Re: Hand counting ballots

[Evan Daniel Ravitz <evan@xxxxxxxx>]

This is a fine idea

On Fri, 14 Nov 2003, Ralph Shnelvar wrote:

> Dear group:
>
> I promised Evan Ravitz that I would do this so here it is.
>
> A couple of weeks ago Evan proposed that all ballots be hand counted.  The
> group rejected this.  I was one of the people rejecting this.
>
>
>
>
> But I gave it some thought and the long-and-short of it is that Evan is
> right.
>
> What changed my mind? My experience being a poll watcher for a Boulder City
> Council candidate.
>
> I watched as ballots were being counted by the scanners and the information
> was pumped into a pc.
>
> Basically, as a person with almost 40 years computer programming experience,
> I know how easy it is to subvert a system.  With a scanner and a pc there
> are now two points of vulnerability.  Folks, it would not be hard to change
> a scanner so that a scanner sends the wrong information to a pc doing the
> counting.  Thus, one could have "proprietary, copyrighted, open source" and
> one still could have the election be rigged by "rigged hardware" (the
> scanner, among other things).
>
> As an example of how hard people work to defraud other people I offer the
> following example.
>
> Most of the participants on this list are too young to remember _mechanical_
> cash registers.  But I remember them and I remember a time when these
> mechanical monstrosities were manipulated so that when, say, your grocery
> bill was more than $23.45 that the _mechanical_ cash register would add an
> extra $0.25 to your bill.  We're talking gears and nuts and bolts here; not
> the infinitely easier to manipulate computer program.
>
> (By the way, that extra $0.25 raised the grocer's profit margin by a
> whopping 30%.)
>
>
>
> My proposal:
>
> The election process proceeds as follows.
>
> 1) "Touch-screen" computers generate printed ballots.  In addition, a direct
> count of the vote is kept by the computer (a so-called DRE computer).
> 2) Those printed ballots are inspected by the voter for correctness.
> 3) The ballot is deposited into a ballot box by the voter.
> 4) The ballots are scanned electronically by a scanner whose vendor is
> different from the vendor providing the computer printing the ballot.
>
> 5) On election day, the results of the count generated by the computer(s) in
> (1) and (4) are reported.  THESE RESULTS ARE PROVISIONAL and have no force
> of law.  The results merely serve to provide a quick - and provisional! -
> result of the election.
>
> 6) After election day the votes are counted by hand using age-old
> techniques.  This may take several weeks. It is this hand-count which is the
> legal basis for all elections.  If the election is close, then the votes are
> hand-counted again.
>
> 7) The result of the election count is now done in three different ways,
> First, the DRE computers generate a result.  Second, scanners generate a
> result.  Third, the hand-count generates a result.
>
> The purpose of this entire procedure is to increase the reliability of the
> voting systems and the confidence that voters have in the election results.
>
>
>
> The voters will get the desired quick election results (assuming that the
> machines are not rigged).  With three - we hope, independent - sources of
> vote counts, the voters will also get a sense that the machines are
> reliable.  In fact, with this procedure, the chances of two manufacturers as
> well as County Clerks colluding to defraud the electorate drops to the
> vanishing point.
>
>
> The only down side is that the voters will get a sense that the machines are
> reliable and will later insist that the necessary hand counting is
> unnecessary.
>
> Of course, it is the hand counting that will help guarantee that the
> machines are kept honest.
>
> The preservation of freedom requires eternal vigilance.
>
>
> Ralph Shnelvar
>