PCs and Liberals neck-and-neck in first Ontario projection

The 2014 provincial election in Ontario was officially kicked-off yesterday, when Kathleen Wynne asked the Lt. Gov. to dissolve the legislature following the decision by Andrea Horwath's NDP not to support the government's budget. The vote will be held on June 12, 2014. If that vote were held today instead, the prediction for victory would be a toss-up between the Liberals and the Progressive Conservatives.

In the first projection from ThreeHundredEight.com for the Ontario election, the PCs are projected to have the lead in voting intentions with 34.4% support, or a likely range of between 33.1% and 37.5%. The Liberals follow closely behind with 33.5%, or between 32.1% and 36.8% support. The New Democrats are in third with 23.9% (or between 22% and 25.6%), while the Greens sit at 7% (or between 4.9% and 8.7%).

The Liberals are projected to pull slightly more seats out of the equation than the Tories, with 44 to 41. But the likely seat ranges narrowly favour the PCs, with between 36 and 51 seats to between 34 and 50 seats for the Liberals. A genuine toss-up.

The New Democrats would likely win between 18 and 26 seats, with 22 being the precise projection. That puts them squarely in third place.

But if we look at the maximum and minimum ranges (designed to capture 95% of likely outcomes), the picture is even muddier. The Liberals could win between 22 and 65 seats, the PCs would take between 25 and 62 seats, and the NDP could win between 12 and 29 seats. The Greens could also win one. This suggests that, taking into account past polling errors, a scenario where an election held today puts the NDP in the Official Opposition role can be envisioned.

Regionally, the PCs hold comfortable leads in eastern and southwestern Ontario, and are narrowly ahead in the 905 area code. The Liberals are well in front in Toronto, but are currently edged out by the New Democrats in northern Ontario.

Full details of the projection, including the riding-by-riding breakdown, can be found by clicking on the chart at the top of the page (or here).

Unless the polls start agreeing with one another, or if one party decisively pulls ahead, this will be a difficult election to call. The polls did a good job in Ontario in 2011, and are coming off two consecutive good calls in Nova Scotia and Quebec, so there is hope for accurate polling results in this campaign. The seat projection model on this site had some trouble with the PC tally in 2011, but that was primarily due to the fact that it was a province-wide model at the time. It currently is a regional model, which in 2011 would have helped capture the disproportionate swing to the PCs outside of the GTA.

This will be a relatively long campaign, stretching to almost six weeks. Hopefully, voters will be more engaged than they were in 2011 when fewer than half tramped out to the polls. As it stands, it looks like it will be a close race.

Ontario NDP drops in EKOS poll

Hot off the wires following the release of the Ontario budget yesterday is a poll by EKOS Research that might give Andrea Horwath pause as she considers whether or not to send Ontarians to the polls this spring.

Update: And just as I hit 'post', news comes out that Andrea Horwath will not support the budget. Looks like we're off to the races. Also, an earlier version of this post said the NDP was at 22.9% in the EKOS poll - they are actually at 22.2%.

The poll by EKOS Research for iPolitics gives the Liberals the lead with 34.7% support, up 2.4 points from EKOS's previous poll of March 27-April 3. The Progressive Conservatives were also up, gaining 4.2 points to reach 31.6%.

The New Democrats, however, fell 6.8 points to 22.2%. 

The Greens were up 1.1 points to 9.4%, while 2% of respondents said they would vote for another party. Of the entire sample, 18.9% were undecided.

Only the drop in support of the NDP appears to be statistically significant, though it also looks a lot like a reset. EKOS's previous poll had the New Democrats at 29%, well above the 22% to 23% recorded in subsequent polls by Forum, Nanos, and Innovative. It does differ, however, from the 27% recorded in the most recent Ipsos Reid poll, but even Ipsos was registering a slip in support for the NDP from their own previous poll.

For the Liberals and PCs, the picture remains muddy. Of the five surveys now conducted since April 7, the PCs have recorded (in order of field dates) 38%, 36%, 30%, 37%, and 32%. The Liberals have recorded 31%, 36%, 39%, 32%, and 35%. Overall, the Liberals and PCs seem to be somewhere in the 30s. That is about as much as can be said with certainty. The aggregate now gives the Liberals 34% to 33% for the Tories.

Of significance, however, may be the Tories' lead among the oldest voters. Of Ontarians aged 65 or older, the PCs held a 40% to 36% edge over the Liberals. Among the next oldest tranche of respondents between the ages of 45 and 64, the Liberals and PCs were almost tied (33% to 32%, respectively). This suggests the Tories may have a turnout advantage (which Ipsos Reid also recorded in their poll).

At the regional level, the NDP drop appears to have taken place throughout the province. But in terms of statistical significance, the decreases were particularly marked in Toronto and eastern Ontario. 

In the provincial capital, the Liberals led with 39%, followed by the Tories at 29% and the NDP at 20% (down seven points). In eastern Ontario, the PCs were ahead with 43% (up 15 points), followed by the Liberals at 31% and the NDP at 18% (down eight points).

Elsewhere, the Liberals led in the suburban GTA (the 905 area code) with 36%, against 31% for the PCs and 23% for the NDP. The governing party also had the edge in northern and central Ontario with 40%, compared to 29% for the NDP and 22% for the PCs.

The PCs led in southwestern Ontario with 33%, followed by the Liberals at 26% and the NDP at 24%. The Greens scored 12% support in this region, their best.

With these levels of support, the seat result would be a toss-up. The Liberals would likely win 45 seats, with the PCs taking 41, the NDP winning 20, and the Greens taking one. (Let's not get overly excited about the Green result - 9.4% is not a particularly reasonable level of support for the party.)

Assuming 10% error in the model, that means the Liberals could win anywhere between 41 and 50 seats and the PCs between 37 and 45 seats. In other words, a coin-flip to determine who takes the plurality of seats.

At these numbers, the New Democrats are not in a good position to make significant gains. Of course, campaigns matter and Horwath does have the best personal numbers of the three leaders, so she may have the most upside. But it is still a gamble. With neither Tim Hudak nor Kathleen Wynne being loved by voters, the NDP cannot rule out the possibility of forming a government of their own if things go very well. But a more plausible best case scenario for the party sees it forming the Official Opposition. The question is whether they would take that role in a minority government, where they would have significant influence, or a majority government, where they would have next to none.