Ohio's Alternative to Diebold Machines May Be Equally Bad

phorest writes "One would have thought the choice of Ohio lawmakers to move away from Diebold touch-screen voting terminals would be welcomed by the American Civil Liberties Union (ACLU). Instead, the group is warning the elections board that their alternative might be illegal under state laws. 'The main dispute is whether a central optical scan of ballots at the board's headquarters downtown would result in votes not being counted on ballots that are incorrectly filled out. The ACLU believes the intent of election law is to ensure voters can be notified immediately of a voting error and be able to make a second-chance vote.'"

Comments

One thng you need: software freedom

[jbn-o]

It's not that difficult. But people in positions of political power are disincentivized from doing the right thing. This includes talking to technical people who advocate for free software voting machines so that we can end up with machines that produce voter-verifiable paper ballots which are stored for manual counting and are built on a free software system so that the county/state can get programmers they can trust when things don't work correctly. Having a choice of proprietors is just picking your monopolist and then hoping they'll do what you want when the contract is signed.

Instead of spending millions on a new proprietary system that will not adequately address local needs issues (and thus cause great embarrassment for the clerks who chose them), they could spend money (even with other states and counties) developing voting machines they can maintain and inspect as much as they like. Counties and states can purchase the required black box testing themselves, they don't need ES&S, Diebold, etc. to do this for them.

In this particular case, the ACLU's fear—voters not being immediately notified that their ballots are invalid—can be dealt with by a computer which scans (but doesn't count) their paper voter-verified ballot. Not only can most voters have an opportunity to read their paper ballot, they could plug in a pair of headphones into the computer and have the computer read them their ballot back and then determine if that comports with their intended vote. Then after this proofing (human and/or computer) each voter has a reasonable expectation that their ballot is valid and accurately reflects their intention.

I was part of the appointed group that recommended a set of voting machines for Champaign County, Illinois' elected County Board. Due to some not-completely-honest measures about only hearing from "approved" vendors, and a bunch of poor choices, I was pushed into picking the least-worst which happened to be a set of ES&S machines (one scanned and/or produced a paper voter-verifiable ballot, the other counted that paper ballot and physically retained it in a locked cabinet). Champaign County ended up with ES&S machines, only one of which had been approved for use by the state (in the state's bound-to-be-bullshit testing regime). The hurdles to overcome aren't ridiculously difficult. It will be hard to get some people to understand that it's beneficial to have local control over the voting machine so the machines can be reprogrammed to meet local needs (including changing the software to accommodate non-first-past-the-post voting, and generally fixing bugs or adding enhancements a county decides they want after the voting hardware contract is signed).

One thing that would really help (nothing like the power of a good example) is a free software voting machine that works just like the ES&S paper ballot scanning machines. These machines have a remarkably simple interface, good and adjustable voice, clear display, and headphone jacks. But these machines run on proprietary software which ES&S isn't willing to relicense (despite being their customer). So you're stuck with them for "support" and that means hoping they'll share your county's idea of what your voting system should do.