Slashdot Mirror
How Would You Design the Voting Technology?
Bob Glickstein asks: "Punch-card ballot machines are now universally reviled, and we techies all know the perils of electronic ones. But I haven't seen anyone talk about a better solution. It's gotta be inexpensive, rugged, reliable, accurate, verifiable, tamper-resistant, simple to use, and secret. Verifying a vote tally should not result in TV news images of rooms full of election officials, squinting at ambiguous marks on a piece of paper. What contraption can possibly meet all these criteria?"
The most transparent technology there is at the moment for recording votes is for voters to tick boxes (or write numbers) on printed ballot papers and put them into ballot boxes. Voting slips are counted by hand based, in the presence of witnesses. If the result is close, the voting slips can be recounted. This system works well in Australia at all levels of government.
OK, we do get problems occasionally. But they are typically things like people impersonating other voters, and people voting multiple times at different polling booths. However, the system copes with this. If the number of voting irregularities detected is sufficient to effect the outcome of an election, a by-election is called in the seats in dispute. It really helps that the courts in Australia are not heavily politicised like they are in the US of A.
(The problems with voter impersonation, etc are also present when voting machines are used. The same solutions could be used in both cases; e.g. requiring voters to present photo ids, and throwing rorters into jail for a long time.)
In 1992 I worked an election as an inspector. Each step of the inspection was signed off by a Republican and a Democratic inspector, after both of us saw and confirmed each step of the procedure. I think it would be much easier to make a mechanical clock run fast after the back was sealed on than it would be for anyone to cheat by manipulating one of these machines.
These are the steps, as I recall:
- The machines are shown with both backs removed. This shows a
matrix of mechanical counters, all of which are shown to read 0. There
are "total for office" and "total for candidate" columns at the top
and right. These also read zero.
- The inner back is fastened on. This covers all but the summary
row and column. These are checked to still read all zeros.
- The outer back is fastened on. This covers the summary numbers.
- The election begins. As each voter comes in, he or she is checked
off, so the number of votes can be compared against the machines.
- Inspectors from both parties are sitting with a view of the back of the machines at all times, to further guard against tampering.
- After the election, the outer cover is removed and the summary
totals recorded on paper. The total voters for each office should be less than or equal to the number recorded at the door.
- Next the inner back is removed and the matrix is recorded on paper.
The totals are checked against the numbers recorded in the previous step. All inspectors sign off on this as well.
I just don't see where such a system leaves any room for cheating. Of course, it also couldn't handle an election among 135 candidates, but that's got to be a first anyway.For anyone who hasn't used these machines, they have mechanical safeguards against voting for more than the correct number of candidates for any office. No hanging chads, no votes for too many candidates, no butterfly ballot confusion, and there's a paper trail that can be verified quickly rather than in a vague and subjective way.
Isn't it easier to trust clockwork you can inspect than code you can't? For one thing, no one's going to "download" you new clockwork when you aren't looking at it... and it's 100 years easier to audit.
--
I don't want to rule the world... I just want to be in charge of mayonnaise.
Oregon abolished the polling place. That's right, we haven't had a voting booth set up for an official election in Oregon starting with the 2000 Presidential Election (don't blame us, we didn't vote for him, and we didn't leave home to vote against him!).
So how do Oregonians vote? In the comfort of their own homes. About six weeks before election day, every residence with a mailbox gets a voter's guide that comes with a voter registration card (if you're not registered and want to vote, you turn it in at least 30 days before the first election you want to vote in). A week or two after that, your ballot, secrecy envelope and return mailing envelope come in the mail. You punch out the appropriate holes on the punch card. Stuff your ballot in the secrecy envelope, stuff the secrecy envelope in the mailing envelope, and put your signature on the back, and either mail it or drop it off at the elections office, or if it's within a week of election day, at any of dozens handy points at various public facilities (libraries, town squares, city halls, courthouses, election offices, etc) staffed by elections officials specifically to collect ballots.
But how does Oregon prevent voting fraud? Easy. We check signatures on the envelopes against the voter registration. Not sure what the sample rate is, but fraud has not been an issue. If you don't get the ballot and you were supposed to, you go down to the elections office, show your ID, they verify your registration and they void out the missing ballot (so even if someone turns it in, when they go to scan the barcode before checking sigs, they see it's void and throw it out). They issue you a ballot and hand it to you and you're on your way.
What does all this mean? Well, for starters, you get three or four weeks with your ballot instead of three or four minutes. Time is on your side in making an informed, well-thought decision without having to stress out that you're missing out on having a life to go down to the polls and vote.
Encourage your state to abolish the polling place
Help us build a better map!
why does 'manual' actual counting of the take so long sometimes?
i mean, here in finland we've used a system where you draw a number in a circle to represent your vote, and return that then(folded).
and we do get election night results, and apart from saving money i just don't see much point in moving to all electronic. can't see it saving much money either since you would need specialised professionals(or maybe not) instead of the common volunteers you have now. true, finland is a much smaller country than usa but that shouldn't really matter because then there would be more volunteers counting with bigger population(the size of a 'voting-area' would be the same population wise anyways). and true, these were for the national big votes, but then again these are the only ones where voting night results really matter.
however, if(that is, when) the votes are submitted through electronic wires.. if you've read harry harrisons stainless steel rats, there's a very far fetched possibility for cheating in which you hijack the transfer of the votes to the main calculation centre(like, why cheat in the votes when you can just cheat in the results).
anyways.. regardless of how the electronic voting system works, i'd think it would be a good idea that the input is on something similar to those used now, so that you end up with something physical proof of what the vote was.
world was created 5 seconds before this post as it is.
Here on Brazil we have this kind of machine, check here:
http://www.procomp.com.br/projesp.asp
Unfortunelly the page is available just on Portuguese, but for the core stuff you can use a web translator.
Its not a perfect system, but it help us a lot here.
This is the way voting is done where I live: Columbia, Missouri. The circles are a lot bigger than the standard scantron, and you bubble them in with a Sharpie marker. It sort of makes sense, with Columbia being a college town, after all, but it may freak out some college students who have seen too many Scantrons!