant, simply, that everybody would start all over again, carefully counting out the votes, one at a time.
No more. It’s impossible to inspect Diebold or ES&S’s computers. How would you do that? How would a volunteer open up the software’s source code? It would take an experienced programmer dozens or hundreds of hours to perform that inspection. Not that it’s possible — it’s a violation of copyright law (the Digital Millennium Copyright Act) to reverse-engineer and inspect that source code. Instead of an inspectable machine, what we have is a private corporation’s black box — we just have to trust it. During the day, it’s impossible to know if it is being tampered with. As I write this blog entry, I am sitting in a classroom, apparently typing out notes when, in fact, I am wirelessly connected to the building’s LAN, which connects to the school-wide network, then out to a network connection in Richmond, then to Waynesboro, then down to Blacksburg, and then to a small, grey box sitting on my refrigerator, the server where this data now resides. There is no reason why the computers tallying the votes could not be similarly, invisibly altered during election day. At the end of the day, how is the vote to be counted? A button is pressed and a number is provided. How do you know it’s right? Trust it. How is a recount done? Hit the button again. (Seriously.)
If the election were “stolen,” or if the results were simply disastrously wrong through an error in the software, how would we know? We wouldn’t, of course. There’s no inspection process. Nobody at any level of the government or citizenry has any method whatsoever of verifying anything that goes on in that black box.
I don’t trust it, and I won’t trust it, until there’s a voter-verified paper receipt and the software is open source. So long as private corporations run our elections in complete secrecy, I will never be comfortable with the results of any election, and I can’t imagine why anybody else would be, either.