[E-voting] Meeting with the Minister/Dept

[Catherine Ansbro]

Tim,

While I agree it would probably not be politically advisable, it is 
absolutely essential that a statistically significant random percentage of 
all ballots be counted at EVERY election.  (What would be statistically 
significant?  I'm not an expert here.  It would have to be enough to mean 
there'd be enough chance of getting caught to serve as an effective 
deterrent to fraud.)

Good electronic fraud would not leave itself open to an obvious needs for 
recount.  It is not just a "very close election" that needs a manual 
recount, if you're dealing with electronic machines.  There's extensive 
ongoing discussion of this in BBV and other websites.  E.g., there are many 
indications of problems with the computers used to tally the optical scan 
counts.  It would be child's play for a programmer to alter a programme so 
that it would produce results that would plausible and not close enough to 
trigger an automatic recount--but not be what the voters voted.

This is what we must protect against now, at all costs.  In the States they 
are now faced with this dilemma--thousands of machines that produced no 
paper trail and cannot be audited.  Even all the original hardware is being 
destroyed; judges are not allowing the evidence to be saved (e.g., audit 
logs), despite the fact that there are ongoing legal challenges and FoI 
requests for the audit logs.

Do you think that here would be any different?  I have seen enough legal 
and judicial irregularities up close here to be less than sanguine about 
counting on fairness from our justice system here or abroad.  There are 
fair judges and solicitors, however it is not uniform.  After all, judges 
here are political appointees and this can affect those who have their eye 
on career advancement.  It's only human.  But that's why the system has to 
be have plenty of non-political regulatory oversight.  We need a system 
that assumes that human beings will try to outwit the system for their 
personal advantage, if they can.  The biggest danger is not individual 
voters, it is people inside the system.

Catherine


At 12:31 18/01/2005 +0000, Timothy Murphy wrote:
>On Monday 17 January 2005 20:34, Catherine Ansbro wrote:
>
> > Also, someplace in the basic presentation should state the need for a
> > mandatory audit of a significant percentage of the votes at every election
> > as necessary to deter fraud.
>
>I don't agree with that, or think it would be wise to suggest such a rule.
>In my view, VVAT should be regarded as a safety-net,
>reserved for cases where there is reasonable cause for doubt,
>as certified in the last resort by a court.
>
>If you put forward a system which seems even more time-consuming
>than the current system it will certainly be rejected.
>
>--
>Timothy Murphy
>e-mail (<80k only): tim /at/ birdsnest.maths.tcd.ie
>tel: +353-86-2336090, +353-1-2842366
>s-mail: School of Mathematics, Trinity College, Dublin 2, Ireland
>
>_______________________________________________
>E-voting mailing list
>E-voting at lists.stdlib.net
>http://lists.stdlib.net/mailman/listinfo/e-voting
>http://evoting.cs.may.ie/