1 MAY: SOCIAL CUE REINFORCES CLOSENESS OF BC ELECTION

Voters indicate that they believe their neighbours, by an edge, are now supporting the NDP.

One of the challenges of polling in any election cycle is to first get behind the horse race numbers but also to get a sense of what voter turnout will be. Further, it is imperative that pollsters improve their ability to triangulate diverse data sources and looking at “necessary and sufficient conditions” to bring depth to the fundamentals of tracking voter intentions. During this writ period, experimenting with an online data collection platform (see note below*), Justason Marketing Intelligence and zinc tank have teamed up to offer regular insights on this Provincial Election. This is a follow piece is Wave 2 on social cues in this election cycle – as mentioned before, some may say that we, as voters, have been socially conditioned to specific crowd dynamics.

Yesterday, we released Wave 2 – Voter Intentions – BC Election 2017 that mirrored some shifts presented in other polls that were covered in the media. (See Ipsos: Liberals and NDP in tight race) At the same time, in this second tracking wave, we fielded another question asking respondents how they thought their neighbours would vote. Likely reacting to reports in media, there was a reversal in the lead – from the Liberals to the NDP – of how they thought their neghbours will vote.

Adopting techniques from psychology and behavioural economics on social desirability bias, crowd dynamics and social conditioning, we conducted two 1Question Google Surveys voter intention polls – one where we asked people who they intend to vote for, and the other with putting themselves into the minds of their neighbours and how they think they will vote. (Note, the unsures/not voting was lower than self-intentions – hence, though the gap is closing, there a belief that voter turnout is actually higher that what it actually is. Here, self was 68%, neighbours perceived to be 70%, and actual turnout in 2013 was 55%. It is imperative that we begin to have this discussion about modeling a voting population.)

We then considered the data from a decided voter perspective (see data tables below). When asked how they will vote in the upcoming election, 37% indicated they would be voting for the NDP, 38% for the Liberals and 19% for the Greens. When asked how they think their neighbours will vote, 38% indicated that their neighbours will be voting for the Liberals (no gap), while 42% will vote for the NDP (+5), and 16% for the Greens (-5). At this point, intentions and perceptions appear to be aligned for the Liberals – with gaps that favour them among male and 55 years+ voters. With the gap opening up between intention and perception for the NDP, there appears to be some upside among male and 35 years+ voters.

     

These two 1Question polls get at that issue where peoples’ perceptions do not necessarily translate into the reality of self-perception, and that there are likely more people out there who actually do seek something different in the governance of BC. And as pollsters are challenged to model a voting population, we need to grasp the alignment of perceptions of social cues (such as lawn signs and media reports) with reported self-intention (that which voters rightfully sense is their private right).

Regionally, there are some differences worth noting (note: caution with regional subsamples).

  • City of Vancouver: The NDP’s lead on perceptions are opening up against Liberals. It appears that the perceptions of the Greens are starting to level off.
  • Balance of Greater Vancouver Region: While level in Wave 1, perceptions appear to be opening up in favour of the Liberals. Both parties will likely focus efforts in this region.
  • Capital Region: While there appears to growth in self-reported intentions for Greens, perceptions are starting to drop off. The NDP has been the beneficiary of this change. There is also a decline in perceptions for the Liberals as well – thus, it is anticipated that Liberals will reallocate any resources from here to key swing ridings.
  • Vancouver Island/Gulf Islands: While the NDP hold a strong lead, there is growing perceptions in the growth of the Greens. This situation may create a few interesting races to watch on May 9. The perceptions of neighbours voting for the Liberals is declining.
  • Northern BC: While there is a stronger perception of the strength of the NDP, this region appears firmly Liberal from a self-reported perspective.
  • South/Rest of BC**: These region looks like it may deliver some surprises. While self-reported intentions are close, there is also now a 3 point gap in perception that favours the Liberals.

     

     

     

All parties should consider this result very carefully to illuminate this fact of perceptions of neighbours, against one’s own beliefs. With some initial strong turnout numbers in advanced polls, we can anticipate some further shifts in these self-reported/perceptual alignments.

Inspirations: System 1 Research Blog – Politics; Mark Earls: I’ll Have What She’s Having; and, Robert Bain: Whatever Next?

For more information, contact Barb Justason, 604.783.4165 or Brian F. Singh, 403.861.9462.

 

VIEW DATA TABLES HERE.

 

* THIS SURVEY WAS CONDUCTED VIA GOOGLE SURVEYS OVER (APRIL 25-28, 2017). GOOGLE SURVEYS USES A COMBINATION OF BAYESIAN, RIVER-SAMPLING METHODOLOGY ONLINE AND MOBILE AND TAPPING INTO GOOGLE COMMUNITIES TO YIELD A POPULATION REFLECTIVE SAMPLE OF BC’S POPULATION. HENCE, NO MARGIN OF ERROR IS REPORTED. OF NOTE, BC HAS A HIGH PENETRATION RATE OF MOBILE USAGE IN NORTH AMERICA, AND BROADBAND INTERNET ACCESS EXCEEDS LAND LINE USAGE. GOOGLE SURVEYS CONTINUES TO BE RATED FAVOURABLY (RATED BY 538: B), AND WAS USED EXTENSIVELY BY BRIAN F. SINGH IN HIS WORK ON BRIAN BOWMAN’S SUCCESSFUL MAYORAL CAMPAIGN IN THE 2014 WINNIPEG MUNICIPAL ELECTION, AND DURING THE ALBERTA AND FEDERAL ELECTIONS IN 2015. GOOGLE SURVEYS IS BUT ONE METHODOLOGY IN A POLLSTERS TOOLKIT WHILE NOT DEFINITIVE (AS HAS BEEN WITNESSED WITH OTHER POLLING METHODS), IT PROVIDES QUALITY DIRECTIONAL DATA THAT IS ABLE TO STAND ON ITS OWN AND IN CONCERT WITH OTHER MODES OF DATA COLLECTION, AND WE CROSS VALIDATE ITS FINDINGS FROM OTHER POLLING FIRMS THAT USE OTHER FORMS OF TRADITIONAL POLLING.

** South/Rest of BC comprises: Central Okanagan, North Okanagan, Okanagan-Similkameen, Columbia-Shuswap; East Kootenay, Central Kootenay, Kootenay Boundary, Thompson-Nicola, Squamish-Lillooet, Cariboo, Fraser Valley, Powell River, Sunshine Coast and Central Coast.

No Comments Yet.

Leave a Reply

Your email address will not be published. Required fields are marked *