- Election Projection: Harper will Lose If and only If, the Moderate Majority Get Out and Vote

Posted: 3:23 PM on April 10, 2011 Globe and Mail
Crunching Numbers
Harper just shy of projected majority as momentum wanes after Week Two, Éric Grenier,
Globe and Mail, Apr. 10, 2011
http://www.theglobeandmail.com/news/politics/harper-just-shy-of-projected-majority-as-momentum-wanes-after-week-two/article1978966/

"the Tories would win 153 seats in the House of Commons, two short of a majority, according to ThreeHundredEight.com’s updated seat and vote projections for The Globe and Mail."

the methodology of these projections seems questionable.

The margins of error are greatly increased when you take a national poll and look at the result on a Province by Province basis, to the point where it is fruitless.

For example, the Nanos running poll.

"In two days, the polling indicates, the party’s support in the province has dropped from 48.4 per cent to 39.3 per cent. "

I would think it is more that:

- restricting the result to a province the size of BC creates too much inaccuracy and I notice that the two results are within the margin of error with each other (±8.1%)

or,

- another possibility is that the polling in BC has simply been inaccurate in and around the period 3-4 Apr. where it took a spike
BC
1 Apr 40.1
2 Apr 42.1
3 Apr 49.7
4 Apr 48.4
5 Apr 41.9
6 Apr 39.9

and the reduction in 9 points is simply the result of better polling in BC.

I find it hard to believe that in BC the Con's would have in reality 50% of the votes, or even 40% - it just doesn't jib with observation.


These seat projections mean very little unless you can project two things:

the degree to which the Moderate Majority consolidate, and in which ridings, to vote Harper and his Con's out of government

and

- voter turnout of the Moderate Majority

To say Harper is at 38.8% without taking into account at the very least the undecided vote, whether they will vote and for whom real means very little.

Harper and the Con's have 33% right-wing extremist die-hard supporters who will both vote and vote for the Con's

the lower the voter turnout the greater percentage of the vote they make up you can throw in a small amount of soft vote but they are subject to the same problems as the opposition parties - voter turn out and voting the same as their prior intentions.

This phenomenon was accentuated in the last election

where voter turnout was down from the previous election (approx 6.6% less) and Liberals received 850,000 less votes (or approx 19% less)

but Con's received approximately the same number of votes (actually 165,000 less or approx 3% less) as the previous election but ended up with 37.7%

and in the Vaughan by-election a few months ago

where voter turnout was way down (32.5%), the Liberals were down approx 10,000 votes and the Con's received approximately the same number of votes


Also,

the best seat projector is the election itself.

if the Con's got 37.7% in the last election and ended up with 143 seats

whereas the projection here is based on 38.8%
(which must be within the margin of error for the poll to the actual result last time of 37.7%)

how is it the projected seats is 153

Lloyd MacILquham cicblog.com/comments.html