* *** POSTMORTEM_2010-05-04 TXT - 5 May 2010  15:26:15 - JKNAUTH

           Post Mortem on 5/4/10 Election at Precinct 01-43
           ==== ====== == ====== ======== == ======== =====

Here are some post mortem comments on the 5/4/10 primary election at
precinct 01-43.

1) When we tested the BOE phone line from our precinct at Monday setup,
   we found that the previously required 9 prefix no longer worked.
   Using a 9 prefix resulted in a recording about long distance charges.
   Not using 9 resulted in normal operation.  By using caller ID when
   calling from the BOE line to another phone, we found that the phone
   number assigned to the BOE line had changed from the number we had
   used in years past; that old number was still marked on a BOE label
   by the phone.  We were concerned that the M100 modem would not work
   because it was probably still programmed to use a 9 prefix.

   We called the BOE and found they were not aware of any change.  The
   church administrator said BOE had successfully tested the line a few
   weeks before.  Because of what we had found, the BOE send a
   technician to check out the setup.  He (Chris?) did a great job,
   first verifying what we found, then tracking down the phone room in
   the church complex (which covers many acres and many buildings, a
   number of which are being either renovated or are being newly
   constructed.  He finally found the phone room, determined that
   someone had switched the lines on the BOE phone, and then restored
   the correct configuration.  He then had us test it by calling his
   phone and showing the 9 prefix was required again (as it should be),
   plus dialing to the BOE line to show the old number had been restored
   to that line.  Later we got a call from the Wake BOE Office; they had
   us run another test, which worked.  Finally (the acid test), when the
   polls were closed, the M100 modem had no problems connecting.  It
   would have failed if Chris had not fixed the problem in the church's
   phone room.

2) At the Help Table class we were given a sample provisional envelope.
   It was special for primary elections in that it had a small section
   for the voter to specify the desired party for the vote.  However,
   all the provisional envelopes we were given in the Help Table bag did
   not have that section.  Instead they looked like normal general
   election envelopes.  We pointed this out to our coordinator.
   Fortunately we did not have many provisional envelopes to handle at
   our precinct.

3) I got no BOE response to my 4/11/10 comments, in particular for the
   REP/DEM/UNA/NP comment on manual page 24 (plus 17, 25, and xx).  See
   http://nbtha.org/WakeBOE/2010_Election_Manual_2010-04-11.txt.

   We had a lot of trouble getting the Registration Table people to do
   what was defined in the manual for this confusing subject -- they
   either marked the field when they shouldn't have, or didn't mark it
   when they should have, or marked it incorrectly.  There was also much
   voter confusion about the difference between an Unaffiliated party
   designation and Non-partisan ballots.  Much of that is caused by the
   erroneous terminology on the ATV.  See my 4/11/10 comments for
   details.  What has the NC State BOE said about this?  I clearly
   pointed out this problem years ago and have pointed it out a number
   of times since then.  It really should be fixed since it wastes a lot
   of time and leads to incorrect settings and unreliable data.

   We had the Ballot Table and Help Table people follow the correct
   procedure, i.e., if the Registration Table people had not handled the
   field, do what the manual specified for the Registration Table people
   to do (or we took the ATV back to the Registration Table and had them
   do what they should have done).  As my 4/11/10 comments noted, the
   manual is incorrect in this area for the Ballot Table.

4) One of our precinct officials sustained an injury the day before the
   election and was not available for precinct work.  It was too late to
   find a replacement.  The BOE was aware of the situation.  This caused
   us to be undermanned.  That together with the requirement to have two
   people work at the Ballot Table (and the need to work on the phone
   line problem) meant the Help Table was not manned full time.  That in
   turn meant some things which should have been handled at the Ballot
   Table had to be handled by a backup person.  A couple of voters did
   not get as good a service as they should have.

5) There was no stated requirement in this election to count a newly
   opened ballot package to be sure it contained exactly 100 ballots.
   Based on our past bad experience in this area, we decided a count was
   required.  Of the five packages opened, three were bad.  The counts
   were carefully verified and crosschecked by two people.  Here were
   the results:

        N-13 #1: One short
        D-01 #1: OK
        R-05 #1: OK
        D-01 #2: One extra
        R-05 #2: One short

   If this is typical (and past experience seems to indicate it is), I'm
   surprised many precincts (who don't do package-check counts) are not
   having reconciliation problems.  Or maybe they are and we just
   haven't been told.

   With the accurate package contents counts used, our reconciliation
   showed the correct totals.  As a further check, after poll closing I
   looked at the M100 tape which reports the number of each ballot style
   used.  That exactly matched the predictions: accurately counted
   ballots when package was opened - unused - spoiled - provisional =
   M100 report for that ballot style.  (We had no challenged or
   emergency bin ballots.)

6) We had one voter problem.  Several minutes after a lady had been
   given a ballot, had marked it, had then inserted it in the M100, and
   had exited the building, she came back and insisted she thought we
   had given her the wrong ballot:  Democratic instead of Republican.
   Since we had properly followed the Ballot Table procedure (two
   people, visual checks, and hand scanner checks) and none had seen a
   problem, we didn't see how this could have occurred.

   We very politely and carefully went over our procedures with the
   voter regarding her concerns and thanked her for her input, but she
   remained unconvinced.

   Shortly after that incident we did a hand-count of the remaining
   unused Democratic ballots.  The intent was to compare the number of
   used Democratic ballots vs. what the ATVs said had been distributed.
   We had a high confidence in the used ballot count since we had
   carefully counted each package contents when we opened each package.
   In fact the calculated used Democratic ballot count exactly matched
   the ATV count (adjusted for the known spoiled and provisional usage).
   This check, done about midday, confirmed that no mistake had been
   made, i.e., that no Democratic ballot had been issued when the ATV
   said a Republican ballot should have been issued.

7) The new requirement to return election material to the BOE in
   downtown Raleigh on election night caused it to be a VERY LONG day
   for our Chief Judge.  There was a long line of cars at the dropoff
   site.  She did not get home until after 10 PM.  The BOE really needs
   to improve this procedure.

Jeff Knauth
Precinct 01-43