During the 2008 election, CTV and its pollster the Strategic Counsel attempted a different look at the state of the race. With an idea undoubtedly inspired by the close red-blue states in recent American elections, they attempted to pick the closest races in three Canadian provinces and argued that the outcome of these races would determine the outcome of the election.

Points for trying something new, but as I've argued here previously, more of the ridings that changed hands in the 2000, 2004 and 2006 general elections had previously been won by more than 5% than had been won by less than 5%. Simply being a close-margin riding last time is no guarantee of being a close-margin riding this time, nor one that changes hands party-wise. Other factors, such as incumbency, which party is the government, the strength of the local candidates, and other local and regional factors, play at least as big a role as previous margins of victory.

In fact once we examine the current set of results, we see that this pattern repeated itself again in the 2008 General Election:

• Of the 42 Ridings that changed hands party-wise in 2008, 19 had been won by <= 5% in 2006, while 23 were won by > 5%.
• In other words just 19 of the 49 "Closest Contests" in 2006 (i.e., ridings won by <= 5% in 2006) subsequently changed hands in 2008, while the other 30 of them did not.
• Moreover, only 13 of the 42 "Closest Contests" in 2008 were also among the 49 "Closest Contests" in 2006 .
• So, only 13 ridings in Canada were settled by 5% or less in both 2006 and 2008. Just under half of them, 6 in fact, actually changed hands in 2008 (West Nova, NS; Saint John, NB; Papineau, QC; Welland, ON; London West, ON; and Vancouver Island North, BC). And only 2 of those (Papineau, QC and Vancouver Island North, BC) had changed hands in 2006 as well.

OK, so "close ridings" are not necessarily "swing ridings", nor necessarily close the next time around. But another claim that CTV and the Strategic Counsel made was that these ridings would decide the 2008 election. Does that assertion hold up?

Well, the change in winners of the "Battleground Ridings" did predict the direction of the election's outcome, but only about one-third of the magnitude of the swing from the Liberals to the Conservatives (and to a lesser extent the NDP).

Category	Cons	Lib	BQ	NDP	Ind
Actual 2008 GE Total	143	76	50	37	2
Non-Battleground 2006 GE (A)	108	86	43	25	1
Battleground 2006 GE (B)	16	17	8	4
Original 2006 GE Total (A + B)	124	103	51	29	1
Battleground 2008 GE (C)	23	8	8	6
Battleground Total (A + C)	131	94	51	31	1

Seven of the Ontario "battleground" ridings changed hands since 2006 (all previously held by the Liberals; 4 went to the Conservatives and 3 to the NDP). Two of the Quebec "battleground" ridings changed hands (the Bloc picked up a Conservative seat, and the Liberals picked up a Bloc seat). And four of the B.C. "battleground" ridings changed hands (the Conservatives picked up 1 NDP and 3 Liberal seats).

In other words, of the 42 ridings that changed hands party-wise in 2008 as compared with their 2006 General Election results, only 13 were found amongst the 45 CTV-Strategic Counsel battleground ridings (CTV counts 14, but they include changes since dissolution such as Halton, ON, where incumbent Garth Turner was elected as a Conservative but crossed the floor and ran for re-election as a Liberal). Another 14 were from provinces other than Ontario, Quebec and British Columbia. And the remaining 15 were Ontario, Quebec and BC ridings not included in the CTV-Strategic Counsel "battleground riding" list.

I do think it's interesting to try and predict ahead of time the battlegrounds on which an election will be fought. Unfortunately, I think the "close margin battleground" approach might be better suited for countries with two-party systems, whose elections turn on the votes of a few swing states. With Canada's multi-party parliamentary system, its 308 ridings spread across very distinct regions, each having different sets of parties in contention, this approach was perhaps not ideal.

If I were picking ridings to watch, I would have selected all the 3-way and 4-way races from last time, the seats with retiring incumbents, floor-crossers, or where previous incumbents were coming back to challenge the first-time incumbents who had defeated them in 2006, along with the traditional bellwether seats such as Peterborough, ON and Ottawa West – Nepean, ON. In this election I would also have added in the likely target ridings of the Conservative Party's ethnic outreach campaign, along with Quebec ridings where the Conservatives placed second in 2006 that had a military presence, and rural Maritime ridings especially those with no incumbent.

For the record, here are the CTV-Strategic Counsel Battleground Ridings for the 2008 General Election, their 2006 (or last by-election) margins of victory and contests, and their 2008 margins of victory and contests.

Riding	2006 Margin	2006 Cont	2008 Margin	2008 Cont
Ontario (20)
Parry Sound – Muskoka	0.1% 28	Cons-Lib	25.2% 10,960	Cons-Lib
Glengarry – Prescott – Russell	0.4% 203	Cons-Lib	10.4% 5,662	Cons-Lib
St. Catharines	0.4% 246	Cons-Lib	17.2% 8,822	Cons-Lib
Hamilton East – Stoney Creek	0.9% 466	NDP-Lib	13.4% 6,464	NDP-Lib
Brant	1.0% 582	Lib-Cons	8.8% 4,793	*Cons-Lib
Thunder Bay – Superior North	1.0% 408	Lib-NDP	8.7% 3,104	*NDP-Lib
Oakville	1.2% 744	Lib-Cons	9.9% 5,483	*Cons-Lib
Thunder Bay – Rainy River	1.7% 658	Lib-NDP	8.0% 2,884	*NDP-Lib
Huron – Bruce	1.8% 971	Lib-Cons	11.8% 5,846	*Cons-Lib
London – Fanshawe	1.9% 868	NDP-Lib	12.2% 5,013	NDP-Cons
Ottawa – Orléans	2.0% 1,231	Cons-Lib	6.1% 3,702	Cons-Lib
Simcoe North	2.1% 1,188	Cons-Lib	22.0% 11,658	Cons-Lib
London West	2.2% 1,329	Lib-Cons	3.7% 2,121	*Cons-Lib
Barrie	2.7% 1,543	Cons-Lib	28.5% 15,195	Cons-Lib
Kitchener – Conestoga	2.7% 1,369	Cons-Lib	24.4% 11,649	Cons-Lib
Halton	2.7% 1,897	Cons-Lib	11.3% 7,850	Cons-Lib
Peterborough	3.5% 2,242	Cons-Lib	15.8% 9,213	Cons-Lib
Algoma – Manitoulin – Kapuskasing	3.7% 1,408	Lib-NDP	13.0% 4,344	*NDP-Lib
Burlington	4.0% 2,599	Cons-Lib	15.4% 9,037	Cons-Lib
Mississauga South	4.1% 2,130	Lib-Cons	4.6% 2,152	Lib-Cons
Québec (15)
Louis – Hébert	0.4% 231	Cons-BQ	8.0% 4,649	*BQ-Cons
Ahuntsic	1.7% 834	BQ-Lib	0.9% 423	BQ-Lib
Beauport – Limoilou	1.7% 820	Cons-BQ	4.2% 2,032	Cons-BQ
Brossard – La Prairie	2.2% 1,243	BQ-Lib	0.2% 102	BQ-Lib
Papineau	2.3% 990	BQ-Lib	2.8% 1,189	*Lib-BQ
Charlesbourg – Haut-Saint-Charles	2.8% 1,372	Cons-BQ	11.9% 5,964	Cons-BQ
Hull – Aylmer	3.3% 1,788	Lib-BQ	15.4% 8,125	Lib-BQ
Honoré-Mercier	3.4% 1,743	Lib-BQ	15.6% 7,673	Lib-BQ
Saint-Hyacinthe – Bagot* (ByElec)	4.6% 1,478	BQ-Cons	26.1% 12,524	BQ-Cons
Pontiac	5.0% 2,371	Cons-BQ	8.5% 3,627	Cons-Lib
Jeanne-Le Ber	6.2% 3,095	BQ-Lib	2.7% 1,303	BQ-Lib
Laval – Les Îles	6.2% 3,312	Lib-BQ	16.9% 9,027	Lib-BQ
Gatineau	7.9% 4,267	BQ-Lib	3.0% 1,577	BQ-NDP
Chicoutimi – Le Fjord	9.3% 4,645	BQ-Lib	6.4% 3,057	BQ-Cons
Brome – Missisquoi	10.4% 5,027	BQ-Lib	2.4% 1,204	BQ-Lib
British Columbia (10)
Vancouver Quadra* (ByElec)	0.5% 151	Lib-Cons	8.6% 4,808	Lib-Cons
Vancouver Island North	1.1% 616	NDP-Cons	4.3% 2,485	*Cons-NDP
West Vancouver – Sunshine Coast – Sea to Sky Country	1.5% 976	Lib-Cons	18.1% 10,848	*Cons-Lib
Fleetwood – Port Kells	1.9% 828	Cons-Lib	18.6% 8,887	Cons-Lib
Newton – North Delta	2.3% 1,000	Lib-NDP	5.5% 2,493	Lib-Cons
Burnaby – Douglas	2.6% 1,244	NDP-Lib	1.7% 798	NDP-Cons
Esquimalt – Juan de Fuca	3.6% 2,166	Lib-NDP	0.1% 68	Lib-Cons
Richmond	4.1% 1,808	Lib-Cons	19.0% 8,138	*Cons-Lib
Pitt Meadows – Maple Ridge – Mission	5.2% 2,721	Cons-NDP	18.8% 9,618	Cons-NDP
North Vancouver	5.6% 3,336	Lib-Cons	4.9% 2,820	*Cons-Lib

Labels: 40th General Election, Swing Ridings