The Coq mailing list

[martin AT vassor.org]

I realised after writing that this is a very long email, so tl;dr: when 
it comes to voting, transparency and trust (i.e. social aspects of the 
votes) are at least as important as the correctness of the result (if 
100 votes are not properly counted, out of 30 millions, it's fine). 
Voting machines, as they are currently done, do not present a satisfying 
tradeoff.

-------------------

Hello,

my two cents on that question, that goes along what Jean-François just 
wrote.

The French polling system does not rely on the fact that ballots can not 
be tempered with to be secure, but on the fact that anyone can verify 
that a given ballot is not tempered. Typically, the law requires the box 
to have at least 4 transparent faces, and anybody can remain in their 
voting office the whole day to check that it is done correctly. In 
addition, candidates can nominate observers that can enter any voting 
office, and also observe that everything is fine.

Said otherwise, the systems relies on the fact that votes can be 
tracked, from the moment the voter enters the office to the publication 
of the results, despite maintaining the secrecy of votes; *and that 
everyone understand what is done*.

Sure, we could go for a formally proven solution, but how much of the 
voters understand what is a cryptographic signature? Going to voting 
machine would mean, for a lot of people, to replace a transparent 
process by a black box (both figuratively and physically); and being 
trust by the population is more important than the speed at which the 
results are known.

Aside, counting the ballots is already pretty fast, despite being done 
manually. See, e.g. that tweet from a journalist : 
https://twitter.com/martinstabe/status/1518380069625942016 .

Also, from experience, the actual counting time is (of course) not 
negligible, but not the only thing that matters: typically, in my 
office, on last Sunday, we started counting at 19h15, counting the 
number of votes and comparing to the number of signature took around 40 
minutes (with voting machine, we would still need to count the 
signatures, which is the error prone part); then we had 3 tables 
counting 1033 votes (which is more than usual, because I'm in a foreign 
office: depending on the elections, there are usually between 400 and 
700 votes), this took approximatelly an hour (this time would be saved 
if using voting machines); and the we had to do the formalities, which 
took 1h30 (writing the report, sending it to the authority, and getting 
approval that everything was correct). Therefore, if we want to speed 
things up, we should rather automate the administrative part.

So, in conclusion, of course I would say that of course, from a 
*technical* viewpoint, we can design a reliable voting machine; but the 
*social* costs would not make it worth it, *in the context of French 
elections*:
 - the cost of transparency is greater (learning formal verification 
vs. observing the ballot);
 - the time gain is not that important, and not that relevant (plus, 
newspaper provide quite accurate estimations before the actual result is 
known: the result presented at 8pm is just an estimation).
This ensures a high degree of trust in the actual result of the vote. 
This article 
(https://www.lemonde.fr/societe/article/2022/04/26/presidentielle-2022-de-trump-a-un-bug-de-france-2-l-entetante-musique-de-la-fraude-electorale_6123706_3224.html 
, in french) actually discuss how frauds are perceived, in particular 
within conspiracy groups, and the outcome is that people that doubt 
about frauds in election basically trust the system when they 
participate in the counting. A quote from the article : "There might be 
some frauds, but in my office, I was here for the counting, Macron was 
first".

Best,
Martin


Le 27/04/2022 à 08:48, Jean-Francois Monin a écrit :
> +1
>
> Low-tech paper solutions are robust and easily understood by everybody,
> as well as attacks and countermeasures, so much better for confidence
> and democracy.
> Complicated technologies are just the opposite.
>
> JF
>
> On Tue, Apr 26, 2022 at 11:19:09PM -0400, Stefan Monnier wrote:
> > The way I see it, the more people who participate (i.e. more eyes),
> > the safer.
> >
> >
> >          Stefan
> >
> >
> > > On Tue, Apr 26, 2022, 18:22 Stefan Monnier <monnier AT iro.umontreal.ca> wrote:
> > >
> > > > > But imagine that voting machines are formally verified from hardware
> > > > > to software,
> > > >
> > > > What problem do electronic voting machines solve?
> > > >
> > > >
> > > >          Stefan