A few days ago, the Ottawa Citizen’s Glen McGregor bragged that “an analysis” showed that a Liberal-NDP coalition wouldn’t win based on the last election and people’s second choice ballots. Of course this “analysis” has no source, and seems to fall nicely into the trash bin that’s become Canadian journalism.
However, the analysis forgot that the Green Party wholly supports the coalition.
Going off raw electoral data, I found that if all votes (from Libs, NDP and Greens) were pooled against the Conservatives in each riding, we find that this coalition could win upwards of 174 seats! This includes 2 for the Greens (Central Nova and Bruce-Grey-Owen Sound). The raw data is available below for a riding-by-riding comparison. In my system, in any riding where the Greens plus Liberals plus NDP win, they get a seat.
However, the only way this result would show (or something close) would be under a STV type system where we ranked our candidates. Since I would absolutely never advocate candidates not run against each other (democracy won’t be helped by less choice people).
Ideally though, what the flawed Ottawa Citizen study shows is that we desperately need electoral reforms, either in the form of STV (like BC is looking into), or a variance of proportional representation (like MMP).
My main point is this though: the “analysis” they perform is crap. I just made a completely different analysis and proved their asses wrong. Someone else may find something completely different. It’s too bad our media portrays all of this as Conservative Party God-given fact.
My progressive-results, there are likely a couple mistakes, but 174 >> 155 (the needed majority number) and I need to study, so I’m not going to double check everything.
These numbers came from the raw elections Canada data.
WordPress
Facebook
Google +
Ian, you’re a bit hard on Glen McGregor, who has done some very good data analysis for the Ottawa Citizen.
I believe by using the latest Ekos numbers for second-choice preference, he was trying to address some problems with the simple additive methodology you propose. But neither his methodology or yours addresses the issue of turnout, who stayed home in the last election, why, and whether they might have shown up to vote under other circumstances.
I blogged about this a bit at the Pundits’ Guide, and would appreciate your comments.
Yes some of those who did not turn out will show up to vote if the system is reformed. I have voted for every major party, to little effect. I have declined ballots but they are never reported and now you have to announce that you are declining so there goes the secret part.
Not voting at all, a first for me, turns out to be the most effective method of registering a protest vote. If you want change don’t vote.
As far as numbers not working…rather than twist yourself into a pretzel trying to undo the will of those that did vote why not take it like a man and accept the results? Oh wait that would be like democracy, which, I guess, is only acceptable when it works as it is “suppose to”. Like capitalism people only like it when it goes their way. Maybe try taking power the old fashion way, steal it! It has worked in the past. grin