A Day As Absentee Ballot Moderator

(Draft subject to improvement)

This election day was spent as moderator for central counting of absentee ballots, once again, in Vernon, CT.  In Connecticut absentee ballots can be delivered to the district (precinct) and counted there or counted centrally for each town.  Most towns with more than a very few polling places count them centrally, with a separate counting team and an Absentee Ballot Moderator.

I tend to learn by reading and doing.  In 2008, I became a Certified Moderator and served my 1st time as an election official.  As I said at the time:

I hear and I forget.
I see and I remember.
I do and I understand. – Confucius

One of my goals this year was to obtain more hands-on experience in the election process by serving as an election official. In August I attended moderator training. I became a certified moderator which, in no way qualified me to actually be a moderator (I would describe a moderator as the czar of a polling place). Reading the manual and taking the training would never be enough for me to grasp such a job, without perhaps several years experience as a poll worker. Thanks to Judi Beaudreau, Moderator Trainer, and Registrar of Voters, Vernon, CT, I worked one of the polls in Vernon this year. I spent about half my time as a ballot clerk and about half as machine tender.

And as I concluded last year:

PS: Tired, but interested in election results, I went to my local party headquarters to follow the election results. One race had a margin of 300 votes and everyone wanted to wait for 2800 absentee ballots to be counted to be sure of the result. The absentee results were predicted to be available at the earliest 3:00am. Few present knew we scanned absentee ballots centrally, they assumed officials were counting 2800 ballots by hand. I wondered: “How can it take that long? Especially when counting and sorting starts at noon? Even if its done by hand, how can it possibly take that long?” I heard some possible issues that I had not considered, but nothing that would convince me it should take that long unless its a very very small counting team. An area for me to study and observe further. Perhaps in the next election some town in Connecticut would let me participate as an official in their central count absentee operation on election day. I know I would learn something.

Be careful what you ask for.  Emailing Judi a couple of reminders that I was interested in working on absentee ballots, so that I could really understand the process, she sent back the “Procedure Manual For Counting Absentee Ballots” and said they were considering hiring me as absentee ballot moderator.  The immediate thought: “Decline, I’m not ready. This time I want to be a counter and maybe next time be a moderator”, but then “I can probably do it, who knows if I would get another chance.  It is not the 18 hour day of a polling place moderator, we start later.  Its not the same as a polling place with real time decisions as voters arrive”.  And not knowing the rates in Vernon, I might make a small fortune.

In the 10 days before the election I read the manual completely, twice.  Judi says it needs revisions and updating, it was written in 2007 just as Connecticut was changing from levers to optical scanners.  It is quite detailed, yet it gives little guidance on the actual hand counting of ballots – but then I realize that almost all absentee ballots are counted by scanner so we are likely to only have a very few ballots to count by hand.

I arrive 15 minutes early.  I note that the office is staffed by young officials, presumably some experienced graduates of the special program allowing 16 and 17 year old’s participate as election officials – Vernon takes full advantage of the program.  Two phone lines are ringing constantly.  The officials answer the calls and perhaps 1/3rd of time consult the registrars.  I recall the many calls I saw last year, from the moderator of my polling place to headquarters.  Many of the calls are checking voter registration, some of which are required by our election system.  It is busy, yet calm.

At 1:00 pm,  Judi swears me in, goes over the procedures and the paper work, and the pre-printed forms with me.  Instead of a single central count optical scanner and memory card, they use one memory card per district starting and closing two optical scanners a total of six times, once for each district.  I will be very familiar with the opening and closing procedures by the end of today.  Judi explains the advantages of this scheme being lower cost for memory card programming, keeping the ballots per district segregated, and that an incorrectly issued ballot would not be  incorrectly counted.  (An incorrectly issued ballot occurs when a ballot for the wrong district is sent to the voter – a separate memory card would detect it, but a combined one would count it for the wrong district, including votes for candidates the voter was not eligible to vote for.  Yet, in Vernon, in this election, there is one ballot for all six districts so the clerk could not mix them up, on the other hand, the machine could not count them by district either.  So the towns using a central count, single card arrangement must have different ballots even if they all have the same contests across the board?)  We will be counting in the town council chambers on the third floor.

The Registrars, Judi and Patricia Noblet, officially deliver the ballots to me, which I count and accept for each of the six districts.  282 ballots in double envelops – easy to count and separate, they do not stick together.  But if there were a much larger number I would need some counters to help.

I distribute the ballots, pre-printed forms, ballot envelopes, and memory cards (sealed in see-through pouches) in six locations around the town council tables.

At 2:15 pm the four election officials arrive who will be the counters, process the envelopes and ballots.  I swear them in.  Judi does the training and provides a one-page list of steps for the processing and counting.  I suggest and decide that we will do the process by performing the steps that logically go together for all districts and then the next set of steps for all districts etc. so that we can concentrate on one part of the process at a time. Judi says to call her when we are ready to start scanning.

The old hand-count system counted absentee ballots in batches for each district as they come in from the post office  – the manual still expects this process.  We will process the envelopes, review each ballot to determine if it should be hand-counted, hand-count those that need to be, but leave final counts open and report everything at the end as one batch.  Because we are using separate memory cards, we won’t scan any ballots until we are sure we have almost all the absentee ballots.  If any come in after we scan a district, then they would have to be manually counted.

Judi explains that we should hand-count all ballots where we think there is any question that the machine would read it correctly or incorrectly.   They have chosen to not override on the scanner to count ballots with overvotes, so we should hand count any ballots with overvotes.  If the machine kicks out any overvotes that we mis then those would be hand counted.

If there are any questions about votes or envelopes etc., the counters bring it to my attention.  As moderator I make the “final” decision and write it on the envolope or ballot.  (I realize that accepting an envelope is final, since once the ballot is separated from the envolope, it cannot be changed.  If I make a decision on a ballot, it could be changed or challenged in a recount or recanvass).

We check each outer envelope against the list of voters – they must have been marked as voting absentee on the list – they all match.

In Connecticut we don’t validate signatures on envelopes, but they must be signed.  But because the law is very specific that we must check to see that the name of the voter on the inner and outer envelopes must be for the correct voter, we need to do what we can to see if the name is different.  (e.g. if a husband and wife somehow put their votes in the wrong envelopes, then the votes do not count.  If both their votes are in one outer envelope, only the one that matches the outer envelope counts).

The counters count the inner and outer envelopes.  When the count does not match what I counted on the delivery from the registrars, I don’t tell them the original count – I just tell them it does not match and they have to count again – it all balance after no more than two counts for each pile.

Over the course of processing, I reject perhaps 6-8 ballots.  Most are straight-forward decisions because the ballot was in the outside envelope but not in the internal envelope. (A little surprising, the frequency of this particular error).  Three interesting decisions.  One ballot has a different first and last name on the inside envelope and the hand written return address, than the official voter name printed on the outside envelope – rejected.  One inside envelope says that it needs to be hand counted as there are corrections – I rule it has to be mixed with other ballots and we need to make an independent decision on how to count it – otherwise we could know which ballot corresponded to a particular voter.  The counters show me a ballot with some votes crossed out with a ‘NO’ and others substituted – it would have not have been problem, but the voter initialed the ‘NO’ – I ruled it could not be counted because it had identifying marks. (Connecticut is a “Voter Intent State”). There were a couple of other rather unique ballots, but since all the marks had to do with canceling some votes, according to my interpretation of the law, they should be counted.

We could have easily missed an overvote and left it to be machine counted.  The counters went back and reviewed some ballots again and found some overvotes they missed.  In the end we hand counted a dozen or so votes by hand and none of those to be scanned were rejected by the scanners.

At about 5:45 pm, the registrars  arrive with four more absentee ballots.  They say that that is it!  We can now go full speed ahead to finish the job.  We process the envelopes and then the ballots.  The counters hand count all the ballots that we have identified as needing hand counting, by hashmarking on a pre-printed sheet for each district.  The counters ask me to make some decisions on the hand counting of individual votes on the ballot.  I call Judi to say we are ready to scan.

Judi arrives and talks me through setting up and opening the first district on each scanner and their closing.  I was going to go through the steps in the manual, but instead followed Judi’s direction – having done it once for real, I can do all the rest without problems or checking the manual.  I find it much simpler actually doing the steps than it seems when reading about them.  Judi says to call her once we are ready to complete all of the paperwork.

We bring all the ballots to be read by the machine and the memory card pouch to each machine, deal with it and them move the district back to the tables.  Before each district is scanned we print three zero tapes.  After all ballots are scanned we print three totals tapes and empty the ballot bins.  On the second district counted, at first we see no ballots.   We pull out the scanner and open the top and find they are all stopped by a small ridge – but, because it is about 50 ballots, there is no jam.  We check that ridge from below for every subsequent district.

The smallest part of the process is the scanning which is accomplished by two officials at each scanner.  The critical factor is my time – the moderator must setup the machine, unseal and reseal memory cards, print and sign six tapes for each district (the printers are slow, especially considering the three copies).  I find myself going back and forth frequently half-way across the room between the two machines and the waiting officials.  Next time I would put the machines closer, but not too close to risk districts being mixed up. (Here is where my personality comes into play – I tend to rush to keep the counters busy – Here is where I need to be especially sure that I am deliberate).

Why three copies of the tapes?  One is the copy for the moderator’s report – the last zero tape copy is is connected to the first closing copy with the vote counts.  Another copy is posted on the wall.  Another copy goes into the bag with the memory card.

At one point in the process one team points out that I printed four copies not three for one district.  I crunch that one up.  Later in looking at the wall that same team asks me why there are three zero tapes for their districts but only one count tape.   One has yet to be printed out – the other is the crumpled tape which we unravel and sign.  The final tapes are posted on the wall after 8:00, for public observers who are allowed to observe the counting, but not close enough to determine any results before the polls close.  Since we are counting with scanners and so few votes are counted by hand, it would be easy to prevent anyone from knowing the results.  Nobody from the public comes to watch.  As far as I can tell nobody comes to see the tapes to report results to the party.

I find the memory card bags very interesting.  They are tough clear plastic, perhaps 7″x9″.  They have a slot for a paper card for recording seal numbers, my name, date etc.  They have a very interesting zipper and plastic seal arrangement that seems a bit better than the usual plastic seal – just a bit of a challenge to remove two little plastic parts left after breaking the seal that must be removed before applying the new seal.  In addition to the memory card, they contain the pre-election test tape.  I add one copy of the zero tape and one copy of the count tape.  I pick a new seal, write its number on the paper card, place the paper card in the slot inside the bag – visible outside but sealed in.  I close the zipper and apply the seal.

Oooooooooops.  We close the last district scammer.  I pack the scanners in their bags.  Opening the top of one of the machines, it has four ballots stuck in a ridge atop the write-in bin.  We processed write-ins through the machine as there were no valid write-in candidates (votes double votes for a candidate by the usual bubble and/or write-in were hand counted).  But on that one machine, apparently the 1st write-in ballot hung up and the others ended up on top of it.  (Next time I will open the top of the ballot box every time, after each district).

Here is where being logical and surfacing, not hiding problems is important.  Since all ballots are the same in every district it is questionable which district the ballots belong with. Reviewing the four write-in ballots, I see that two ballots only have write-ins for the town council and the other two only for the board of education.  The tapes show the town council write-ins are all in one district so those two ballots can be filed with that district.  Unfortunately, the write-ins for board of education are split between two other districts and the vote counts on the rest of the two ballots are different, so we cannot classify them into a district. We keep them separate and I write a note explaining the problem.  If there is a recanvass or a recount then the totals will need to be combined for the two districts for the purpose of comparison.  Later, Judi will have us rubber band all the ballot envelopes for each district, and rubber band the two districts together with the two write-in ballots between.

The next morning I learn that there will be a racanvass based on a close margin between the lowest winning town council seat and the highest losing seat.  I rest peacefully believing that I handled this correctly.  (But I learn the newspaper reports were wrong – Judi emails me that the candidates are farther apart than initially reported in the paper – there will be no recanvass) .  Yet, part of me knows its easy to make a logical error in handling a situation like this, especially for others with less focus on the details of vote counting or with less of a logical bent – and its always possible that there is some mistake that one of us made that we have yet, if ever, to realize or be discoverd.  There is a lot to get right and a lot to coordinate for the moderator – with the toughest, most detailed part at the end of the night.  Starting at 1:00pm rather than 5:00 am, I am more rested than any polling place moderator.

I call Judi on her offer to help with the paperwork.  For each district I read the hand counts and the tape, while she fills in the form and adds the two numbers.  She reminds me to let the counters do the rest of the work, sealing  and signing envelopes and bring us one district at a time.  In the process there are some envelopes mixed up by district and we have to straighten that out.  I also do the paper work by district reconciling the absentee ballot envelopes delivered with the ballot count by hand, plus machine count, plus rejected – I am a little surprised and greatly relieved that for each of the six districts, everything balances perfectly the 1st time.  After some of districts are complete Judi calls in one of the young officials to replace her, to help me complete the rest of the districts.

I also do the paper work reconciling the envelopes delivered with the ballot count by hand, plus machine count, plus rejected ballots by district – I am a little surprised that for each of the six districts, everything balances perfectly the 1st time.  This is where every step done correctly, every bit of organizing by district and keeping things straight paysoff.  If anything is out of whack then it is a lot of recounting by hand, just when everyone is tired and anxous to go home – the last thing you want to do is have to recount and reconcile when you are tired and think you are done.  This is where experience is a great, yet very expensive teacher – I am lucky to have Judi available for my 1st time – this would be unusual for a polling place moderator except in an a single district town.

We pack everyting up and head downstairs to the registrars’ office.  Judi reads the counts and a young official types them into a spreadsheet.  He prints the page.  I read the totals from the paper work again and another young official checks the printed sheet.  We find one error of one vote which is quickly corrected in the spreadsheet.

What would I do differently next time?  I would pull the scanner and check the bins from the top after each district was scanned.  I would make up the envelopes for each district ahead of time (Judi gave me pre-printed sheets to attach to the envelopes).  I would keep all the totals for each envelope on stickies on the labels until we were sure we had the final counts, so that I would not have to replace the labels.  I would have a talk with my team at the beginning and explain two things:  1) The importance of keeping each district completely separate and maintaining that separation without fail – we had a couple of close calls;  2) The potential for boredom, especially at the end – there are many critical steps for the moderator that require concentration and the rest of the team may have nothing to do.  Next time I would consider having the counters do the forms and totaling while I supervised and later re-checked their work. Having done the process once, with help from Judi, it would be much easier for me next time.

My official role was complete. As I left, I believe, I overheard that the first polling place moderator was arriving with results, ballots and paperwork  for checking.

PS: My next stop was party headquarters in my own town about twenty minutes away.  When I arrived they had all the results available, except the absentee results.  They likely got the results from each polling place before the moderator packed up and left for town hall.  But where were the absentee ballot results?  A couple of the races were close.  The results came in at about 10:30 pm.  Soon, one of the counters from the absentee ballot team arrived.  I asked her what happened? Why did it take so long?  She said they did not get the final absentee ballots until very close to 8:00 pm and did not start scanning until shortly after 8:00 pm.  That pretty much explained it all to me – why it took until 10:30 pm.  They did nine districts in perhaps 1/2 hour less than we did in Vernon.  Now, I am left wondering where the process is differnent in the two towns between the post offices, town clerks, and registrars in the timing of the availablity of the final set of absentee ballots?  In the grand scheme of things, 10:30 pm is not that long to wait for final results.  Especially knowing the team is one mixup or transcription error away from quite a bit of delay and extra work at the end of the day.

Facebooktwittergoogle_plusredditpinterestlinkedintumblrmailFacebooktwittergoogle_plusredditpinterestlinkedintumblrmail

Leave a Reply

You must be logged in to post a comment.