Who Is This?

Frank, a forty-something software engineer in Southern California.
Weblogs Of Special Merit
Reciprocal Links
Referrers
Powered by
Movable Type 3.34

October 15, 2003

Government

How to destroy a democracy, part 2.

Today's Salon has an interview ("Bad grades for a voting-machine exam") with a 28-year-old programmer named Jeremiah Akin, who observed a "test" of the Sequoia touch-screen voting machines used in Riverside County, California. He was not impressed, to put it mildly. His conclusion was that

"statements made by the Registrar of Voters indicated to me that she is not qualified to assess the reliability and security of such systems, and that she misunderstands some essentials of computer programming and operation. Her deputies refused to answer some important questions. Some statements made by officials at the Registrar's office, and found on the contractor's Web site, I learned on the test day were misleading or inaccurate. Further research after the test day has turned up several other reasons to doubt the reliability, security and accuracy of the system."

From his description of the so-called "test," it wasn't a test at all. The observers were unable to use the touchscreen or even to see its contents, the "test" was done entirely from a cartridge inserted into the rear of the device and it appears that no one hung around to see the "test" complete. The only evidence that anything like what should happen in an election did happen was an assertion by Mischelle Townsend, the Riverside County Registrar of Voters. In addition, the software that was running the devices was apparently running in "pre election mode," a test mode of some kind, and was not operating as it would in a real election.

I suppose that to someone not knowledgable in the field of computer science, programming or software engineering, this might seem sufficient. It's not. A real test would have had the software operating in full election mode, the testers would have been able to actually "vote" using the devices and they would have been able to verify that their vote was processed properly. That is how software testing works: You run it as if it were running for real and see how it works. If it fails, you fix it. In this case, this being a supposed verification before an actual election, if it failed, the devices should have been disqualified for use in the election. Unfortunately, the devices neither failed nor succeeded. They were not verified at all; whether or not how they performed in their "pre election mode" might match their performance in a real election (a seriously problematic situation itself), the "test" was a fraud. If I were to demonstrate software in the way Akin describes the demonstration that he witnessed, I would be very deservedly fired.

I will repeat my assertion, with emphasis:

The "test" that Jeremiah Akin witnessed was a fraud.

The only evidence as to the proper operation of those devices presented at that "test" was the assertion by Mischelle Townsend that they worked correctly. All observers except Akin himself simply took her word for it. Akin refused to accept her assertion and tried to see for himself. For his efforts, Townsend has attacked him as (from the interviewer):

"a young man who had a chip on his shoulder when he came in here." She said that you came into the test with a "closed-minded" attitude and that you didn't want to "listen to the facts."

So rather than addressing Akin's concerns, Townsend attacks his intellect, ethics and objectivity. I'm afraid that the person in this situation, though, who has difficulty with their intellect, ethics and objectivity is one Mischelle Townsend.

Go read the article. It is both illuminating and very disturbing.

There are real and very, very serious concerns being raised by people like me, people who know software and who know how this stuff works. Rather than having those concerns be addressed, these people are being vilified, denigrated or simply ignored. The people pushing the touchscreen technology, whatever their motives might be, are behaving deceitfully, are hiding the actual operation of the devices and are providing no way at all to verify that they operate correctly. They want us to simply take their word for it that the things work like they say they work and like a voter might expect them to work.

Their word isn't good enough for me. It shouldn't be good enough for you, either.

Posted by Frank at October 15, 2003 10:11 PM
Comments

Nice summary. As a programmer, I think this is a dangerous travesty. If Akin's account is accurate, the registrar of Riverside County should immediately lose her job; Sequoia should lose the contract and have to start from scratch; and depending on how the state regulations for voting system verification are written, legal action might not be out of the question against the state or Sequoia or both. And I have no trouble imagining that things were done equally badly elsewhere.

I'm not really surprised that some officials don't get it, but I'm pretty sure many of them are just playing dumb and counting on other people's ignorance.

BTW, your Salon links are broken - left out the http://.

Posted by: Eli at October 16, 2003 11:16 AM

Thanks. The Salon articles are cached locally; please let me know if you can't get to them. I've fixed the links in the cache to properly point to Salon, however.

Posted by: Frank at October 16, 2003 11:35 AM

I got here via your comment on Jeanne's site. I like your site, and I fully agree with your conclusions in this and your previous post.

Re the operating system Diebold uses in its equipment in the field: I've read (where? BBV? I can't remember) that it is a build of Windows CE, which as I understand it must be custom-built for every device it runs on. The fact that it is CE is redacted from all public info Diebold makes available about its systems, possibly to conceal the fact that they are using software that is neither off-the-shelf nor certified as required by law in many states. Something really, really stinks here.

Keep up the good work, Frank!

Posted by: Steve Bates at October 17, 2003 1:03 AM

All Rights Reserved