Closing polls significantly underestimated the Conservative Vote

I have seen a great deal of speculation about what happened at the end but what you can't see are any real signs of movement at the end of the campaign.  Here are the closing campaign polls (I added a couple of polls since the previous posting):

Average of Closing Polls
	C.P.C.	Liberal	NDP	Green	Bloc	Other	Total
CANADA	37.1	20.0	30.9	6.2	4.6	1.1	100
Atlantic	32.2	28.4	34.3	3.5		1.6	100
Quebec	15.2	15.2	39.7	2.9	24.3	1.8	100
Ontario	38.1	27.4	28.5	5.0		1.1	100
Man/Sask	50.5	14.2	28.7	5.6		1.0	100
Alberta	63.6	11.1	17.5	6.5		1.3	100
B.C.	40.8	16.6	34.6	7.6		0.6	100

Here are the average vote shares based on the preliminary count of votes by Elections Canada:

2011 Preliminary Results
	C.P.C.	Liberal	NDP	Green	Bloc	Other
Canada	39.6	18.9	30.6	3.9	6.0	0.9
Atlantic	37.9	29.3	29.5	3.0		0.3
Quebec	17.1	14.1	42.6	2.1	23.3	0.9
Ontario	44.4	25.3	25.6	3.8		0.9
Man/Sask	54.8	12.7	28.9	3.2		0.4
Alberta	67.0	9.2	16.7	5.2		1.9
BC	44.7	13.6	33.0	7.8		0.8

And here is the difference:

Difference 2011 results compared to average closing polls
	C.P.C.	Liberal	NDP	Green	Bloc	Other
Canada	2.5	-1.1	-0.3	-2.3	1.4	-0.2
Atlantic	5.7	0.9	-4.7	-0.5		-1.2
Quebec	1.8	-1.1	2.9	-0.7	-1.1	-0.9
Ontario	6.3	-2.0	-2.9	-1.2		-0.2
Man/Sask	4.3	-1.4	0.2	-2.5		-0.6
Alberta	3.4	-1.9	-0.8	-1.2		0.5
BC	4.0	-3.0	-1.6	0.2		0.3

The Conservatives did much better than the polls suggested.  The difference is the difference between majority and minority. Note that NDP shares were overestimated in Ontario and Atlantic Canada and underestimated in Quebec.  TC doesn't consider differences in the 1-2 point range to be of any significance.  For the Conservatives that only held true in Quebec.

Final polls - the NDP tide continues to rise

The closing polls don't change the story of the election.  Conservatives will finish first, close to their 2008 mark with no majority, the NDP second making spectacular gains, Liberals third and falling, a bad night coming for the Bloc Quebecois.

Average of Closing Polls
	C.P.C.	Liberal	NDP	Green	Bloc	Other	Total
CANADA	37.6	19.8	31.1	6.4	4.3	0.7	100
Atlantic	30.9	28.6	34.8	4.0	0.0	1.7	100
Quebec	15.0	15.4	39.5	2.8	24.5	2.3	100
Ontario	38.4	26.7	29.1	5.1	0.0	0.7	100
Man/Sask	50.9	13.3	29.7	5.4	0.0	0.7	100
Alberta	64.7	11.1	17.2	5.7	0.0	1.2	100
B.C.	40.8	16.1	35.4	7.4	0.0	0.6	100

Gains and Losses since 2008:

Closing Polls
Change since 2008
	C.P.C.	Liberal	NDP	Green	Bloc	Other
Canada	-0.0	-6.4	12.9	-0.4	-5.6	-0.5
Atlantic	1.2	-6.5	8.8	-2.0	0.0	-1.6
Quebec	-6.7	-8.3	27.4	-0.7	-13.6	1.5
Ontario	-0.8	-7.1	10.9	-2.9	0.0	-0.1
Man/Sask	-0.3	-3.9	5.0	-0.7	0.0	-0.0
Alberta	0.0	-0.3	4.6	-3.0	0.0	-1.3
BC	-3.7	-3.3	9.3	-1.9	0.0	-0.1

Turnout will be up

The evidence suggests turnout will be up this year.  Audiences for the TV debates were up, advance voting is up and there is a party that is on a roll, the NDP, and that is likely to bring additional voters to the polls (along with opponents who fear they might form a government.)

Increased turnout is likely to be helpful for the NDP and not for the Conservatives.  Older voters, who are more likely to vote Conservative, already have a high turnout.  Increases are most likely to come from younger cohorts who are much less inclined to support the Conservatives.

Look at the difference, for example, between the Under 25s and Over 65s from Friday's Ekos poll:

	Cons.	Liberal	NDP	Green	Bloc	Other
Under 25	16.4	21.3	33.0	14.3	6.6	8.5
Over 65	48.2	23.0	20.6	3.4	3.5	1.2

I suspect all the youth voting turnout initiatives, the vote mobs and so on, are probably making Mr. Harper nervous.

Liberal Decline

The post mortems on this election are beginning.  Two topics are of particular interest: the NDP wave in Quebec (and, one should add, Atlantic Canada) and Liberal decline.

Too much of the blame for Liberal decline is attributed to Liberal leader Michael Ignatieff but he deserves some of it. The roots of Liberal decline are long-term and complex. TC wrote a series of posts about Liberal decline last year in June and July. Much of the analysis stands up reasonably well. Here are links to that series:
The Liberal Party in Decline: Part One - The Early Years
Part Two: The Chrétien Era
Part Three: The NDP Revives
Part Four: Conclusion

NDP gains in Quebec

It is clear now the NDP will win a huge number of seats in Quebec on Monday, likely more than 50.  In the process the orange wave will wash over some very big names. Here are a few examples of well-known MPs certain to lose:

• BQ Leader Gilles Duceppe in Montreal's Laurier St. Marie
• Conservative Foreign Affairs Minister Lawrence Cannon in Pontiac
• Liberal and former Canadian astronaut Marc Garneau in Westmount-Ville-Marie

No Tory Majority Come Monday

The overall shape of the final outcome is now clear.  The election next Monday will not produce a Conservative majority.  The Conservatives will finish first, followed by the NDP, with the Liberals third and the Bloc a distant fourth.

Here is a chart showing shifts in support from 2008, compared to polls released since Monday, April 25.

Polls - Last week of April
Change since 2008
	C.P.C.	Liberal	NDP	Green	Bloc	Other
Canada	-2.3	-3.3	10.2	-0.6	-4.2	0.1
Atlantic	2.7	-4.4	4.7	-1.4	0.0	-1.7
Quebec	-6.8	-7.3	25.9	0.0	-12.8	1.0
Ontario	0.4	-4.3	5.3	-1.6	0.0	0.3
Man/Sask	-3.3	1.1	2.6	-0.7	0.0	0.3
Alberta	-5.2	2.5	4.9	-0.7	0.0	-1.4
BC	-5.4	2.4	2.4	0.1	0.0	0.6

Apart from a tiny gain in Ontario and a small increase in Atlantic Canada, the Conservatives are losing ground, not making gains. The story of the campaign is the spectacular success of Jack Layton's NDP, with the possibility that their final total will top 100 seats. A perfect symbol is all the commercials the Conservatives are running attacking Michael Ignatieff who no longer poses a threat.

Come Monday night, there will be disappointment for Stephen Harper, celebrations from coast to coast for the NDP, and devastation for the Liberals and Bloc.

The East is Orange

It is becoming clearer that there is an NDP wave not confined to Quebec. Comparing polls in Atlantic Canada released from April 15 to17 (shortly after the debates held April 12 & 13) to those released on April 21 & 22, we see an average NDP gain of ten points, most of it coming at the expense of the Liberals:

	CPC	Liberal	NDP	Green	Other
April 15-17	36.0	38.6	17.0	4.4	4.1
April 21-22	36.7	30.7	27.6	4.2	0.9

This is the point in the campaign where a clear direction is established.  The story of the campaign has been the NDP's growing wave of support in Quebec, with a parallel move happening in Atlantic Canada.

This commentary by former CBC Ottawa bureau chief Elly Alboim captures the significance of the moment:

More often than not, these sorts of break outs cannot be reversed. They represent a collective decision making process that sometimes builds on mounting evidence or sometimes catches media by surprise after events or debates — although this would represent a very slow reaction to a debate. There are notable exceptions like the PC’s beating back the resurgent Liberals in 1988 but they are rare.

Often, the final results overshoot the initial wave. Momentum builds and begins to sweep into ridings that most think are not in play. I’ve been involved in dozens of CBC projection meetings where seasoned political reporters said that it was inconceivable that certain ridings and personalities were lost. And yet they were. Canada is littered with former cabinet ministers who never should have lost.

There have been some small NDP gains in provinces west of the Ottawa River, but so far they lag the progress Mr. Layton's party is making in the east. Conservative support so far is holding but not growing.  Here is another table comparing early national polls to the last few days:

	CPC	Liberal	NDP	Bloc	Green	Other
March 24-28	38.8	25.0	18.0	9.7	7.5	0.9
April 21-22	38.0	24.7	23.5	6.7	4.9	2.0

The Conservatives' hoped for majority remains elusive but possible. Vote-splitting by the opposition may help. The open question is: how far can the NDP go?