(A note to my subscribers: After a long hiatus, I’ve resumed this blog and will be posting weekly on various political topics. I hope you’ll continue to follow me and will share any pieces you find intriguing with your friends and contacts.)
In my previous post, I projected that Mark Carney would win the Liberal Party of Canada leadership race in a landslide on the first ballot. I based my projection on donation numbers, rather than other factors, for reasons that I explain in that post.
On March 7, Elections Canada published the second set of campaign financial returns, providing another 4 weeks of donation data. I’ve crunched the numbers, and… not much has changed. Here’s my final projection, along with last week’s projection for comparison:
| Candidate | Initial Projection* | Final Projection |
|---|---|---|
| Baylis | 1.4% | 3.0% |
| Carney | 81.9% | 71.2% |
| Freeland | 6.8% | 12.3% |
| Gould | 10.0% | 13.6% |
No, this isn’t a matter of Carney faltering or the other candidates surging. The differences you see are mostly due to the numbers coming into sharper focus, rather than any real change in standing.
(At the time of the earlier projection, some candidates were still working toward paying the $325,000 entry fee. Because of the rules of the contest, it was impossible to know how much of the fee each candidate had already paid, which added some uncertainty to their donation numbers. By the time the second financial returns were submitted, all candidates had paid the full $325,000 fee, making it possible to estimate more closely.)
Regardless, what we have here is still a landslide first-ballot win for Mark Carney.
The Race for Second
Intriguingly, the numbers still suggest that Karina Gould will place second. If this comes to pass, it will be a surprise to anybody who based their expectations on caucus endorsements or polling, both of which strongly favoured Freeland. (See my previous post for discussion on this.) Meanwhile, as far back as February 18, the donation numbers had already alerted us to the possibilty of Gould outpacing Freeland.
(Yesterday, I surveyed major Canadian media for their take on the contest. I found only two pieces that even mentioned the idea of Gould potentially placing second: this CTV News interview of March 8 with Scott Reid, who mentions the possibility explicitly, and this National Post Article of March 8, which states that “some Liberals have been wondering if Gould might end up second instead”. By contrast, this CBC summary of March 7 still presents the race as being between Carney and Freeland.)
What about the 100-point system?
I haven’t said anything yet about the 100-point system the Liberal Party uses for its leadership elections.
Here’s how it works: each of the 343 ridings across the country is worth 100 points, regardless of the population in that riding and regardless of the number of Liberal Party members there. Those points are assigned to candidates according to how the members in that riding voted.
(For example, let’s say a riding with 3000 Liberal Party members voted 2000-500-400-100 for Carney, Freeland, Gould, and Baylis. The points earned by each candidate would be 67, 17, 13, and 3.)
Does this affect our projection? If we look at where each candidate’s donations came from (excluding earlier donations, which typically come from a candidate’s personal network), there’s only one pattern that stands out as significant.
Carney, Freeland, and Gould were all weak in Quebec, drawing 9-11% of their donations from that province, despite it having 23% of the ridings. And they were all over-represented in Ontario, drawing 49-58% of their donations there, despite it having only 36% of the ridings.
This isn’t a great sign for the Liberal Party, which usually forms majority governments by winning in Ontario and Quebec. As far as the leadership race is concerned, all three candidates share this same pattern, so none of them gains an advantage.
Apart from the Quebec-Ontario imbalance, the top three all had fairly balanced campaigns. Freeland had the most balanced base of donations overall, Carney was over-represented in BC, and Gould was over-represented in Ontario.
To test the significance of the regional imbalances, I did a rough simulation of the 100-point system, calculating on a province-by-province basis rather than riding-by-riding, and the result was as follows:
| Candidate | Total Points | % |
|---|---|---|
| Baylis | 1139 | 3.3% |
| Carney | 27806.7 | 81.1% |
| Freeland | 2475.0 | 7.2% |
| Gould | 2879.2 | 8.4% |
This falls in line with the other two projections. There simply isn’t enough difference in the candidate’s donation patterns for the 100-point system to have a big impact this time around.
Some trivia
Buried deep in the numbers, I found a few pieces of trivia you might enjoy:
Voting Bottleneck? Based on a series of public statements by the Liberal Party from February 28 to March 7, about 15,000 members appear to be getting verified and voting each day. This is strange, because the typical pattern for an online vote is an early rush at the start of the voting period, then a much smaller tail. There have been complaints about the party’s strict ID verification process, so this may be evidence of some kind of bottleneck.
No Debate Impact? The debates on February 24 and 25 don’t appear to have changed the trajectory of the race at all. Donations were up slightly for all candidates on the day after each debate, with an extra bump for Baylis after the French debate and Gould after the English. But the overall ratio of support wasn’t affected.
Freeland’s Lost Week? The Freeland campaign’s financial returns have a gap of 14 days where no donations are reported. The gap starts on January 30, which was the deadline for paying the second $50,000 deposit and gaining access to the membership list. Keeping in mind that donations held by the party are not reported in the campaign’s financial return, these 14 days may represent the time it took the Freeland campaign to raise the remaining $225,000 of deposit money.
However, no other campaign has a gap like this, despite all of them needing to raise the same deposit money. It’s possible that the Freeland campaign simply chose a more aggressive approach, withholding 100% of donations until the fee was accumulated, while other campaigns chose to keep receiving a portion of their money while withholding another portion to accumulate their deposit fee.
Conclusion
There are more tranches of data still to be released by Elections Canada in the coming weeks, but I don’t plan to write any more about this race unless a particularly interesting storing emerges.
Instead, I’d like to tackle a broader question: are leadership races like this actually democratic? Does the outcome directly represent the members’ preferences, and if not, whose preferences does it represent?
I hope you’ll join me next week when I take a look at how leadership races really work, and which forces influence a party’s final choice for leader.
See you then.
Saul Bottcher 
Leave a comment