POSSIBLY IMPORTANT UPDATE BELOW – SCROLL DOWN!
Rest assured, ITQ readers, that I am doing my darnedest to get to the bottom of this story, which seems to get more confusing with every additional tidbit of information that comes out. Thanks to Wai Young’s decision to ask the court to order a second judicial recount, we have a better idea of what happened during last week’s now controversial partial recount – according to the Conservatives, that is:
The Elections Act gives the judge supervising the recount the option of counting all the valid votes, plus spoiled and rejected ballots, or only those from a sampling of ballot boxes.
In the Vancouver South recount B.C. Supreme Court Associate Chief Justice Patrick Dohm opted to recount votes from 28 out of 184 ballot boxes.
An automatic judicial recount was triggered because Dosanjh led by only 33 votes when results were tallied on election night Oct. 14.
The recount gave Dosanjh, a former NDP premier of British Columbia, 16,109 votes to Young’s 16,087.
Leitch says that all the ballot boxes should be reopened and their votes counted, noting that a full recount Friday in a Quebec riding overturned a Bloc Quebecois victory in favour of the Liberal challenger.
The recount in Vancouver South found four vote changes in 22 of the 28 boxes that were opened, raising the possibility of 30 or more vote changes in the unopened boxes, Leitch says.
It’s worth noting that the only people who seem to be willing to go on the record are Young and Leitch. As yet, there is no official statement from Elections Canada on the results of the recount, which is listed as “in progress” on the website, which is frustrating, although understandable – the matter is before the courts and all. The putative winner, Liberal MP Ujjal Dosanjh, has declined to comment for the same reason.
Also, as pointed out by several commenters in the last thread on this controversy, it’s not clear how, exactly, the Canada Elections Act gives the judge the option of conducting only a partial recount, using a “sampling of ballot boxes” – although during an interview with CanWest News, Leitch himself seems to concede that point:
“The judge has the prerogative to decide how many ballot boxes he’s going to look at, and he decided after the 28 that he felt it was enough and there wasn’t a trend there,” said Leitch.
Leaving aside the issue of whether a partial recount is allowed under the law, however, the numbers that have come out so far just don’t seem to add up.
According to Canadian Press, the recount “found four vote changes in 22 of the 28 boxes that were opened,” which Leitch says “raises the possibility of 30 or more vote changes” in the other 156 ballot boxes. But it doesn’t seem possible to reconcile that statement – which is annoyingly ambiguous; does he mean four changes, total, or four changes in each of 22 boxes, for a total of 88? - with a recount result of 16,109 to 16,087 (which has yet to be officially confirmed by Elections Canada). If accurate, that would give Dosanjh eight more votes and Young an additional nineteen, for a total of 27 new/changed votes – we don’t know if some or all were the result of previously rejected ballots that the judge found to be valid, although it’s possible that the remaining vote changes were distributed amongst the other three candidates.
Anyway, I’m still on the hunt for answers, and I’ll post whatever new information I can track down. In the meantime, just as a point of reference, here are the preliminary and validated results for Vancouver South, as provided by Elections Canada, with the changes listed in brackets. I’ll add the official recount result as soon as it becomes available.
UPDATE: Okay, according to someone who was in the room – I’m not trying to be cagey, and I’m going to be calling Ray Leitch just as soon as it is a slightly more civilized hour on the West Coast – what this story fails to mention is the fact that the recount included – in fact, began with – the examination of each and every one of the rejected ballots – 259 in total – which seems to me to be a rather glaring omission. Of the 259, 41 were declared valid and counted, which reduced Dosanjh’s margin of victory from 33 to 26. That, plus the four changed votes from the opened ballot boxes, adds up to 22 – exactly the difference between the candidates’ final tallies, as provided by CP.
Preliminary results (Source: Elections Canada)
Ujjal Dosanjh (Liberal)– 16,774
Wai Young – (Conservative) 15,995
Ann Chambers (NDP) – 7,288
Gulyas Csaba (Green) – 2,209
Charles Boylan (Marxist Leninist Party) – 213
Total votes – 42,479
Results as validated by the returning officer (Source: Elections Canada)
Dosanjh – 16,101 (-673)
Young – 16,068 (+73)
NDP – 7,369 (+81)
Green – 2,067 (-142)
Rejected – 259
Total votes – 42,076 (-403)
Results after judicial recount (Source: Canadian Press)
Ujjal Dosanjh – 16,109 (+8)
Wai Young – 16,087 (+19)
Print
Comment
Food for thought:
If they reviewed 28 ballot boxes out of 184, this corresponds to 15.2% of the total ballot boxes.
During the recount, they found that 4 votes were turned over in favour of Young out of the 28 boxes examined (randomly selection assumed or perhaps each party chooses 14 or an equal number + random that may be biased in their favour).
This represents a *net* rate of vote conversion to Young (there is a chance that ballots will go to Dosanjh upon re-examination too).
Given that, if they continue to examine the remaining 184-28 = 156 boxes (i.e. 6.6 times more), the expected number of conversions would be 26.3 in Young’s favour, which is not enough to exceed the difference of 33 found on election night.
This may the reasoning behind stopping the recount.
Austin
BTW…one could even go by total ballots cast in each of the 28 boxes examined. That will create a better sense of the confidence in this outcome.
But based on statistics alone, it is very unlikely that the outcome will change.
Austin
I see I’m lagging way behind in this post…beg pardons…
Austin
Yeah, but Austin, doesn’t using statistics to determine the votes strike you as similar to using pollster results to decide the election?
All I know is my ballot is potentially being recounted as we speak. I damn well want them to look at the mark I personally put on the piece of paper, rather than “assume” I did something based on what a bunch of people at another voting station did.
Jenn: “I damn well want them to look at the mark I personally put on the piece of paper”
See, that was your first mistake. You should have written “is the man” in the little blank space. (Or, if your preferred candidate is female, “got it all goin’ on,” though you have to write quite small for that; better to go with the male candidate.)
You’re right Jenn…absolutely.
But, in point of fact, they already did just that (i.e. personally examine the mark on the piece of paper) on voting night, right?
This is about verification of verification, isn’t it? In other words, another layer of quality control placed on the quality control performed on voting night.
I’m just answering the question about how they could reason halting the recount.
That is all.
Austin
Now I hear that the recount in Kitchener-Waterloo could go on into the weekend.
It’s amazing that the one in Vancouver South could be wrapped up in a day!
[...] [Continue reading] Related: Vancouver South recount: Don’t shoot the (electoral) messenger, y’all. [...]
Bring on the full re-count!!!
All this, fortunately, pales in comparison to the registration chaos going on down south.
Send money , guns, and lawyers.
TWO POINTS:
-a mandated recount is different from a requested one. In a mandated one Elections Canada pays. In a requested one the requester has to pay all costs if it does not go in his/her favour (thousands of dollars). If he stops before the end and gives up he doesn’t have to pay. That is why the the Martin recount was stopped just before the end.
-this election saw a new form called the ‘bingo’ form. Every 30 minutes the unique number assigned to each voter that voted was recorded on a paper and given to a central registrar who then gave it to all scrutineers. That way scrutineers did not have to sit at polling stations all day. In fact in my polling center (I was a DRO) we only had 3 or 4 scrutineers all day for 50 polls. This was problematic during the count because there were no scrutineers around to observe and the one I finally found to sign my form wouldn’t have had a clue whether it was valid or not. I would say that this could lead to all kinds of errors that would have been caught in previous years.
Cheers.
Interesting insight, Sandra. It seems there are a lot of issues that need to be investigated from this last election.
I’m not sure who should be alerted, but lack of sufficient scrutineer oversight seems to be one that should really get some attention.
Getting scrutineers out is the job of the individual parties, and VanSouth hasn’t been close enough for many years to energize enough volunteers. I think that after this close call, it may be easier to turn out enough volunteers next time.