Monday, February 13, 2012

What U.S.-style primaries might look like for the NDP

The excitement surrounding the NDP leadership race pales in comparison to the headline-grabbing contest for the Republican nomination in the United States. But what if the New Democrats adopted U.S.-style primaries to choose their next leader? 

You can read the full article on The Globe and Mail website here. Once you do that, come back to read the details of this hypothetical NDP primary.

Politicos in Canada are spoiled being by so close and having so much access to the political drama that unfolds every two years in the United States. The mid-terms are an appetizer but the presidential race is a true meal, lasting over a year. The 2012 Republican primaries are about as roller coaster as they come, yet here in Canada the race to choose our next leader of the Official Opposition is not nearly as exciting.

Because of the way our parties do things, leadership races here are very much aimed at the party and their members and little else. Whereas in the United States some primaries allow registered Republicans, independents, and even Democrats to vote (i.e., everyone), here in Canada the privilege is limited to members only. The chances that a party will choose someone that doesn't have wide appeal always exists, because a leader has to hit the right notes among the party faithful in order to be chosen. Whether a message that resonates with the party faithful will also resonate with a plurality of Canadians is, sometimes, a secondary consideration.

So what if the New Democrats held primaries in each province in order to choose their next leader, allowing anyone who supported the NDP to vote? I thought that would be a fun and interesting exercise.

How did I calculate the results? Four factors were weighted equally to come up with the support each candidate would get in each province. The four factors were the recent Forum Research poll of NDP supporters, the recent Abacus Data poll of which candidates would make people more likely to vote NDP, ThreeHundredEight's endorsement rankings divvied up by province, and the endorsement rankings adjusted by the Q3 and Q4 fundraising totals for each of the candidates. Those were combined together to give a vote share for each province. Delegates were divided up proportionately across the 13 provinces and territories to total 1,000, and delegates were awarded to each candidate proportionate to their vote share.

The Globe piece provides a narrative with a few US-inspired touches, but let's look at the actual results this exercise gave for each of the provinces, in the order that the primaries are deemed to have been held in this hypothetical scenario. The schedule was determined mostly by the debate schedule of the NDP leadership race. And if these graphics resemble those of a certain American 24-hour news channel, that is not exactly by accident.
Thanks in large part to good endorsements in the province and decent polling numbers in the region, Thomas Mulcair wins the first primary in PEI (scheduled January 24) with 43%, with Brian Topp trailing with 28%, Paul Dewar with 13%, and Peggy Nash with 8%. Niki Ashton heads the bottom tier with 4%. With this first win, Mulcair takes two delegates to one for Brian Topp and one for Paul Dewar.
With the debate in Halifax on January 29, primaries are held in Nova Scotia and New Brunswick on January 31. Brian Topp has the strongest polling numbers in Atlantic Canada, and with good endorsements in New Brunswick he wins 57% of the vote in the primary, followed by Mulcair at 17%, Dewar at 12%, and Nash at 5%. Ashton comes up fifth with 4%.

In Nova Scotia, Peggy Nash's endorsements give her the win with 31%, followed closely by Topp at 29%, Mulcair at 19%, and Dewar at 13%. Again, Ashton comes up fifth with 4%.

With these two primaries, Brian Topp moves into the lead with 21 delegates, followed by Mulcair with 11, Nash with nine, Dewar with seven, Ashton with two, Martin Singh with two, and Nathan Cullen with one. Romeo Saganash has zero, and in real life he pulled out of the race after the dates I scheduled for the first three primaries. So he was dropped off for future primaries.
After unofficial debates in Saskatoon on February 7 and a forum in Edmonton on February 8, the next primary is held on February 14 in Alberta and Saskatchewan.

Topp's good polling in Alberta and Nash's labour support make them the two top finishers in the province, with 35% for Brian Topp, 25% for Nash, 19% for Dewar, and 13% for Mulcair. Ashton finishes fifth with 6%. In Saskatchewan, Topp's endorsements allow him to walk away with it with 51% to 25% for Mulcair, 10% for Ashton, and 9% for Nash. Dewar finishes with 7%.

This puts Brian Topp further in the lead with 75 delegates. Peggy Nash is second at this point with 40, followed by Mulcair with 32, Dewar with 30, Ashton with 12, Singh with three, and Cullen with two. One can already imagine the grumbling in the Mulcair and Dewar campaigns after a few bad finishes.
The primaries in the territories and Newfoundland and Labrador are then held on February 21, with Peggy Nash winning both of them thanks to her endorsements in both areas. Nash wins 31% in Newfoundland and Labrador, with Mulcair taking 25%, Topp 23%, and Dewar 12%. Ashton finishes fifth with 4%.

In the territories, Nash takes 38% to Topp's 24%, Mulcair's 20%, and Dewar's 12%. Nathan Cullen finishes fifth with 2%.

At this stage, Brian Topp would still lead in the delegates with 79. Peggy Nash closes the gap a little with 46, while Mulcair (37), Dewar (32), Ashton (13), Singh (3), and Cullen (2) trail.
After the Winnipeg debate on February 26, Manitoba holds its primary on February 28. In the Prairie provinces, Mulcair, Topp, and Ashton have polled the best but in Manitoba most of the endorsements have gone to Dewar and Ashton. The result is a narrow Mulcair win with 28%. He's followed closely by Paul Dewar (one can imagine this being Dewar's first "big effort" province) at 23%, Brian Topp at 22%, and Niki Ashton with 17%. Nash follows in fifth with 9%.

This win gives what would have been Mulcair's sagging campaign a boost, but Topp would still lead in the delegate count with 87. Peggy Nash follows at this stage with 49, with Mulcair (47) and Dewar (40) not far behind. Ashton lands six delegates in Manitoba and is at 19 at this stage, with Cullen and Singh at three apiece.
The first big primary is then scheduled for March 6, after the debates in Quebec City (February 12) and Montreal (March 4). Obviously, this one is Mulcair's to win and he does with 69% of the vote. Brian Topp finishes second with 17%, with Nash at 8% and Dewar and Singh at 2% apiece.

With this big win, Mulcair moves to the front of the pack with 206 delegates to 126 for Topp, 68 for Nash, and 45 for Dewar. Ashton follows with 22 delegate to Singh's seven and Cullen's five. The race would change complexion here with Mulcair now the frontrunner.
And then comes what I dubbed Super Tuesday in my article, the primaries in British Columbia and Ontario on March 13. With the majority of the delegates at stake, the race would be won or lost here.

In British Columbia, Brian Topp and Nathan Cullen lead in both the polls and the endorsements, meaning that Brian Topp wins with 45% of the vote to Cullen's 19%. Mulcair trails in third with 13%, followed by Nash (11%) and Dewar (7%). It's a big win for Topp and gives Cullen a boost.

In Ontario, the polling is far more favourable to Nash and Dewar and they are the two frontrunners. Nash edges out Dewar with 32% to 26% of the vote, followed by Topp at 18%, Mulcair at 15%, and Cullen at 4%. Dewar or Nash had to really win big in Ontario to have a shot at overtaking Topp or Mulcair.

After the last primary, Thomas Mulcair still leads the delegate count with 282, but he is far short of the 501 needed to win. Brian Topp takes 256 delegates to the convention, with Peggy Nash claiming 208 and Paul Dewar 154. After his good performances in Ontario and British Columbia, Nathan Cullen pulls ahead to stand fifth with 47 delegates, while Niki Ashton has 35 and Martin Singh 18.

There you have it - a plausible description of what might happen if the New Democrats held primaries to choose their leader and opened up the race to all supporters, rather than just members. Does it tell us anything about the race itself? Not really, though it does give an indication of the regional pockets of strength each candidate might be expected to have. It is really more of a fun little exercise that shows how much more interesting primaries would be than the current model our parties use here in Canada.


  1. This is a really cool scenario Eric, well done. I only have one objection. The Republican Primaries have been moved significantly by the debates. Most media reports state that Nathan Cullen performs very well in debate. In addition, if the voting was open to "supporters" and not just members Cullen would probably draw those interested in a progressive coalition. I think he would hop over Ashton, Saganash, and Singh.

    If, as Mulcair put it, Cullen has a Gingrich momenet he could skyrocket and win Alberta, or over-perform in other provinces.

    Admittedly, I am biased, but I thought I'd offer my two cents. Thanks!

  2. I agree with you, but I based my figures on polls and endorsements. In the few polls we've seen, Cullen hasn't done very well outside of British Columbia. Of course, if we did have these kinds of primaries more people would be paying attention to the debates and Cullen would likely gain in popularity, but I limited my analysis to what the numbers show.

  3. That was just cool, Eric. Well done.

  4. Sadly for Mr Mulcair, his good performance in the primary scenario, fueled by Québec numerous delegates, won't happen this way in the one-member-one-vote system implemented for the actual race.

    Using your numbers and the last november figures of NDP membership (latest I could find, making up about 350 members for N-B and 200 for the North) and the regional breakdown of this article, Topp emerges first with over 300 delegates, with Mulcair second with little more than a quarter of the convention floor, trailed by Nash and Dewar ranking below 200 and the other below 100.

    I found it relevant since Topp strategy is clearly focusing on heavy card-carrying province to secure the nomination with establishment connexion assuring national visibility, while Mulcair has to juggle with boosting his home province weigth and campaigning country-wide.

  5. This is pure fiction.

  6. I think primaries are not necessarily a bad idea. One of our three major parties should try this method out. But only the Liberals are in the mood to experiment at the moment.

    One of the main reasons why the U.S. primary system is getting bad rep here, is due to the Republican circus of 2011-12. The primary system has worked efficiently for both major parties in the U.S. Other countries like France uses this system to elect leaders.

    1. Great articles, as always.

      One of the biggest problems with the primary system is that it gives a select number of states (or provinces in the case of Canada) an enormously disproportionate amount of power in the process. If we applied the American model to the Canadian model which you have set up, Atlantic Canada would have an absolutely enormous say in deciding who will become the next leader of the NDP. For example, Dewar, Ashton, and Cullen would be forced out of the race long before they are able to prove their strength in Ontario, Manitoba, and BC, respectively. One of the best ways to solve the Iowa/New Hampshire/South Carolina problem is to ensure a rotating calendar so that no one province gets a stranglehold in the same way that those states do. For example, with BC and Ontario last in this scenario, perhaps come time for the next leadership election, one or both of those provinces would vote first.

      That having been said, it isn't all bad. Having small states go first does allow for a large degree of retail politicking, which I think is good for democracy. It provides less established candidates to come to the fore (think Obama's win in Iowa in the 2008 caucus).

  7. I don't see why opening things up to supporters requires you to phase the voting over several weeks.

    IMHO any political party would be insane to adopt the US primary system. It may be entertaining, but all the bickering is hardly good for the Republican party's prospects in the next election. Moreover, the phasing of the primary system gives early voting states disproportional weight in the nominating process. I suspect that the later-voting provinces in Canada would not take kindly to this.

    Still an interesting what-if, but I think if the NDP wants regional balance it should adopt the CPC system of weighting members' votes by riding.

  8. Though for US-style primaries, most of the provincial contests should be winner-take-all, rather than divvying up the delegates proportionally.

    Winner-take-all would make Ontario a vastly more important a contest.

    1. Ira,

      I believe the Democratic Party awards all delegates on a proportional basis. Presumably the NDP would imitate the Democrats rather than Republicans, though I guess this is all just academic.

    2. I thought it differed from state to state, much like with the electoral college itself.

      Éric also ignored the possibility of some privinces using caucuses rather than primaries, which can skew the results.

      The US party system is just broken. I'd feel sorry for them if they gave any indication that they knew they had a problem.

    3. I think the Republicans have a rule that to go early a state has to allocate its delegates proportionally (or be non-binding). In practice the delegate numbers don't mean too much compared to the momentum from winning a state.

    4. Ira,

      The rules differ between the two parties (and are constantly evolving). In 2008 the Democrats allowed states to use the method of their choice, provided delegate allocation was proportional with a 15% threshold.

      And agreed completely. :(

    5. Actually, it's even more complicated than that. The 15% threshold was only in place in Iowa, and that was only to allocate delegates to the state convention (as opposed to the big national convention).

      A good rule of thumb is that Democrats require some form of proportional representation when awarding delegates, whereas Republicans largely leave it up to the state parties to decide. It doesn't matter whether states go early in the process or later. In fact, of the ten or so states that have voted so far in the Republican primaries, only Florida is a winner-take-all state. The confusion may be that Florida was sanctioned by the Republican party for holding its primary before Super Tuesday -- reducing the total number of delegates from that state from 99 to just 50.

      And this craziness is just the tip of the iceberg. Believe me my Canadian friends when I tell you: stick with what works, because the primary system does not.

  9. This made my day. This would actually be very interesting and I wish it would happen in Canada. The CNN graphics are very appropriate.

  10. I'm not sure how likely this senario is. From what I've seen and heard from rank and file members, Topp seems to have a fair bit less support than endorsements would suggest. I think he could have second or third ballot support but not those kind of numbers off the top. Cullen would also have a fair bit more support I would argue as would Mulcair

  11. I realise you're concerned with the statistical trends and numerical analysis of the political scene, and not political issues themselves, but I think it's really important to take a step back for a moment to reflect on the implications of your thought experiment.
    The “excitement” of the Primary system in US politics is the product (and I mean that word precisely) of a colossal expenditure of money that effectively strips the electoral process of any substance and undermines democratic participation. It may sound good that “in the United States some primaries allow registered Republicans, independents, and even Democrats to vote (i.e., everyone),” but that very fact contributes to the dissipation of party platforms into virtually interchangeable platitudes and verbal. And the delegate system itself is inherently less democratic than one member-one vote – as delegates take their voting power wherever they decide (or are enticed to), not wherever people prefer. The Primary system – like the US election in general – is a media spectacle in which the profound issues of the day are rarely mentioned, concrete policy proposals are rarely raised, and the vast amount of talk is spent on character assassination. It has the same morbidly fascinating qualities as Survivor or American Idol. Sadly, unlike so-called reality TV, this system has a genuine impact on the health of the nation – in political and policy terms.
    The NDP, whatever its shortcomings, at least makes a claim for democratic inclusiveness and for a substantive political programme. Given its declared platform, it makes no sense to enlist the votes of non-members and non-fellow travellers in its leadership campaigns... why should supporters of other parties want to endorse leadership candidates for a party that represents a programme they disagree with? And since we’re speaking of the NDP, it’s worth recalling that their conventions – much more than those of other parties – are also policy conventions, not merely exercises in producing a party leader. All to say that, unless something were to go terribly amiss, the NDP would be the last party to adopt a Primary System.
    It’s true that the leadership campaign has been by fairly dull – at least as it has appeared in the mainstream media. The NDP is partially responsible for that, presumably, but it’s also the nature of a leadership contest in which the candidates are necessarily fairly close in political orientation, and in which there’s little place for the cynical fabrication of adversaries that comes with a massively moneyed and corporatized process, as in the US. The dullness of the NDP leadership race is also the result of political pundits’ and journalists’ own lack of interest in real political issues, and their inability to come to grips with them, to engage them, or to convey their significance to the public.

  12. sorry, should read, "the dissipation of party platforms into virtually interchangeable platitudes and verbal attacks."

  13. Eric,

    Is there any polling of second preferences? It'd be interesting to know how likely each candidate is to grow their support on successive ballots.

  14. The Dewar campaign came out with a poll on first and second choices today.

  15. Hmm, interesting. He didn't release the cross tabs I see... (

    So if we combined first and second (I know I know) as a ceiling for the first two ballots we have:

    Mulcair: 42.2%
    Dewar: 36.3%
    Nash: 36.2%
    Cullen: 27.2%
    Topp: 25.1%
    Ashton: 20.2%
    Saganash: 7.2%
    Singh: 5.9%

  16. Aaron, what you identified is actually one of things I was trying to emulate. PEI is a tiny province, but as the first province it would have a lot of importance - just like Iowa. Is that ideal? No, but I wanted to do an NDP version of the US primaries, including all of the problems with them.

    Ira, I wanted to choose something simple and uniform. I considered doing the winner-take-all option, particularly if a candidate got more than 50% of the vote in the province, but figured the NDP would probably choose proportional.

    1. I understand. I was simply attacking the idea of having a primary system in Canada, rather than the hypothetical situation which you set up in your article. In fact, I think the way you laid it out is very plausible -- and closely mirrors what we have in the US. While it won't happen this time around for the NDP, it is very informative to ask the question of what might have been, and you did a great job of doing that in your article.

  17. @chimurenga, completely agree with your post. Spectacles lasting years like the ones in America don't come cheap, and do we really want the NDP, of all parties, to become beholden to moneyed interests above all else?

    On the other hand, both the NDP and the news channels have dropped the ball on making this leadership race visible to Canadians. On the NDP's side, why don't they at least have a website specifically dedicated to the leadership race, with news and debate videos? Provide a single location for anybody who wants to find out more.

    As for the media, this race has simply been invisible. There's no comparison at all to the heavy coverage of the last real Liberal leadership race (when Dion won). Don't know if it's lack of ability or lack of will on their part.

  18. Really liked it

    But then what would happen hypothetically?

  19. I'd be curious to see what hypothetical Liberal and Conservative primaries would be like? And would the Bloc split up by individual region in Quebec?


COMMENT MODERATION POLICY - Please be respectful when commenting. If choosing to remain anonymous, please sign your comment with some sort of pseudonym to avoid confusion. Please do not use any derogatory terms for fellow commenters, parties, or politicians. Inflammatory and overly partisan comments will not be posted. PLEASE KEEP DISCUSSION ON TOPIC.