Topp Spin

There's no positive spin to put on yesterday's by elections for the Liberals, which may be a good thing, because trying to spin by elections too much can make you look rather, well, like Brian Topp:

The Liberal vote in Montreal dropped by about 30 per cent (the Michael Ignatieff-led Liberals got slightly more than 14 per cent in that riding, compared to the Stéphane Dion-led Liberals, who got 20.7 per cent in the 2008 general election).

The Liberal vote in rural Quebec dropped by about 15 per cent (Ignatieff a little over 13 per cent, Dion 15.4 per cent)

The Liberal vote in British Columbia dropped by about 10 per cent (Ignatieff a bit more than 10 per cent, Dion 11.3 per cent).

[...]these by-elections will give the NDP under Leader Jack Layton new energy, new credibility, and an opportunity to focus on providing Canadians with their best alternative to Mr. Harper. As the results show, Canadians are taking a careful look at Jack Layton and the New Democrats.

The Conservatives are holding their rural franchise and filling it in around the edges. But they appear to be in big trouble in British Columbia, and going nowhere in urban Canada. There is no Conservative majority in these tea leaves.

[...]

33,608 Canadians voted Tory in these by-elections, which was 35.72 per cent of the total vote cast.

22,783 Canadians voted New Democrat, 24.22 per cent of the total.

19,709 Canadians (perhaps we should say, Quebeckers) voted Bloc, 20.95 per cent of the total.

And 13,914 Canadians voted Liberal, 14.79 per cent of the total.

Ergo, in terms of the absolute national poll, Mr. Harper got in the range of what he had in the 2008 election. Mr. Layton is a solid second, improving nicely. And Mr. Ignatieff led his party to fourth place in terms of total vote, behind the Bloc Québécois.

So, to recap, the Liberal vote drops from 11.3% to 10.3% in a BC riding, and, ergo, this translates into disastrous news for the grits across British Columbia. Some people look at a smudgy window and see Jesus' face - Brian Topp looks at a smudgy window and sees NDP victory.

Now, I don't want to dwell on this too much because, to be honest, if there was any good news, the Liberals would be doing the exact same spin-job. However, I would like to point out to Topp that in the March 2008 by elections:

1. The NDP vote in British Columbia dropped by 10% (from 16.1% in 2006 to 14.4% in the by election).

2. The NDP vote in Ontario collapsed by nearly 50%.

3. The Liberals were in majority territory with a total of 48% of the vote in these by elections, while the NDP were just 2 points up on the Greens (12% vs. 10%).

And even I'll concede the Dippers had a fairly good 2008 election, gaining seats. So, you know, maybe those by elections weren't the best predictor of what was to come 7 months later.

I'm not saying by elections are meaningless - my post yesterday, which looked at the numbers, shows they're about twice as useful as past election results when it comes to predicting electoral outcomes.

But trying to predict how an entire city or province will vote, based on a few shifting percentage points (when only a quarter of people are voting)? Or simulating a national election based on four ridings?

In the words of Gob Bluth - Come On!