Stats on the State of Play, and How We Got Here
[Welcome, National Newswatch readers!]
If you want to know where we're going, it's always worth taking a second to remember where we've been.
In the 2008 general election:
- The 4 vacant seats, all of which had been in the throes of a by-election campaign when the election was called, were all retained by the retiring MPs' parties (3 Liberals and 1 Bloc Québéois).
- 11/33 open seats changed hands
- The Conservatives won 6 of them (picking up 5 from the Liberals, and 1 from the NDP)
- The NDP won 4 of them (picking up 3 from the Conservatives, and 1 from the Liberals)
- The Liberals won 1 of them (from the Conservatives)
- 31 incumbent MPs were defeated (21 Liberals, 4 Conservatives, 2 Bloquistes, 2 NDPers, 1 Green, and 1 Independent)
- The Conservatives won 18 of those seats (picking up 16 from the Liberals, 1 from the NDP, and 1 former Liberal from the Greens)
- The NDP won 6 of them (picking up 5 from the Liberals, and 1 from the Conservatives)
- The Liberals won 5 of them (picking up 2 each from the Conservatives and Bloc Québécois, and 1 from the NDP)
- The Bloc won 2 of them (picking up 1 each from the Conservatives and a former Bloquiste running as an Independent)
Here is how the seats moved from the 2006 General Election, by category, all the way through to dissolution and the 2008 GE, and from then right up until today.
| Lib | NDP | Grn | BQ | Cons | Rest | Vac | TOT | |
|---|---|---|---|---|---|---|---|---|
| 2006 GE | 103 | 29 | 0 | 51 | 124 | 1 | 308 | |
| Minus: | ||||||||
| - Expelled from Caucus | 1 | 1 | 2 | 4 | ||||
| - Floor-crossers -> Cons | 3 | 3 | ||||||
| - Retired | 9 | 3 | 12 | |||||
| - Deceased | 1 | 1 | ||||||
| 13 | 0 | 0 | 5 | 2 | 0 | 20 | ||
| Plus: | ||||||||
| + Floor-crossers <- Lib | 1 | 3 | 4 | |||||
| + Floor-crossers <- Cons | 1 | 1 | 2 | |||||
| + Floor-crossers <- BQ | 1 | 1 | ||||||
| + Byelxn Holds | 4 | 2 | 6 | |||||
| + Byelxn Gains <- Lib | 1 | 1 | 2 | |||||
| + Byelxn Gains <- BQ | 1 | 1 | ||||||
| + Vacant | 4 | 4 | ||||||
| 5 | 1 | 1 | 2 | 5 | 2 | 4 | 20 | |
| Dissolution | 95 | 30 | 1 | 48 | 127 | 3 | 4 | 308 |
| Minus: | ||||||||
| - Open seats lost -> Cons | 5 | 1 | 6 | |||||
| - Open seats lost -> NDP | 1 | 3 | 4 | |||||
| - Open seats lost -> Lib | 1 | 1 | ||||||
| - Incumbents lost -> Cons | 16 | 1 | 1 | 18 | ||||
| - Incumbents lost -> NDP | 5 | 1 | 6 | |||||
| - Incumbents lost -> Lib | 1 | 2 | 2 | 5 | ||||
| - Incumbents lost -> BQ | 1 | 1 | 2 | |||||
| 27 | 3 | 1 | 2 | 8 | 1 | 42 | ||
| Plus: | ||||||||
| + Open seat gains <- Lib | 1 | 5 | 6 | |||||
| + Open seat gains <- Cons | 1 | 3 | 4 | |||||
| + Open seat gains <- NDP | 1 | 1 | ||||||
| + Gains <- Lib | 5 | 16 | 21 | |||||
| + Gains <- Cons | 2 | 1 | 1 | 4 | ||||
| + Gains <- BQ | 2 | 2 | ||||||
| + Gains <- NDP | 1 | 1 | 2 | |||||
| + Gains <- Ind | 1 | 1 | ||||||
| + Gains <- Grn | 1 | 1 | ||||||
| 6 | 10 | 0 | 2 | 24 | 0 | 42 | ||
| + Vacant seats held | 3 | 1 | -4 | 0 | ||||
| 2008 GE | 77 | 37 | 0 | 49 | 143 | 2 | 308 | |
| Note: | ||||||||
| - Vacant at elxn | 3 | 1 | 4 | |||||
| + Vacant seats held | 3 | 1 | 4 | |||||
| - Retired at elxn | 11 | 3 | 4 | 15 | 33 | |||
| + Open seats held | 5 | 2 | 4 | 11 | 22 | |||
| + Open seats lost | 6 | 1 | 4 | 11 | ||||
| 2008 GE | 77 | 37 | 0 | 49 | 143 | 2 | 308 | |
| Minus: | ||||||||
| - Expelled from Caucus | 1 | 1 | ||||||
| - Retired | 1 | 2 | 3 | 3 | 1 | 10 | ||
| 1 | 2 | 0 | 3 | 4 | 1 | 11 | ||
| Plus: | ||||||||
| + Floor-crossers <- Cons | 1 | 1 | ||||||
| + Byelxn Holds | 1 | 1 | 1 | 3 | ||||
| + Byelxn Gains <- Lib | 1 | 1 | ||||||
| + Byelxn Gains <- NDP | 1 | 1 | ||||||
| + Byelxn Gains <- BQ | 1 | 1 | ||||||
| + Byelxn Gains <- Ind | 1 | 1 | ||||||
| + Vacant | 3 | 3 | ||||||
| 1 | 1 | 0 | 1 | 4 | 1 | 3 | 8 | |
| Current Standings | 77 | 36 | 0 | 47 | 143 | 2 | 3 | 308 |
| Note: | ||||||||
| - Now Vacant | 1 | 2 | 3 | |||||
| - Retiring at elxn | 4 | 1 | 2 | 3 | 10 | |||
Below are the average 2008 margins and the average previous margin for the major categories of seats mentioned above. The table shows that my earlier close-race threshold of a 5% previous margin in the Close-but-no-cigar analysis should be tested against a 10% previous margin threshold for predictive strength (as suggested by Bryan Breguet of the new 2CloseToCall blog).
Still, looking at the relationship between previous vs. current margin for just the 31 defeated incumbents, there is virtually no relationship between previous and present margin of victory, (though, granted, incumbency issues were in play in 4 of the outlier ridings).
[Click on graph to open full-sized version.]
Of the seats that changed hands in 2008 which were not Open Seats (i.e., seats with retiring incumbents) — in other words, in the 31 seats in which an incumbent was defeated — the average margin of victory in the previous election was 6.6% while the average margin of victory in 2008 was 8.0%. The same figures for Open Seats that changed hands party-wise were 9.5% and 11.6%.
[It would make the table a little messy to show the standard deviations as well as the means; but they're very high, as you can guess from the above chart, because the values are pretty spread out.]
| Seat Chgd Hands? | Gain | Open | Vacant | TOT | |
|---|---|---|---|---|---|
| Yes | N Avg %Mrg Avg Prev %Mrg |
31 8.0% 6.6% |
11 11.6% 9.5% |
42 9.0% 7.4% |
|
| No | N Avg %Mrg Avg Prev %Mrg |
22 25.8% 32.9% |
4 10.3% 19.7% |
26 23.4% 30.9% |
|
| TOT | N | 31 | 33 | 4 | |
Finally, let's take a look at how the 2008 seats broke down in terms of Contest and Margin:
Count of seats by Winner, Second-place finisher, and % Margin of victory in 2008
| 2008 GE | 2nd place finisher in '08 | |||||||
|---|---|---|---|---|---|---|---|---|
| Won '08 | %Mrg '08 | Lib | NDP | Grn | BQ | Cons | Rest | TOT |
| Lib | %Mrg < 5% %Mrg < 10% Rest |
1 4 9 |
2 - 5 |
9 11 36 |
||||
| All | 14 | 7 | 56 | 77 | ||||
| NDP | %Mrg < 5% %Mrg < 10% Rest |
1 5 11 |
5 2 12 |
|||||
| All | 17 | 19 | 36 | |||||
| Grn | %Mrg < 5% %Mrg < 10% Rest |
0 |
||||||
| BQ | %Mrg < 5% %Mrg < 10% Rest |
4 3 20 |
1 - 3 |
0 3 14 |
||||
| All | 27 | 4 | 17 | 48 | ||||
| Cons | %Mrg < 5% %Mrg < 10% Rest |
11 10 58 |
4 1 46 |
- - 5 |
3 - 7 |
1 | ||
| All | 79 | 51 | 5 | 10 | 1 | 146 | ||
| Rest | %Mrg < 5% %Mrg < 10% Rest |
1 |
1 |
|||||
| TOT | 123 | 69 | 5 | 18 | 92 | 1 | 308 | |
As you can see, the largest group of seats in the last election were Liberal-Conservative (79) and Conservative-Liberal (56) contests. According to some commentators, they believe the Liberals have given up trying to win those back from the Conservatives and are now fishing elsewhere. The Conservatives, on the other hand, are said to be eyeing further gains here.
The next largest group of seats in the last election were Conservative-NDP (51) and NDP-Conservative (19) seats. The NDP has consistently said over the last year that it is primarily targeting Conservative held ridings, and of course we know that the Conservatives have been targetting a few NDP incumbents over the long-gun registry issue, (though primarily in formerly Liberal held seats, as we noted at the time).
A much smaller group of seats in the last election was the NDP-Liberal (17) and Liberal-NDP (14) contests. If you believe some of the commentators in the gallery now (and their sources), these are the seats the Liberals are targetting.
A similar-sized pool of seats last time was found in the BQ-Conservative (17) and Conservative-BQ (10) contests.
Finally, there were also BQ-Liberal (27) and Liberal-BQ (7) contests, and a few other smaller groupings as well.
Tags: 40th (2008) General Election, Close Races, Retiring Incumbents, Vacant Seats
I suspect the government has done a great deal of research on what resonates with voters.
I believe many of their issues on the law and order file will help with the marginal seats.
The NB and Rob Ford results in my opinion demonstrate a simple message with a good ground game can over come a hostile media or incumbency.
Are the Conservatives prepared to fight or invest in 20-30 close close contests?
Given their past performance I think the party is capable of poaching seats from every party in every region.
Does that hold true for every party?
CS, some of the contests they invest in (as with other parties) will not have been “close” last time, but will be cases where they believe, with the right candidate and taking other factors into account, that they can challenge for first place. Like Vaughan.
Alice, the main reason the Tories could form a majority is how few seats they give up once they won them.
In 08 only 4 conservatives did not get re-elected. Take Danny Williams and ABC out of the equation and you would only have had 2 (Jaffer and Luc Harvey) actually defeated in straight up races.
If you look at how well the CPC has done in holding beach-heads you will understand why a majority is very possible.
Every election since 04 the Conservatives have grown their available universe of winnable ridings. Not only have they grown their universe, they have become better and better at winning those ridings.
How we do it is a trade secret.
Experienced, you draw out an important point, but one I think we could elaborate on a bit further.
We’re at one stage of the electoral cycle now, and the Conservative Party has been — and may still be — the party on the ascendancy. Their major competitor to this point, the Liberals, have been moving in the other trajectory.
But like the business cycle, the political cycle has its ups and down, as I’m sure you’ll (reluctantly) agree. [Never great to think about in the wind up to an election, I grant you that.]
So this trend you’ve identified is of only two or three elections in duration, and its effect can’t be statistically isolated from the effect of (a) the Liberals having been in decline, nor particularly (b) the new conservative party being able to increasingly unite the core bases of its predecessors’ support, and of course certainly not from (c) incumbency, which now works in the Conservatives’ favour more so than for the Liberals.
ABC in Newfoundland, and a few of the other issues around incumbency, all make for outliers, I’ll also agree.
But while you might not be trying it with me here, it’s also a recognized trick in the bag of political operatives to try and show their war paint and bare their teeth, in the hopes of scaring off their opponents a bit. Certainly, I had very confident boasts that the Conservatives could siphon off Liberals votes in Winnipeg North to the benefit of the NDP … or perhaps even to win the seat outright … when in fact we saw just the opposite occur once all the ballots were counted.
So clearly there are some exceptions that prove any rule. You have to give the Conservatives points for trying a lot of different gambits. They don’t all work, but the party’s strategists have no shortage of them up their sleeves, for sure.
We are definitely entering a very interesting phase of the pre-election / pre-budget war dance, where the parties are starting to watch one another extremely closely for signals of fear, complacency, moxy and so forth, and to send out as confusing a set of signals in return as they can muster.
It should make for a very enjoyable time on this blog, enlivened by contributions such as yours, for which I thank you.
Alice Winnipeg North is interesting to be sure but the by-election that is really driving the conversation seems to be Vaughan.
IN terms of seats they are targetting the CPC still seems to be in an expansionary phase.
In political circles we are discussing which ridings in the GTA the Tories could win as opposed to which ridings in Alberta the Liberals could win.
(This is telling.)
The gun registry issue has created a few targets in rural NDP areas. Candidate recruitment has created Montreal and Atlantic seat targets.
Of course campaigns and actual voting behaviour matters but structurally speaking the advantage going into this seems to still be with the CPC.
(Although I must agree with your point that when looking at the ability of the CPC to hold on to seats during an election their upward tragectory is a confounding factor.
However, I imagine EXPERIENCED must be on to something by virtue of the CPC having such a large monetary advantage.
Consultants must be good for something right ?)
Well, Vaughan is very interesting in this context, I agree, Shadow. It also helps my point that close margins have to be read together with incumbency, and the phase of the political cycle.
Let’s step back and look at a couple of non-Vaughan cases for a second. Parry Sound–Muskoka, ON (Tony Clement’s riding), was home to Stan Darling of the PCs for dogs years (as my mother would say), but the Liberals won it on the split in 1993, in spite of some very strong candidates coming forward in the meantime (for example, Lewis Mackenzie in 1997). Once the cycle started to turn (due in large part to the merger of the two conservative parties), the margins started getting closer in 2004, and then Clement defeated incumbent Liberal Andy Mitchell by the smallest margin in the country in 2006 (28 votes).
Now, would we expect this seat to be close again the next time? No. The Liberals weren’t running Andy Mitchell again, Clement was by then the incumbent, and the Conservatives were on the up cycle. But I can’t tell you how much commentary I read at the time, saying the riding would be a nail-biter because of the previous margin. All of it hogwash.
Let’s take another example: Thunder Bay–Superior North, ON. A traditional Liberal-NDP contest, the riding was narrowly won by the NDP for a term from 1984-88, but returned to the Liberals in the Free Trade election of 1988 under Joe Comuzzi, who kept it until he retired in 2008 (the last few years sitting as a Conservative). The NDP was in the down part of their cycle from 1992 forward, but started to get some wind in their sails again in 2004.
That party pretty ruthlessly targets its resources to seats it can win, however, and so Bruce Hyer only spent 37% and 43% of the limit in 2004 and 2006 (what he could raise locally at the time). He wound up missing out on the seat by 408 votes in a case of heightened turnout in 2006, and sure enough in 2008 he got an influx of $25K from the central party (you can see all these stats on the “Finances” tab of the riding profile here). Spending 86% of the limit in 2008 was enough to put him over the top by a more comfortable 8.7% margin, which probably also wasn’t hurt by a stronger Conservative vote following the retiring Mr. Comuzzi’s example.
So, there are a number of other factors besides “closeness of the previous race” that one needs to pay attention to. I don’t like to predict seats here, and think it’s unwise to do so before all the candidates are in place, but the significance of the Conservatives’ victory in Vaughan is that it demonstrated Toronto was not impenetrable. Now Mr. Fantino is the incumbent … and he’s also Minister Fantino, a signal that’s undoubtedly been received loud and clear where it mattered.
Even if all the Conservatives have succeeded in doing is to tie the Liberals down defending their incumbent seats, that alone would be a strategic objective achieved.
As to the financial resources question, it’s an obvious advantage. But it takes wise and experienced strategists to spend their resources effectively. I’ve seen campaign resources squandered to little effect in some situations as well, as I’m sure you have from time to time, Shadow.
Thank you for raising some interesting points.
I know that Angelo Persichilli is not popular with some of my Liberal readers, but one way to read his column from today would see it as reflecting the natural consequences of one part of the electoral cycle for any political party.
The factor Angelo doesn’t identify is what wing of the Party the MPs come from. The Blue mostly prolife Libs that took old PC seats in 1993 mostly has come back to the Conservatives. The closer one gets to old Lib seats in Toronto the harder it will be to take those seats away even those that were old PC seats. Albina’s seat will be interesting and Pat O’Brien’s seat.
Hey Alice,
Just fyi, the Halton Liberals are preparing for a nomination contest early next month, with apparently four “interested” candidates and one who has already been green-lit, but no names given yet. Just thought I’d make you aware of it, though.
Alice I think your comment illustrates one of my chief complaints with election predictions.
To determine the ridings that will be close in the next election we need to look at candidates, at how ridings and regions are trending, and at party strategy.
Looking at ridings that were close the last go around is the equivalent of fighting the last war.
Unfortunately that’s basically what election forecast models tend to do. They translate shifts in %vote into riding results based on the past relationship between %vote and riding results.
Sometimes as few as 3 cycles are used.
The problem here is obvious. Over the past 3 cycles the CPC have been in geogrpahic expansion.
If we assume they are still expanding into the GTA than any modelling based on past elections would be COUNTER to this trend.
It would stack up their vote in old ridings (where it is wasted) and short change them in new battlegrounds (where it is needed).
Basically it wouldn’t capture the increasing efficiency of the CPC vote as they’ve gone from a Western regional powerhouse to a national party.
Volkov, thanks for the heads-up. As soon as you have a date, I’ll add it to the list on the side.
George, I think that’s a useful addition to the analysis for sure. But I thought Pat O’Brien’s seat was London-Fanshawe, and that he left he Liberals over same-sex marriage to sit as an Independent before 2006. Am I mixing up my Liberal MPs?
Shadow, that’s a pretty good summary of the challenges a quantitative prediction methodology faces, in light of the dataset we currently have.
On the bright side, if we keep having elections at this pace, we’re going to be building up a much bigger dataset, eh! ;-)
Has Elections Canada released the Winnipeg North riding results by polling station yet?
It will be interesting to see if the vote was split on racial lines with the southern part of the North End voting for aboriginal Kevin Chief while the northern part flocked to the Liberal candidate Lamouroux.
With the Liberals “quintupling” their vote in the riding, something is amiss.
Oops, it was Paul Szabo I was thinking of. Derek Lee and John MacKay can also be added to the list when they leave.
Boy, you are really pushing this racial thing. Of course even if the results have been published (I’ll check in a second, but it appears it should be soon), they will tell us regionally who voted for who, but that won’t absolutely prove anything about the ethnicity of the voters.
Are you actually interested in any other dynamic of election results, or just in proving this one particular pet theory of yours? You certainly are “consistent”, I’ll give you that.
PS, still no “Official Voting Results” (what they call the poll-by-poll results) at Elections Canada yet. Sorry to ruin your weekend.
Those last two comments weren’t directed at you, of course, George. Sorry about that.
I agree that the Conservatives are heavily targeting both the open seat of Mississauga East-Cooksville being vacated by Albina Guarnieri, and Paul Szabo’s seat of Mississauga South. Though in the latter case, it seems that some of the travails of the Conservative association have spilled out into the public. Nevertheless, candidate Stella Ambler (a former aide of Jim Flaherty’s as I understand it) would appear to have the confidence of their central campaign.
There hasn’t been the same evidence of heavy Conservative targeting in the Scarborough seats yet, at least not from a distance. It’s really hard to see past the incumbency factors there, and figure out what could come next, though I did notice that Michelle Simson took over Scarborough Southwest fairly smoothly last time after the retirement of Tom Wappel, and even notwithstanding a little local frisson over her appointment by Stéphanie Dion.
The Liberals defeated 5 incumbents in 2008 (picking up 2 each from the Conservatives and Bloc Québécois, and 1 from the NDP). Only two in Quebec? But this disguises the fact that Dion picked up votes in Quebec in 2008. In fact his campaign was a success in two major ways. Harper called the election early to get a majority; Dion stopped him. And Dion was a winner in Quebec. He must have been startled when his non-Quebec advisers counselled an instant resignation, prematurely. That story has not yet been written, unlike “How we almost gave the Tories the boot.”