Hot Take on ProRep Referendum Results

Last week, I wrote that First-Past-the-Post (FPTP) would prevail by “a whisker, a few grey whiskers”.  It turned out that it was a full-on Santa’s beard of whiskers as it prevailed with 61% of the vote, far exceeding projections by public pollsters.

The Regional Story

What’s striking about the win for FPTP is its dominance in the Lower Mainland.  It took 63% of the vote in BC’s most populated area.  The Interior delivered a resounding 67% verdict, but it represented just over a quarter of the votes for FPTP.  The Lower Mainland did the heavy lifting.  The Island was a 50-50 split, a disappointment to ProRep supporters, especially the Greens.  However, as I wrote last week, there are a lot of ‘experienced’ voters (hint: old) on the Island and experienced voters don’t care much for ProRep.

Feast your eyes on this table then I will break it down for you…

Screen Shot 2018-12-20 at 8.23.52 PM.png

  1. The South Island (south of the Malahat) was the best area of the province for ProRep, delivering a 55% win across seven ridings in the Capital region.  Those efforts were cancelled out by the seven ridings north of the Malahat.  As was the case up north, rural, resource-producing areas are very skeptical of ProRep, which partially explains the difference on the Island between north and south.
  2. In the Lower Mainland, Vancouver-North Shore’s 15 ridings were close with FPTP edging ProRep 53% to 47%.  This stands in stark contrast to the rest of the Lower Mainland and should serve as a wake-up call to the NDP.  Richmond-Delta-Surrey voted 72% for FPTP and Burnaby to Mission went 70% FPTP.  This is exactly where Christy Clark’s government met its demise in 2017.  In these areas alone, eight BC Liberal seats flipped to the NDP.  In this referendum, they were clearly not buying what the GreenDP were selling.  The Fraser Valley went strongly for FPTP, as expected, even then, 75% is emphatic.
  3. In the Interior, the northernmost ridings, including the Cariboo, strongly backed the current system with 74% support.  I return to the previous point – it is surprising to me that the suburbs of the Lower Mainland matched the North.  Frankly, it’s shocking, to a guy like me, who obsesses over numbers and ridings.  The Okanagan and Kamloops region were in lockstep at 68-69%.  The Kootenays demonstrate, yet again, that they march to the beat of a different drummer – at least those in the West Kootenay.  Both NDP-held seats there voted ProRep, the only two of the 24 Interior seats to support the proposal.  The region overall was 55% for FPTP due to the East Kootenay BCL-held seats.

Results by Party 

The BC Liberals elected 43 seats on Election Day 2017 and those 43 seats voted 70% for FPTP, nine points above the 61% average province-wide.  The NDP’s 41 seats leaned toward FPTP with 54% compared to 46% for ProRep.  The Green held-seats slightly favoured FPTP, by a margin of 51% to 49%.

Screen Shot 2018-12-20 at 8.43.12 PM.png

Let’s look at the NDP seats.  ProRep was blown out in Surrey seats with Panorama leading the way at 74.5% for FPTP.  Newton 73%.  Green Timbers 73%.  Fleetwood 72%.  Delta North 70%.  Hop over to Maple Ridge where it was 68-69% for FPTP.

I don’t want to overstate the importance of this referendum to the next election but it’s not great when your initiative, which sucked up a lot of political oxygen, is thumped.

In the Interior, the four NDP held seats did OK, relative to the overall result.  Stikine was 60% for FPTP which is better than Surrey, I guess.  North Coast was close (53% FPTP) while, as mentioned above, the NDP seats in the West Kootenay backed ProRep.  Is there anywhere to grow for the NDP?  When you see 25% for ProRep in the Cariboo, where the NDP had seats not long ago, and 26% in Fraser-Nicola which was held up until 2013, it’s hard to see this referendum as an Interior growth strategy for the NDP.

Here are the 16 of 87 seats that voted ProRep:

  • Vancouver Mt. Pleasant (NDP)
  • Victoria-Beacon Hill (NDP)
  • Victoria-Swan Lake (NDP)
  • Vancouver-Hastings (NDP)
  • Vancouver-Fairview (NDP)
  • Vancouver-West End (NDP)
  • Nelson-Creston (NDP)
  • Powell River-Sunshine Coast (NDP)
  • Oak Bay-Gordon Head (GREEN)
  • Vancouver Pt. Grey (NDP)
  • Esquimalt-Metchosin (NDP)
  • Saanich North & the Islands (GREEN)
  • Vancouver-False Creek (BCL)
  • Kootenay West (NDP)
  • New Westminster (NDP)
  • Langford Juan de Fuca (NDP)

The top four on the list are probably the top four safest NDP seats in the province.  For the BC Liberals, it stands out that ProRep won in False Creek.  Sam Sullivan hung on by his fingernails last election and this riding is evolving.  The BC Liberals held Pt. Grey, Fairview, Saanich North, and Oak Bay up until 2013.  The BC Liberals, strong proponents of FPTP, continue to go against the grain of these type of ridings.

Another riding, which was very close between FPTP and ProRep was West Vancouver-Sea-to-Sky.  This is a riding that should be on the BC Liberals’ “watch list”.  It’s changing, as is North Vancouver-Seymour.  Ridings that used to be safe, aren’t as safe anymore.    The key is the suburbs. There are a lot of seats out there and, like Quebec in federal elections, they can go one way or the other, en masse.

In this referendum, ProRep did not have a broad enough coalition.  It did poorly in the suburbs and the regions, and in ridings with higher proportions of people who do not have English as a first language.  This pretty well sums up the Green Party actually.  While it won in Andrew Weaver’s riding (which hosts UVic) and in Saanich-North & the Islands, where it has two elected Green representatives (Adam Olsen and Elizabeth May),  it lost in Cowichan Valley.  It’s more rural, and it’s less typical of the Green base.  The Greens like PR because they haven’t been able to break through in FPTP because of their limited appeal.  The referendum reinforces the Greens’ weaknesses.

ProRep also lacked meaningful support from BC Liberals.  NDPer Bill Tieleman was extremely important to the FPTP campaign, along with other traditional FPTP supporters in the NDP.  Glen Clark’s support for FPTP was an important signal to an element of the NDP base.  Former Premier Ujjal Dosanjh also made interventions into the campaign in favour of FPTP. There was no signalling from iconic BC Liberal / Socreds-of-old to the free enterprise base that ProRep was ok.

The process

At the end of the day, this was a process that was nakedly designed to support the partisan interests of the Green Party, and to a lesser extent, the NDP (who were more interested in keeping the Greens happy).  There was no secret about it.  The consultation had all the appearances of a sham, especially when compared to the much-admired Citizens’ Assembly process prior to 2005.  Here was a Premier, Gordon Campbell, who had 77 of 79 seats.  He put the power of recommending a new system in the hands of two citizens per riding.  They spent months learning and deliberating.  I didn’t like their recommendation but I admired their efforts and their example.  It was democratic jury duty.  That process also had a much better result for electoral reform than what just transpired.  Is this a lesson learned for David Eby? Too clever by half?

Another sign of a flawed process is that 40% of the voters did not even vote on Question 2. There were 1.391 million voters overall, but only 832,000 chose one of the three PR options, while 559,000 skipped the question. Of the options, MMP had the most with 343,000. That would not have been a resounding mandate and such a gap in responses between Q1 and Q2 would have undermined a positive ProRep outcome.

The players

FPTP supporters, like Bob Plecas, Suzanne Anton, and Bill Tieleman, got started early.  They branched out to bring in other voices in multicultural communities. They had the experience to focus their resources and messaging effectively.  They worked on persuading likely voters and punched through with clear arguments.  The ProRep campaign appeared to focus a lot of effort on getting non-voters to vote.  That is harder to do.  As with many campaigns, followers can get fooled by their echo chamber.  If your social media galaxy is made up of people just like you, then you can be led to believe there is more support than really exists.  Clearly, the ProRep campaign was spanked once it left the cozy confines of Victoria and Vancouver.  It did not have a ground game or a regional game.  This was a similar challenge in the Transit Referendum.  The suburbs rose up and defeated the initiative (though there wasn’t a great deal of support in Vancouver either).  What’s the answer?  Show up, listen, engage.  Get out of your comfort zone if you want to grow your cause.

This is why the system works

As the results indicate, there is dynamism in the system.  I can see in these results how the electoral landscape continues to shift.  It’s not a given how the province will vote next time.  Big-tent parties have to get their arms around the largest swath of voters possible in order to govern.  That requires overarching vision, compromise, and brokering, but all within a unified structure that brings forward a coherent program that that party is accountable for if they win.  Moreover, it is highly competitive.  Elections between two or more big-tent parties bring clarity and a clear choice for the electorate.  Smaller parties play an important role too.  They push issues on to the agenda that big-tent parties must respond to.  If they don’t, big tent parties can disappear and smaller parties can emerge to take their place.  As a Teenage Vote Splitter, that was my story – that of the BC Liberals supplanting the Socreds after 1991 and the Reform Party overtaking the Progressive Conservatives in 1993.  It happens all the time.  We have a highly competitive, dynamic electoral system.  It’s not perfect, but it’s not static either.

The best argument I heard during this debate was that FPTP is the most effective system for throwing out a bad government.  It’s decisive.    In most cases, these parties are the better for it.  Out goes the old and in comes a new generation of leaders that rebuild. ProRep would allow power structures to linger like last year’s salad dressing.  So, I’m obviously happy with the outcome.  Our system largely works.  There are a few changes that could be made that would help, not requiring a referendum, but that is a post for another day.

Now that this is over, next stop for BC politics: Nanaimo.

Advertisements

FPTP vs. PR – and the winner is?

If this was a conventional election, it would appear First-Past-the-Post (FPTP) will prevail over Proportional Representation (PR).  Is it a conventional election? Let’s take a look.

electoral-reform-envelopes.jpg

It’s all over but the counting. (photo: Times Colonist)

It appears, according to public polls, that it’s a competitive race between the two options. Various polls have it in the 50-50 range.  There is also a consensus among public polls that PR is strongly favoured among young people and FPTP is strongly favoured among older people.

There are regional variations.  Areas where PR appears strong are Vancouver Island (where the Greens are strongest) and the City of Vancouver. FPTP appears strongest in BC’s Interior.

Ethnicity and language also appear to play a role.  Public polls do not show breakdowns by ethnicity (and I wonder if they are properly represented in sampling), but we can see that turnout has been lower in ridings that have lower English-first language populations, such as Richmond and Surrey.

There has been some interesting analysis conducted by Star Metro reporter David Ball and Langara College professor Bryan Breguet (an unabashed PR supporter) who has a blog called “Too Close to Call”.

Ball tracked the rate-of-return of the mail-in-ballots and compared it to the HST and Translink referendums.  The seemingly slow rate of response at the outset of the process was typical of these processes.

_2_bcballots_dec6_davidpball.jpg

Brequet ran regressions and other analysis to determine who might be voting.  He’s also conducted Google polls, that show the race very tight.  I’m a bit more primitive in my approach so take the following as you will.

Here’s what we know:

1,361,000 ballots were received, and Elections BC has published the riding-by-riding totals for over 92% of those ballots, providing a fairly clear picture of the regional breakdowns.

Turnout

On the spectrum of turnout for province-wide votes over the past seven years, the PR referendum was low.  Take into account population growth and the PR turnout was even weaker when compared to the HST referendum.

Screen Shot 2018-12-13 at 10.08.38 AM.png

Regional Differences

One of the wrinkles with looking at these results by region is to look at raw votes rather than ridings.  There are 24 seats in the Interior and only 15 on the Island/Sunshine Coast.  Yet the average riding population on the Coast is higher.  In the final analysis, there are about as many votes from the Island/Sunshine Coast as there are from the Interior.

The Interior received ballots earlier, and returned them earlier.  Over the course of the balloting period, the Interior’s share of the overall pile of votes diminished to about its share of registered voters.  The FPTP advocates may have hoped that the Interior would punch above its weight in terms of turnout.  The region that did over-perform its share of the electoral pie was Vancouver Island / Sunshine Coast.  By the time the final 8% of ballots are allocated to ridings, the Island will likely surpass the Interior in terms of overall votes, and will have over 20% more influence on the process than compared to its share of registered voters.  The Lower Mainland will have about 8% less influence on the process that its share of registered votes.

Screen Shot 2018-12-13 at 10.15.43 AM.png

Partisan Differences

We can also track where the votes came by held-seat.  In the HST referendum, every NDP seat voted against the HST, while half of the BC Liberal seats voted for the HST.  Who represents a seat can be an indicator of support – not because of the MLA, but because of the underlying attitudes that got that party and MLA elected in the first place.  One could reasonably assume that BC Liberal seats will lean FPTP and NDP/Green seats will lean PR.  That will probably be the case writ-large, though there will be exceptions.  (For the purposes of this analysis, I am basing it on held seats as of election night, 2017).

Screen Shot 2018-12-13 at 10.17.44 AM.png

What does this tell us?  The three Green seats over-performed on turnout, while the NDP under-performed.  NDP held-seats in Surrey, for example, had very low voter turnout.  Apparently, voters in Cowichan, Oak Bay, and, Saanich North & the Islands were more eager to receive their ballots than to have a visit from Santa.  But there may be other reasons for that, stay tuned (see below).  The BC Liberal seats held their own, but no advantage in turnout.  Again, those 20 of 24 seats in the Interior held by the BC Liberals did not generate as much in terms of turnout percentage as I would have thought.

In the Lower Mainland, it was very close in terms of raw vote turnout between the NDP-held seats and the BC Liberal-held seats.  The NDP’s 26 Lower Mainland seats account for about 50.5% of all votes in the Lower Mainland processed already versus 49.5% for the 22 BC Liberal-held seats.  There are no Green seats in the Lower Mainland.  So, in the Lower Mainland, the BC Liberal seats did better relative to the NDP.

In the Lower Mainland, the turnout was highest in Vancouver-North Shore and the Fraser Valley, relative to Burnaby to Mission corridor and Richmond to Surrey corridor.

Older people vote

It is a consistent truth in Canadian elections that turnout is higher among older people than younger people.  In the 2015 federal election, there was a bigger turnout among all age groups, but the old geezers still outpaced the Millennials.

Let’s look at turnout-by-age in the 2017 provincial general election.  Work with me here.

The 18-24s represent 11.1% of the eligible voters (the red bar).  However, they represented only 6.3% of those who actually voted (the blue bar).  Meanwhile, the 65-74s are busy voting on the first day of the advance polls following their coffee and muffin at Tim Horton’s.  This group represents 12.7% of eligible voters (red), but 17.6% of those who voted (blue).  There are only about 10% more 65-74s than 18-24s in terms of eligible voters, but almost a 3X difference between them in terms of who voted.

Chart 1: Turnout by Age, 2017 BC provincial election (source: Elections BC)

Screen Shot 2018-12-13 at 9.28.36 AM.png

So, the question is, given that public polls indicate a strong preference among seniors for FPTP, did old voters rule the roost in the PR referendum too?

Bryan Brequet from Too Close to Call has undertaken some analysis and he sees a trend toward more younger voters.  I think what he means is that the gap between younger and older voters in the PR referendum may not be as pronounced as the 2017 general election.

I’m not so sure.  When you have a lower turnout, the age discrepancy is usually bigger.  When you have a higher turnout, more younger people (and other less-likely voters) are showing up to the polls.

So here’s what I did…

I ranked the ridings 1-87 in terms of turnout in the PR referendum.  Parksville-Qualicum is #1, at 46% turnout. Surrey Whalley is #87, at 18%.  (This is, as of, 92% of ballots being screened.  Turnout by riding will increase but the ranking of ridings 1-87 will probably not change that much).

Then I took BC Stats data and looked at the 18-44 population per riding.  I ranked the ridings 1-87 in terms of their proportion of 18-44s relative to the overall adult population of the riding.

The green shading indicates the ridings that are in the lowest quartile of 18-44s (the oldest) ,and the red shading indicates ridings that are in the top quartile of 18-44s (the youngest).

What did I find?  The top 8 ridings in terms of voter turnout in this referendum are also in the oldest quartile when it comes to age by population.  Six of the 8 lowest turnout ridings are in the youngest quartile.  In fact, as you go down the list from highest turnout to lowest turnout, you will see 19 oldest-quartile ridings before you hit the first youngest-quartile riding.

The Rosedeer Decision Desk calls it:  Older people vote more.

Screen Shot 2018-12-12 at 3.12.17 PM.png

Older People vote more … but what does that mean?

If I’m right, and a similar pattern exists in this referendum as it does in most general elections, then it’s good news for FPTP.  But there are definitely some mixed messages.

One of the oldest ridings is Saanich North & the Islands represented by the Greens.  Will older people deliver this riding for FPTP or will the Greens deliver it for PR?  I think it’s both.  I’m sure the Greens will have worked hard to deliver votes there (they get elected there for a reason), but age will be a bit of a mitigating factor.  This could be the story of the Island – an inexorable pull to PR by the Greens (and to a lesser extent, the NDP), restrained by a sizeable population of old geezers.

In older, BC Liberal-held Lower Mainland ridings like Delta South, West Vancouver-Capilano and White Rock, you might see some of the largest margins for FPTP.

Even if there are more older people voting than younger people voting, the question is, where is support at for FPTP and PR?

If it is 50-50 in the polls, and if public polls are correct in so much as older people favour FPTP and younger people favour PR, and IF an age turnout factor is present is as above, then 50-50 becomes 47-53 or even 45-55 in favour of FPTP.

However, if PR has burst through, and if it has weakened opposition among older people down the home stretch, and made a breakthrough with 35-54s, then overall support of 55% among eligible voters may translate to just enough (50% +1) among those who actually voted.

Language

As mentioned, some of the Lower Mainland ridings have the lowest turnout of any seats in the province. The Lower Mainland also has the highest share of non-English (first-language) households in the province as well.

Similar to age, there is a correlation between turnout and English-language skills.  This table, like the previous, is ranked by turnout. The green shading indicates ridings in the lowest quartile when it comes to non-English households.  The red shading indicates ridings in the highest quartile of non-English households.

What does this tell us? We can hypothesize that election materials were not accessible to some voters, or was not debated as extensively in their language (via media) compared to English-speaking media.  Unlike turnout-by-age data supplied by Elections BC, we do not have comparable data for ethnicity or language.  I can merely point out the correlation.  Anecdotally, the HST referendum appeared to have had a high level of engagement in the Chinese community, owing in part to mobilization of Chinese restaurateurs who opposed the tax.  It’s fair to say, I think, that the PR referendum did not hold the attention of the Chinese media (or Punjabi media) like the HST issue.

Screen Shot 2018-12-12 at 3.14.51 PM.png

While Chinese voters tend to favour the BC Liberals, historically, and voters from Punjabi-speaking households have leaned more to the NDP over time, the Greens are clearly weaker in these communities relative to support they have elsewhere.  Therefore, one could argue that lower turnout in parts of Surrey, Richmond, and other Lower Mainland areas might be good news for PR, if you think you are more likely to find PR voters where the Green Party finds more fertile territory, generally.

The Upshot?

The voting is done.  I made my case for FPTP (“I was a Teenage Vote Splitter“) and have also pointed out how PR advocates should be careful what they are wishing for.

I think the final result will be close.  I think there was momentum toward PR in the final weeks which helped close a (perceived) gap.  For PR to win, it will have had to have done quite well with 35-54s (assuming young and old cancel each other), and have won the Lower Mainland (assuming Island and Interior cancel each other).

If PR does win, then it will have likely done so with less than 22 to 23% of registered voters.  We can debate the legitimacy of that if it happens.  Gordon Campbell had set a 60% threshold which seemed to be a reasonable threshold for a fundamental change to the electoral system.  The current government obviously thinks otherwise!

I do not think 42% turnout is enough for PR to win with a majority.  I think they needed more voters to flood the polls.  I expect FPTP to win by a few whiskers, a few grey whiskers.  We will know soon enough.

Merry Christmas and Happy Holidays to PR and FPTP supporters alike !