04 September 2014

Voter Choice #nlpoli

When Kathy Dunderdale jumped or was flicked out of office in the first part of 2014,  CRA boss Don Mills issued a release covering his company’s February 2014 self-promotion poll that claimed that Tom Marshall was doing wonders for the Conservative party because public satisfaction with the government was up in the poll.

NL government satisfaction improves with new leader” said the headline. Unfortunately for Mills and CRA,  that headline connected up two things  - government satisfaction and new leader – in a way the poll data didn’t support.  You see, satisfaction went up the quarter before that as well, with the old leader.

There’s just no connection between “satisfaction” and the public choice for best party to form government or for best premier.  The Conservatives have strong satisfaction numbers and yet a clear majority of respondents want to vote for some other party to run the government and someone other than Tom to be Premier.

Skip ahead six months and Mills is at it again.

Who needs data when you can speculate?

”Tom Marshall is a very respected individual as you know,” Mills told CBC about the results of his latest poll, “and clearly as an interim leader for the government he has done a pretty good job for the Conservatives based on the satisfaction numbers.”

Mills simply abandoned his own data there and decided to substitute speculation.  There’s still nothing to connect Marshall and the government satisfaction numbers.  Satisfaction is about the same as it was in May.  Marshall remained in August where he was before:  a second choice but to make matters worse for Mills’ claim, the gap between second choice Marshall and top choice Dwight Ball is actually greater now than it was three months ago.

Tom’s level of support as “best choice for Premier” is a basically back in the same neighbourhood Dunderdale was living in for the quarter before she jumped/was pushed.  The best you could say based on the CRA numbers alone is that people were really frigging happy to see the backside of Kathy Dunderdale. That’s it.  You can’t even really say, as Mills did, that eligible voters in the province have some sort of extra regard for Tom that they don’t have for others. 

We went through this in a post last winter – “The Satisfaction Delusion” - for those who may have missed it.  But you have to admit, it’s pretty bad when a pollster decides to ignore his own data and substitute speculation instead.

The Long Slide

Just so that we keep this approach going,  here is a chart showing the CRA party choice numbers with the undecideds calculated as a valid response.  CRA tosses them out to give you the shares of “decided” responses.  To compare the poll results more readily with everyone else and to remove the skew that CRA’s method throws in,  here’s the longer term trending.

CRA Q3 2014

That skinny line is the undecideds plus the people who say they won’t vote and those who refused to give an answer to that question.  Some people will look at that and think there are tons of potential votes in that pile.  Experience over the last decade shows us that such ideas are misguided. CRA doesn’t poll voters.  They poll people who are eligible to vote.  That isn’t the same as people who will turn out.  There’s also not a simple correlation that says all those people just won’t vote but won’t tell the pollster.  In CRA polls, we don’t know what they really are so it is safer to put the numbers in there anyway so that we have them.  The party choice responses – as a share of responses including the “undecideds” – gets us close to what everyone else reports in their polling. That’s what we go with around here.  Your mileage may vary, as the old Internet hands used to say.

Let’s just look at the fat lines, which are the party choices. Except for a brief blip in the latter part of 2012, the Conservatives have been on a steady downward slide since the last election. They hit bottom, below 20%, in the second quarter of 2013 and have been basically stock steady ever since.  Those quarterly variations are well within the margin of error for the poll, so basically things haven’t changed to any degree worth noting for the Conservatives. 

Note, that there was a small uptick in February 2014, when Dunderdale jumped out the political window.  Unfortunately, for the entire time of the Conservative leadership campaigns, the party support has actually declined. It actually dropped more as we’ve gotten closer to the leadership convention in a couple of weeks’ time.  That’s not just a lack of any kind of bounce in the polls, that’s a pretty firm public rejection of what’s been going on.

Premier Kim Campbell

Translate these numbers out to seat projections and basically you have the Liberals sweeping everything west of Gander on the island and all of Labrador.  That 30 seat strategy some Conservative insider supposedly proposed would be an exceedingly optimistic thought based on these CRA numbers.

According to the latest assessment by labradore,  the Conservatives can only count on two seats and are theoretically in contention in another 10.  The problem for the Conservatives is – as labradore notes – that if you look at the vote swings, then the Tories likely won’t even hold those 10 seats if the current trends continue.  If that’s the case, we are looking at the next Premier pulling a Kim Campbell:  two seats. 

Things are that bad.

Bill 29

In 2013,  we got a look at another set of polling numbers and they show us something really interesting.[MQO%252012%2520month%255B3%255D.jpg]

MQO was tracking a Tory decline in 2012.  They poll every month so they are perhaps a bit more sensitive to shifts than CRA’s polling might be, leaving aside any other differences between the two companies as to how they conduct their polls.

Note that the NDP actually surged into the lead in this polling immediately after the spring session of 2012.  That’s the time of the two filibusters but the one everyone remembers is Bill 29.  It’s been a long time since any one piece of legislation stuck in the public mind the way Bill 29 did.  In fact, there’s probably never been such a bill here. It rivals the infamous Bill 101 in Quebec, in fact.  People may not know what was in Bill 29 but odds are they can tell you that they didn’t like it one bit. 

It’s that visceral public rejection of Bill 29 that makes the CRA polling for that period a bit hard to swallow. The MQO polling, on the other hand,  seems to fit intuitively with what happened in 2012.  Since we don’t have MQO polling results in public for the period through 2013 and 2014,  we can’t go any further with their trending. Put MQO and CRA together, though, and you might have a picture of what’s happened since 2010.

Think Rob Ford

In Toronto these days,  voters seem to have settled on John Tory as the next mayor of Toronto.  Voters want to get rid of Ford and they have settled on Tory as the one to do the job.  They looked at Olivia Chow and decided against her.  That’s one explanation for the current poll results in Toronto.

Now look at this province.  The Tories have actually been on a slide since early 2010.  Neither the Liberals or New Democrats presented a clear alternative so voters went with the Conservatives in 2011.  But by that time, a significant number of  voters defected from the Conservatives to the NDP in St. John’s.   The Conservatives won, but with the smallest share of eligible vote in the party’s history for a majority Conservative government.  And of course, they lost seats in their traditional bedrock in St. John’s.

The Bill 29 fiasco turned voters off in droves.  They opted for the New Democrats who looked like a fresh and potentially solid alternative to the arrogant Conservatives.  By 2013,  the bloom was coming off the New Democratic rose.  The party was stalled as Lorraine Michael and her supporters didn’t really do anything to develop the party and its lead.  Meanwhile, the Liberals were getting their act together.  The leadership convention attracted new supporters as the Liberals began to look like the party to replace the Conservatives.  Meanwhile, the Conservatives – like the NDP – just stayed pat.  They didn’t change.  In fact, they  just carried on doing more of the same sort of thing that got them into trouble in the first place. 

By the fall of 2013,  the Liberals had surpassed the NDP in the polls.  That brought problems inside the NDP to a head.  Lorraine handled the caucus revolt badly:  sure she stayed in office, but she fractured the party support in the process.  Huge chunks of the party support in the way of workers and elected MHAs headed off to the Liberals.  At the same time,   the NDP voters, including people who were backing the NDP merely as a way to get the Conservatives out of office, also started to shift to the Liberals.

That’s where they have stayed since the early part of 2014.  The string of by-election victories for the Liberals, including the personal and undeniable defeat of Danny Williams in Virginia Waters confirmed what had already been underway for some time.

The Turning Point

If you want to accept this explanation of what has been going on politically in Newfoundland and Labrador for the past four years.  The decisive point was the Bill 29 debate in the House and the filibuster.  The NDP may have initially gained from it politically, but the record  shows that it was the Liberals’ skill in the House that sustained the fight. 

Voters decided against the Conservatives at that point. They went initially to the NDP but shifted to the Liberals during the Liberal leadership as it became clear – in the public mind – that the Liberals were more likely to take out the Conservatives and form a credible, stable government.  That’s similar to what happened in 2001, incidentally.  Voters appear to have made a shift away from one party well before the election came in 2003.  In that case, it was against the Liberals and for the Conservatives in 2001. 

In the more recent case, the decline started in 2010 and carried through 2011.  But the decisive point came 2012. The Conservatives didn’t “run out of steam.”  Voters made a choice.

-srbp-