H.R. 811, the e-voting bill originally introduced by Rep. Rush Holt, is reportedly up for a vote of the full House of Representatives tomorrow. Passing the bill would be an important step in securing our elections. I have supported H.R. 811 from the beginning, and I am still firmly behind it. I hope it passes tomorrow.
H.R. 811 gets the big issues right, requiring a voter-verified paper ballot with post-election audits to verify that the electronic records are consistent with the paper ballots.
The bill is cautious where caution is warranted. For example, it gives states and counties the flexibility to choose optical-scan or touch-screen systems (or others), as long as there is a suitable voter-verified paper record. Though some e-voting activists want to ban touch-screens altogether, I think that would be a mistake. Touch screens, if done correctly – which no vendor has managed yet, I’ll admit – do offer some advantages. Federalism makes sense here: let localities make their own choices, as long as basic standards, such as the paper-trail and audit requirements, are met. Down the road, we may be glad that we left room for better touch-screen systems to develop.
The current version of the bill allows inferior paper-trail systems, such as ones storing ballots on a continuous reel of relatively fragile thermal paper, to be used through 2010, in places where they were already in use. The full requirement of a durable, permanent, privacy-preserving paper record takes effect everywhere in 2012, but starts immediately in places not already using a paper trail. Though less than ideal, the grace period is the best reasonable choice under the circumstances. A change of this magnitude takes time, so some kind of grace period is necessary. We could argue over whether it should be two years or four years, but at this point the most important thing is to start the clock ticking, by passing a bill.
If your representative is on the fence, this is a good time to call and urge a vote for H.R. 811.
[UPDATE (Sept. 6): The schedule has slipped so the bill will not be up for vote today. So there’s still time to call your congressperson.]
I just don’t get it. How can we be willing to risk our democratic republic to a computer program?
There is no transparancy. If we think we have trouble with apathy now…
If we use paper ballots, store and count them publicly, everyone can see that thier vote is being counted. There should be no need to check it if no one can change it.
This is the worst solution ever. I have been a republican for 32 years, and nothing has made me more ashamed of it than this. For the sake of decency, trash these machines.
I’ll have a post up responding to Daniel’s ITIF report within a day or two.
I wanted to share with you the report ITIF just released today on electronic voting. We look at the limitations of paper audit trails, and what some of the alternative technology options are for audit trails. Moreover, we highlight the fact that cryptographers have come up with universally verifiable voting protocols whereby every voter and election observer can independently verify the results of the election. I think this is where we should be focusing our attention.
http://itif.org/files/evoting.pdf
When is the bill scheduled for a vote, now that there’s been a delay?
@ mumu & Chris
That argument already went through a few cycles ago and it turns out that Australian elections are more expensive (per capita) than US elections by approx 2x or 3x (somewhere round that mark). Thus the USA can use the cash that they save on Democracy to contribute just a tiny fraction more to their Oil War campaigns. On the other hand, Australia spends so much money on elections, there’s nothing left to train the police in how to read security passes at a checkpoint.
More importantly, paper is (how best to put this), well it’s old. Like anything old, the high-tech industry must replace it with something new. After unsuccessfully creating the paperless office, the paperless financial industry, and the paperless medical industry and the paperless legal industry (all of which consume many times more paper now than they ever did in the world of mechanical typewriters and handwritten chits), it seems inevitable that technocrats must unsuccessfully design and build paperless voting. Only to discover that in order to keep their design stable, they will employ twice as many people and twice as much paper.
I’m also in Australia and I can’t see what the real benefit is in using electronic machines. We sometimes get technical people talking about using electronic voting, but thankfully the political parties are more than comfortable with the existing paper system. Paper gives them some visibility over what is really happening at the grass roots level.
Our Federal system is a two house system. The upper house ticket is usually huge and needs to folded twice to get it into the voting box, but our voters seems to be able to work out how to vote ok. The first round of counting is basically done by hand in the voting booth after voting closes, with the slips being forwarded to a central location for allocation of preferences – we have a preferential voting system. It usually doesn’t take all that long to get a result in the lower house (which usually only has a small number of candidates in each electorate). Unfortunately the upper house can sometimes take a few days for the final preferences to get allocate to the candidates – but the system is has lots of advantages that out way the disadvantages the US seem to experience.
Paper ballots mean the parties get to scrutinize the votes during counting and recounting is relatively easy, even though it can be time consuming.
I think I would be comfortable in saying the majority of voters in Australia believe that vote counting is not subject to corruption. I can’t remember ever hearing anyone that wasn’t a ‘fruit loop’ claiming someone stuffed the ballot box.
In Australia we all submit our votes exactly the same way, on paper ballots. There is no confusion, no complicated software or machines that can stuff up or be hacked. Its much better than the chaotic mess the USA elections appear to be.
Note that the bill as it will hit the floor does not require undisclosed source code… only executables and “configuration files”. The bill points to the VVSG for the definition of “votin system software” and that definition explicitly excludes source code.
The whole tool chain should be disclosed; in fact the compiler should have to be a stock version of gcc, or more generally the tool chain should have to be common off-the-shelf. Otherwise there’s a danger of a Ken Thompson hack to invisibly corrupt the elections.
It offers provision for any voter to decide they want to go back to paper ballot. Thus, it is possible to democratically measure the percentage of voters who trust the machines (and even to correlate trust against other demographics).
I can predict the vendors getting cranky about their software ending up “disclosed”, they won’t go down without a fight. There’s going to be a few tricky issues with regard to making sure that the running binaries really are compiled from the disclosed source code. They will need an “official standard” compiler — and then there’s the question of whether the compiler itself need to be “disclosed” when you consider that elements of the final binary code are actually originating from the compiler. Despite such details, I agree that the gist is in the right direction.
I similarly support this legislation. It’s been a long time coming. There is a lot of last-minute action with proposed amendments and no end of back-room intrigue concerning how the bill has changed during markup, but all of that is beside the point. This bill will have a lasting, long-term impact on the conduct of elections in our country. While many of the provisions may not kick in until 2010, I’ll take then over never.
Barry, from what I understand it does require public disclosure of source code. This does not imply that the source code is required to be open source. The code can still include copyrighted or proprietary parts that would not be licensed freely.
Many voting systems today are closed/hidden source despite many rules on the books requiring public code review. From what I understand since there are very few of these types of vendors the contracts receive a waiver for these code source review requirements. I may be wrong here but something to look into if you want to seek additional information.
Hope this helps. I called my rep today about it.
I noticed in the linked pdf that “undisclosed software” would be banned from use in voting machines. What does disclosure entail? Does it require open-source access to the software for public review, or does it only allow certain people to review the software?