Save democracy from a broken standards committee!

Cory Doctorow, being the Outreach Coordinator for the EFF — the Electronic Frontier Foundation, the good guys who campaign for civil liberties online — asked to make a mention of this important issue. The issue may seem merely technical, but it affects us all, and I’ve put it in layperson’s terms.

What is the IEEE?

IEEE is short for Institute of Electrical and Electronics Engineers, a non-profit professional organization of engineers who work with electricity, electronics and computers. Their mission, as put forth on their web site, is “The IEEE promotes the engineering process of creating, developing, integrating, sharing, and applying knowledge about electro and information technologies and sciences for the benefit of humanity and the profession”. They do a lot of technical publishing, host conferences and quite often help to define standards (one example, the “FireWire” standard for high-speed computer interfaces, also know as IEEE 1394).

What are they doing with voting machines?

One standard that they’re currently working on is for electronic voting machines. Work on this standard arose from the voting debacle during the 2000 U.S. Presidential elections in Florida. Most of the work on this standard is nearly done, and the draft for it is currently out to ballot by voting members of the IEEE. Once finished and passed by the IEEE, the standard will be forwarded to ANSI — the American National Standards Institute — for final validation.

The IEEE sits on an advisory committee to the forthcoming Election Assistance Commission established by the Help America Vote Act. This means that this standard could ultimately be adopted broadly throughout the United States. The EFF summarizes: “In a very real sense, the future of democratic systems in the U.S. and around the world are implicated by this standard — the stakes couldn’t be higher.”

Okay, so far, so good. Where’s the problem?

The IEEE working group for the voting machines created a design standard instead of a requirements standard.

(Those of you who are software engineering types are probably nodding your heads and saying “ah, I see.” I’ll explain for everyone else now.)

A requirements standard, simply put, is a document that describes what the end result should be. For instance, the requirements for a voting machine might be:

• It must store a log of each vote, and attach a timestamp so we can tell when the vote was made
• There must be some kind of mechanism or way for an independent verication process to determine whether or not the votes have been tampered with
• The storage system used to store the votes must be 99.9999% reliable (that is, it should fail only 0.0001% of the time)

A design standard, on the other hand, describes how the end result shall be acheived. the deesign for a voting machine might be:

• It should have a touch screen
• It should be built with standard PC components
• Votes should be stored on a hard drive, with another hard drive “mirroring” the main hard drive in case it fails.

As you can probably tell, a requirements standard, while being more general, tends to be valid for a much longer period of time. Technologies change often, but most of the time, the needs addressed by those technologies don’t.

Simply put, they didn’t write a standard to address the problems with voting machines, they wrote one that simply says how they should be built.

Cory reports that in concentrating on the “how” and not the “what”, the standard fails to require or even recommend that voting machines be truly voter verified or verifiable. How this could’ve been missed by actual certified electrical engineers, but caught by me — a guy who failed out of electrical engineering at Crazy Go Nuts University (ranked 430 out of the 431 student in the class of ’91) — boggles the mind.

Not only did the IEEE write a design rather than a requirements standard, according to the EFF, they also followed the basic plans of current voting machines. They also say that they’ve heard disturbing things:

• claims “that the working group and committee leadership is largely controlled by representatives of the electronic voting machine vendor companies and others with vested interests.”
• reports of “serious procedural problems with the…Committee processes, including shifting roadblocks placed in front of those who wish to participate and vote, and failure to follow basic procedural requirements.”

The EFF is concerned about this and is asking people to get involved. Go take a peek at the page devoted to this issue on their site to find out more. They’ve even provided a form that makes it very simple for you to voice your concerns about the voting machine standard.