2017 British Columbia election

[DL]

It's been a while since we have had any polling in BC but apparently Mainstreet will have a release tomorrow. I'll bet that after Horgan did so well in demolishing Christy Clark in the radio leaders debate we may start to see a bigger NDP lead

[Lotuslander]

Another new poll out by Justason Market Intelligence - opt-in online panel methodology:

BC NDP: 39%
BC Liberal: 36%
BC Green: 19%
Other: 5%

http://www.justasonmi.com/?p=5074

To put matters into context, at this point in the 2013 BC election campaign, Justason Market Intelligence also had an opt-in online panel poll showing the BC NDP with a 22% lead:

You must be logged in to read this quote.

https://thetyee.ca/News/2013/04/29/NDP-22-Point-Lead/

[DL]

The latest JMI poll uses a different online methodology than in 2013. This time it's based on google analytics based panel. Not saying it's better or worse than in 2013 just that it's different. But they were also in field April 18-20 meaning they would not be capturing the impact of the radio debate last week where Horgan crushed Clark

[Lotuslander]

New Mainstreet Research poll is out:

BC NDP: 44%
BC Lib: 34%
BC Green: 22%

Gotta love IVR/robo polls.

[DL]

Comparing apples to apples these polls by Mainstreet show a clear trend. It has to be particularly distressing to the BC Liberals that Green voters who have a second choice pick the NDP over the Liberals by a 74 to 7 margin. That means that if the Green vote declines at all the NDP picks up TEN votes for every one that goes to the BC Liberals

[136or142]

Mainstreet Research has a pretty good track record (unlike, say, Forum) but I find the sub-samples a bit hard to believe.  

For the NDP to be up over the Liberals 41-27% in Greater Vancouver (this does not include the 20% they find undecided) makes no sense to me given the unpopularity of the NDP in such places as the Eastern Fraser Valley (though the NDP is a little more popular in Chilliwack), Richmond and most of the North Shore.

As far as I'm concerned, for that to be accurate, the NDP would have to be up by 20% or so in Burnaby.

Apparently, the Mainstreet polling analysts predicted a minority NDP government on the basis of this poll, with NDP gains in Greater Vancouver offset by losses on Vancouver Island.  However, this poll also shows a tie in 'rest of province' (i.e The Interior) which if that was the case suggests the NDP could very well win such ridings as the two Kamloops ridings and Kootenay East.

If the NDP does win by 10%, I find it hard to believe they wouldn't win a majority government.

As I wrote a while back, this election was looking more and more to me like the situation in 1972, where after losing an election they were expected to win in 1969, nobody ever thought the NDP could ever win an election in British Columbia and this made those afraid of the 'socialist hordes' complacent in 1972.

So, although there is polling for this election, given what happened with the polling in 2013, the 'Anti NDP' coalition also appears to have gotten somewhat complacent this time around (I realize the advertising from the 'concerned groups' and the amount of donations to the B.C Liberals don't suggest that, but on the other hand, the Liberals themselves have once again put Christy Clark at the forefront of their campaign and are basically running the a repeat of 2013, even though Premier Clark is now one of the most unpopular political leaders in Canada.)

[DL]

Yeah, if the NDP wins the popular vote by 10 points, i would say that an NDP majority is guaranteed and likely a super-majority. In 1972 the NDP won the popular vote 39% to 29% and in seats they took 38 to 10 for the Socreds. In 1991 the NDP took 41% and the BC Liberals 33%...in seats that translated into NDP 55 and Libs 17!

As for those numbers for the Lower Mainland, i would like to know what the geographical definition of Greater vancouver is? Would Fraser Valley places like Chilliwack and Abbotsford even be considered Greater Vancouver or would they be lumped into the Interior?

One thing to keep in mind is how some of these places rapidly changing. Not too long ago Abbotsford was considered Bible Belt and ultra conservative...but in recent years it has been bursting at the seems with young lower middle class families fleeing the high cost of housing in Vancouver...and parts oif Abbotsford are getting to be as heavily South Asian as parts of Surrey. it might not happen this election but I would be surprised if the NDP starts to become competitive in one of those seats soon.

[DC Al Fine]

Yeah, if the NDP wins the popular vote by 10 points, i would say that an NDP majority is guaranteed and likely a super-majority. In 1972 the NDP won the popular vote 39% to 29% and in seats they took 38 to 10 for the Socreds. In 1991 the NDP took 41% and the BC Liberals 33%...in seats that translated into NDP 55 and Libs 17!

An interesting wrinkle in this pattern is that the Greens don't really fit the mould of "right wing vote splitter" that helped the NDP form government. I mean obviously if they win by ten points they should have a solid majority, but the shape of said government would look different from past NDP governments I'd imagine.

[136or142]

Yeah, if the NDP wins the popular vote by 10 points, i would say that an NDP majority is guaranteed and likely a super-majority. In 1972 the NDP won the popular vote 39% to 29% and in seats they took 38 to 10 for the Socreds. In 1991 the NDP took 41% and the BC Liberals 33%...in seats that translated into NDP 55 and Libs 17!

As for those numbers for the Lower Mainland, i would like to know what the geographical definition of Greater vancouver is? Would Fraser Valley places like Chilliwack and Abbotsford even be considered Greater Vancouver or would they be lumped into the Interior?

One thing to keep in mind is how some of these places rapidly changing. Not too long ago Abbotsford was considered Bible Belt and ultra conservative...but in recent years it has been bursting at the seems with young lower middle class families fleeing the high cost of housing in Vancouver...and parts oif Abbotsford are getting to be as heavily South Asian as parts of Surrey. it might not happen this election but I would be surprised if the NDP starts to become competitive in one of those seats soon.

I'm pretty sure their dividing line is Hope.  The Mainstreet Research site doesn't say so, but that is usually the geographic division.  If the NDP is tied with the Liberals in 'rest of province' and the rest of the province includes Chilliwack and Abbotsford, it's almost certain the NDP would win ridings in Kamloops, East Kootenay, The Cariboo and even Prince George. (Along with Penticton, Boundary-Similkameen and Fraser-Nicola)

[Lotuslander]

Entertaining that some folk here take this Mainstreet Research poll with more than a grain of salt. A key metric/number buried therein should jump out as a "RED FLAG". Right away. Analytically. But I will let others attempt to figure it out. Doubtful though.

[DL]

Check out this hard hitting ad by the BC NDP. It takes a sledge hammer to the BC Liberals and smashes them into a million pieces. It will be particularly effective with blue collar workers in the interior

https://www.facebook.com/bcndp/videos/10150864124154978/

[ProgressiveCanadian]

Check out this hard hitting ad by the BC NDP. It takes a sledge hammer to the BC Liberals and smashes them into a million pieces. It will be particularly effective with blue collar workers in the interior

https://www.facebook.com/bcndp/videos/10150864124154978/

Masterful ad! In this populist age of politics talking about Christy Clark's big donors is a fantastic argument. There's no way she can defend it.

[Lotuslander]

I think the NDP is more likely to win Kamloops North than either of the Cariboo seats

Interestingly enough, the first riding poll of the campaign is Kamloops-North Thompson by JMI, which was commissioned by the Kamloops Voters Society. IVR methodology and n = 239 with 2013 changes in brackets:

BC Lib: 47% (-5%)
BC NDP: 31% (-8%)
BC Green: 15% (+15%)
Communist: 8% (+8%)

https://armchairmayor.ca/2017/04/28/election-poll-shows-milobar-ahead-doctor-shortage-biggest-issue/

Why high Communist intention? Likely due to IVR methodology & smaller sample size. Still surprised BC Libs even leading with IVR here.

PS. Kamloops (now Kamloops-North Thompson) is BC's bellwether riding. Since 1903, whoever wins this seat wins gov't.

PPS. Also note that this "poll" was conducted pre-TV leaders debate.

[adma]

Sounds to me like you're cherrypicking a mediocre-quality poll in order to reinforce your vested anti-IVR/anti-NDP interests.

[Lotuslander]

Sigh. IVR in BC has a heavy NDP bent/bias. Anyone who diligently follows BC politics is fully cognizant of same. No question about that.

To wit, not only 2013 BC provincial election polls but 2015 federal riding polls in BC as well. It almost seemed IVR methodology was from another planet.

Frankly, I also suspect that IVR pollsters pumped out numerous polls for specific ridings until they obtained their desired result (or their clients) -  akin to IVR riding polls showing fed Con leads in downtown Toronto and Ottawa, fed NDP leads in Leeds-Grenville, and fed Lib leads in all northern Ontario ridings.

Just not a plausible scenario. Period.

In that vein, some 2015 BC fed riding polls (with actual results in brackets)

1. Cariboo-Prince George (Environics)

NDP: 36% (25.8%)
Con: 30% (36.6%)
Lib: 29% (31.5%)
Green: 5% (3.5%)

2. Coquitlam-Port Coquitlam (Environics)

NDP: 34% (27.3%)
Con: 31% (32%)
Lib: 29% (35.3%)
Green: 6% (3.8%)

3. North Okanagan-Shuswap: (Environics)

NDP: 37% (25.6%)
Con: 33% (39.3%)
Lib: 22% (30%)
Green: 8% (5.2%)

4. Pitt Meadows-Maple Ridge: (Environics)

NDP: 41% (29.6%)
Con: 35% (31.4%)
Lib: 19% (33.9%)
Green: 6% (4%)

And so many more examples as well. Again, corroborating/confirming that IVR has a heavy NDP bias in BC akin to opt-in online panel polls. Caveat emptor.

As for those IVR polls by Mainstreet Research? Well, an April, 2017 press release from MRIA - the Canadian governing body of polling companies - confirming that Mainstreet Research has been suspended from the organization:

You must be logged in to read this quote.

https://mria-arim.ca/sites/default/uploads/files/Mainstreet%20Research%20Sanction%20Posting.pdf

All cheap polling junk. Rinse, spit, and repeat.

[adma]

Yes, we know your beef about IVR; you've been saying it a zillion times as if that's the *only* thing that matters in this election.  My cherrypicking/mediocre poll comment was not about IVR in general; it was about JMI's Kamloops poll in particular.  Look; if we're dealing with an 8% Communist figure, we're not even talking about a poll that operates on an Environics/Mainstreet level.  (And please, don't reopen the "what about the BC Cons?" issue.)

[Lotuslander]

New poll by JMI - opt-in online panel. Good that they posted an appropriate caveat therein:

You must be logged in to read this quote.

BC Lib: 38% (+2%)
BC NDP: 37% (-2%)
BC Green: 21% (+2%)
Orther: 4% (-1%)

http://www.justasonmi.com/?p=5122

[DL]

One thing to bear in mind is that polling errors and methodological biases from one election rarely carry over into the next. Polling companies refine their methodologies and factors change. There is a hypothesis that in the 2013 BC election the online polls overestimated NDP support and underestimated Liberal support because a ton of people joined online panels during the anti-HST campaign and they were mostly people who were angry at the government...

However, in the 2009 BC election it was a very different story. In that election all the phone based polls forecast a BC Liberal landslide with a lead of 8 to 11 points over the NDP. Only the final online poll by Angus Reid had it close at 44-42...and the results were 45-42. So in 2009, the one online poll was almost dead-on while all the phone polls overestimated the Liberals and underestimated the NDP.

https://en.wikipedia.org/wiki/British_Columbia_general_election,_2009#Opinion_polls

What will happen this time - only time will tell.

[Lotuslander]

During the 2009 BC election, Mustel (CATI) was in the field for one week from April 29 to May 6. Election was May 12, 2009. The TV debate was on May 3, 2009. Moreover, then BC NDP leader Carole James "won" the 2009 TV leaders debate. Even then, both Mustel's CATI numbers as well as Ipsos (then CATI) numbers were still within their margin of error at extremities.

Angus Reid (opt-in online) was a newcomer to the BC political scene both as a pollster as well as its new methodology - first opinion poll was in late, 2008. Prima facie, their panels  were more in tune with the electorate wayyy back then. Since then, it has been obvious that online panels have had relatively major recruiting problems - folk dropping out/new folk coming in skewing their results.

2013 election opt-in online pollsters Angus Reid & Ipsos had final results holding a BC NDP lead at 9% and 8% respectively. OTOH, the BC Libs won by ~4% margin - a 12% to 13% reversal.

During the 2014 Ontario provincial election, the Toronto Star retained a pollster to test the 3 different methodologies (CATI, IVR, opt-in online panel) v. the actual ON election result. Their opt-in online methodology finding:

You must be logged in to read this quote.

https://www.thestar.com/news/queenspark/2014/06/15/polling_the_electorate_three_different_ways_with_differing_results.html

[Lotuslander]

Speak of the devil. Another new opt-in online panel poll by Ipsos:

BC Lib: 43% (+4)
BC NDP: 41% (-3)
BC Green: 14% (+2)

Green numbers are low with Ipsos.

[Pragmatic Conservative]

Speak of the devil. Another new opt-in online panel poll by Ipsos:

BC Lib: 43% (+4)
BC NDP: 41% (-3)
BC Green: 14% (+2)

Green numbers are low with Ipsos.

Care to link that to me, because I can't find that poll anywhere.

[RogueBeaver]

Here. http://www.cknw.com/2017/05/01/exclusive-poll-shows-bc-liberals-edging-slight-lead-after-televised-debate/?sc_ref=twitter

[Pragmatic Conservative]

Here. http://www.cknw.com/2017/05/01/exclusive-poll-shows-bc-liberals-edging-slight-lead-after-televised-debate/?sc_ref=twitter

Thanks.

[DL]

I notice something interesting in the crosstabs of the Ipsos poll - they asked people who they voted for in the 2013 election:

54% say they voted Liberal (Libs actually took 44% in 2013)
37% say they voted NDP (NDP actually took 40% in 2013)
8% say they voted Green (same as 2013)
no numbers on reported past vote for BC Conservatives/Other.

So contrary to the theory that this online panel might over-represent New Democrats - if we look at reported past vote - the sample is actually wayyy more Liberal than it ought to be. I'm not questioning the overall findings, just pointing this out for people to interpret as they wish

[136or142]

I notice something interesting in the crosstabs of the Ipsos poll - they asked people who they voted for in the 2013 election:

54% say they voted Liberal (Libs actually took 44% in 2013)
37% say they voted NDP (NDP actually took 40% in 2013)
8% say they voted Green (same as 2013)
no numbers on reported past vote for BC Conservatives/Other.

So contrary to the theory that this online panel might over-represent New Democrats - if we look at reported past vote - the sample is actually wayyy more Liberal than it ought to be. I'm not questioning the overall findings, just pointing this out for people to interpret as they wish

This could be nothing more than the well known psychological factor that many people can't remember how they voted in the previous election, and so say they voted for the winning party.

It's related to but not quite the same to the polling during the 1980s showing up to 20 million people claimed to be one of the 600,000 or so who attended Woodstock.