Leadership races are tricky beasts to project, due to their insular nature. There are few meaningful polls, the media is being spun in twelve directions, and the air war rarely corresponds to the ground war. Name recognition and a spiffy social media campaign are a lot less important than having organizers who can deliver hundreds of membership forms.
So when some compared Brian Topp’s leadership bid to the Paul Martin 2003 juggernaut back in September, it was premature to say the least. It would be equally premature to call Topp’s campaign dead after a few weeks of bad press, when we haven’t reached the membership deadline yet.
The fact is, there’s very little for anyone (especially for those of us outside the NDP) to go on when it comes to handicapping this race – but here are how the candidates stack up on a few metrics:
The “donations” and “donors” columns come from the most recent fundraising numbers, with “media” merely being the number of news stories that pop up on a google news search under each candidate’s name. The “poll” column refers to an IVR poll of NDP members released by Paul Dewar’s campaign yesterday. If you don’t know what Facebook and Twitter are, then get with the times.
With five different candidates leading these seven metrics, it’s hard to know what to think. Clearly, this won’t be decided on the first ballot, and we shouldn’t be surprised by anything short of an Ashton or Singh victory.
Inspired by Pundits Guide’s look back at the 2003 NDP leadership race, I’ve decided to go back and see how useful these different factors have been in predicting first ballot support in past leadership contests. Behold the table of correlation values!
The numbers are all over the place, but that’s to be expected when you consider these races all had different rules, fundraising restrictions, and voting systems.
Still, there are a few take-home messages.
1. MP (or MPP/MLA, as the case may be) endorsements and fundraising totals are both moderately useful at giving a sense of the race, but they’re hardly perfect. After all, Stephane Dion was sixth in fundraising in 2006, and you could count Christy Clark’s caucus support on one amputated hand in 2011.
2. Social media may be an important element of leadership campaigns, but there isn’t enough data out there yet to suggest it’s a good barometer of a candidate’s strength.
3. Polls among party members are likely the most useful predictor, though they remain rare.
It should be noted that polls among the public are worthless (correlations generally between 0.1 and 0.4), but every leadership poll among party members I’ve found has tended to stack up fairly well. Still, the Dewar poll should be read with caution given the source – after all, the opportunity for massaging the data exists, and we know he wouldn’t have released the numbers if they didn’t look good for him. I’d be a lot more confident in the numbers had someone leaked a membership list to a polling company instead.
But at this juncture, Mulcair has a 9-point lead in the campaign’s only poll, has three times the MP endorsements of anyone else, and has the most donors, if not the most money raised. He might still be too polarizing a figure to win, but at this point I’d put down $20 (or $10,000 if Mitt Romney comes calling) that Mulcair will be ahead of Brian Topp on the first ballot.
