One of the great clichés of quantitative analysis is "In God we trust. All others bring data." So let's apply some data to Thursday's Ontario provincial by-elections.
First, let's test this spin that we saw a major shift in public opinion last evening because of the results in Ottawa West-Nepean.
We have had a number of by-elections in different environments around the province. Some assumptions can be drawn:
» The weakness in the PC vote in rural Ontario that saw John Tory defeated in Haliburton-Kawartha Lakes appears to have firmed back up with the change in leaders. They are no longer in danger of losing more of their rural base seats.
» The Liberal vote remains similarly solid in the urban seats and among new Canadians, given the shape and scope of the results in St Paul's and Toronto Centre.
» There is a swing to the PCs and from the Liberals in suburban seats of around 7.1 per cent, given the results in Ottawa West-Nepean.
Now these are assumptions, not facts. It will be impossible to prove they are real until the general election, and even then, the assumptions only apply to current conditions, not conditions in October 2011. But they are the best data sets we have to date.
So what is the impact of a 7 per cent weakening of the Liberal vote on the seat count? And what is that impact if found particularly in seats like Ottawa West-Nepean? That is, swing seats that are affluent and have only a smattering of new Canadians, not the ethnic enclaves like Malton or the modest-income 905 like Durham.
Here are the seats that flip to the PC Party with a 7.1 per cent swing. (Liberal to PC swing required to win the seat follows each.)
Northumberland 7.1
Prince Edward Hastings 6.8
Richmond Hill 6.6
Mississauga South 6.1
Oakridges Markham 6.1
Willowdale 5.9
London Fanshawe 5.6
Don Valley West 5.4
Stormont Dundas 5.1
Lambton Kent 3.5
Ancaster-Dundas-Flamborough-Westdale 3.4
Algoma-Manitoulin 2.8
Eglinton-Lawrence 2.7
Kitchener Conestoga 2.4
Barrie 1.5
Nipissing 0.7
Only six of those seats are anything like Ottawa West-Nepean: Barrie, Kitchener-Conestoga, Ancaster-Dundas-Flamborough-Westdale, Oakridges-Markham, Mississauga South and Richmond Hill. These are all affluent and relatively ethnically homogenous areas in the suburbs.
I wouldn't want to draw any assumptions about how a Northern Ontario seat like Nipissing, an urban Toronto seat like Eglington-Lawrence or a southwestern rural seat like Lambton-Kent-Middlesex would vote based on the swing in Ottawa. In fact, based on the results in St Pauls or Toronto Centre, I would doubt Eglinton-Lawrence, Willowdale or Don Valley West are likely to move at present.
However, if we follow John Baird's sweeping claim and transpose these results across the entire province, we still only see 14 seats change hands, a far cry from "celebrating a Tim Hudak majority government."
That leaves the PCs stuck at just 39 seats, and Dalton McGuinty with a third majority of 56 seats. That's about the same size majority as the 59 seats Mike Harris had after the 1999 election.
And let's keep in mind that is even being generous and transposing the figure across the province, not just in the six identified "like" seats that would switch with a 7 per cent swing.
So the idea that this was a moral victory because the vote shifted to Tim Hudak is false. It was a loss. Even if we transpose the OWN shift across the province, it still leaves Dalton McGuinty with a majority.
More to the point, transposing by-election results is useful but the battle is for seats, not votes.
After all, the man himself cited a massive victory in the seat left vacant by Mike Harris's resignation, based on a nine-vote margin. A CP wire story from that 2002 by-election: "'It's a real kick in the teeth to the Liberals. Dalton McGuinty's going home with his tail between his legs,' said cabinet minister Tim Hudak."
Spin keeps one warm at night, and kudos for the PCs for searching out every silver lining in this dark cloud, but data is fact. And the fact is even a swing of this proportion leaves Tim Hudak facing another Dalton McGuinty majority.
The second thing to test is why there was a 7 per cent swing.
By-elections are very different from general elections because turnout is lower. When there are fewer voters, the electorate changes its composition. It tends to become more partisan and deviate more from the political mean.
The need for an effective get-out-the-vote machine also goes up, because fewer total votes means the marginal gains from knocking on doors and driving people to the polls is magnified significantly.
The hypothesis therefore is this: recent by-elections saw party turnout impacted by proximity to political staff and other organizers.
Both by-elections where the Liberals swept to near majorities of the vote were in areas where they could best consolidate their election machine, that is downtown Toronto. The 416 and inner 905 region is chock-a-block with Liberal activists, and it is also home to the 500-plus paid political staff who work for the party, caucus and ministers.
The situation in Ottawa is the opposite. Not only is Ottawa a four-hour drive from Toronto, limiting the ability to get their ground troops in place, it is also home to the federal Conservative government. The federal Conservatives enjoy almost 1,000 full-time paid political staff, between the party office, MPs offices, PCRS and the ministers and secretaries of state offices.
While Thursday was budget day - which likely limited the overall number of organizers free to pull vote - even a 10 per cent turnout of federal Conservatives would have meant a massive infusion of 100 bodies.
If turnout, as opposed to vote switching, was the culprit, then we can attribute at least part of the 7 per cent swing to a by-election organization effect.
Sure enough, the turnout in strong Liberal polls is down about 50 per cent, compared to a decline of just 25 per cent among strong PC polls. This is based on checking turnout in five random strong Liberal polls from 2007 vs turnout in five strong PC polls in 2007, using unofficial election night data.
Basically, the PCs were able to get their vote to the polls, the Liberals got enough to win, but had some trouble maintaining their general election base, likely because of less ability to apply superior numbers to the get-out-the-vote challenge.
This is not an uncommon phenomenon in by-elections. Liberal voters tend to be centrists, new Canadians and skew younger. All three groups are harder to mobilize. Centrists don't hold the hot-button political views of base Conservatives or New Democrats. New Canadians can be less plugged into both mainstream media and local politics. Younger voters are notoriously hard to get to the polls unless gripped by an issue. To do well in by-elections, Liberals need to almost drag their voters to the polls, as they clearly were able to do in St Paul's and Toronto Centre.
I'm not prepared to write off the entire 7 per cent swing to voter turnout, but some subset of the swing is certainly a function of a temporary get-out-the-vote phenomenon based on the Conservatives ability concentrate significant numbers in the Ottawa region.
Third, the on-going struggle between the two third-place parties continues unabated.
The Greens did improve their margin of victory in Leeds-Grenville by a couple points, and almost passed the NDP in Ottawa West-Nepean.
But the results of this election should be troubling for both parties.
For the NDP, there is some hope in that the drop in their vote was just around a single percentage in each seat. Extrapolated out to the general election, and incorporating the results in the other by-elections, the NDP narrowly remain a party that is viable, but only in a handful of seats. Everywhere else they are in danger of falling into fourth place.
There hasn't been a Northern Ontario by-election this cycle, and that would test the most fertile ground for the NDP. But the result of these by-elections confirms the assumption that the NDP is somewhat akin to the Bloc Quebecois or the old Reform Party; a regional party of the Northern-Industrial-Inner City axis with no room for growth outside that redoubt.
In fact, the NDP appears to be aiming for just a dozen seats in the next election, an achievable goal but one that underlines how marginal they are in the vast majority of Ontario seats.
The Greens have the opposite challenge. They have the potential for growth, but no geographic anchor to win seats in. The NDP at least has a corporal's guard of ridings where they can win a plurality of the vote. The Greens have no seat within reasonable striking distance where they could even pool their resources behind a single candidate.
A two per cent increase in the vote on their eight per cent base, while positive, still leaves them about sixteen election cycles short of a workable majority of seats. That means a Green Premier by 2074 at the latest.
Parliamentary democracy is about seats, not votes. I would rather have the NDP's problems than the Greens'.
Fourth, there is the spin that the HST was the issue of the campaign and responsible for moving vote to the PCs.
There have been four by-elections since the HST was announced. Two were massive Liberal victories, one a solid Liberal victory, and the fourth a rout for the PCs in a seat they have held since time immemorial.
A month ago, pundits were concluding that the HST was a squib of an issue because of the victory in Toronto Centre.
To suddenly conclude that province-wide public opinion about the HST has shifted - in a single month and in the absence of any new information - is a bit absurd.
The HST may have been finding slightly more purchase among suburbanites than urbanites and new Canadians all along, but going back to our swing analysis, a 7 per cent swing to the PCs doesn't get them anywhere near forming a government.
If this by-election were held after July 1,2010, when the sales tax reform takes place, I might buy an argument around HST increasing in saliency. But I certainly do not for explaining Ottawa West-Nepean. Proximity to political staff appears to be a more robust cause.
Finally, there is some potential for the Liberals to relax after this string of wins. "Resting on your laurels," my mom used to call it when I talked about past achievements instead of future plans.
This is about the worst thing the Grits could do.
The last twelve months have been the most activist and dynamic of the Dalton McGuinty era. From the harmonized sales tax to the Green Energy Act to full-day kindergarten to last week's announcements about a clean water tech industrial strategy and chromite mining, the Ontario government is making news and shaking things up far more than their rather stately (and at times sluggish) first five years.
Those relaxed first five years saw the Liberals lose seats in by-elections in Hamilton East, York South-Weston and Parkdale-High Park.
These dynamic years saw them pick up a seat in Haliburton and hold vulnerable ridings with solid margins.
If there is any hesitation around downplaying dramatic elements of the coming Throne Speech or Budget, the lesson is simple: voters want you to put your foot on the gas and keep going.
The data says to not even tap the brakes.