[EM] Re: touch screen voting machines

[Eron Lloyd]

On Tuesday 11 November 2003 8:00 pm, David GLAUDE wrote:
> This list is about "advance voting technique"...
>
> My question to this list is wich of those technique can be used in large
> scale election using only manual counting and computing of the result.

A very good question. For the most part any advanced voting method can be 
captured on paper, but how it *scales* is indeed something to look into.

> I see too many peaple that push e-voting in order to introduce "complex"
> counting system. Where I think for election complexity is the opponent
> of transparency, security and clear understanding on the effect of my
> vote on the potential result.
>
> Eron Lloyd wrote:
> >>In Belgium when paper voting, I have the right to draw a little house or
> >>a mickey mouse (without beeing affraid of Disney corporation for
> >>copyright violation since my vote is anonymous). Of course my vote is
> >>not valid... but I expressed my fealing. This is important in a country
> >>where voting is mandatory.
> >>
> >>In Belgium when e-voting (magnetic card), it is not possible to draw
> >>anything or vote in an invalid way. Actually you can only vote blanc
> >>(wich is technicaly the same but not from a psychological point of view).
> >
> > Error-free simply meaning it should accurately reflect the voter's
> > intent.
>
> If my intent is to fill every option (even if they contradict each
> other)... You want to use ratio button where only one choice is
> possible... I want check-box for all option (actually I want paper and
> pen).
>
> > There should still be mechanisms to voice dissent, which can come in the
> > form of write-ins (also easy using a computer interface) or what should
> > be required on every ballot line, the "None of the Above" option.
>
> "None of the Above" is "White" or "Blanc" voting.

Not true. When a ballot item is left blank, it can be interpreted several 
ways. One is the way you mentioned, but it could also mean that they either 
missed that item or were unsure. By explicitly selecting "None of the Above", 
it sends a clear message. In a three way single-seat winner-take-all race, if 
the winner wins (with say 37%) but more votes are cast for NOTA, it sends a 
clear message to the candidate that they will be operating without the will 
or mandate of the majority, which could only be inferred otherwise (like here 
in the US, where Bush won without a majority vote or clear lead).

> > This also
> > would be recorded on the resulting paper reciept. If you have other
> > ideas, like walking up to an official and putting a lighter to your
> > ballot, that's fine by me.
>
> This actually one of the only way in Belgium to make your vote counted
> as NULL. You have to make it not anonymous (like writing your name on it).
>
> I was working at the pooling station where 1100 peaple had to vote. One
> hangry computer aware citizen did not want to vote... and I gave him the
> only possible solution to his problem (He wanted to vote NULL to protest
> against the use of computer for election).
>
> I was the second one to protest... but I had a formal lettre with me and
> a carbon copy to keep a trace. ;-)
>
> Now out of the 1100 to come, only 1000 came.
> So it is actually 0,2% that were against e-voting!

Interesting.

> > You mentioned that voting is mandatory. This is very interesting. What
> > penalties apply if you don't vote, and what is the resulting turnout
> > percentile?
>
> For the exact number of peaple voting... please check (in french):
> <<Pascal DELWIT, Des élections sans électeurs ? Causes et conséquences
> de l'abstention aux élections européennes de juin 1999, Les Cahiers du
> Cevipol, vol. 99, n°3, 1999. en format PDF, en format ZIP (405 Ko)>>
> http://www.ulb.ac.be/soco/cevipol/cahiers/cahier99-3.pdf
>
> Belgium, Grece, Luxembourg have mandatory voting.
> [In Italy it is "almost mandatory"(???).]
>
>                               1979  1984  1989  1994  1999
> CE-UE (vote non obligatoire) 61,29 58,99 54,19 55,19 48,00
> CE-UE(vote obligatoire)      91,23 86,21 84,94 79,71 79,36
>
> EU 1999 election % of vote / citizen supposed to vote
> Belgique   91,29 92,19 90,73 90,56 90,96
> Grèce            80,54 79,97 71,24 70,27
> Luxembourg 88,90 88,80 87,60 88,54 86,63
>
> EU 1999 election % of valid vote / citizen supposed to vote
> Belgique   80,04 82,09 83,13 82,74 84,74
> Grèce            80,54 79,97 71,24 70,27
> Luxembourg 80,30 80,60 78,60 79,74 79,13
>
> For Belgium I think.
> Arround 10% of the citizen do not present themself to vote.
> Arround 5% of the citizen voting do vote 'blanc' or 'NULL'.
>
> In Belgium (and in Greece) no penalty are practicaly applyed...
>
> But in Belgium penalty are as far as I know...
> If you get notified that you did not vote for the 3rd time, then
> financial and potential prison. I think that if you continue not to
> vote... then you loose your citizen right to vote. ;-) LoL.
>
> It is like punishing tentative scuicide by death penalty.

Good analogy. Thanks for the stats, as well.

> > Yes, along with banking, grocery shopping and automated telephone
> > systems, things are getting more impersonal. There are still issues that
> > need to be addressed, regardless of the system being used.
>
> There is no secrecy between me and my bank.
> The know how much is in my account and I have paper copies of every
> transaction. (Paper audit trail).
>
> Same for my shopping list and my telephone call. In order to pay the
> bill, they need to know what I consume...
>
> It is not the same for election where secrecy is the internationnal rule.
>
> >>So think twice, when I go voting (physicaly) my ID is checked, my face
> >>is compare to the picture on my id card.
> >
> > There are a variety of software and hardware solutions that can aid in
> > this area.
>
> Give me biometrics and ADN test to make sure my vote is anonymous and my
> twin will have no right to vote after me. (all of that remotely)
>
> There are well known, well tested procedure and technique available
> since ever to do that. Paper, Pen, Printed ballot, ID card, printed
> elector list, witness, ...
>
> > Not in my system. Python (and C) is the language, Linux is the operating
> > system, Qt is the GUI toolkit, MySQL is the database backend. All are
> > available in GPL or looser licenses, and could be built against reference
> > sources. Short of consumer PC hardware used, the entire system would be
> > completely transparent.
>
> How do I check that the binary code running is the compilation of those
> reference source code? (without interupting the system).
> How do you make a security/logic analyse of so many lines of code?
> How does a normal citizen trust the system??? trust an expert?

It would be quite hard to do (see my previous post). However, these issues 
must be addressed, as such systems are being deployed all over the place. By 
developing an open alternative will at least serve as a good reference model.

> >>>Relying on voters to audit their votes is unacceptable,
> >>
> >>Who else can audit their vote???
> >>Expert? Whitness? Big Brother?
> >
> > The machine would keep an electronic tally as well.
>
> Now you get 2 differents result... the paper result and the electronic
> result... what do you do? Cancel election? Use the paper everywhere? Use
> electronic everywhere?

No...both are used to cross-validate the other. A random ID on the ballot and 
in the database will retain the same votes, so if a recount is called, both 
are there for auditing.

> Also if what is on paper and what is on the screen (or where I did
> click) is not the same... what do you do?
>
> >>>That might just mean pen and paper for a long time.
> >>
> >>PAPER RULES.
> >
> > Unfortunately, we don't use paper now. We use lever machines with a
> > mechanical counter. The county is looking at DRE machines for 2005, and
> > we really can't turn back. We have to only hope for as open a DRE system
> > as possible. One
>
> Your only hope is to print a paper audit trail and to count those paper.
> The push for DRE was made to avoid recount like in Florida...
> Where the use of DRE make it even more important to have recount
> (including full recount at random place)...

DREs aren't designed to avoid recounts. Recounts can and must be possible in 
any voting system. Agreed.

> > last though, a computer interface would be able to assist voters in
> > casting STVs, like Approval, IRV, or Condorcet. People don't properly
> > read paper ballots. You should see how many mistakes happen on voter
> > registration cards! I'd be very interested in any studies looking at the
> > usability issues of purality-majority vs. STV or other types of paper
> > ballots. For advanced voting methods I'm not sure paper would succeed
> > well.
>
> If it is not applicable on paper... then do not call them "advenced
> voting methods". You are going backward I am affraid.

It's not me, just my government ;-)

Eron

> David GLAUDE
>
>
>
> ---
> [This E-mail scanned for viruses by Declude Virus]

---
[This E-mail scanned for viruses by Declude Virus]