Election Writ 10/15: Are the B.C. Conservatives falling behind?
Most polls say yes, but the race remains close.
Welcome to the Election Writ, a special-edition newsletter to get you up-to-date on the ongoing campaigns in New Brunswick, Saskatchewan and British Columbia.
With days to go in British Columbia’s provincial election campaign, have the B.C. Conservatives lost their chance to win it?
Most polls now suggest the Conservatives will come up short. But not all of them.
Polling in the campaign has taken one of two forms: individual polls published by various pollsters at largely irregular intervals, and the Mainstreet Research tracking poll that has come out every day.
The difference between the two sets of polls has been stark. The infrequency of the one set vs. the drumbeat of the other also makes for competing narratives that feel like they might carry the same weight. But the preponderance of evidence suggests that the B.C. New Democrats are in the better position with just four days to go before Saturday’s election.
Recent polling by the Angus Reid Institute, Ipsos and Léger has given the New Democrats a lead of five points over the Conservatives, and Research Co. puts the gap at four points. With the realignment that has taken place after the collapse of B.C. United and the emergence of the Conservatives, it is difficult to know for certain what popular vote margin would deliver an NDP victory. But, in all likelihood, four or five points is more than enough.
Polling by Pollara and Pallas has shown a smaller lead of one or two points, which would make the seat outcome a toss-up. But what we have from these five pollsters is a range of NDP-leaning toss-up to an NDP victory.
Then there is Mainstreet, which has been publishing every day. Mainstreet’s polls have been consistently stronger for the Conservatives, and the toplines from the latest iteration published yesterday put the Conservatives five points ahead. That’s probably more than enough to give them the win.
The difference between the Pollara and Pallas polls and the Mainstreet polls isn’t huge, but the 10-point spread between the ARI, Léger and Ipsos polls and the tracker from Mainstreet is quite significant. Unless the numbers converge between now and Saturday, one set of polls will be very wrong (or, if things end up in the middle, both will be somewhat wrong).
There are likely some mode effects here — it’s perhaps no coincidence that Pallas and Mainstreet, who conduct their polls with automated telephone calls, are at the Conservative-friendly bottom of the list, while the online panel pollsters are at the NDP-favouring top. Research Co. has never given the Conservatives a lead and the Angus Reid Institute has only once (in August, by a single point), but Léger had the Conservatives ahead in two polls (by two and three points) at the end of September. From Léger’s perspective, the Conservatives have faltered.
With the exception of one poll, however, Mainstreet has had the Conservatives either leading or tied throughout this campaign. That one poll coincided with the field dates of the last Léger poll, so that could have been a rough period for the Conservatives that they’ve since recovered from. But the ARI and Pallas polls are very recent, and have not shown the Conservative strength picked up by Mainstreet.
As a survivor of the 2013 B.C. campaign, this is triggering some traumatic memories of that catastrophic polling miss for me. We’re not on track for something like that this time, however, since the polling margins are tighter and there isn’t a consensus on who is in front.
There’s also the fact that the polls did pretty well in the 2020 campaign in B.C. and there wasn’t a consistent mode problem in that election. Mainstreet over-estimated the NDP’s lead over the then-B.C. Liberals by five points, but so did Ipsos (by three points). Research Co. was off by just one point. The ARI and Léger under-estimated the NDP’s lead by four and three points, respectively. In other words, the polls were pretty close to the mark and pollsters landed on both sides of the target.
If that happens again this time, then perhaps the Pallas and Pollara surveys will prove closest to the mark. A five-point lead for either the NDP or the Conservatives might be too much and, in reality, the gap is far smaller, but with multiple pollsters pointing toward an NDP lead rather than a Conservative lead, we have to conclude that the NDP is probably ahead by a few points.
The map might also be favouring the NDP. Mainstreet has its regionals paywalled, but the regionals from the other polls put the NDP ahead by about seven points in Metro Vancouver and 10 points on Vancouver Island, while the Conservatives lead by five points in the Interior and North. The NDP leading in two of three regions, including the biggest, gives them the edge.
They also probably have a turnout advantage. The Angus Reid Institute gives the NDP a lead of two points among British Columbians 55 years old and older, while Ipsos puts the lead at 21 points and Research Co. at 18 points among this age group. Pollara and Pallas have the NDP ahead by 16 and 19 points, respectively, among those 65 and older. That should add some confidence to the view that the NDP has the advantage — even if they are tied with the Conservatives in the provincewide vote, that their support is higher among older voters (who actually vote) should bump up their share by a couple of points.
We’ll see what the last few days of the campaign have in store for us on the polling front. For now, the New Democrats probably have the advantage.
Now, to what is in this instalment of the Election Writ:
News on advance voting in B.C. and New Brunswick and the candidate slate in Saskatchewan.
Polls show a tight race in B.C., less so in Saskatchewan.
Eby and Moe would hold on if the elections were held today.
This week in New Brunswick, Saskatchewan and B.C. election history.