[EM] STV can be hand counted...

[Stephane Rouillon]

James Gilmour a écrit :

> Stephane, sorry if I sounded a little harsh in my one-word response.  My problem is that, as an
> active campaigner for practical reform, I encounter comments about the complications and "problems"
> of STV-PR counting almost every week from those who are opposed to any reform at all.  (We have a
> "First Past The Post" Campaign here in Scotland that is very actively doing everything it can to
> wreck the STV-PR reform that is now going through the Scottish Parliament.)

No offense taken, you are right.

> The STV-PR elections you describe above are amongst the most frustrating to have to count by hand.
> Usually you end up transferring small numbers of ballot papers of ever-decreasing values in a most
> time-consuming process.  The effort seems out of all proportion to the numbers of candidates or the
> numbers of electors.  For such elections I would strongly recommend that you punch all the
> preferences from the ballot papers into a computer data file and then use one of the (freebie) STV
> programs to do all the sorting, counting and tabulation of the results and intermediate
> calculations.

Yes but it is still a long work to do alone... I did it like that and it took me around 1 hour,
some would find that fast, I though it would go faster... Typing 15 (ballots) x 15 (preferences)  x 12
(districts) preferences at more than 1 second a preference...

> For public elections (FPTP and MMP) we already use an army of enumerators in one counting centre in
> each local government area.  Northern Ireland experience shows it will be no great problem to train
> them to handle STV-PR.  The main thing is make sure that the senior officials really understand the
> STV rules and understand what their staffs have to do.  Here in the UK we have a stupid obsession
> with getting the results out as fast as possible.  So we start counting about one hour after the
> polls have closed and count all through the night.  Constituencies compete to see which can be the
> first to declare!  And we have through-the-night TV programs where pundits predict the total outcome
> based on projections from only one or two results!!  I am pleased to say that ever since 1973 a more
> sensible approach has been taken in Northern Ireland for their STV-PR elections.  They start fresh
> the morning after polling day, having had a good night's sleep.  For large public elections most of
> the counting is completed by mid-day on the second day after polling, with all of the results
> through by mid-afternoon.
>
> Ironically, one complaint about the electronic voting and computerised counting in the 2002 Dáil
> Éireann election that produced "instant" results once the button had been pressed, was that
> established politicians who were booted out by the voters had no time to adjust to their impending
> fate.  With manual sorting and counting, they would have seen for some hours that their piles of
> ballot papers were not likely to be high enough to secure election and that transfers were passing
> them by, stage by stage.  This pandering to over-inflated political egos was brought forward as a
> "problem" during the Stage 1 oral evidence sessions when the the Local Governance (Scotland) Bill
> was being considered by the Local Government and Transport Committee of the Parliament.  Such
> considerations seem a long way from most of the topics discussed on this list, but (sadly) we do
> have deal with them if we want to win practical reform.
>
> James

As for any system, both electronic ballot entrance and paper trail should be used.
If you can ask voters to enter their ballot in a console then print their ballot and send their vote
once they check it is identical to the paper they put in a box after, you have both benefits.
You could have separate computers output the results in a matter of hours and keep the papers in case
the election office suspects any fraud by statistical analysis...
STV is a good model. Could you elaborate on the debate aspects?

Stephane