Only a little more than 200 days remain before Ontarians cast their ballots in the next provincial election. The polls suggest that Patrick Brown's Progressive Conservatives are in a strong position to defeat Kathleen Wynne's Liberals and bring her party's 14-year reign in Ontario to an end.
But can the polls more than seven months out from election day really tell us anything about what the result might be?
In six surveys conducted by four pollsters since the beginning of September, the PCs have averaged about 40 per cent support, followed by the Liberals at 30 per cent and the New Democrats under Andrea Horwath at 23 per cent.
That average 10-percentage point lead for the Tories is the widest any party has enjoyed this far out from an election campaign since 2003, when Dalton McGuinty's Liberals held a polling lead of that size over Ernie Eves's governing PCs. The Liberals eventually won by about 12 points.
This is a good sign for Brown. But a lead in the run-up to a campaign is far from concrete — in two of the last five elections, the party ahead in the polls with seven to nine months to go before the election ended up losing.
Clearly, campaigns matter.
In 1999, the PCs gained six points and the Liberals lost three over the last months before the election, turning McGuinty's polling lead into Mike Harris's re-electon.
In 2011, Tim Hudak's Tories dropped six points, with the Liberals gaining two and the New Democrats soaring by eight points in the run-up to the vote. That ruined Hudak's hopes of defeating McGuinty that year.
And in the last six months before the 2014 election, the Liberals picked up five points while the PCs dropped one and the NDP fell three. What was looking like a toss-up became a Liberal majority.
But voting intentions can also be locked-in and immune from the influence of an election campaign. In 2003 and 2007, the difference between each party's polling average this far out from an election and their eventual result was three points or less.
It is impossible to know whether the 2018 provincial election will follow the more unpredictable patterns of the 1999, 2011 and 2014 campaigns, or the steadier trend lines of 2003 and 2007. But Brown's PCs will be hoping for the latter.
Liberals have a history of rebounding
On average, over the last five elections the gains and losses of both the Progressive Conservatives and the New Democrats have evened out. There is no particular trend in one direction or the other: the Tories have gained in one, lost in one and held relatively steady in three election periods, while the NDP has gained in one, dropped in two and held steady in two.
The Liberals have tended to see their support increase in the months and weeks before an election. Their election results have beaten their polls seven to nine months earlier by an average of two points. In four of the last five campaigns they have experienced a notable increase in support.
This suggests that the gap between the Liberals and the PCs is more likely to narrow rather than widen. The last time the Tories were able to use the pre-election period to their advantage was 18 years ago.
PC lead leaves wider margin for error
However, the Liberals would likely need to match their 2014 pre-election gains to have a hope of staying in office. That five-point increase would put them at about 35 per cent support. All of those gains would need to come from the PCs to drop them to the same level and even the odds of either party winning the most seats.
Applying the shifts in support over the last five election campaigns to each party's current standing in the polls, the PCs could hope to end up with between 34 and 46 per cent support, compared to 27 to 35 per cent for the Liberals and 20 to 28 per cent for the NDP. In contrast to Brown, that does not provide Wynne with a big margin for error.
But every campaign is different and five elections make for a small sample size. There is no reason to assume that the polls will change between now and the 2018 provincial election. Support could very easily hold, putting Brown in the premier's office.
Alternatively, the Liberals could beat the odds and secure their fifth consecutive election win, something no party has done in Ontario since the days of the PCs' Big Blue Machine when that party held power without interruption from 1943 to 1985. But the polls suggest Wynne's Liberals will need a significant feat of electoral engineering to pull that off.