As CTVotersCount readers may recall, last September I blogged on a morning spent at a recanvass in Cromwell. Yesterday I observed a recanvass in Hartford. It was a very educational day for me. Hopefully my participation was good for the candidates, the officials involved, and election integrity.
Let us start with some basics. Connecticut has a recanvass after a close election. People, including yours truly, often refer to the recanvass as a recount. It is quite different. See Recanvass Procedure Manual from the Secretary of the State’s Office and our post, Minnesota Recount vs. Connecticut Recanvass. And as Ted Bromley from the Secretary of the State’s Office reminded me during the recanvass: A recanvass is not the same as a post-election audit. There are different purposes and different rules.
So that no misimpressions are left, let me say that based on my observations and discussions everyone there was working hard to do a good job. And I have no reason to question the integrity of the recanvass. During the course of the recanvass I had several discussions with the Registrar, Olga Iris Vaquez, Tony the Head Moderator, Ted Bromley, and Lucien Pawlak from the Secretary of the State’s Office. I often brought up points of disagreement or concern. Most ended in agreement with my opinion or my education to their opinion. In every case discussions remained cordial with gracious acceptance of officials when they had to change or redo work and my acceptance of their authority when differing opinions remained.
At the end, I was left with several areas where I would want to change the law/regulations/procedures to increase the integrity and transparency of the recanvass process. I also have a few suggestions for election officials as well.
I am also left with a new appreciation that one of the primary benefits of a recount, recanvass, or audit is for the loosing candidates. A properly conduced recanvass that results in the loosers accepting the result and moving on is an important benefit – even if the election result is changed and the winners and loosers change.
Background and overview of the day
The election was between two slates of candidates for the Democratic Town Committee in Hartford. The election covered four voting districts and involved two slates of candidates with ten candidates each, slate “B” and slate “C”. Hartford is a one party town, so the Democratic Primary often determines the Town Council, Mayor, and State Legislative elections, while the Democratic Town Committee determines the endorsed candidates. So these elections are perhaps more important than the general election, and about as important as the primary.
To my knowledge, I had only met four people in the room prior to the recanvass: Ted and Lucien from the Secretary of the State’s Office and two other citizens who came to watch and support voting integrity. I knew none of the candidates on either slate. My understanding was that slate “B” was the challenge slate and the original election result left them with one winner and nine loosers – with three votes separating the highest looser from the lowest winner. The election was less that 400 or so voters and ballots. (I have a theory: the more candidates and the more that can be voted for in a race, the more likely two candidates in the middle will be close and a recanvass will be mandated)
A recanvass is not a manual count of ballots – it is a slightly improved machine recount. A recanvass is not a thorough, adversarial review of ballots like the recounts we are familiar with from Florida in 2000 and Minnesota in 2008 – it does involve an assessment of voter intent by officials – as I understand it, the public, parties, and candidates have no standing to object to procedures and official judgments. And the ability to observe the process is limited.
The recanvass was scheduled for 9:00am. I arrived just a bit early and met or was introduced to the Registrar and several of the members of slate “B” and their supporters who were present. The Registrar indicated that they were waiting for representatives from the Secretary of the State’s office to arrive before starting. I think the representatives Ted and Lucien arrived about 9:30 and left sometime before 11:00. The audit concluded at about 1:00pm. In addition to the Registrar and Moderator, I counted seven other election officials.
The room was the Town Council meeting room in town hall. The council sits in a raised semi-circle. There is a wooden rail separating the front of the semi-circle from the public seating, which is perhaps a dozen feet from the rail. Most of the activity occurred behind the rail. There were two optical scanner boxes – one was used for the absentee ballot optical scanner and the other for the four district optical scanners, in sequence. Some counting activity occurred on the council table, other activity on a table between the rail and the public seating.
The Moderator conducted the recanvass and did all the work associated with testing, feeding ballots through, and printing tapes for the four district optical scanners. Another official supervised the absentee optical scanner and fed ballots with the help of two other officials. Four additional officials did the manual ballot counting at the front table and two of them reviewed ballots for questionable markings at a table behind the rail.
Even though the officials said they were waiting for the Secretary of the State’s staff, it suddenly dawned on me that the four at the front table were opening envelopes and counting ballots – yet I had not noticed unsealing of ballot bags etc. I decided to observe what they were doing. This resulted in the first discussion and controversy. The Moderator asked me to observe from the seats – that I could not stand there. After some discussion with the “B” slate, they appointed one candidate as an observer – the procedures say:
each such [party] chairman may send two representatives to be present at such recanvass. Such representatives may observe, but no one other than a recanvass official may take part in the recanvass. If any irregularity in the recanvass procedure is noted by such a representative, he shall be permitted to present evidence of such irregularity in any contest relating to the election.
Everyone seems to interpret that similar standing applies to slates or independent candidates. Yet, it is quite ambiguous what exactly “Such representatives may observe” actually means.
As for voters and other members of the public, the procedures say:
All recanvass procedures shall be open to public observation, however, no member of the public can actively participate or interfere with the conduct of the recanvass.
Again, the meanings of “observe” and “interfere” are open to wide interpretation.
After further discussion, slate “B” asked me to be one of their two observers and they conveyed that to the Registrar and Moderator. I attempted to observe the four counters, Tony did not want me so close. I wanted to be close enough to actually see the ballots and forms the counters were using, as observers are allowed in post-election audits. I wanted to walk around the table and get a view from the position and angle I chose, also as observers are allowed to do in audits. Tony insisted I stay along the rail which allowed me to get, perhaps four feet from the backs of the counters on one side of the table – I could observe some parts of the ballots but not everything, and not all that clearly. I agreed to disagree and accept his authority.
Soon I realized that the team was not counting cast ballots. They were counting envelopes of spoiled ballots – which apparently were not sealed in carrying cases – just in envelopes with unnumbered “tamper-evident” seals. (As readers know I have concerns about the chain of custody of ballots, one of the least secure ways of protecting ballots is sealing in envelops with unnumbered “tamper-evident” seals – just create a new envelope labeled like the original and apply a new unnumbered seal.)
After Tony had read the first district through the scanner, I pointed out that the procedures called for two opposing officials (not slate representatives) review each ballot for questionable marks before scanning. Tony felt that because the machine had read all the ballots originally they were all ok – an assumption also held by several registrars in each post-election audit. It took quite a bit of discussion involving me, Tony, the Registrar, and Ted before everyone agreed to follow the process in the manual. Tony completely accepted the need to redo the entire process for the district with a new memory card based on the manual requirement. I was impressed with the diligence of the two officials in carefully reviewing each ballot. In each of the four districts there were about a handful of ballots set aside for hand counting. I think going through the process Tony understood the reason that a properly functioning machine may not accurately count voters’ intent.
As in most observations, over time, officials became comfortable with observers and tended to not object when I observed more closely. I moved around and closer to the officials as they counted the questionably marked ballots a the table in front of the rail. I noted two or three ballots where there assessment of voter intent on the ballot was that it was an overvote – there were were too many bubbles filled in, and likely the scanner missed one or two because of the way the ballot was filled out. I noted one of the ballots had markings that were highly inconsistent between the bubbles – they counted it as an overvote – I suppose it could be argued in court that some of the marks should not count and it was actually not an overvote – I’m not sure how I would come down – I would probably agree with the officials decision to treat it as an undervote.
At the end of the day, from unofficial calculations, most of slate “C” lost from one to three votes. This was a learning/confirming experience for me. One of my criticism’s of the post-election audit process and reporting requirements is that it only anticipates positive ‘questionable votes” that might not have been read by the scanner. I have contended that there are also negative such votes and the lack of recognition of that possibility causes some of frequent officials’ confusion in accounting for such votes in the audits. Here was dramatic proof that such concerns are real, and in this election those represented the major difference in the count. (Presumably the ballots were read by the machine, the machine did not count an overvote, so it did not classify some of the marks as votes. But the voter clearly overvoted, resulting in an actual count of -1 for several candidates.
Here I also had a short debate with Tony. He had the counters “spoil” the overvote ballots and put them in the spoiled ballot envelope. I tried to reason with Tony and offered to provide an extended explanation, yet he insisted and I quickly yielded to his authority.
Here is the problem: Voters are allowed to overvote and undervote by their choice. These ballots actually represent the wishes of a voter who voted and is checked off. Putting them in the spoiled envelope has several potential problems. 1) If a later recount or court action is required, people can be confused why the counts are off. 2) Even if they remember the change, there is no record of the change -which district, how many ballots etc. 3) And its only the judgment of the officials that makes the ballots overvotes – a court review might change the assessment. But with such ballots left unidentified, there is no way to find and evaluate them. 4) Even if the officials were correct, without the ballot there is room for charges, counter charges, and bad feelings on the part of loosing candidates.
When the third district questionables were being counted I noticed another issue. One ballot had the apparent signature of the voter writ large on the ballot. It only had one bubble filled in, for one candidate on slate “C”. I noted the candidate and the exception. Even though Connecticut law would exclude such a vote where the voter might be identified, the recanvass manual makes no mention that I know, that officials need to take such identifying marks into account.
This is my third recanvass as an observer. The first was a lever machine recanvass for the 2006 Courtney/Simmons race for the U.S. Congress. I went to a mass training given by National Democratic Party lawyers – they interpreted Connecticut law to be that observers could note issues that could later be used to dispute results in court and cause an actual recount. This was just one of those types of issues.
I did not know why the officials pulled that ballot for hand counting. Maybe because of the signature, maybe not. Could I have missed more by not observing the officials as they reviewed ballots for questionable marks?
I moved to the table behind the rail to observer the officials reviewing ballots for the fourth, and last, district to be counted. To my surprise the officials tried to stop me, and then insisted that I should have an escort everywhere I observered, (a requirement not mentioned or in effect earlier). I argued and also mentioned that that was their responsibility if they insisted, not mine. The officials complained that they could not do their work because I was bothering them – my interpretation was that it was the other officials insisting on a escort and questioning if I could observe from behind the rail that was disrupting them. In any case, I observed where I wanted to, they supplied no escort and everything went smoothly thereafter. I observed no voter identifiable ballots for the fourth district, yet the question of the first three districts and all the absentee ballots remains open.
I photographed the machine tapes posted on the wall. The officials agreed to let me photograph all the hand count hash sheets and later the single sheet providing totals for all the candidates in each district. The totals for each district were read aloud. Candidates on slate “B” and other members of the public added the four districts together and saw from their unofficial totals that the election results would hold. Tony left with four sheets to add them officially on a spreadsheet, to file with the town clerk. The slate would get a copy later, and I headed for a bus home.
As far as I know, slate “C” had no representatives present. If I were on slate “C” I would make sure we were represented.
Some unsealings and sealings involved two officials checking the seal numbers and public announcements, others involved a single person and were not all that apparent. Much of the process was not initially obvious to the public. I had to fdetermine what what was going on by observation or asking the leaders.
What I would change in the law/regulations/procedures:
I would rather have a recount like the ones in Minnesota and elsewhere:
- Manual, hand counting
- Reviewing each ballot for voter identification
- Adversarial review of all ballots, and absentee ballot materials
Failing, that and continuing the spirit of the current recanvass:
- I would allow several observers per slate/independent candidate. One more than enough to have one per concurrent activity: Machines being setup/used, counting teams, questionable mark reviewing, tallying area etc,
- I would allow public observers to the extent there were insufficient slate/independent observers, selected by lot if necessary.
- I would allow observers to go anywhere and see anything just like we can based on procedures in audits. And get reasonably close, to actually see well. Standing just behind any person should be fine.
- I would have the votes with questionable marks include any stray marks that might identify an individual – and have the counters determine and reject such ballots.
In summary, I would like to have legitimate, reasonable things specified in the law/regulations/procedures, rather than have nagging questions of accuracy/integrity, and rely on legal action to establish/enforce such things that seem to be common sense.
Suggestions for officials conducting recanvasses:
- Go beyond the procedures and allow everything suggested above
- Invite reasonable levels of observation and transparency
- Don’t set the room up in a way that separates officials from observers or makes in difficult to observe
- Describe or at least announce each part of the process before it is done
- Invite observers to see seals and verify numbers for themselves
- Train officials to expect and to be comfortable with reasonable, close observation
