Gradual NDP Erosion to Liberals Continues in Ontario

While Canadians continue to be surprised by a virtual three-way tie in national popular vote estimates, there has been a less obvious trend occurring that seems to have slipped below the radar. The aggregated data presented in this projection basically follow the same pattern of stability as the national contest tightens even further. In fact, only two provinces, Ontario and British Columbia, register any change in seat distribution from the previous projection. However, there has been a gradual trend occurring over the past three months that has been barely perceptible on a week to week basis. Since the NDP reached its zenith of support in Ontario in mid-June, the Liberals have moved from a six-percentage-point deficit to the NDP there, to an 11-point advantage in the most recent set of polls. During this time, the Conservative support level in Ontario has hardly budged. The effect of this has translated into a 15-seat gain for the Liberals in the province over the three months, which represents three-quarters of their improvement nationally during that period. The data below are drawn from a collective sample of over 7000 respondents interviewed between Sept. 9-20 from surveys by Ipsos, Nanos, Ekos, Environics and Forum.

Projected distribution of seats by party and region compared with actual election results (in brackets), released Sept. 22, 2015.

Canada 114(166) 117(103) 106(34) 0(4) 1(1)
Atlantic 4(14) 6(6) 22(12)    
Quebec 5(5) 60(59) 13(7) 0(4)  
Ontario 49(73) 20(22) 52(11)    
Prairies & North 15(26) 9(3) 7(2)    
Alberta 28(27) 3(1) 3(0)    
British Columbia 13(21) 19(12) 9(2)   1(1)

Note: The "regional swing model" is more fully explained in a paper presented by Dr. Barry Kay to the 2009 annual meeting of the American Political Science Association, entitled "A Regional Swing Model for Converting Canadian Popular Vote into Parliamentary Seats 1963-2008." It should be noted that the application of the model above does not make use of the "incumbency effect" described in that paper. In tests for past elections, using late campaign polls to project electoral outcomes, the model has proved to be accurate within an average of four seats per party since 1963. Readers interested in post-dictions for past federal elections dating back to 1963, for projections using pre-election polls dating back to the 1980 federal election and for three Ontario provincial elections, may contact me at[email protected].

Level of Government: