Told You So: More E-Voting Problems

I (along with lots of other people) said that electronic voting — “black-box voting” — was going to be a problem.

Forget questions about deliberate tampering; the complexity of these systems and the hidden nature of the software that runs it means glitches may happen without anyone noticing. And even if they do notice, voters can’t be quite sure that the person they voted for actually got their vote.

Guess what? Problems surfaced in Alameda and San Diego (California) on Tuesday (Mar. 2, 2004). (Big thanks to Liz in California for pointing this out.)

The first paragraph of the article is supposed to make people feel better: “County officials said Wednesday that they do not think significant numbers of people were deprived of their right to vote Tuesday…” Of course, if you’re an election official the last thing you want to have is an election foul-up, so what are the chances you’ll say, “Yeah, we had big problems”?

Another article, this one from the San Jose Mercury News, had a rather creepy headline:

Election officials report some e-voting glitches
But equipment worked better than expected

“Better than expected”? I expect something as important as a voting machine to work perfectly. When the “glitches” involved lost ballots or the inability to vote at all, that’s not a glitch. It’s a failure.

And don’t forget that, thanks to its quirky Diebold machines, Alameda County had problems once before, during the recall election. Somehow, the machines reported that not a single person in the county skipped the recall question. (At least a half percent of voters in every other county did just that.) That means either the machines discarded thousands of votes (those who abstained) or cast a vote for them. Which do you think is better?

The election officials may try to spin what happened as “glitches,” but the aforementioned Liz was there. What she has to say is long, but worth the read:

The election was total fiasco although mininized in the press.

Not all of the polls had trouble but many did. People were turned away or told to vote elsewhere. Some polls were gotten to work in a couple of hours and others half the day. Some people were leaving town and had planes to catch or work schedules that did not permit a return to the polls. Of course some could not vote in their district since ballot measures differ. Nobody wants to say whether there were hundreds of people turned away or just a few .

In a nutshell the County and the Registar of Voters wanted to go with the Diebold system and spent $31 million dollars. They spend a lot a money but did not seem to think it through, and rushed to get it online.

They gave the workers a little bit of training but on election day it proved to be not enough. (but remember most of the poll workers are the nice eldery ladies who are just using to being name checkers and not IT guys. ) Many precinct workers opened their computers that morning and the computers booted up to an unfamilar screen. It was supposed to automatically open to the program but didn’t. They had some IT troubleshooters but they were spread thin. (The talk show I listened to on election day had some MSCE certified people unoffically volunteer their services and some things were fixed. Although it was not easy.)

Of course then others were worried about unoffical volunteers and thought they were a security problem. It seems the system is easily broken into.

It is just on a Windows CE platform and supposed to open right to the program. Instead the program just opened to the Windows desktop and poll workers were totally stumped and did not have any instructions on what to do in this instance. Meanwhile people arrived to vote and there was no back up plan. They were turned away and told to come back later. Some areas had alot of polling places out.

[...]

Even thoroughly, well thought out, well planned for new software always have bugs to work out. Not all of the polls had trouble but many did. People were turned away or told to vote elsewhere. Some polls were gotten to work in a couple of hours and others half the day. Some people were leaving town and had planes to catch or work schedules that did not permit a return to the polls. Of course some could not vote in their district since ballot measures differ. Nobody wants to say whether there were hundreds of people turned away or just a few . Then there were issues of how the system counts and whether one can even trust it to be accurate. Some people had issues with the touch screen not registering the vote correctly having to go back and redo.

How many people didn’t get to vote? How close were some of the races? How many people thought they voted for Candidate A, but thanks to a lack of a paper trail will never know that they voted for Candidate B?

This is a bipartisan issue — it cuts across all affiliations. No one is served by having faulty or even questionable voting machines that are installed simply to give the appearance of a solid system.

Until the multitude of problems are worked out, these systems need to be removed from service, period. This is democracy we’re talking about — there’s a lot more at stake than the egos of election officials.