[E-voting] Belgium?

David GLAUDE dglaude at poureva.be
Thu Jun 2 23:37:03 IST 2005


The reason we had VVAT in 2003 and not after in 2004 is because it was a 
one time law suppose to proof that the system without VVAT was OK.

It might be reintroduce for 2006 and 2007 as there are law proposal. We 
did not decide to fight this fake attempt at VVAT since it can not be 
generalised without paying even more than what we do already without 
printer.

Catherine Ansbro wrote:
> Could you clarify why it was that Belgium began to ignore the VVAT paper 
> ballots and not count them?

We always had counted the paper we had. But VVAT was attempted in only 
two electoral "zone" for one twin election only.

 From the oral report of one of the expert at one of our conference, 
peaple had difficulties to count the paper (more likely the kind of shop 
receipe you get). That expert also said he was happy that ticketting 
days are over (since he does trust the computer and believe it is 
unnecessary and he knows it was a mess counting those).

Various reason (mostly human factor):
* There were two elections per paper or the two simultaneous election 
were in the same paper ballot box.
* Because many elector wanted to change their vote when discovered on 
paper, many ticket had to be render void (I think there was a special 
line printed for saying it was to be not counted). The trick is that the 
law say: "If the paper does not match the screen call the president." 
but if "The paper does not match what you want or what you think you 
voted." then you are not suppose to invoque this procedure to revote... 
however president accepted in most case to redo the vote.

One could compare the end of those two document:
* The one we have digitalised from the original a parliamentarian gave 
us and then re-encoded: 
http://www.poureva.be/IMG/pdf/RapportExperts2003.pdf linked from 
http://www.poureva.be/article.php3?id_article=32 where there is an 
additionnal confidential page (the last one).
* The one later available from the official web site:
http://www.poureva.be/IMG/pdf/RapportExpert20030605.pdf and linked from 
http://www.poureva.be/article.php3?id_article=87 where the same last 
page is not confidential anymore.

[[
L’expérience n’est pas concluante dans la mesure où la conception et la 
forme des tickets n’a pas permis d’effectuer le comptage manuel 
conformément aux prescriptions légales.

Le collège est d’avis que les résultats des opérations de dépouillement 
manuel des tickets ne peuvent être considérés comme fiables et que les 
résultats de la totalisation automatisée peuvent être considérés comme 
les plus fiables.
]]

Earlyer in the part of the report that never ever was confidential there 
is this saying that counting ticket is nut easy but it prooved that our 
system is relyable...:
[[
L’expérience de « ticketing » s’est révélée satisfaisante
en ce qu’elle a permis de montrer la fiabilité du
vote électronique (émission et comptage des votes). Le
dépouillement manuel des « tickets » est extrêmement
difficile et fastidieux.
]]

So if there was no journalist to uncover the count discrepency, then 
there would be no trace of the fact that paper and eVoting result did 
not match... just a conclusion to say: "Ticket gave us the proof that we 
can trust the computer... don't try ticket again because it is not funny 
to count".

Less than a month ago someone at the head of the administration taking 
care of Voting and eVoting said that if ticketing was to be used again, 
then he would recoment using the bar code (that were printed dispite the 
fact that it was not in the law) to fastly count the ticket.

> When there was a discrepancy, I believe you mentioned that the election 
> officials decided not to bother with counting the paper ballots, perhaps 
> since the discrepancy wasn't large, and eventually they dropped having 
> paper ballots altogether.  Is this accurate?

I think they did count, it was not funny for various reason... size, 
hard to read, ...

Also the counting team wanted to have the eResult before the end of the 
paper count. That way when you know the number you should reach it is 
easyer to get exactly the same number... however it was refused to them.

Another story is that they opened the bier before before the end of the 
count... because they forgot to count one ticket ballot box. ;-)

But at the end of the night, they may have give up with badly checked 
number. None of the two location had exactly the expected result (the 
eResult).

> Were there ever any legal challenges to the fact that the paper ballot 
> was no longer being treated as the legal record of the vote, even though 
> the Belgian law says that the paper vote should have precedence?

No (except that one of the elected parlementarian that had is count 
affected by this story raised the question the day of the election 
validation... but he was told to read the addendum to the report... the 
"confidential" one).

One trick is that to challeng the result of the election, you must 
submit that the Senate/Chamber at most 10 days after the election (and 
we made some attempt that way on other ground) but the rapport of the 
expect is available to the parlementarian 15 days after the elections 
(and month later for normal citizen).

I don't know if this a bug or by design. Maybe this is just to give time 
to the "expert" to write the report.

> Why was/is there no public challenge?

The main reason might be that we (PourEVA) do not believe in 
"Ticketting" in a few location to be the solution to make eVoting 
democratic. We believe it was a marketting attempt to make in invalid 
demonstration that "it work". I don't have the number here but it costed 
way too much money in pure lost.

We might reuse that story if they plan to reintroduce Ticketting once 
again. But currently it is totaly not exploited.

David GLAUDE

PS: I hope this does not contradict anything I might have explain 
earlyer about this story. I tryed to be as complete as possible. For me 
that story is like the cosmic ray... a proof that pure paper rules (and 
that optical reading is an acceptable alternative).



More information about the E-voting mailing list