BC Election 2013

In an Angus Reid public opinion poll (their link, local link) conducted from March 16 to 17, sampling 807 people has the following voter intentions:

BC Liberal – 43%
NDP – 38%
Greens – 10%
Conservatives – 5%
Other – 4%

Metro Vancouver: BCL 44 / NDP 38 / GRN 10 / CON 5 / Other 3
Vancouver Island: BCL 43 / NDP 38 / GRN 11 / CON 4 / Other 4
Interior: BCL 37 / NDP 43 / GRN 9 / CON 6 / Other 5
North: BCL 55 / NDP 24 / GRN 9 / CON 5 / Other 7

This survey has results very similar to the February survey, before Premier Christy Clark got elected. I still maintain that until both BC Liberal and NDP leaders have been elected and have had a sufficient time to lose their “honeymoon” with the public, the polls are not to be trusted with respect to the projected result of a general election.

The media is portraying this as a “honeymoon” period for Clark, which is somewhat correct, but the real fireworks will begin with the HST referendum.

Mike Farnworth leads Adrian Dix in terms of preferences for NDP leadership, and continues doing so with NDP supporters. However, the margin amongst NDP supporters is close (54% for Farnworth and 41% for Dix in a non-mutually exclusive vote) so that race is still wide open and will head up in the next few weeks.

On the HST question, when you normalize out the 13% undecideds, you have a 62% rate to abolish the HST, while 38% support it. This is also similar to the December 2010 survey. It will be very interesting to see how this number will change leading up to the HST referendum, assuming there is one.

In an Angus Reid public opinion poll (their link, local link) conducted from February 15 to 17, sampling 811 people has the following voter intentions:

BC Liberal – 41%
NDP – 38%
Greens – 11%
Conservatives – 4%
Other – 6%

Metro Vancouver: BCL 43 / NDP 38 / GRN 12 / CON 3 / Other 4
Vancouver Island: BCL 39 / NDP 38 / GRN 14 / CON 4 / Other 5
Interior: BCL 40 / NDP 41 / GRN 6 / CON 5 / Other 8
North: BCL 37 / NDP 26 / GRN 12 / CON 9 / Other 16

This poll should not be taken with much relevance given the context of the two leadership races that are going underway. Indeed, until the NDP leader is selected in the middle of April, and when both of the leaders have had sufficient limelight after their respective honeymoon periods have ended, I will not be taking voter intention polls too seriously.

Indeed, the only piece of information anybody should be gleaning out of this poll is that out of 308 NDP supporters, 18% of them think Dana Larsen would be a good choice to replace Carole James and 7% support MLA Nicholas Simons. This poll just might save MLA Simons his leadership entrance fee!

Disappointingly, the HST question was not asked, but hopefully the next poll could include the question.

An Angus Reid Poll (Local copy, their copy) conducted on December 20 and December 21, 2010, sampling 806 people have the following voter intentions:

NDP – 38%
BC Liberal – 38%
Greens – 12%
Conservatives – 7%
Other – 5%

Metro Vancouver: NDP 35 / BCL 43 / GRN 14 / CON 5 / Other 3
Vancouver Island: NDP 43 / BCL 31 / GRN 11 / CON 9 / Other 6
Interior: NDP 39 / BCL 35 / GRN 10 / CON 10 / Other 6
North: NDP 35 / BCL 31 / GRN 2 / CON 19 / Other 13

I won’t read too much into these tea leaves other than to say that speculating on election polling before the BC Liberals and NDP have selected their leaders is pointless.

The most interesting information to be gleaned from the poll dealt with the HST question. When you normalize the numbers to remove the 11% that were undecided on the matter, 61% of the respondents wanted to abolish the HST, while 39% wanted to keep the HST. This is the true stealth result of this poll, and the next poll will be interesting to see if this trend continues. This is hardly the 80% opposition that was seen before. Are the people of BC thawing to the idea of the HST?

An Angus Reid Poll (Local copy, their copy) conducted on December 7 and December 8, 2010, sampling 804 people have the following voter intentions:

NDP – 36%
BC Liberal – 36%
Greens – 14%
Conservatives – 6%
Other – 8%

Metro Vancouver: NDP 37 / BCL 39 / GRN 13 / CON 5 / Other 6
Vancouver Island: NDP 36 / BCL 34 / GRN 15 / CON 4 / Other 11
Interior: NDP 34 / BCL 31 / GRN 18 / CON 8 / Other 9
North: NDP 42 / BCL 38 / GRN 5 / CON 6 / Other 9

Analysis

This poll is going to be often quoted by the media (similar to the earlier Mustel poll), but the political relevancy of such a poll is questionable. In fact, such polling results are going to be downright misleading, so it is amusing to see people quote such polling with strong convictions. Polling done after both major parties have selected leaders are going to be much more relevant. So I won’t comment on the NDP/BCL 36/36 result other than that a lot can be won or lost by the parties by choosing the right or wrong leader.

What is more interesting is the subset of readers that indicated their preference for party leader.

Out of the BC Liberal side, of the BC Liberal voters in the 2009 election, if asked who would make a “good choice” favour Christy Clark with 51%, Kevin Falcon with 42%, Mike de Jong 40%, and George Abbott with 35%. Moira Stillwell trails the pack with 8%.

Assuming 45% of the sample is eligible for this question, this gives you a standard error of 5.3%. This would suggest the top four candidates quite possibly have a shot at winning the BC Liberal leadership.

On the NDP side, when asked the same question for 2009 NDP voters, we have Mike Farnworth with 43%, Gregor Robertson 35%, Jenny Kwan 29%, Adrian Dix 33%, and the rest trail with 18% and under (Moe Sihota, Bob Simpson, John Horgan, Derrick Corrigan, Bruce Ralston, George Heyman).

Assuming 41% of the sample is eligible for this question, this gives you a standard error of 5.5%. I find it unlikely that Gregor Robertson and Jenny Kwan will run for the leadership. It appears right now that it will be between Farnworth and Dix, although if you threw in a guy like Derrick Corrigan, it might make things more interesting.

The HST question is showing slightly more support for the HST, when normalizing the numbers – 68% opposed, 32% supportive. This is a 10% change from previous polling, but could be well within the confines of “white noise” given all of the other political material going on in the province.

A new Mustel poll (their link, our link) conducted from November 4 to November 15, 2010, sampling 502 people (with 16% undecideds) has the following voter intentions:

NDP – 42%
BC Liberal – 37%
Green – 10%
BC Conservative – 9%
Other – 3%

Premier Gordon Campbell‘s approval/disapproval is 32/60, while Carole James is 33/45.

My quick analysis here is similar to the previous Mustel poll – they are working with a sample size too small, too spread out over time, and these results should be treated with effective 15% error bars instead of the statistical 4.4% margin of error. When you toss out the 16% undecided people, this leaves 422 responses. The only real “blip” in the data compared to the Mustel survey two months ago was a spike up in Carole James’ approval/disapproval from 42/36 to 33/45. This is probably explained by her leadership woes internally within her party, while it had no effect on the NDP value.

But with a 422 response sample, it is so close to looking at random noise that I won’t give it much more text. I, along with others, will likely wait for the next Angus Reid poll to come out; this will provide a better snapshot.

Thanks for BJ for alerting this to my attention. He’s faster than Google Alerts.

Éric Grenier of Threehundredeight has a seat projection from the last Angus Reid poll (November 2/3, just before the Premier’s resignation). His 70 seat NDP prediction is quite close to my first-pass model (69 seats), but I am awfully curious to know how he managed to sneak the Green candidate (presumably in their best-performing riding in the last election in West Vancouver-Sea to Sky) as a victory. The numbers still do not support it.

I also do not think this poll can be taken too seriously, nor any other poll regarding voting intention until the BC Leadership Race is settled.

An Angus Reid Poll (Local copy, their copy) conducted on November 2 and November 3, 2010, sampling 807 people have the following voter intentions:

NDP – 47%
BC Liberal – 26%
Greens – 10%
Conservatives – 10%
Other – 7%

Metro Vancouver: NDP 44 / BCL 31 / GRN 13 / CON 6 / Other 6
Vancouver Island: NDP 51 / BCL 25 / GRN 8 / CON 10 / Other 6
Interior: NDP 45 / BCL 20 / GRN 8 / CON 18 / Other 9
North: NDP 57 / BCL 26 / GRN 2 / CON 7 / Other 8

As the Premier’s resignation was done on November 3, 2010, it is likely that this poll is already contaminated (Update: BJ reports that the poll closed before the 11:30am resignation). Completely ignoring this event and the impact of his earlier speech, there is little change to be reported in the political support polling.

There is some data to be had with the HST and the 15% income tax cut announcement.

The HST question continues to be 76% Yes, 24% No in response of the question to get rid of it (after normalizing the figures to get rid of 13% undecideds).

The income tax cut was supported by 65% of respondants, and 14% opposed, leaving 21% not sure. Normalizing the numbers, that is 82% supportive, and 18% not supportive.

More interestingly is the question “At this point, which of the following scenarios would you prefer for British Columbia after the 2013 election?”, of which 32% indicated they want the NDP to form the next government (with 47% saying they will vote for the NDP in general, which is odd), and “The BC Liberals with a different leader form the next government” at 28%.

I don’t think these numbers should be read too deeply in light of recent events.

An Angus Reid Poll (Local copy, their copy) conducted on October 13 and 14, 2010, sampling 804 people have the following voter intentions:

NDP – 49%
BC Liberal – 24%
Greens – 13%
Conservatives – 8%
Other – 6%

Metro Vancouver: NDP 45 / BCL 29 / GRN 14 / CON 6 / Other 6
Vancouver Island: NDP 60 / BCL 14 / GRN 17 / CON 5 / Other 4
Interior: NDP 48 / BCL 18 / GRN 11 / CON 16 / Other 7
North: NDP 44 / BCL 31 / GRN 6 / CON 9 / Other 10

The geographical split has been consistent with prior surveys from Angus Reid, with the minor exception of declining BC Liberal support in the Interior – 18% in this survey; and increased support in the North with 31%. A curiousity is that the Green party was more popular than the BC Liberal party on Vancouver Island. The sample sizes for the geographical split is quite small, so big swings are to be expected and the data should be treated with very high error margins.

The leader approval ratings were given a bit of headline space, with Premier Gordon Campbell at 9%. Carole James was also slipping, at 27%.

Finally, 72% of the respondents said they would vote to abolish the HST, and 21% will vote to keep the HST. Normalizing the numbers to excluded undecideds, we have a 77.4% “Abolish” and 22.6% “Keep”, which is consistent to prior measurements and intuition.

I am still working on the seat model, and until then, I will not give a seat projection. This has been progressing slower than I anticipated. However, you can look at Bernard Schulmann for some interesting projections which, at initial glance, look reasonable.

Sean Holman of the Public Eye Online reports the following (and apologies to Sean for just ripping the entire article off, but I will try to add some value):

A union-commissioned poll shows provincial Conservative support has surged into double digit territory, Public Eye has exclusively learned. The poll, which surveyed 1,000 British Columbians between August 16 and 31, puts the Conservatives at 15.8 percent – just two percentage points below the governing Liberals. By comparison, the New Democrats are at 34.0 percent, the Greens are at 12.1 percent and the undecideds are at 13.1 percent. That leaves 7.2 percent of survey respondents who refused to say who they’d vote for, won’t vote or would cast their ballot for another party. With the undecideds factored out, the New Democrats are at 42 percent, the Liberals at 23 percent, the Conservatives at 21 percent and the Greens at 15 percent.

There is no mention of methodology (although the two-week period suggests telephone polling) and geographical spread, and questions asked. So this cannot be used as a standardized poll, but rather a good point of gossip. The rest of the discussion assumes the poll is completely valid.

A 1,000 person sample has a standardized error (normal distribution), assuming you exclude the 7.2%, of +/- 3.2%. Excluding the undecideds and “others”, you have an error of 3.5% at 95% confidence.

Sean Holman did get the normalization of the polling numbers slightly incorrect in his last sentence – assuming the undecideds/others intended on voting in proportion to the four main parties, we would have the following result:

NDP – 42.7%
BC Liberal – 22.3%
BC Conservative – 19.8%
Greens – 15.2%

If this was indeed the case, this would be an excellent performance by the NDP – conceivably you would see the BC Conservatives splitting votes in BC Liberal dominated ridings, and you could see the NDP pick up seats in areas that were not otherwise to be considered. Very informally (still working on the seat model), I’d estimate the BC Liberals would be left with roughly 10 seats, and the Conservatives would likely pick up a few. The NDP would sweep with roughly 70-75 seats. The Greens would have an outside shot of one or two, but I still project zero – they would be close with 16%, but not quite there.

A Mustel Poll (their link, local link), conducted August 26, 2010 to September 7, 2010, surveying 502 people (13% undecided, thus 437 responding) has the following voter intention:

NDP – 42%
BC Liberal – 33%
Green – 12%
Conservative – 11%

I am not going to dive into this result simply because the sample size is smaller than my liking. If I was going to guess what the “true” numbers would be in an election, it would likely be closer to Angus Reid‘s prior poll rather than Mustel’s.

Also, the poll indicated that Premier Gordon Campbell‘s approval went from 28% to 34%, and his disapproval went from 61% to 57% from the prior (May 2010) poll, so this is likely a sign of the roll of the dice that was received when you have a 437-person sample responding in a poll (i.e. I do not think public sentiment for the premier right now is more favourable than it was back in May 2010).

Thanks to BJ for bringing this to my attention. The guy has his brain hard-wired to the internet looking for poll releases!