One last look at the ON vote x interview mode

Alright, Ontario, we’re on day 27, 2 more to go. My polling is holding steady. The PCs are at 47%, the OLP are at 28%, the NDP are at 16%, and the Greens are at 9%. But what are other polls saying? 

I’m revisiting some analyses I did a couple of weeks ago, which looked at how the Ontario polls vary based on polling methodology. I pulled data from the polling aggregator 338Canada, which currently models the vote at 45% PC, 27% OLP, 19% NDP, and 6% Green. In my analysis, I standardized the data so that the vote for a party other than the PCs, OLP, NDP, or Greens has been suppressed, allowing for apples-to-apples comparisons.

In my earlier analysis, I found that pollsters, regardless of what their baseline is, were showing a pretty stable race. That continues to be true for some pollsters but not for others. Pallas Data, Nanos Research, Mainstreet Research, and Liaison Strategies are showing pretty consistent gaps between the PCs and the OLP week over week. Both Innovative Research and Abacus Research have shown a pretty dramatic narrowing of the race—10 points for Innovative and 9 points for Abacus. Ipsos (-6), Léger (-4), and my firm (-6) have also shown the gap closing, but not quite to the same degree. 

I once again plotted polling results by three interviewing modes: online panel, phone, and interactive voice response (IVR). It’s important to remember that interview mode is only one factor in polling methodology. There are also differences in sample size, sample frame, weights, etc, which helps to explain why we still see some volatility in the numbers even if the same interview mode was used.

Online polls have tended to show a larger gap between the PCs and the OLP. The PCs have mostly been around 48-49%, while the Liberals have been more around 26%, although they’ve tended to poll a little better throughout the second half of the campaign. Just one online poll has found the OLP to be above 30%. The NDP has mostly been around 17-18%.

Nanos Research’s polls, which use live caller interviews, have been quite consistent. The PCs tend to be around 45%, and the OLP have been in the low 30s. They also show a slight growth in the NDP over the past week.

The IVR polls showed some early PC momentum, to the detriment of the NDP, but have since remained fairly consistent. The PCs have tended to be around 44%, the OLP around 30-31%, and the NDP have been around 19-20%. 

Broadly speaking, we’re looking at three different dynamics: PCs having an advantage in the high teens, mid-teens, and around 10%. Of course, it all depends on who actually shows up to cast a ballot.

Tomorrow, I’ll be digging into some different voter turnout scenarios. The last provincial election had record-low turnout (43.5%). Elections Ontario has reported that 6.14% of Ontarians have voted in advanced voting so far (over 3 days), compared to 9.92% in 2022 (over 10 days). That, combined with the PC’s dominance in the polls, voters’ limited attention to the campaign and… winter… makes me think this isn’t going to be an emblem of democratic participation. I’ll map what that could mean for each party tomorrow.