[E-voting] Re: terrorism and electronic elections

Michael McMahon michael at hexmedia.com
Tue Jul 12 14:50:37 IST 2005

paul.holden at icsmember.ie wrote:
> For me, questions remain: If we end up with VVAT, and we decide to check by
> sampling, what happens if the sample audit uncovers a discrepancy between the
> paper and the electronic votes? Is the paper taken as the true vote in that
> constituency only, or is the entire electronic election declared invalid in
> favour of the paper? And if a margin of error is allowed, how big is that
> margin? What triggers a full manual count in all constituencies?

In theory, even one unexplainable discrepancy means the paper ballots have to be
manually counted (or at least all checked), and it would apply to that constituency only
because a particular fault/fraud would not necessarily occur
in more than one location.

Because you are only sampling a small percentage of ballots,
one error discovered could mean dozens or maybe hundreds of
errors which were not detected.

Having said that, you would have to allow for the (remote) possibility
that fake ballot papers could be dropped into the ballot box.

The real printed ballots would have to
have some authenticating information like printer marks
equivalent to the old stamp, and pre-printed serial numbers on
the back. In that situation, fake ballots would just be removed
without affecting the audit. However, if a ballot paper turns
up that looks authentic, but has the wrong data on it, then
that is evidence of a serious problem.

- Michael.

More information about the E-voting mailing list