Friday, November 29, 2013

The Federal By-elections: What Stories Are They Telling Us?

Most media interpretations of the by-elections lumped them together. It is true that in all instances Conservative support dropped substantially. TC's view is that there are different stories that need to be told to understand what happened.  Since half of the by-elections were in Manitoba, it makes sense to start there.

Manitoba - Provencher and Brandon-Souris

Although the Conservatives won both Provencher and Brandon-Souris, their margins were sharply reduced in both places (the Senate scandal was no doubt an important factor) and the Liberals actually came close to winning in Brandon. Indeed this analysis by Brandon analyst Deveryn Ross asserts that, but for a series of strategic errors including not having leader Justin Trudeau campaigning there last weekend, the Liberals could have overcome their 391 vote margin of defeat. Many have called Brandon a safe Conservative seat but that is an over-simplification. First, while many analysts referred to it as rural, more than half the population lives in Brandon, a large urban centre.  In provincial politics its two ridings have delivered more votes to the NDP than the provincial PCs in the last four elections. Brandon also was never won federally by the Reform Party or the Canadian Alliance. A potential to challenge Conservative dominance has been there for some time and as the constituency becomes more urban over time that is likely to remain. Provencher, despite being won three times by the Liberals since 1968, is currently the safer seat because of its large socially conservative population.

In Manitoba there is a strong reciprocal relationship between Liberal and NDP levels of support both federally and provincially: when one of the two parties is strong, the other is weak. Over the past forty years, voters have quite comfortably switched back and forth between the two. It is clear from voting patterns that the provincial NDP has depended on federal Liberal voters in provincial elections. A spike in Liberal strength in 1988 when the Liberals came close to winning a provincial election was accompanied by success in the 1988 federal election later that year. This strength was matched by NDP weakness from the late eighties into the mid-nineties before reversing.  Below is a graphic on Manitoba federal politics; there is a similar pattern provincially.

Recently polls conducted by Probe research, a Manitoba polling firm, have reported declines in NDP strength both federally and provincially. Liberal support has increased.

In particular, the provincial NDP government of Greg Selinger has been experiencing political weakness since increasing the provincial sales tax in last April`s budget. The NDP government, with the exception of the virtual one-party state in Alberta, is the longest serving Canada and the strains of age are beginning to show.

The Manitoba-Saskatchewan sub-samples of National polls also report increased Liberal strength. The results of the federal by-elections appear to fit these larger patterns. Overall the Manitoba by-elections were a good news story for the Liberal Party.

Toronto Centre

Toronto Centre was less competitive than has been generally perceived. As TC noted in a previous post:
The impending by-elections in the Liberal held seats are being touted as a "test" for Justin Trudeau. However, these were ridings held by the Liberals in 2011 at their lowest point ever. If the recent polls mean anything at all the Liberals should be in a position to easily retain them. I think this will be true of Toronto Centre, where the latest numbers would put the Liberals twenty-five percentage points ahead.
As it turned out Liberal Chrystia Freeland finished 13 points ahead of New Democrat Linda McQuaig. Relatively speaking, this was a strong performance for the NDP, and a weak one for the Liberals despite widespread perceptions to the contrary. The fact that the Liberals were at a reported 37% in Ontario in the new Ipsos poll only reinforces the argument that the party under-performed in Toronto Centre relative to what it should have done. The Liberals only received 19% in the 2011 federal election in Ontario. Given that their poll support is now almost double we should have seen a much greater vote share for them in the by-election.

The NDP performance was partly a product of a strong effort by the NDP. But it always struck TC that the NDP had no realistic chance of winning. Nonetheless it seemed they came to believe they could overcome the odds and the party allowed public perceptions to see a race that wasn`t actually close as winnable. I found much to agree with in this analysis by Robin Sears who is referring to both Toronto Centre and Bourassa:
If they had not allowed themselves to be pushed into the “close NDP/Liberal fight” narrative before election day, their message could have been, “That the Liberals couldn’t win back their traditional support in these party bastions, even in a byelection, should make them a little scared about the next election. We are very proud that we held onto the high-water mark that Jack Layton established for us for the first time in these Liberal fortresses.”
The NDP`s unduly optimistic assessment of the race apparently also led to the adoption of churlish and ineffective tactics such as attacking Chrystia Freeland for not having lived in Canada for 10 years. This tactic distracted from the core NDP message on inequality. Although all parties do this from time to time, they are also all hypocritical in this regard and as far as one can tell, wasting their time and energy.

However, the debate over income inequality likely helped the NDP overall. Linda McQuaig had a more focused and aggressive position on this while Freeland was cautious and platitudinous. In one way the issue may have helped Freeland, albeit in an ironic sense. The Forum polls taken during the race (about which more in a future post), while highly inaccurate in Manitoba, did come close to the mark in Toronto Centre.  If you look at the table on page 3 in this press release from Forum, one notices a steady decline in the Conservative support over the five surveys taken during the race. It would be my guess that some Conservatives saw Linda McQuaig as a threatening figure who should be stopped, and therefore voted for Freeland. If this is true (I am only speculating), it would make the Liberal performance appear that much weaker.


This by-election was held in a long-time safe Liberal riding.  However, the pattern of Quebec politics was upset by the orange wave in 2011 so it was worthwhile for the NDP to take a stab at winning here despite the long-term pattern of voting in this constituency. If one looks at recent province-wide polls in Quebec one sees that the Liberal Party has more than doubled its 2011 support.  However, in the Bourassa by-election, Liberal support rose from about 41% in 2011 to 48% in the by-election, a much smaller overall increase. The NDP`s support slipped a bit but did hold close to its 2011 support. At the very least this suggests the NDP will remain an important player in Quebec in the next election. It is nonetheless a weak (and no doubt disappointing) showing for the NDP given the effort and resources the party was able to focus on this race. Its support in 2011 was part of a wave election and required no special effort then. However, as was the case in Toronto Centre the large increase the Liberals have achieved in recent Quebec polls should, in TC`s view, have given them an even better outcome in Bourassa.

The BQ also slipped a bit further in this by-election (they actually won the riding in 1993) and there may be some larger potential significance there. I have been struck by the weakness of the BQ brand that appears in the recent series of polls conducted by Nanos Research designed to measure "composite federal party brands".  If you look at the Quebec section of the most recent Nanos release (on page 9), you will see that the BQ is running behind even the Green Party in Quebec. It was clear in 2011 that the huge losses suffered by the BQ went mainly to the NDP. Is this survey an indicator that the Bloc might suffer additional losses next time (they did retain 23.1% of the vote in 2011)? If so it is likely the NDP that would benefit most.

Overall, there were a variety of patterns evident in the by-election campaigns and results and they provided a great deal of information to be considered about the state of Canadian politics.