Random Drawing

Updated: 9/14 and 9/16

Yesterday I attended and participated in the public random drawing of districts for audit. Hopefully, in the next day or two CT-N will put up the video. I did manage to catch some of the rerun late last night. (Note: I will keep updating this entry as more information and the video becomes available.)

Read the Good News, The Not So Good News, and What We We Can Learn, below:

The Hartford Courant article: <read>

The good news:

It was pleasing and thrilling to be a part of the Transparent, Public, Open, Random drawing. The Secretary of the State and her office are to be commended for the event and the commitment behind it to fair elections and following the law. During the process it was realized that following the letter of the law required that eleven districts rather than ten be selected – this was handled quickly and smoothly without hesitation.

The Secretary expressed a desire to increase the level of audit. As summarized in the Courant:

Voting rights advocates took turns drawing the names in a ceremony in Bysiewicz’s office, even though one, Luther Weeks, complained that the audits should be more extensive.

Bysiewicz agreed, saying she had sought audits in 20 percent of the polling places, a requirement deemed too expensive by municipalities.

Both my comments and Secretary Bysiewicz’ comments were a bit more extensive. Unless and until the CT-N replay is available let me quote my remarks as well as I can recall from the press conference:

This process today is great, transparent and open! The whole process needs to be transparent and open including the programming of the memory cards. We need stronger audits. By my calculations in a municipal election in smaller municipalities or a state house race the probability of catching errors or fraud is in the range of 2%-4%.

7 of 23 municipalities will be audited: Hartford (3 districts); New Haven (1) ; Derby (1); Bridgeport (3); East Hartford (1); Cornwall (1); and Cromwell (1).

Because of close races Hamden, Hampton, Lisbon, and Plainville will have a full recount under the law.

Because of multiple races, municipal clerks in the following towns will need to randomly select races for audit: Hartford and Bridgeport.

When the press conference started Bridgeport had not fully reported results. This would have been a problem, however later in the day Deputy Secretary of the State Lesley Mara e-mailed me that Bridgeport had actually reported at about 11:18 – during the press conference but before the actual drawing.

The not so good news:

12 of 23 towns will have no audit and have races that will not be recounted.

Eight districts will be exempt from the audit via the recount loophole: West Haven (2); New Haven (2); Bridgeport (4);

I count 41 offices (there seems to be an ambiguity in the law, I’ll explain later) and of these, I count 14 offices that will be audited or recounted, half by automatic recount and half by the random audit.

The level of audit remains inadequate. The probability of detecting error and deterring fraud is too low. The law allows too much leeway in interpreting audit results.

Update: 9/16: The audits cannot start until September 26th by the law 07-197 which is way out of line with the suggestion of the Brennan Center and others that they should be prompt to prevent mucking with the results to make the paper match the machine. Much worse than that is that the districts to audit were chosen well in advance of the audit. There are at least three problems:

  • Knowing which districts will be audited gives at least thirteen days for mucking with the results in districts that will not be audited. Gives the same thirteen days to have the paper match the machines where there will be an audit, where the results have been previously fixed.
  • Gives time for friendly candidates to contest races which will eliminate the audit and provide cover for mucked with results.
  • May provide a headache for the SOTS because the law requires the random drawing of another district in the case of any contest covering one of the previously selected districts.

What we can learn:

Most of the recounts were in town council races. Such races with multiple candidates are, by their nature, much more likely to have two candidates separated by 0.5%, causing a mandatory recount which invokes the loophole. So, to add to the already woefully inadequate audits for municipal races we have this additional statistical reality that more will be exempt from the audit. Worse it suggests several creative ways that software can be used to alter votes taking specific advantage of this loophole.

I want to determine the time and place of the municipal clerk’s random drawing of races to be audited, and if the event is open to the public:

  • Hartford – I spoke to a staff member, she suggested I leave a voice mail to the Town Clerk, which I did.2:
    Update: 9/14 : mid afternoon 9/14 Town Clerk, Daniel M. Carey returned my call. He was unaware of the responsiblity until I discussed it with him. He said he would likely determine the details of the random selection on Monday and I could call him back late Monday or early Tuesday.
  • Bridgeport – I spoke with the Town Clerk, Hector Diaz, he said he had just gotten in to the office and had not yet heard that Bridgeport had been selected for random audits. He asked that I call back another time. Both he and a staff person in the City Clerk’s office told me that a room had been booked by the Registrar for the recounts on Monday Sept 17th, noon to 8:00pm.

Update: 9/16: The municipal clerks have more time than I thought to figure this out. See the remarks above RE: the audits cannot start until Sept 26th. The law needs to be changed to have them start as soon as possible after the election and after the drawing by the SOTS. Eliminating the loophole for recounts would likely eliminate the only possible reason for the fifteen day delay. This proves once again, that many heads reading the law are better than one (my one at least) and it would be a lot better if the legislature and the SOTS involved others in writing the law, earlier and more often.
An accurate reading of the law indicates that the drawing by the municipal clerk needs to be random. Unlike the district drawing by the Secretary of the State it does not specify that the drawing must be public. They should be public. The law needs to be changed to assure that.

There is some ambiguity in my mind what drawing an office to be audited means. I had always assumed it was for example Mayor, Town Clerk, Town Council, President, Senator, State Representative etc. But what about towns that have multiple council races in multiple districts? Or the state which has multiple U.S. House members, State Senators, and State Representatives? Is the Secretary of the State actually required to select a generic office or could, within the law, all individual races involved be considered as a separate offices? In this municipal election would each council race in Bridgeport be considered a different office? Also can the municipal clerk do a single drawing city wide or choose to do a separate drawing for each district selected in one town?

The law allows the drawing to proceed before all results are in. The problem is that once it is known that a district will be audited or not, the results can be changed to be accurate or changed to be inaccurate based on the selection or non-selection of a district. In an informal discussion with members of the Secretary of the State’s Office after the press conference it was clear that they understand the implications. The drawing could be scheduled a little later, however, it is important to hold drawings and audits soon after the election to allow less time for tampering. If a district is late in reporting it should be subject to a full recount of all races. Late reporting indicates a likely problem with the voting equipment or possibly ongoing tampering or trying to reconcile incorrect results – in any of these cases a full manual recount of the paper is very advisable along with close independent scrutiny of the circumstances and the voting equipment.

Also, in an informal discussion with members of the Secretary of the State’s Office after the press conference, they requested feedback on the random drawing process. The asked what I thought of the cards and using the historic oak box rumored to be made of the Charter Oak. I felt the process met the requirments of the law but had a couple of suggestons: The cards were difficult to shuffle and would be quite a challenge if we there were anything close to 769. It would also reach the limit of a pile within the oak box. I suggested smaller cards, perhaps the size of playing cards – actually since the cards each had stickers with the districts printed, actual playing cards with stickers could be used. I also suggested that a barrel with a crank, as many of us have seen in raffles, would provide a more convenient way of mixing up the cards.

Facebooktwittergoogle_plusredditpinterestlinkedintumblrmailFacebooktwittergoogle_plusredditpinterestlinkedintumblrmail

Leave a Reply

You must be logged in to post a comment.