Sunday, May 2, 2010

Canada: We're number two, but we try harder

Jack Layton addresses the 2003 NDP convention ...Image via Wikipedia
EKOS has released their weekly poll on the Canadian political landscape. This one includes two different indicators, both of which serve as kinda-good-news for the NDP. One is the non-news that Jack Layton gets better personal ratings than either Stephen Harper or Michael Ignatieff (I think Mahmoud Ahmadinejad might poll better than those two lately). The other, and for our present interests more noteworthy, item regards voters' 'second choice'. In other words, if for some reason they were unable or decided not to vote for their main party, who would they vote for instead?

In this category, the NDP do incredible. Their total 'second choice' numbers are 17.4%, which puts them in a virtual tie for first with the Liberals, at 17.5% ('no second choice' leads in almost every category I'm about to consider, so I'll not discuss it). The Conservatives rank below even Green for second choice - a sign of how polarised opinions of the CPC are, and also a sign of how fluid support among the traffic-jam of left-of-centre parties could potentially be. Anyway, let's see how the NDP do stat-by-stat.

They're preference #2 (behind the Liberals) for Tory supporters and Green supporters, and preference #1, by large margins, for Liberal and BQ supporters. Oddly, they rank low among supporters of 'other' parties.

They're the most popular second choice in Québec as a whole, preference #2 behind the Liberals in BC, Sask/Man (considered together in EKOS polls for some reason), Ontario and Atlantic Canada, and #3 behind a Liberal-Green tie in Alberta.

They're second-choice preference #2 for men nationwide, behind the Liberals, but #1 for women nationwide. They're behind the Liberals in the <25 and 25-44 age groups, but #1 in 45-64 and 65+ age groups. They're #1 for university graduates and 'high school or less', but tie Green for #2 in college grads.

Impressive stats, all-round. Except, of course, that EKOS is polling which party people don't plan to vote for. It's a bit of an always-the-bridesmaid thing, really. It shows that the NDP are well-liked, especially outside of the Conservative base, but not necessarily trusted to cast a vote for (this seems especially true in Québec, where the NDP's stock have at times been so low that they do not even have a provincial party, uniquely so among party-based legislatures in Canada). It's tough to know how much tactical voting enters into an opinion poll, but you might say that there are people presently telling EKOS they plan to vote for the Liberals or, arguably, the BQ as opposed to the NDP merely in order to keep the Tories out (in Quebec, possibly to keep each other out too). But I get the feeling that that's not really all that high a percentage. I think that 17.4% of the electorate 'like the NDP, but like another party better'. And in that context, it's not all that great news. It suggests that the NDP would stand to gain votes primarily through some large-scale shattering of confidence in the LPC or the BQ.

One thing that is interesting, though, is that this poll suggests that perhaps the NDP should reconsider which alternative to First Past the Post to support. As far as I know, the NDP by and large supports Proportional Representation - something that, at the moment, would afford them approximately 17.6% of seats in the House of Commons (yes, that's correct: the second-choice numbers for the NDP almost exactly equal their first-choice numbers). Perhaps they'd do best to consider a Preferential Ballot, where, confronted with a list of candidates, people enumerate them in order of their preference, '1' being their principal vote and '2' being the candidate to whom that vote is transferred if vote #1 was cast for a losing candidate. In this system, we can see that some 35% of the electorate would write either a '1' or a '2' next to the NDP.

But so what? Well, the poll shows that roughly as many NDPers choose Liberal second as vice versa, and as many NDPers choose Green as vice versa. This is also true for NDPers and Conservatives, mind you, but to a lesser degree. I think that, by and large, what we'd find in a Preferential Ballot system, is that in ridings where the NDP polled third or lower, the NDP vote would be broken up, largely in favour of the Liberals, in many cases pushing the Liberals ahead of the Conservatives. In ridings, though, where the NDP polled above the Liberals in second place, most of the Liberal vote wold be distributed to the NDP, in many cases pushing them above the Conservatives. What we might find, in this case, is a large number of ridings with a nominal Conservative plurality (the most votes but less than 50% of the vote) changing hands in a Preferential Ballot arrangement to either the NDP or the Liberals - in Québec, perhaps also to the BQ. Emboldened Green supporters in some ridings might push Green to second overall in some ridings (Bruce-Grey-Owen Sound, for example, or Central Nova, or some ridings in BC or Alberta), which could then, upon redistribution of the Liberal and NDP votes, push them above the Conservatives, producing a parliament with Greens in it.

In any case, in a Preferential Ballot system, the Conservatives - 44% of whose electorate say they would vote for no-one else - would be the losers. Yes, they get double-digit second-choice support from each party, but that still remains substantially less than the second-choice support the other four parties enjoy from each other. The famous image from the leaders' debates in 2008 of four party leaders sat next to each other, all directly confronting Stephen Harper on the other side of the table is grounded in reality, and is a feature of the Canadian landscape that on the one hand registers great animosity towards the Conservatives but on the other hand, under FPTP, returns more Conservatives to parliament than any other party.

Which is a reality that all the good news Jack Layton gets in this EKOS poll won't change.
Reblog this post [with Zemanta]

No comments:

Post a Comment