Wednesday, October 28, 2015

Cariboo-Prince George and Edmonton -Griesbach Illustrate the Difficulty--Even the Perversity--of so-called "Strategic Voting"

On the whole, the October 19 election was a good day for democracy. I had feared—more than once in these columns—what a re-elected Conservative Harper majority might mean in terms of creating a “new normal”. Instead, we had a 7% increase in national voter turnout, and a government with more women and aboriginal people than ever before. The Liberals won by having an optimistic and upbeat leader who  seemed to embody “Real Change” . The Opposition is strong, and , aside from the complete Liberal monopoly of Atlantic Canada,  each of the major political parties is well-represented throughout most of the country.
Of course, electing the country's first NDP government, with proportional representation, Senate Abolition and a million daycare spaces on the policy docket, looks more like "Real Change" to me. But elections are about politics, not policy.

Let's not forget that the Conservatives, the NDP and the Media all unwittingly conspired to give Mr. Trudeau his Golden Opportunity. The Conservatives, whose extra-long campaign made people forget about C-51 and the Senate Scandal , but which also made people forget about Trudeau's poor performance on those issues and Mulcair's strong performance; the NDP , by weakening its Quebec base and then taking its balanced-budget theme too far, creating more policy space for Trudeau; the media, in particular the Globe and Mail and the Munk Centre, who could have crowded the stage by adding Elizabeth May and GIlles Duceppe to the mix but instead gave Trudeau tons of visibility.

One sad feature of the election   was that,  all too often, so-called “strategic voting” just gave way to a  blind, convulsive bandwagon effect, of course amplified by our first past the post electoral system.  Cariboo-Prince George is a case in point. Voter turnout in 2015 was 53,590 or 68,9% of the electorate--- an increase of over 10,000 votes or nearly 11% higher than in  2011.  This was clearly a determined vote for “change” , since the Conservative share of the total  dropped from 24,324 to just 19, 668 votes—a loss of  nearly 20%.  So why weren’t the other 63.3% of local voters able to dislodge them?

A week before the election, LeadNow, a national organization dedicated to coordinating the efforts of all “progressives”, released a local Environics poll showing  the NDP’s  Trent Derrick was in the lead with 36% of the vote, the Conservatives’ Todd Doherty  well behind at 30% , the Liberals’ Tracy Calogheros running third at 29% and the Greens polling at 5%. Accordingly, LeadNow recommended to local progressives desiring a change to unite behind the NDP candidate. 

Edmonton - Griesbach was another perfect example. Janis Irwin of the NDP had long been in the lead locally and was seen as being the NDP's second-strongest riding in Alberta. A creditable survey of 509 people , considered accurate within 4 percentage points, was published on August 19. It showed Irwin to be in the lead with 48% of decided voters(!), Kerry Diotte of the Conservatives well behind at 32%, and Liberal Brian Gold third at only 15% of local voters.  Liberals were being urged to do their bit and hold their noses and vote NDP  in order to "heave Steve".  But a problem for the NDP was that 60% of those surveyed said they would be willing to vote differently in order to defeat the Conservatives. As the Liberal surge began to wash across the country in October , many of those ABC voters, either indifferent or ignorant of local circumstances and conditions,  began falling off the fence and voting Liberal (in effect, pinning their tail on the wrong donkey because it appeared to be the right donkey nationally). The final result: Irwin got an unexpectedly low 34.04%; Gold got an unexpectedly high 21.67%, and Diotte won with 39.91% of the vote. Conservatives deserve some credit for getting their vote out and perhaps for persuading a large chunk  of the 11% who were undecideds, but it was that vote-splitting defection from the NDP to the Liberals that sealed their victory.

The problem was that  a lot of people were watching what was happening  nationally and decided to climb on the Liberal bandwagon.  In those constituencies where the Liberal base was already big enough, (e.g Edmonton Centre, Vancouver Granville) that switch managed to elect a Liberal MP.  But in  several constituencies in western Canada,  the Liberal wave had the opposite, unintended effect: of splitting the vote and electing a Conservative MP instead.  In Cariboo-Prince George, Todd Doherty was the beneficiary of that vote-splitting.  In Edmonton-Griesbach, it was Kerry Diotte.

On the whole, however,  it was still a good day. I look forward to Canada presenting a fresh face to the world , at the U.N. and at the Paris Climate Conference. I wouldn’t be surprised if, when being introduced in Paris,  Justin Trudeau received an ovation…just for not being Stephen Harper.

But for me the big lesson to take away from this election was that given the difficulties of information sharing and voter coordination in the context of  increasing electoral volatility, in the end there is really no alternative to electoral reform, and despite what the prime minister seems to think, that means some degree or kind of proportional representation (PR). While many arguments can be marshaled against PR, nearly all can be met to a large extent through modifications of the PR formulae. So please, let's do it.


CuJoYYC said...

Blaming the media is a CPC speciality. I'm surprised at the number of NDs that have been proffering those same talking points.

As for your main thesis, whichever opposition party was perceived to be ascendent after Labour Day would have benefitted from the coalescing of the 'change' vote and the vagaries of strategic voting within the FPTP system would still have been in play. The Dippers had far more issues with their campaign than a claimed biased media and strategic voting gone wrong. The time for the NDP and CPC to reflect is now. Now is not the time to look for scapegoats. Finger-pointing is so passé now that Harper is gone (or at least going)

"In both ridings the majority of voters—especially the more knowledgeable voters—did NOT want these men to represent them in Parliament."

An assertion without evidence. Just how do you assess who are "more knowledgeable voters"? And whose definition of more knowledgeable is correct? I expect these kinds of unsupportable assertions from CPC trolls. I expect more from my opponents in the NDP.

buckets said...

With respect, I think you're making conclusions that are a little beyond our evidence. Leaving aside the percentages (which I think obscure things), the Conservative vote fell from 19.8k to 19.1k; the NDP vote increased from 14.1k to 16.3k; and the Liberal vote increased from 2.5k to 10.4k. What proportion of which group were 'strategic'? It seems to me that there are several groups that are important here: loyal voters for each of Lib, Con, and NDP; some CPC voters that migrated to the Liberals (and probably not available to the NDP); strategic voters who voted NDP (rationally); strategic voters who voted Liberal (irrationally, imo), and new voters (who apparently broke towards the Liberals).

Mark Crawford said...

Dear CuJoYYC : My point about the media is precise and accurate. They played their part--most notably, in the respect that I stated in the article, by giving the third party more room and space on stage--a situation that Gordon Wilson profited from in 1991 and Bob Rae in 1990. I could have talked about the roles of Althea Raj and Jeffrey Simpson and other journalists, but my point wasn't to blame the media as much as to point out that the blame was shared among all the relevant parties. All credit to the Liberals for taking advantage.

My "main thesis" was to point out the difficulty of strategic voting when even the most knowledgeable and determined voters--those equipped with creditable local polls as well as the published national ones, which they are looking at--have difficulty discriminating between local and national information for the purposes of achieving what they have clearly indicated is their desired result. In the case of Edmonton-Griesbach, the local poll data I referred to was old (August)but it showed a huge 33% lead for the NDP candidate over the Liberal. The problem is that that same survey identified 60% of non-conservative voters as "Anybody but conservative"---which, in a way, was the problem, because if they had ignored the national polls and stayed with Irwin, the Conservative candidate would have been defeated, even if we grant a majority of the margin of error and a majority of the undecided to Diotte.

AS for Cariboo-Prince George, the polling data showing the NDP as the best vehicle for defeating the Conservatives was only a week old! And still there were Liberals writing letters to the Prince George Citizen pointing to the national polls instead.

As for the passage you highlight--I do not rigorously define what I mean by "knowledgeable voter" , but my article does not hinge on that quote. I could erase it without affecting my thesis at all. All I meant by the term is those armed with both local and national polling data--which, if you have spent any time on the doorsteps, you will know that that is more knowledgeable than most.

But if you like, I'll remove the highlighted sentence. Nothing much hangs on it.

Mark Crawford said...

Buckets: Be careful about comparing numbers of voters between elections in Edmonton-Griesbach which is a riding which considerably different boundaries from Peter Goldring's old riding. WHile percentages can be misleading (e.g. an increase from 1 vote to 2 votes is 50%), the survey done in August by Environics is a creditable survey of over 500 people. While I could have parsed the voters in the way that you suggest, it can be reasonably inferred that Kerry Diotte must have gotten a large chunk of the margin of error and undecided to go from 32% to over 39.9 %

Of course, the August poll could have been the 1 in 20 that is wildly inaccurate (outside the margin of error), but there is no reason to jump to that conclusion, when the 3.2 k increase in Liberal support over that predicted by the August poll captures about half of the 6.68 K drop from what was predicted by the August poll for the NDP --and about 60% of the NDP & Liberal voters surveyed in August self-identified as strategic (ABC) voters!

All of which is to substantiate the rather obvious: that in ridings like Edmonton-Centre , where the Liberal candidate had a high profile and the party was more competitive with the NDP, the switch from NDP to Liberals caused by the national wave had the intended effect of electing a Liberal MP. BUt in ridings like Edmonton-Griesbach and Cariboo-Prince George, voters influenced to switch from NDP to Liberal by the national wave, despite the weak historical base of Liberal support and the best available local polling, caught a wave that wasn't quite big enough to achieve their hoped-for result. They split the vote and elected a Conservative instead.