One of the more interesting election websites is “Too Close too Call” – an election forecast site produced by Bryan Breguet, a Ph.D student in Economics at UBC.  I haven’t met him but admire his work.

It seems like he’s beating himself up a bit about his predictions today.  I took a spin on his seat forecasting model and inputted the actual popular vote numbers from BC.  His model extrapolates the final results fairly accurately.  Seat forecast models basically extrapolate the new pop vote numbers on a platform of the previous pop vote numbers (2011).

Here is the result:

His accuracy is 81% (34 out 42).  Not bad.  That’s a B in most schools.

What interests me are the differences – why do some ridings break the pattern?

In BC, those eight ridings – based on Too Close too Calls model – are:

• Cloverdale-Langley City
• Coquitlam-Port Coquitlam
• Kelowna – Lake Country
• Kootenay-Columbia
• Mission-Matsqui-Fraser Canyon
• Pitt Meadows-Maple Ridge
• Richmond Centre
• South Surrey-White Rock

I will offer a theory:

• These were all CPC seats to begin with
• Five of eight seats did not have an incumbent seeking re-election.  (Conservatives lost 4 of 5 with new candidates)
• The Conservatives won two of the seven outliers
  • Richmond Centre – bucked the trend for the CPC because of incumbency advantage.  Possibly retaining stronger support among Chinese.
  • Dianne Watts bucked the trend in South Surrey because she was a much better candidate for CPC than her predecessor.
• Kootenay-Columbia was going to be close under these circumstances, though the NDP outperformed there, defeating a CPC incumbent in a squeaker.
• Kelowna-Lake Country was a Liberal surprise, perhaps reflecting changing demographics in BC’s most urbanized Interior City.

Candidates make a difference, especially incumbents.  A closer look at popular vote per riding based against the model would show this too, but it’s the winning and losing that matters most.

Let’s start my election prediction with a little bit of Hamlet:

Now whether it be
Bestial oblivion, or some craven scruple
Of thinking too precisely on th’ event—
A thought which, quarter’d, hath but one part wisdom
And ever three parts coward—I do not know
Why yet I live to say this thing’s to do,
Sith I have cause, and will, and strength, and means
To do’t.

In other words, I’m pretty sure I am overthinking this election!

I have written about many aspects of polling over the course of this election and will apply them to what we may see tonight.

I’m going to borrow from the poll aggregator 308 to show the trends in this election and the aggregated outcome.  The Liberals separated from the NDP around September 20th and never looked back while the Conservatives held steady but never able to grow their market share (according to the polls).

Polls schmolls.  They are often wrong.  Even the blended numbers provided by the aggregators  have been wrong time and again.

Here are factors I apply to these numbers:

• Turnout by age – older people vote at a higher level.  This is a traditional Conservative advantage.  Some polls (EKOS, ANGUS REID INSTITUTE) show an age advantage, but others (NANOS) are showing the Liberals with big gains among seniors.  Moreover, with increased voter turnout, which is likely to happen this election, the proportion of seniors as an overall share of the electorate will likely drop.  It will still be significant and have the highest turnout rate, but the gap between seniors and younger voters won’t be as dramatic.

• Shy Conservatives –we saw this in spades in the UK where no one detected a David Cameron majority.  All the polls suggested ‘Hung Parliament’ yet the Cameron Conservatives triumphed in stunning fashion.  Part of this may relate to “won’t says” – the estimated 8% of voters who just won’t cooperate on surveys when asked the ballot question.  My view is that they probably skew Conservative.  There is also a theory that people stick with the incumbent and “hold their nose” if they are fearful of the alternative.  In BC, we used to call them “10 second Socreds”.

• Cultural Bias – My view is that media polling does not accurately reflect Canada’s diverse multicultural population.  Those who struggle with English are much less likely to cooperate in a telephone, IVR, or online panel.  In fact, online panels are the least representative, culturally speaking, unless they are done in-language.  There can be major differences in perspective culturally and they, ideally, should be reflected in polling.  I wrote about this recently with regard to the Chinese-Canadian community.  In BC, Chinese make up about 10% of provincial electorate and about 20% in Metro Vancouver – a huge factor.  My view is that the under-representation of Chinese in BC likely means Conservative support is under-represented.  I have seen no data in this election that tells us what’s going on in this community specifically.

• Ambivalent respondents –
  

  Some survey respondents really don’t have a clue… and some still use flip phones!

  Innovative Research Group has done interesting research on those who answer survey questions.  There is a continuum between strongly consistent to strongly ambivalent.  The Conservatives do better among consistent respondents (ergo likelier voters) and weaker on ambivalents.  Angus Reid Institute had similar findings along the lines of vote firmness.  See my post on “Why Conservatives have hope”.  Liberal support has been growing among consistent respondents.

• Gut – What does the tummy feel?  Hamlet would have had a real quandary figuring out these poll results.  The problem with the gut is that has a built in bias based on the echo chamber it lives in.  I am wondering if I’m drinking too much of my own bathwater, but my gut is pulling me in directions based on conversations with family and friends, and seeing indicators that may show how voters are behaving that are not reflected in the polls.  A big gut check is momentum.  That is something that Nanos has detected in the final three nights of polling.  Liberals on the rise.  Is it real?  I’m consulting my gut.

When I think about election surprises over the past few years, I’m comparing this election to those.  What is lacking in this election for the incumbent Conservatives is an overall narrative that gives voters a positive vision.  Negativity is an important part of campaigns – or you can call it contrast.  Parties are fighting over market share and the market has a ceiling of 100%.  It’s a zero sum game.  Parties must contrast themselves from their competition.  I have no problem with that and you can go back to the 1800’s for examples that would make today’s ‘attacks’ pale by comparison.  However, there must be a either a strong vision for the future, a stark choice, or a sense of renewal and change.

In both Alberta 2012 and BC 2013, these were female premiers in their first general election.  They were new, different and both were offering a proposition to voters.  In Alison Redford’s case, she was contrasting against a more conservative Wildrose brand and she was proposing to address issues on the progressive side of the ledger.  Christy Clark put forward an economic vision based built on a platform of fiscal responsibility that contrasted sharply with the NDP.  In the UK, David Cameron morphed fear of the Scottish Nationalists and the perceived weakness of the UK Labour leader to drive English Liberals to the Conservatives.  The reasons, in hindsight, are apparent for those election surprises.  I don’t see it here.  The case for Conservative re-election has seemingly slipped away over the last week.  While it may have been too late in the campaign to change much, the Ford family-Stephen Harper photo op on the weekend may have provided an added push for “time for change” Tories to jump ship.

The Upshot

While there are many factors – turnout, cultural, and polling bias – that mitigate in favour of a better Conservative outcome – which I am accounting for, my gut pulls me in the other direction.  I believe “red Tories” and soft NDP voters are going Liberal and the polls are seeing a glimpse of that.

Prediction

I’m throwing away the calculators and the models and the spreadsheets.

Hearkening back to Hamlet, three parts cowardice, one part wisdom would have led me to predict a Liberal minority.  I have put aside the cowardice and perhaps the wisdom, and decided to be bold.

I don’t think Canadians are going to be wishy washy tonight.  I think it’s a Liberal majority… by a hair (which would be ironic… “nice hair” they’ll say).

And though 38.5% has been the minimum to attain a majority in past 60 years, I think the Liberals may do it with less this time because of the vote splits.

The math for the Libs:

Atlantic and West/North:  50 seats

Ontario and Quebec:  120 seats

Libs 170

In BC, I think the NDP have jumped the shark.  Bringing out warhorse Stephen Lewis, last seen promoting the Leap Manifesto, to stump shows they are doubling down on their core vote.  The Liberals are going to win some seats that no one, including them, thought they had any business of winning.

BC seat count:

Libs 15

CPC 14

NDP 11

GREENS 2 (yes, not sure if it’s Victoria or Nanaimo)

This post will not impact a single vote so it’s all just fun and games.  When we are sitting around tomorrow reading about a shocking Conservative win, I will begin my tour of shame with my Conservative friends who will no doubt remind me of this post forever more.