the one thing I notice, is the top 20 is split 10-10. I also notice the bottom 10 has 3 tories, and only 2 NDPers.

the myth that only rich people vote PC and only poor people vote NDP is proven untrue by this and other similar stats.

I took a second look at the numbers. the Liberal-PC are not only tied at 20, but also at 30, and near tied at 40 and 50 as well. It’s only after you reach the median, that the PC numbers drop off.

This reminds me of a discussion I had with my poli sci prof after he commented he found it hard to beleive more women voted for Harper then for Layton.

all of these things are in proportion. That means if the tories are at 40% and the NDP at 10%, then you are going to see things skewed. More poor people might vote NDP. even if you double the support for the NDP when you get into the poor, that’s only 20%. The tories would have to half thier vote to get 20%. putting it another way, men are in general, taller then women. If you take 100 people, you’ll find that. You will still find women who are taller then some men though. There will always be dirt poor voting PC and some very rich people voting NDP. I myself make $12 an hour, and I voted PC (I voted in advance polls)

You are right that more rich people will vote PC, but that is all in comparison to the total number. maybe the tories get +10% amongst the rich – that is, when they are at 40% provincewide, they are at 50% with the rich. but if the party is very high, or very low, you start to see it get skewed. Remember too this is the average income, not the median income. Also remember, it’s by household, $80,000 is going to go farther in a hosuehold of 2, then it will in a household of 6. I suspect it’s one of the two reasons above that explain ridings like Trinity-Spadina and Parry Sound.

Don Valley West is only leaning PC on the strength of the leader running there. Any other candidate and it wouldn’t even be close (LIB federally too).
Thornhill is currently LIB provincially and federally, and is only swinging PC on the schools issue and a weak LIB candidate, otherwise would be LIB.

I agree with Alpha, plus the first post-debate poll has been released by SES, which shows the liberals at 41% and the PC at 33%. Most of the strength for the liberals are coming from the 416/905 areas; its pretty clear, with numbers like that the wealthier ridings will go liberal, not PC.

The essential flaw is that by assigning each seat a single projected party, you are losing considerable information concerning the strength of that party in the riding. In a close race, it makes little sense to try to relate wealth to partisan preference, especially, since it ought to be the degree to which a party is preferred that is the meaningful result.

You are all committing the ecological fallacy, and assuming that because a riding is, on average, wealthy, the people in it that vote for party X are also wealthy.

That is LIKELY but not really proven by the data.