Slashdot Mirror
Touch Screen Voting Industry Circling Wagons
bhoman writes "Salon has an interesting article/interview with the author of a forthcoming book, Black Box Voting, by Bev Harris, that looks at electronic voting machines, especially Diebold touchscreens. The story includes incriminating internal memos, cease and desist orders from Diebold, transcripts of an industry teleconference where Harris Miller of the ITAA brags of his lobbying experience, and documentation of a backdoor via an Access MDB with no password. This is for software currently being used in 37 states. "
An open invitation to election fraud
The U.S. government seems to me to be becoming more and more corrupt. As David Letterman recently said, "When you make out your check for the Iraq war, there are two Ls in Halliburton."
Money seems to be everything, the health of the country nothing. McCain is right, we need campaign finance reform.
Every software in government, which is paid for from citizens taxes, should be open source. So that every citizen (at least the one which is a programmer) could check whether the code is good and fair, especially in elections.
Of course the code actually used in voting machines should be double checked by government professionals, but everyone should have an access to read the code.
I love high tech as much as anyone on Slashdot, but paper ballots make a whole lot more sense: with even a modicum of security you have the originals for recount (recounts being actually pretty straightfoward Florida FUD not withstanding).
"Everyone is entitled to their own opinion, but not their own facts."
... for anything important such as voting. I'm a programmer, I do that for a living I've *never* seen a software project that didn't include quick hacks, known vulnerabilities by the dev team, ,a lazy programmer and a PHB.
The fact the matter is, EVERY software project has stuff like that.
I wouldn't trust a software (much less a closed source software) written by anyone (including NASA, govs, whatever) to do anything like this. And personally, I can't believe anyone who has worked in the industry would.
And that is, regardless of the project management techniques, reviews, whatever.
IP Therefore I am.
Predictably, a bunch of /. responses focus on the fact that the source isn't available for public review as the primary problem, but that's irrelevant, and Bev Harris explained the correct solution quite clearly in the article.
Open source wouldn't be a bad thing, mind you, but why bother auditing the code? What you really want is to audit the *results*, and the easiest, best solution to that is also the simplest: Have the touch screen machines print paper ballots with a nice list of races and selected candidates. Then the voter can verify that they actually voted the way they wanted to, and the paper ballots can be counted and compared with the computerized tallies by anyone who wants to question the system.
As Harris points out, the fact that the manufacturers sem so dead-set on avoiding paper printing seems almost sinister... the solution is so obvious, and so simple that it makes you wonder what their true motivations are. They make a lot of noise about printers being too error-prone and difficult to operate, but that's just silly. Take a look at the thermal printers used by retail systems -- they work day in and day out for years with no more maintenance than replacing rolls of paper. Designing a workable printer for a voting booth wouldn't be trivial, but neither would it be an impossibility. The requirements are very simple: Be able to run for an entire day without jamming or running out of consumables, and print paper ballots that are easy to read and remain clear and legible for at least three years.
There are various minor improvements that can be made to this idea, such as a machine-readable section of the ballot to make automated verification easier, etc., but at bottom paper achieves a level of transparency and reliability that no purely automated system can ever achieve, no matter how many geeks have pored over the code.
Note to ACs: I usually delete AC replies without reading them. If you want to talk to me, log in.
Not the whole answer, at least.
We need to check, not only that the software has no obvious backdoors, but that
I'm not that paranoid; there are probably any number of other things that could be screwed with and still have the code pass any kind of review with flying colors.
Paper ballots are the only answer.
Welcome to the Turing Tarpit, where everything is possible but nothing interesting is easy.
You will want to ensure that the machine accurately registers and tallies votes. Verifying the source alledgedly used in all the machines is not sufficient: you'd need to inspect the (sufficiently large) CRC of the binaries on each and every of the voting machines. You'll want to verify that they are indeed running the software that you have inspected, not some doctored version.
Even if all machines produce accurate data, that will do little good if anyone can edit the resulting data file, or if the totals are communicated to a central counting facility through a means which allows easy forgery of the results.
The problem with any electronic voting system is its intransparency, not of the program source, but of the voting and tallying process. Once the job of vote registration and counting is delegated to a machine, it becomes invisible. It is like handing a box of paper ballots to anyone in the streets and asking him to tally up the votes without any supervision. You'll have no idea of the accuracy of the resulting count, unless you are able to recount yourself... and for that, you need a paper trail.
I firmly believe that any electronic voting needs to be accompanied by a paper trail, and that the counts must be subject to verification of a recount using this paper trail. An electronic voting machine should either produce a paper ballot which the voter can inspect and post in a lockbox, or it should scan a paper ballot on which the voter has indicated his choice by hand. There arer very good reasons to trust paper ballots over electronic ones that are hidden inside some machine:
- The voter has tangible assurance that the vote that is deposited is the one that he has cast
- The counting rersults are verifiable: the counting can take place in a group of people from all stakeholders in the election, who will all watch each other.
- In case of doubt, a recount can take place using the original ballots counted by a different group of people.
- Most importantly: paper ballots are incredibly hard to forge in bulk, and it is very hard to introduce a significant amount of them into the counting process.
If construction was anything like programming, an incorrectly fitted lock would bring down the entire building...
It strikes me as incredible that the "technical" people writing these emails are engaged in such Mickey Mouse chatter, and so interested in just cranking out something, anything that will work. I just don't see how electronic voting is really all that hard to engage in...as long as you have your priorities straight.
There are two primary things we want to accomplish with EVotes -- first, we want to make the voting process easier to engage in. Second, we want to make the counting process more efficient (less costly). We would also like to reduce the error rate, to the extent that we are able.
A touch screen voting interface, big and clear and nice, is exactly what we need to help walk people through the process. We can't, though, rely on the software in these machines. One read through the memos above should convince you as to why -- these people just have no idea what they're doing. Basic? Access databases? Windows? My god.
What this says to me is that we simply cannot get away from paper. So what we want is a system that makes paper easier to use, leaves a paper trail for auditing and verification purposes, and provides ample opportunity for error checking by the voter and by election officials.
We use the touch screen to answer questions. At the end of the voting session, the system prints a "vote" and electronically tabulates the results. The voter verifies that his printed vote matches what's on the tabulation screen. The voter then folds his paper vote and deposits it with election officials in a good old fashioned ballot box.
We can then use the electronic tabulation to check quickly on the results -- this is quite efficient. We will also engage in a substantial amount of verification, by counting the paper votes by hand and verifying this against totals learned electronically. The paper always wins, in this system. We do not necessarily need to count all of the paper votes -- we can use random sampling.
It seems like a win in both directions, for me. Risks include unacceptable printout quality (printer wear), and insufficient random verification.