Without a doubt, the reason for the major shocker on the election night were the NDP voters going Liberal in the last minute out of fear of the big, bad politician from Alberta. The NDP fell from 20% to 15% and the Liberals increased by a similar amount. But was this a smart move? Did they save the government from Harper?
To test this, I took 5% of each riding's vote and shifted it from the Liberals to the NDP. I ignored Quebec and simply gave 1 Liberal seat to the Bloc. Now, since voter shift didn't happen in Quebec and the "key" ridings voted strategically more than, say, Red Deer Alberta, this is likely minimizing the impact of the strategic voting. But it gives you an idea of what the effect was.
So, as a result of this shift, the following changes occured on election night:
8 NDP ridings went Liberal
5 NDP ridings went Conservative (mostly in BC)
7 Conservative ridings went Liberal
With our 1 Quebec seat, this gives us election night results of:
Wow! Good thing those NDP voters jumped ship, eh? All the strategic voting did was cost the Tories 2 seats, at a price of 13 NDP seats. Once again, the two coalitions have similar numbers but the NDP lost well over a third of their seats due to strategic voting. Prominent victims include Lorne Nystrom and Olivia Chow.
I think the NDP should comision someone who knows what they're doing (unlike myself) to do a study and find out just what the strategic voting likely cost them. Since we're in for an election in another year, they need to stress the point that it doesn't work before the next election.