Another Ontario provincial poll

The Ontario provincial Liberals are still slightly ahead according to the latest Ipsos-Reid poll conducted during the second half of June. Based on the poll numbers, it would suggest that the release of the Conservative platform did not give the Tories a boost, perhaps because it was vague and contained the controversial religious school funding proposal. The poll result showed the Tories drop one point, and the Liberals drop two points which seemingly moved to the Greens. The result is a 3 point lead for the Liberals, 39% to 36%. Noticeable is that the Tories are hardly any higher than their 2003 election finish. If these numbers held on election day the Tories would pick up seats but only because of a vote split due to a surge in third party support. Although the article wants again to talk about a possible minority government, I have again put the numbers into an election forecaster and found it predicting a narrow Liberal majority government. This proves that the supposedly “magic” 40% citied both federally and provincially needed for a majority government is an arbitrary number. It is apparently theoretically possible for a majority government to be formed with 39% of the vote, only 3 points ahead of the nearest opposing party. A 39% to 36% result could produce a minority government of course, but a majority government is apparently also quite possible. It is important to remember that a large spread between the top two parties in a first-past-the-post general election is not needed for a majority government. For example, in the 1999 election the Tories beat the Liberals in the province overall by just 5%. Yet this produced a moderately large majority government. A party winning a majority government yet beating its nearest competitor by only 3% in the popular vote is entirely possible. It is true that the chances of a minority government in Ontario is the most possible it has been since the 1985 election (which did produce a minority government in the end), but at this point such an outcome is far from certain. The media who keeps bringing up the strong possibility of a minority government may be jumping to conclusions. The Liberals should not be counted out yet, but nor should the PCs.


There is little point at looking at the regional breakdowns given for the poll, because I believe the sample size is too small for them to mean anything. For the poll to show only a single point lead for the Liberals in the GTA, the GTA surely has to include places like Burlington, Oakville, and maybe even Hamilton. Otherwise that regional breakdown wouldn’t make sense. What makes even less sense is the result outside Toronto. Supposedly, according to the poll, the Liberals hold a stronger 39% to 34% lead outside the Toronto area. I suppose this is believable, but only if very conservative regions of the GTA held great sway in the poll. I suppose it is possible that the poll may have over-sampled GTA places such as Whitby, Oakville, Oshawa, and Burlington. If so than the Liberals would in reality have a larger province-wide lead than the poll indicated. The strangest result is in central Ontario, which gives the Liberals a 5 point lead of 41% to 36%. If this result were actually true I’d say the Liberals have a very strong chance of re-election. Central Ontario includes the very Conservative Haliburton/ Kawartha Lakes electoral district, the county of Northumberland (The Northumberland-Quinte West riding is a key swing riding that usually goes to the party who wins the election both federally and provincially), the key bellwether riding of Peterborough, as well as Conservative Simcoe County. Although the region contains a couple swing ridings, most parts of this region are rock solid Conservative and therefore I would find a lead of 41 to 36 for the Liberals to be highly questionable, yet very good news for the Liberals if true. At the very least if that result is true it means the Liberals would hold Peterborough and Northumberland-Quinte West. And the Liberals would be hard pressed to lose the election if those two ridings are in their column. However, the low sample size makes me have to take that Central Ontario result with a grain of salt.