26 September 2005

Mail-in voting: more details; more evidence of major problems

Candidates and their representatives received a briefing today on how the mail-in ballots will be counted on Tuesday.

Here are some details:

1. Candidates are unable to scrutinize ballots as they would in any other election. Since ballots began arriving at City Hall on 12 September, city officials have already made their decisions about validity of specific ballots and included them or excluded them based on their own policy.

2. To date, about 700 ballot kits have been ruled invalid. These ballot kits either had no voter declaration, had one that wasn't signed, had multiple signatures on the declaration (presumably including a notary's signature attesting to the identity of the voter [!!!]) or some other technical failing.

3. "Spoiled" ballot kits are running at a rate 10 times higher than the provincial election despite a turn-out thus far that is one tenth of the total provincial votes cast. One tenth the votes; 10 times as many "spoils".

Of the approximately 25, 000 mail-in ballots cast thus far, almost 3% have been ruled "spoiled". In the last provincial general election, out of more than 278, 000 ballots cast only 790 were ruled "spoiled". That's .28%, compared to the 2.8% for the city election.

4. The actual number of rejected ballots (each kit actually contains four separate ballots for different races) is actually around 2, 800. The total number of possible ballots (four per voter) is 75,000 times four, or 300,000. Even if we allow this figure as being the number of "ballots" then the spoilage rate thus far is double the last federal election and three times the number of "spoils" in the last provincial election.

5. In the last federal general election, the number of rejected ballots for Newfoundland was 0.5%.

6. The St. John's municipal vote will set a new national record for spoiled ballots, once the actual number of spoiled ballots will be determined on 27 September.
All that has been determined right now is the number of kits that have been rejected as invalid. Within each ballot kit that will be counted on 27 September (accepted as valid by officials), individual ballot sections such as the one for mayor may be blank, double-voted, written on or otherwise spoiled using conventional definitions.

7. Ballot counting will begin shortly after 0800 hrs on the 27th, or 12 hours before the polls close. In every other election, ballot boxes are not opened until after the polls have closed. In this election, the election results will be known (except for the handful of ballot kits received on the official counting day) by around noon. Results will not be released until after 2000 hrs (8:00 PM) but city officials appeared to have no firm policy on this point.

8. Candidates have had no opportunity to challenge the credentials of voters, as provided in the Municipal Elections Act. If any candidate detects a problem on polling day, there is no way to determine which fraudulent ballot has been cast or how many such ballots may have been cast. City officials did not consider this to be an issue.

9. City officials have not been crossing names off their voting list, another way of detecting fraudulent voting in cases where ballots have been sent to the same person under two slightly different names.

10. Dominion Voting System Corporation [DVS Corp] has been contracted to count the ballots using their computerized counting system.
Likely problems don't stem from their system; problems come from the completely inadequate procedures established by city officials and approved by city council.