Voting machines ditch ballots in Scotland

More than seventy thousand votes in the recent Scottish elections were rejected by the electronic counting machines, with no human oversight. The news has dismayed observers, with new First Minister Alex Salmond describing the news as "astonishing", and deeply disturbing. He told the BBC that he had been under the impression …

[Richard Tobin]

There's no evidence of anything wrong here

This story is absurdly exaggerated. It was entirely reasonable to vote in one column and not the other (in fact, in the previous election there were two separate forms instead of two columns). A ballot paper with a cross in one column and not the other is not spoilt, and should be counted as a vote in one ballot and none in the other, which is what the computer did. The only question is whether in some of these cases there was a cross that it didn't notice.

[Chris Collins]

Useless tech

I have yet to see a reason why all this electronic voting shenanigans is necessary, let alone any better or cheaper, than applying your X with a stub of pencil in the local school hall and having it counted by old ladies. It appears to have worked relatively faultlessly for the last 250 or so years and gives some coffin dodgers a reason to keep breathing.

[Alan Donaly]

Did it make a difference

There is already information on why these electronic voting setups

are by their very nature insecure but for the purposes of this

story it would be nice to know if all it did was kick out random

votes or did it favor one side in other words was this intentional

or was it one of our little electronic buddies having fun with the

humans as they put it.

[Anonymous Coward]

Does it matter?

All the bloody politicians are the same anyhow....

Dunno why I bothered, I suppose it was nice to have a bit of a walk to somewhere other than Haddows.

[Anonymous Coward]

@ Chris Collins

I guess one benefit to the "politicians" in introducing this technology is blatantly obvious... it makes it far easier to "manage" the outcome of an election and then blame the technology or service owners when the fraud comes to light. It's not as easy to do that when you have a hand-raulic vote count.

By the time that the "mistake" or "error" *cough* has come to light, the relevant person/party that has been fraudulently elected is then well established in his position, and with posession being 9/10ths of the law... ultimately the argument gets diluted and lost in a quagmire of allegations and retterick. The public then switch off, the media gets bored and moves onto the next celebrity "shocka!!!" and ultimately nothing changes apart from us being robbed of another little piece of our democracy.

This is the new politics. Just look what Bush did first time round by using database tables to exclude certain ethnic groups to win the Florida count - and he got away with it, the redneck fu**wit!!!

Just thankful I don't live there.

[Chris Cooke]

So many problems...

The electronic counting of votes was considered necessary because of the complexity of the vote counting mechanisms: the Scottish government election votes were counted both by first past the post and by the D'Hondt system and the council votes were counted by single transferable vote. It was felt by the Scotland Office that results could take days to appear if counted manually.

One big problem though, at least the one reported at the time, seems to have been that the vote counting machines could not cope with ballot papers of more than a certain size - and so many candidates were registered in some regions that there wasn't room on the ballot paper for both the candidates and the instructions on how to fill in the ballot paper correctly. Instead of falling back to manual counting at this point, Douglas Alexander's Scotland Office decided at the last minute to remove the voting instructions from the ballot paper where necessary to make room for the candidates. So in some regions the instructions were left off, or severely curtailed. Those were the regions that had the highest number of spoiled ballots.

Another problem was the bizarre decision, again by Douglas Alexander's Scotland Office, to amalgamate the two separate Scottish government votes on the one piece of paper. Nobody had seen such a ballot paper before and it baffled a lot of people; particularly where the instructions had been removed!

By contrast, the single transferable vote was used in the same election for the very first time for the council ballot, and voters seem to have taken to it like a duck to water; there were almost no spoilt council votes, as I recall. Which is interesting, when one big reason for adopting the bizarre FPTP/D'Hondt system in the first place was that voters were said to be incapable of coping with STV - or so I remember from the late 90s, anyway...

[Steve]

STV

STV has been used in N. Ireland for decades, with manual counting. Despite all the bad jokes it isn't difficult to use, nor to count. It takes time, certainly, with a few days before the results are available, but so what? Isn't it worth waiting 3 days to get a reliable result that you'll be stuck with for years?

[Anonymous Coward]

RE: Useless tech

Old Ladies who volunteer to count votes are hard to hack, I can't even start to formulate a method that doesn't involve failure. How do you socially engineer, social engineers.

[Charles Manning]

Hacking votes is nothing new

Back in 1980 I was at Univeristy of Cape Town. Student body election voting was done by punch cards. The puch cards were then run through the computer which tallied them all up.

Since it was a Univac, the end of a data set was specified by a card marked with @EOF. Of course a reasonable % of computer science students punched out this pattern which meant that only a few cards would be read before the end of the data set was reached.

It would typucally take quite a feew iterations before all the @EOF cards were found and removed, making a half-hour exercise into an all-nighter.

[Anonymous Coward]

Title

I was at DRS (the company behind the electronic counting machines) for a training course just before the elections. According to them, the reason the Scotland Office chose to go the electronic counting route was because of the single transferable voting system they chose to use. It was possible to work it out by hand, but the algorithm was overly complicated. Having said that, when the count was over, all the hard drives from the counting machines were to be surrendered to the returning officers and they should still have the paper ballot slips. Wonder why they just didn't simply order a manual re-count?

[Anonymous Coward]

Even hand-count isn't foolproof

Bleh. Watch the 2006 Mexico Elections to see how even hand-counting can be botched if the system allows it. The counting process is like this:

1) Each polling station counts the votes,

2) Results are jotted down on an official act,

3) Same results are posted outside the polling station,

4) all ballots are put into sealed envelopes, put into a package that itself is sealed,

5) the act is put in front of said package.

Then the "district counting" consists in basically feeding the numbers on these acts to the central counting system, and the ballots are never, ever touched again. The pretext on the last election was that "opening the packages was illegal and useless".

Now anyone can see that the act could have been tampered with, and in fact, in some cases this was proven: the "outside polling station" sign and the act had, in some cases, 200 or 300 count "errors". Yet most of these were not even checked, even during the "recount" (which only recounted 9%, and didn't even open most of the packages).

E-voting only does the process easier, but the 2006 Mexico elections prove that even the manual system is fraud-prone.

[Lani Massey Brown]

Statistically Improbable Result Indicate Broken Laws

May the Colonies have a word here? The question is, are these election results statistically improbable? One piece to the US puzzle is our need to revamp election laws to be on par with technology. In the US and perhaps with you as well, until we fix our election laws to protect us from machine and human error, and HUMAN INTERPRETION of election results our election process will continue to be broken. The courts should not decide the people's choice. In 2006, it was the failure of Florida’s revised election laws that permitted an election with statistically improbable results to stand (18,000 undervotes). Our 2000's debacle with the pregnant chads resulted from failure to maintain the voting equipment properly. However it was the failure of Florida's election laws that permitted the chaos that followed. Had Florida’s election laws caught up with technology, both elections would have been an automatic re-do.

In the US we need tighter, better, wiser controls. What would the profit-conscious business world do? We must be more business wise. If we are to secure one voter, one vote…every-time integrity in our election process, we must implement high-bar guidelines for voting machine providers and elections officials to uphold. Moreover, until we fix our election laws we can be assured of continuing unacceptable results.

While the heroine in A MARGIN OF ERROR: BALLOTS OF STRAW scoffs at the notion of a silent coup marching across America in her fictitious voting machines…. It could happen more easily than any of us want to believe.

Lani Massey Brown, MARGIN OF ERROR: BALLOTS OF STRAW, political intrigue of a stolen election, available on Amazon.com.

[A. Merkin]

ROTM

In related news, a suspicious number of ballots had "Deep Blue" in the write-in section.

[Roy]

Even hand-count isn't foolproof (2)

Yes, hand counting is not infallible. It's done in each polling place by only a handful of persons--in some cases, persons from a single party. Who ensures that they act wisely and honestly?

Greater reliability may require that we widen the circle of hand counters--to include all the voters in the precinct.

How can that be accomplished with reasonable thrift?

If the ballots can be reliably and legibly videotaped, the videos can be easily published on the Internet, and every voter in the precinct can double-check the vote count for himself/herself.

Any interested citizen should be allowed to do such videotaping and publishing. This would allow private citizens to check the government's vote count, and also to check the authenticity of each other's videos.

To expand hand-counting in this way should create no additional cost and effort for the government, beyond the need to provide an easel for ballots and to ensure that space is available for videotapers in each polling place.

And it will be much easier for pollwatchers to see whether videotaping is being properly allowed, than to see whether the ballots are being properly counted.

More-detailed considerations at http://e-grapevine.org/citizensaudit.htm