This seems quite comprehensive and ambitious. Two questions about assumptions. First, why do you assume that all the delegates in a riding will go to one candidate? We know outright that this is not the case. Second, why have you not used membership numbers to also estimate support? Geographic origin of financial support seems a bit silly, because candidates may concentrate their fundraising where the money is while focussing their campaigns where the votes are.

Finally, an observation: who believes that Brison will actually win 21 ridings in Montreal?

What you have is essentially a four way race for third place. Any of these candidates could come in third within a margin of error. Given the repulsion most Liberals feel for Rae, if you anaylsis is correct, very few delegates will swing to him no matter who their candidate chooses. I would say this means Rae will be the first of the top six to go after the bottem four candidates delegates go to other candidates. I also do not think the delegates will split evenly. Many Liberals also do not want Ignatieff as leader very strongly. I do not think he will gain many of the bottom four’s delegates. If Dryden gains more delegates than Brison, which he probably will, then Brison is the next to go. If many of Rae’s people have gone to Dryden, which makes sense ideologically, then, depending on the split of Brison’s delegates Dion or Dryden is the next to go. Dryden’s campaign manager is from the East coast and he is very popular there. He could pick up many of Brison’s dlegates. If he is ahaead of Kennedy at that time, who will Kennedy’s delegates go to. Not Iggy, which makes Dryden the winner. He is a consensus candidate and everyone likes him. I think he will take it.

Nice analysis.

However, as in so many elections, the key is turnout. This is the first election of a new leader of a party under the stringent new rules, which have cut back funding through limits on individual donations.

Having delegates on your side is good. But to win, you have to get your delegates to the convention. And this takes money.

So, the better funded campaigns stand a much better chance than your analysis would indicate.

My understaniding Curiosity is that Candidates cannot fund delegates as it is against the Canada Elections Act. The funding of the campaigns won’t matterin whose delegates get there. The dedication nof the delegates will. Rae’s money comes from very few people and it can’t help delegates so there again he’s at a disadvantage. Ridings in Southern Quebec and Southern Ontario should be given more weight because these delegates are more likely to come.

Personally, I would prefer to see your breakdown name ridings as a first priority and communities as a secondary item. Name all the ridings, not just the few communities where individuals have given donations. A bit more work perhaps but then you are already a political analysis junkie….. what’s a bit more when it would be so much more enlightening?

Maatje — I’d love to break down the donations by riding, unfortunately, we don’t have consistent data. The June quarterly report from Elections Canada contains postal codes (although you have to click on every person to see that data, not very convenient), which allow us to map the particular riding of a donor, but the Liberal contribution reports (the only thing that will document donations for July and August) only contain the city name. So we don’t know when someone is listed as “Calgary” whether their riding is “Calgary West” or “Calgary Centre”, for example. The September quarterly reports from Elections Canada should show postal codes, but by that time we will know the results from the delegate selection meetings. So in the meantime, we are left approximating support at the regional level, and listing only the communities.

I realize it would be pretty subjective, but is there any way you can weight the number of delegates by place, since some areas will likely not send the full slate. For instance, on Vancouver Island you have 5 locations (6 ridings),but in all except Victoria (which covers one and a bit ridings at the most) there is only one donation noted. Therefore to assume that all 6 ridings will send a full slate of delegates seems to be an overstatement. I realise that by using the percentages you are evening it out to some extent, but the problem is that there are a number of areas that seem to be completely uninterested in the race, and will likely not want to get involved, especially considering the cost of travel from BC or Alberta for instance.

HAHAHAHA
Volpe sells 35,000 memberships, 4,400 of 9,000 in Quebec and you have him under 2% and at 0% in Quebec.
HAHAHAHAHAHA

Frank — feel free to look at the June quarterly Elections Canada report yourself. There are no Quebec donations listed. And, unfortunately, Volpe’s team has reported no donations for July as of yet. As noted above, I will update the tallies as information is reported. If you are so concerned about Volpe’s campaign, perhaps you could ask them to get their act together and file their July contributions report.

DS:

I origially misunderstood your exact simulation method, but now understanding it (I think) I still have two questions. First, why is money a good proxy for support? Why not memberships? Second, do you really think Brison will win 43% of delegates in Montreal. Doesn’t the absurdity of this make you question your model?

Again, I really do applaud your efforts; I am just seeking clarification on some of your modelling choices.

Peter

SB:

I am pretty sure campaigns will find plenty of ways to get delegates sent to a convention without the money flowing through the books. It’s a cash business, as they say.

Peter

Peter - this analysis doesn’t look at the money. But what the donations identify are *known supporters* (admittedly, it rests on the assumption that you support someone to whom you give money). Likewise, the ex-officios tell us who supports who (in that case, those get directly translated into delegates). That is all this analysis tells us — i.e. the geographic distribution of people we know support for sure support the various candidates. In many cases, there simply aren’t enough supporters for it to be reliable, but in other cases, we have a good idea (for example, in Toronto, there are 687 supporters for a region with 308 delegates — I think when the number of supporters hits the number of delegates in a given region, then the above analysis is a reasonable approximation of how the delegates will break down. So, this analysis will become more reliable as we can identify more supporters in August and September).

About memberships: a) the numbers are notoriously unreliable, b) they really don’t contain any information on which candidate they support, and c) there is a high likelihood that someone who was recruited by X candidate will support Y candidate after looking at the various options.

I don’t know whether Brison will win 43% of Montreal delegates. All I know is that Brison has received double the donations that Dion has in both Westmount and Montreal. Whether that translates into delegates, of course, we do not know.

Well, I think there are some significant flaws in your reasonings or estimates, no offence. First, the idea that every delegate will go to a candidate is going to be unheard of outside of Eglington Lawrence. Secondarily, you dont seem to be taking into account any of the support that already existed. The amount of support does in no way translate into the amount of delegates. Polling might be a more accurate way of looking at things, or looking at memebership sales. I cite specifically the estimate that Brison will take 16 delegates in the Calgary region, or, similarily, that Dion is riding one delegate about MHF in Alberta, or that Dryden will have ALL of Newfoundland. These estimates do not seem to be reflective of the actual membership sales, and, therefore, the number of delegates. Volpe will continue his enexorable treck into the 4th place kingmakers spot, based on his inability to grow and the massive amount of memberships he signed up, but this dosent seem to be reflected. I think the problem is that membership sales are more spread out, while the donations are concentrated.

your numbers for volpe make me laugh. if you only used financial contributions, how can the candidate who is 3rd in fundraising overall get shutout everywhere? instead of shooting back at frank, perhaps you can get your numbers straight. Why do your alberta numbers not have any mention of cities in the peace river riding, or fort mcmurray? i urge you to keep this original post up, and come talk to me on the convention floor. i shall bring a very dead crow for one of us to nibble on.

Volpe will get closer to 10% on first ballot. Hall and her short skirt on a stool will get 2%.

You are right Brison will do very well with Aboriginal delegates. The APC has 245 Delegates through its provincial and territorial organizations. This is not including Aboriginal ex-officio delegates, past candidates/MPs, and the delegates who would go through constituencies. I can say Scott has the majority support in western Canada where a good chunk of the APC delegates will be.

Another update and another model that makes no sense. Do you believe that Brison will win Quebec? And that Volpe will not take a single delegate there?

I appreciate the difficulties of modelling with scarce data as much as anyone else, but when you don’t have enough data to make sensible predictions the best thing to do is wait until you get more. I think everyone knows this model to be flawed, and I am not sure how it’s helpful to publish it.

My prediction. The race will be largely over by the 4th ballot. The 3 women will be gone and at least 1 of the 3 middle men (Brison, Dryden, Volpe) will have moved to one of the top four while he still has some support to deliver (several candidates will see their vote drop as first round supporters abandon the losers). Going into the fourth ballot there will be no more than three people seen as having a chance of winning and the delegates will give the nod to one of the top two. Iggy will take it with the support of Brison and Volpe.

Greg,

Your predictions in the subsequent ballots is based on the assumption that once the candidate drops out, all his support goes to the person he moves to. In previous conventions, the max number of candidates would be two-thirds and can go down to one-third. If Volpe drops out and supports Iggy, I see Iggy at the 29-30 range and would be dependent on Brison to take the bulk of the Atlantic pro-military delegates to him on the third. Dryden’s support would be crucial as some of his remaining supporters will go to either Rae or Dion. The arm wrestling tactics would be critical in either the second or third ballots. Rae and Dion would need to gain initial momentum to become credible for the fourth, as either one would be faced with elimination. Kennedy will try to stay in second and make sure that Rae and Dion cannot catch up to him.

This scenario is based on the assumption that some of Dryden’s supporters will have defected before the second ballot to determine whether Rae, Dion, or even Kennedy will have the best chance to stop Iggy. Since you have an interest in Kennedy winning the race, the greater margin of the lead GK has from Rae and Dion are crucial in convincing Dryden supporters that Stephane will not have the necessary momentum to stop Iggy.

Peter, the fourth ballot will determine who the stop Iggy candidate will be. I suspect that Dion and Kennedy has the momentum with Rae fighting hard to stay ahead of Stephane or being in danger of being left behind by a GK juggernaut. In that case, Iggy may need to rely on some of Rae’s supporters may take him over the top.

first of all, volpe is still underestimated grossly. but if your numbers for him are accurate or not, he and his supporters would never go to iggy, nor would iggy welcome them. the two have taken shots at each other through the race and will not heel the wounds on the convention floor

What will happen if Kennedy, Dryden and Brison make deals together to become the winning slate or the definite King Maker?