Explaining the 2011 Federal Election I: Who Switched to Whom, and When
[Welcome, National Newswatch readers!]
The final data is in from the 2011 Canadian Election Study, and it shows the Liberals, Bloc Québécois and Green Party held even less of their 2008 vote in 2011 than earlier thought.
The provisional "voter migration matrix", published at last year's Canadian Political Science Association, used CES survey data about vote intention from the post-debate 2011 campaign period and compared it against the respondents' reported 2008 vote.
The final matrix, reported Thursday to a workshop at this year's CPSA annual conference at the University of Alberta in Edmonton, was able to plot respondents' actual reported 2011 vote, as collected after the election, against their reported 2008 ballot.
Comparing the two versions of the matrix, it's clear that support for the three waning parties had continued to slide between vote intention and final voting behaviour — in some cases quite significantly.
The final version of the matrix also includes values for two categories of previous non-voters: (i) Abstained – those who were eligible but did not vote in 2008, and (ii) Entering – those who were not eligible to vote in 2008, whether by age, citizenship, etc., but were eligible in 2011.
The data is reported separately for Québec and the Rest of Canada (ROC). Here is the final 2011 matrix in full (the earlier version can be found in my blogpost from last year's conference).
Sources of the 2011 GE Vote, Rest of Canada vs. Québec (Canadian Election Study, Fournier et al., 2012 CPSA conference)
|Reported 2008 Vote (%)|
|Rest of Canada|
Starting with the ROC, we see the following:
- The NDP was able to win over a greater percentage of vote switchers in 2011 than any other party did, claiming 33% of 2008 Green support, 24% of 2008 Liberal voters, and 5.3% of the 2008 Conservatives. In addition, they held some 80% of their own vote from the previous election.
- The Conservatives held an impressive 87% of their 2008 vote outside Québec in 2011, and added around 10% from each of the Green, Liberal and NDP 2008 electorates.
- The Liberals actually held just 56% of their already-depleted 2008 vote outside Québec in 2011 (not 76% as earlier estimated based on the late-campaign vote intention surveys), and they bled 24% of their 2008 vote to the NDP in 2011 (not 13% as earlier estimated). As mentioned, another 10% or so switched to the Conservatives and 3% to the Greens, while 7% just stayed home ("abstained").
- The Green Party held less of its own 2008 vote in 2011 (30%) than defected from them to the NDP (33%). The earlier estimates had placed their retention rate at 42% outside Québec, rather than 30%; and showed a 25% bleed rate to the NDP, rather than 33%.
Amongst the categories of non-voters in the ROC, there are some other interesting findings, that obviously could not be shown using the available data at this time last year, before all the post-election surveys were compiled.
- The NDP and Conservatives posted the best performances in attracting 2008 abstainers to their fold in 2011: 23% of 2008 non-voters showed up in 2011 to vote for the NDP, while 18% showed up for the Conservatives. Another 14% of 2008 abstainers returned to the polls to vote Liberal, 3.7% voted Green, and just over 40% stayed home again in 2011.
- The NDP won nearly half of all the first-time 2011 voters ("Entering") outside Québec who showed up to vote (36%), followed by the Conservatives (20%), Liberals (12%) and Greens (8%). A further 24% of this sample of folks newly-eligible to vote in 2011 reported they did not cast a ballot (i.e., they abstained).
The findings on 2008 non-voters (the abstainers) are not a surprise to me, but should be paid close attention, given the results in many North Toronto and 905 West seats where the Liberal incumbents either held their raw vote or lost less of it than the Conservatives themselves gained. I argued at the time that in cases such as Ajax-Pickering, where Liberal incumbent Mark Holland obtained a virtual identical number of votes in 2008 and 2011, and the NDP did not run a campaign but gained a bit from out of the ranks of the Greens and previous non-voters, the only way to explain the very large gains in raw vote by Conservative challenger Chris Alexander, was to realize that his campaign succeeded in finding a lot of new supporters from out of the ranks of 2008 non-voters.
My thesis is that the Conservatives converted 2008 non-voters (aka "Abstainers") in many of the suburban ridings where the "Liberals stayed home" in the Dion-Green Shift election; while the NDP found new votes from 2008 abstainers in different kinds of ridings where either they had not been seen as a viable choice before, or where their improved on-the-ground organization was better able to motivate the kind of low-voter-efficacy, low-attachment, lower-income voters who often form their base.
Of course, the lion's share of the "Entering" category would be first-time youth voters, and new immigrants. That the NDP could win 36% of them is certainly consistent with data we've seen on both youth and recent immigrant voting from the 2011 exit polls, as would the Conservatives' share of that category.
[Meanwhile, the fact that 60% of 2008 ROC abstainers showed up to vote in 2011 makes it a teensy bit harder to argue in favour of massive widespread vote suppression. No doubt some voters felt discouraged and some felt very discouraged, and others were being actively discouraged by competing campaigns. But overall, turnout was up. The highest reported abstention rates were for the Conservatives in Québec (9.7%), the Liberals in Québec (8.8%) and the Liberals in the ROC (7.1%). I understand some other research has found some very small effects in ridings said to have been affected by misleading calls, but the study using poll-by-poll data may not have taken changing polling division boundaries into account from my initial reading of it. Obviously I need to take a closer look, though.]
Moving on to the Québec results:
- Here the NDP not only kept 87% of its much smaller 2008 raw vote, but gained 31% of the much more sizeable 2008 Bloc vote, 33% of the 2008 Liberal vote, 26% of the 2008 Conservative vote and fully 65% of the 2008 Québec Green support.
- The Conservatives held 59% of their 2008 support, and apart from the 26% bleed to the NDP, saw 5% switch to the Liberals or Bloc, and 10% just stay home; while at the same time they picked up 12% of the 2008 Liberal vote, and 9% combined from the NDP and Bloc.
- The Bloc Québécois retained just 55% of its 2008 vote in 2011 (not 65% as earlier estimated), with virtually all the rest bleeding to the NDP (31%, up from the 24% estimated last year).
- The Liberals held even less of their 2008 vote in Québec, at 43% (not 55% as earlier estimated), with as we've seen another 33% switching to the NDP, 11% switching to the Conservatives, and 9% staying home.
- And the Greens retained just a lowly 13% of their 2008 Québec vote in 2011 (not 29% as first thought), with fully 65% switching instead to the NDP (a stronger result than the earlier 57% estimate).
- Only 51% of 2008 Abstainers from Québec voted in 2011, but the NDP won about 40% of them (21% overall), while the Conservatives claimed another 15% overall, and the Bloc and Liberals split the other 15% evenly.
- Meanwhile, the NDP also won 65% of the sample of all first-time voters in Québec, while the Liberals and Bloc were not identified even once in this category. 6% each of the remaining new voters picked the Conservatives and Bloc, while 23.5% reported that did not vote in 2011.
Obviously there's a lot to be gleaned from these results, and from the other data that has been reported in the election sessions so far. But, given that those sessions all start up again in a few hours, perhaps we'll end here for now and pick up again in the next blogpost.