[EM] Voting debate - missing the target

[Dave Ketchum]

Much has been written recently about problems using computers in voting 
machines.

Some say to somehow use paper backup copies of ballots.  I doubt this is 
worth the pain, and it does not address the heart of the problem.

The HEART of the problem is that we, the public, should DEMAND the right 
to check whether the computers have been told to do what we, the public, 
expect of them.

Computers have been around for decades and we have learned to build and 
program them such that they DEPENDABLY do what they are told.  If not, 
would banks dare use ATMs?  Would we dare enter a hospital, where much of 
the more complex activities depend on computers fulfilling their 
responsibility?  There is MUCH around us for which we now depend on 
computers being DEPENDABLE.
      Of course, for most of this, the computer builders and programmers 
want to get correct results, and will do their best whether or not we see 
the details of how they go about it.
      Voting is a special case.  We NEED the ballot to be secret, so that 
no one can find out and then pay those willing for voting "right", or 
punish those who vote "wrong".  If the computers count wrong, accidentally 
or on purpose, AND checking is restricted by means such as trade secrets:
           There is little incentive to use care in making sure they count 
right.
           Some could be tempted to deliberately support counting wrong to 
control who gets elected, or whether initiatives, referenda, or recalls 
pass or fail.
           Note also that the voting machine task, while it needs careful 
design, is not a challenge as to the amount of work to get done.

Thus the HEART of the demands of us, the public, should be the right, as 
poll watchers, to check the equipment to be used before election day - not 
that many of us, individually, are qualified to do this checking - but 
that, collectively, we can see to the checking being done.  Actually, 
attending to the right is the major detail because, with the possibility 
that we may check, we can expect machine builders and programmers to pay 
more attention to correctness.
      We do recognize desire to protect trade secrets - fine, BUT let it 
be done in some way that does not interfere with validation.

Note:
      Some machines have been bought, with trade secret clauses attached, 
and I understand that judges have agreed that those who signed such 
clauses have made the sellers immune to checking as to quality of the 
equipment.  Seems like this should be a black mark against those who 
failed to protect their voters.
      Some are still considering buying equipment.  We, the public, should 
demand that NO MORE contracts be signed that restrict the right of poll 
watchers to see to validation.
      I read of "an independent testing group".  I have no idea how 
thorough an effort they invest in.  Still, what I read of a hacker about 
10 years ago says the more eyes looking for trouble, the more likely all 
the serious problems get found:
           Hacker was copying password files - a big project back then. 
BUT - why such a waste of time - passwords were encoded and you could not 
reverse the encoding?  BUT - he had demonstrated enough smarts that such 
wasting of his time was not believable!  Turns out he could identify 
enough passwords for his needs without decoding - he made up a list of 
likely passwords and had his computer encode each of them - got enough 
matches to be useful.

As to extended capability:  Most of us are used to Plurality voting, where 
we vote ONLY for a single candidate.  Sometimes we add a separate rerun at 
a later date.  With computers better capability is possible and should be 
demanded, such as:
      Ranked voting, where we indicate first choice plus second, third, 
etc.  Voters can indicate their true desire, even if not expected to win, 
AND their preference among those with a serious chance.  Also means less 
enthusiasm for reruns.

As to paper backup ballots:  I am not enthused but, if something is done 
here, let the paper be the real ballot, rather than being called a backup, 
and pay attention to protecting secrecy of the ballot.

As to the internet:  Seems to me that a voting machine SHOULD NEVER be 
connected to the internet from the time preparation for opening the polls 
starts until the polls close and the results are documented.  Would not 
bother me to have it connected later for convenient reporting results to 
wherever, but this is simply reporting the already documented results.
      Reason for this is both to prevent an internet connection from 
destroying secrecy, and to prevent hacking via internet.

Diebold gets mentioned as one vendor with trouble.  Some then say all 
vendors are equally bad.  I question this, though I have no proof that 
Sequoia is better, beyond not reading the same kind of comlaints against 
them - and knowing they claim a 100 year history of learning to do it right..
-- 
davek at clarityconnect.com  http://www.clarityconnect.com/webpages3/davek
  Dave Ketchum   108 Halstead Ave, Owego, NY  13827-1708   607-687-5026
            Do to no one what you would not want done to you.
                  If you want peace, work for justice.