... his case that computers are "more accurate" than hand-counts....

A common misconception, but in fact that has never been the case, nor even the purpose of "computer-counting".

Machine-tabulation is faster than human counting, that's all; the error rate is in fact expected to be slightly greater as an acceptable effect, because most elections are not so close that the error would change the outcome.

But this is exactly why, when the results are very close, recounts may be done by hand to assure a more accurate count. The law (based on experience) recognizes the greater accuracy of the slower human hand-count with bipartisan observers double-checking.

As opposed to, say, single-party, single-person county clerks "discovering" 14,000 "missing" votes on their PCs while alone in their offices....

(I should add that my career was in data processing during the Hollerith computer-card heyday, and I spent most of a decade working a major corporation's card-sorters/tabulators, readers, output-punches, printers, and programmable duplicators, clearing jams, replacing ruined cards, and dealing with all the other problems computing evangelists omitted to mention when preaching their products.)

Great questions Brad!

Although, I suppose I have to give Mr. Norden credit for coming anywhere near your program, his answer concerning why the Brennan center went missing for years on the subject of computer voting concerns is very weak.

This may seem harsh, but it looks to me like he is doing just what congress did to us in the beginning of this mess: giving the computer corporations a massive paycheck.

Does he think only scientists are capable of casting simple votes on a ballot. Why can't we trust the American people to do the easiest thing on earth, and put a check mark next to the name of a person we care about, and then count it.

I've had enough of this everything-imaginable-except-what-will-work approach to democracy. There just isn't a single good reason to add all this complexity to something we need so badly.

Plus Lawrence should probably have that cough looked into.

Since I don't have the authority to give you an Edward R. Murrow award, which you deserve more then Edward R. Murrow for your tenacity on this subject, I've already sent you another $20!

You and Bernie Sanders are draining me dry.

The only "advantage" of computer voting is making it is easier to cheat. Why else ignore the obvious?

Paper ballots, count where cast, in front of the public, the press, and the politicians, which is what is nice about living in a small town. Proud to be primitive.

Mavin Jones says:

Paper ballots, count where cast, in front of the public, the press, and the politicians, which is what is nice about living in a small town. Proud to be primitive.

And SO much less expensive in this great "conservative" country of ours.

Interesting comparison between the Volkswagen test mode and the computer test mode also. But like Brad says, the majority of the machines aren't tested, as if that would make any difference anyway. The whole thing is a bad joke.

Please join me in commenting on today's Boston Globe tech story on VW and relying on computers ( http://www.betaboston.co...ng-is-digital-deception/ ).

In addition, I've been meaning to give a short report on the Massachusetts Democratic State Convention last Saturday. After the convention itself, I went to a two-hour breakout session on Racial Injustice and the Voting Rights Act, which was well worth my time.

One of the speakers works on voting issues for the DNC (sorry, I was a little late and didn't get their names, and there were no handouts and nothing on the internet that I can find). He gave the same line as the Brennan Center guy about just needing better machines.

I finally got to ask a question (after about half the people had left, unfortunately). I started with a request that all of the speakers and audience members change their vocabulary and say "photo ID restrictions" rather than "voter ID". The ACLU and DNC guys were nodding their heads, but the moderator from MassVOTE asked what I meant, so I got to explain.

Then I raised the issue of the machines, saying none of us can follow the electrons to know that the machines are counting our votes properly, so why not advocate for hand-marked paper ballots counted in public at the precinct? As I recall (long day), the response focused on audits as the answer, or just said that wasn't going to happen so give it up.

I may have confused the panel's answer with the person who came up to me afterwards to talk about how Rockport, where she lives, just switched from hand-counts to opscan machines. She told me the one thing that I can see as an actual issue: finding people to do the counting. She said that Rockport has relied for decades on the volunteerism of its senior citizens for tons of things, as well as their availability to work the polls, but that well is starting to run dry. It seems to me that, for what machines cost us, we ought to be able to pay people (high school and college students, for example, because they have a much easier time than we old people do staying up late) to do the counting if we actually care. I sure haven't seen my neighbors in Kendall Square and at MIT coming up with any high tech solutions I can believe in.

Thank you, Brad, for that interview and for continuing the effort to secure US elections so that our democracy really is a democracy. As for a response, well, where to start?

First, I have never spoken with Larry Norden at the Brennan Center and need to rectify that now. No, he did not ask the Columbia County Board of Elections about our process of hand-counting paper ballots (all of them - not just the ballots from 3% of the machines we deploy).

Second, not being able to find replacement parts is a problem if your technology of choice is a computer. It's not a problem if the counters are human.
Third, poorer areas are said to be particularly at risk because they don't have the funding to purchase new computers. My response is, poorer areas have plenty of people, and those people can read and count. There are no racial or class disparities involved in hand-counting paper ballots. And it's far less expensive than buying new computers. Our costs have been quite modest, amounting to a few thousand dollars to count an entire election. (We have 40,000 registered voters and have hand-counted as many as 29,000 complicated ballots.)

Fourth, the concerns that hand-counted ballots might not be secured and processed properly, and that personnel are not well trained, have easy solutions. Implement chains of custody that are secure and bipartisan. Improve election-worker training. And involve more, and more capable, people in the process. Columbia County has done all those things.

Fifth, if your technology is optical scanners that read paper ballots, use the scanners' results as a check against your hand-counters' numbers. When our counters' numbers conflict with the scanners', it doesn't take long for us to figure out why. Sometimes it's counter error. Sometimes it's because humans can detect voter intent, which by NYS law must be followed, and which we do. And, as everyone knows, local elections can be won or lost by one, or two, or four, votes. Two weeks ago in our primary elections we had (1) a tie, (2) a win by two, (3) a win by five.

But that's just a start. There's so much more I could say.

CambridgeKnitter @ 7 said:

I may have confused the panel's answer with the person who came up to me afterwards to talk about how Rockport, where she lives, just switched from hand-counts to opscan machines. She told me the one thing that I can see as an actual issue: finding people to do the counting. She said that Rockport has relied for decades on the volunteerism of its senior citizens for tons of things, as well as their availability to work the polls, but that well is starting to run dry.

First, apologies for the slow reply. It's been that kind of a week!

To the substance of the point above, remember a couple of things. For one, those who work the polls must have a LOT of training in advance, be there for the entire day (from dawn until very late at night) and, thus, must be largely compromised of retired folks who don't have to work on a Tuesday!

Hand-counters don't need to show up until after the close of polls (out here in CA, that's not until 8pm), they can receive all of their "training" just before counting begins, and they only need to help out for a few hours --- unlike poll workers.

You, correctly point out that there would be a lot more money to pay those folks if it wasn't wasted on terrible, unverifiable voting systems and incredibly expensive "maintenance" contracts. But, that said, if you look at the towns in NH where they hand count, the entire community shows up and seems to find it an honor to participate in their democracy. From the folks I speak to in NH, it seems they have no trouble at all finding folks to do the hand-counting.

We are already detached enough from our own system of democracy. I suspect folks would welcome the opportunity to have more of a hands-on chance to work for it.

Also, too, "market forces" and "competition" would likely to come to play here as well. Republicans don't want those cheating Dems to do all the counting. And Dems don't want those cheating Repubs to do all the counting. To that end, I'd think, both parties (as well as "third parties") would have an incentive to make sure that their people were on hand to participate to make sure "those other guys" don't cheat!

Virginia Martin @ 8:

Thank you very much for ringing in here! As well as for your important work in Columbia County, NY and instance (along with your Republican co-Election Supervisor) that every ballot is counted by hand!

We need to get you on the radio show very soon! I'll be in touch! And thank you again!

It seems to me that the NYU study about the inadequacy of polling equipment is designed to raise public support for the electronic internet connected machines.
it is my opinion that votes sent into the internet can and will be tampered with. Vote theft is too attractive an option.