While the three-way split at the national level is regularly repeated in most polls, the polling situation in Ontario is markedly less consistent. Some polls show a tight race between the Liberals and Conservatives, while others have a large lead for the Liberals.
The latest Nanos three-day tracking poll had the Conservatives slightly ahead of the Liberals, 37 per cent to 34 per cent, in the province. The NDP trailed in third at 24 per cent. The latest Ipsos poll had the Liberals ahead of the Conservatives 41 to 32 per cent. Meanwhile, the latest Forum poll gave a narrow lead to the Tories, ahead as they were of the Liberals 35 per cent to 34. Both Forum and Ipsos also placed the NDP in third.
How would each of these scenarios play out on election day? Using the assumption of uniform swing (where the change at the provincial level is reflected in each of the individual seats), let's imagine one where the parties do exactly as well in Ontario as they did last time. Here, we'd end up with a huge victory for the governing party in terms of seats won, with the Tories picking up 83 of the 121 seats on offer, while the New Democrats and Liberals trail with 24 and 14 respectively.
What about a three-way race? Here we'll assume a 31.5 per cent split in the vote for each of the three major parties, with the remainder going to the Greens and other candidates. In this scenario, the Conservatives still win the most seats, but suffer heavy losses almost exclusively to the Liberals. In this case, the Conservatives would win 52 seats, the Liberals 41 and the New Democrats 28.
However, what about the scenario currently suggested by Ipsos, where the Liberals would run up a margin of nine percentage points over the Conservatives? Here, the Liberals would net 60 seats (slightly less than half of the available 121), while the Conservatives would win 43 and the NDP would pick up the remaining 18.
These quite different pictures painted by the polls points to the fact that poll aggregation can be helpful, and point the way Ontario can be used on the path to victory. When looking at all of the polls released in the last few weeks, a clearer picture emerges: The Liberals are perhaps a percentage point or two ahead of the Conservatives, with the New Democrats behind by about 10 percentage points.
If the NDP are able to make a true three-way race, they can hold their position from 2011, and even add a few seats to their count. If they're able to hold their vote throughout the rest of the country steady, their chances of victory increase tremendously.
The Conservatives currently appear to have little chance of repeating 2011 this time. Two things went very well for the party in that vote: they won a very strong 44 per cent of the popular vote, and their major competition almost evenly split the remainder, the NDP narrowly edging the Liberals 25.6 per cent to 25.3 per cent. Even if the split between the Liberals and NDP happens again, it seems unlikely that it will be combined with a strong Tory vote share. Nonetheless, the party's greatest hope in the province at the moment seems to be in narrowly winning a large number of seats through Liberal-NDP vote splitting.
For the Liberals, Ontario can serve as the jewel in the party's electoral crown. In the victories in the 1990s, Ontario was at the heart of the party's majorities. In their win in 1993, the province delivered 98 of its 99 to the Chrétien-led party. While a united right and divided left makes recreating this landslide highly unlikely, winning 50 or more seats in Ontario seems like a prerequisite for a Liberal minority government.
In the next three-and-a-half weeks, it seems likely that focus will shift strongly to Ontario, not because it is seat rich (this is basically just a way of saying "has a lot of seats") but because many of those seats are particularly competitive. In Atlantic Canada, the Liberals have consistently polled strongly, just as the New Democrats have in Quebec, and the Conservatives have done in the prairie provinces. In a tight, three-way national race, shifting vote shares by even three or four percentage points might end up making the difference between placing first and trailing in third.
Paul Fairie is a political scientist at the University of Calgary. He designed the Globe's Election Forecast.