In terms of seats, the Liberals are now in majority territory - again for the first time since the beginning of the campaign. They are projected to win 55 seats, or between 46 and 62 seats. The PCs are projected to win 37 seats, or between 31 and 46 seats. The NDP is projected to win 15 seats, or between 13 and 16 seats.
The Forum poll has the same sort of sampling issues I have highlighted before. The proportion of respondents 55 or older is almost twice as large as it should be, and the proportion of respondents 34 or younger is almost a quarter of what it should be, if this was a representative sample. Weighting can correct for this, but as explained before that has the potential to magnify the errors that can creep in with small sub-group sample sizes.
Those three points came equally from the other three parties, with the PCs down to 34%, the NDP down to 20%, and the Greens down to 4%.
None of these shifts are outside the margin of error, but of note is that the New Democrats have now fallen in five consecutive Forum polls, while the Liberals have gained in three consecutive polls.
It will be interesting to see what other pollsters say, as it does seem that the Liberals are improving their position (either among all eligible or just likely voters) in every poll that has been out since the campaign began. Whether they really hold a lead of this magnitude, however, is another matter entirely. The normal margin of error of a random sample of this size would be around +/- 3%, roughly reducing the Liberals to as little as 38% and the PCs to as much as 37% (or, conversely, to as high as 44% and as low as 31%).
The Liberals led in every region of the province in this poll except in eastern Ontario. A Liberal lead in the 905 area code and, especially, the southwest is somewhat unusual. The only large shifts that appear statistically significant was a swing between the NDP and Liberals in Toronto: the Liberals were up 12 points to 51%, while the NDP fell 10 points to just 14%. This echoes the latest poll from EKOS, but not the one from Abacus.
Forum showed little change in approval ratings, with Kathleen Wynne and Andrea Horwath steady at 38% and 35%, respectively. Their disapproval ratings fell to 46% and 40%, respectively. Tim Hudak's approval rating increased to 25%, while his disapproval rating was unchanged at 59%.
In the second release from Abacus's poll this week, similar numbers were shown in terms of favourability - or at least relative numbers, since Abacus provides the option of people having a 'neutral' opinion. But 29% said they had a favourable opinion of Horwath, with 28% having a favourable opinion of Wynne and just 22% saying the same for Hudak. Negative views topped out at 42% for Hudak, against 37% for Wynne and just 21% for Horwath (she had the highest 'neutral' score).
On who would make the best premier, Abacus gave Wynne 26% to 20% for Hudak and 18% for Horwath. Among likely voters, however, Wynne's score improved to 31%, against 24% for Hudak and 20% for Horwath. Those numbers were very similar to Forum's estimate of 34% for Wynne, 22% for Hudak, and 17% for Horwath.
Also on the plus side for the Liberal leader is that Abacus found Wynne polling ahead of her rivals in three key swing groups: OLP/PC voters (32% to 17% for Hudak), OLP/NDP voters (43% to 25% for Horwath), and three-way swing voters (19% to 10% for Horwath and 7% for Hudak). Hudak, meanwhile, beat Horwath among PC/NDP swing voters by a margin of 36% to 23%.
