Tuesday, October 30, 2007

Saskatchewan election poll confirms SK Party to win

A poll on the Saskatchewan election out today gives the Saskatchewan Party a commanding lead in public opinion. It gives the Saskatchewan Party a lead of 50% to 35% for the NDP and 10% for the Liberal Party.

My forecast model suggests this would produce a majority SK Party government of 35 seats to 23 for the NDP and none for the Liberals.

The television debate is on now and can be seen on the net on CBC, but it is unlikely to make any difference, especially as the format seems to encourage a free for all.

Seat projections in Hill Times

This week's Hill Times has a story on seat projections made by political scientist Barry Kay.

Here is an excerpt:
Prime Minister Stephen Harper's minority governing Conservatives would have won another minority government in early October, says a leading public opinion expert.

Barry Kay, an associate professor of political science at Wilfrid Laurier University in Waterloo, Ont., said according to the seat projections he came up with by combining Ipsos Reid and Strategic Counsel polls conducted between Oct. 9 and Oct. 14, the Conservatives would have won 138 seats, the Liberals 101, the NDP 27 and the Bloc 41 and one independent MP would have been elected.

This is consistent with TC's views. Another Conservative minority still seems to be in prospect based on the latest polling by Ipsos.

Monday, October 22, 2007

All Hope of Harper Majority Lost!!

This slightly tongue in cheek header should startle most readers at least slightly. That's because the lines the conventional media have been peddling: the Harper can do no wrong/ Dion can do no right/ pay no attention to the other parties/ an election is just around the corner/ might make it seem a tad confusing. This conventional wisdom has been reinforced by the Ipsos polls, which keeping preaching the notion that the Harper majority is at hand (See the release here titled helpfully, "Canadians Prefer Harper Majority").

The first and most important point about this is that the Ipsos record, while not terrible, is weaker than other pollsters (see this evaluation of how pollsters did in 2006). And so far none of the others are producing polls these days as favourable to the Harper Tories as Ipsos, which just happens to conduct its polls for the blatantly pro-Conservative Canwest press.

However, what I find interesting is that the last two Ipsos polls (I wrote about the earlier one here), as favourable to the Conservatives as they are, still don't produce a majority in my seat estimation model although the most recent one gets close. Another poll out last weekend, done for La Presse by Unimarketing, also showed Liberal decline (to 24.6%) but found much weaker support for the Conservatives (36.3%) with the NDP at 18.9%. This produces a much weaker Conservative minority - 139 C, 81 L, 35 NDP & 53 Bloc. What continues to amaze TC is that even with a 13 point lead, the Ipsos poll doesn't actually produce a Tory majority, hence the header. The Conservatives do have a real uphill struggle to win that elusive majority and it is abundantly clear, poll after poll, that they are not going to find it in Quebec.

There is no doubt the Liberals have had a bad period but as the default choice of so many, especially in Ontario, it would be unwise to simply write them off. I would also like to see the Ottawa media begin to recognize the complexities of a system with extremely strong regional differences, four parties represented in the House, and another on the outside (the Greens) starting to win significant numbers of votes. It is time to move beyond the Harper/Dion soap opera and begin to recognize that Canada is a diverse country with diverse political trends and experiences.

Friday, October 19, 2007

Saskatchewan Election

The Saskatchewan election is underway and as noted on September 6 by TC it looks like the Saskatchewan Party (essentially the provincial equivalent of the federal Conservatives without the nominal connection to the scandal ridden regime of Grant Devine) will win. This time series of polls by Environics conducted over the past four years makes it clear they went into the election ahead. However, it is my impression now that it will be fairly close. The result last time was NDP 30 & the Saskatchewan Party 28 with the NDP winning the cities and the north, and the Saskatchwan Party the rural areas - all nicely visually displayed in this map on the CBC election web site.

The key to the election is Saskatoon (Regina is the real NDP stronghold). The Saskatchewan Party picked up three seats there in 2003 (two on the semi-rural edges). It needs two more to put them over the top. An early local poll puts them ahead (with lots of undecided). However, the campaign has yet to acquire a clear dynamic. There is also one Regina seat and a couple of rural ridings they could gain as well.

The Saskatchewan Party is finally behaving like Ross Thatcher did in 1964 when he ousted the CCF regime of Douglas/Lloyd. It is tilting to the centre, for example, by offering a campaign promise of drug benefits following one from the NDP. TC suspects, however, it will also resemble the conservatism of the Thatcher regime once in office.

I hope to see some polls on this race but Saskatchewan seems to have fewer party preference polls than other places. In the end what seems to matter most is that the province, which gave Tommy Douglas five election victories, appears to think simply that it is time for change.

Tuesday, October 16, 2007

Throne Speech - No election in sight

Watching Michael Ignatieff respond carefully to the Throne Speech makes it very clear that there will be no early election. The punditi talk endlessly about traps, etc but there are none that I can see at the moment. The Liberals can vote for all the prospective bills on second reading. Most of the delays can then be created in the committee if they so desire. And they really only have to get through eight weeks of debate before Christmas.

Parliament comes back the first week in February. That means no election before we see an end of February/ early March budget.

Sunday, October 14, 2007

Al Gore's "Errors"

You have all seen the news stories about the ruling by a British judge about the "errors" in An Incovenient Truth.

Well, the stories are wrong. Consider this one in the Toronto Star:

A British judge has ruled that Al Gore's Oscar-winning environmental documentary, An Inconvenient Truth, contains nine scientific errors or omissions.

High Court Judge Michael Burton was asked to rule on a challenge from a school official who did not want the film shown to students.

In his ruling Wednesday, Burton said Gore's film "is substantially founded upon scientific research and fact ... albeit that the science is used, in the hands of a talented politician and communicator, to make a political statement and to support a political program."

However, it just ain't so. The judge, named Burton, was quoted by this blog:

Let's look at what Burton really wrote (my emphasis):

Mr Downes produced a long schedule of such alleged errors or exaggerations and waxed lyrical in that regard. It was obviously helpful for me to look at the film with his critique in hand.

In the event I was persuaded that only some of them were sufficiently persuasive to be relevant for the purposes of his argument, and it was those matters - 9 in all - upon which I invited Mr Chamberlain to concentrate. It was essential to appreciate that the hearing before me did not relate to an analysis of the scientific questions, but to an assessment of whether the 'errors' in question, set out in the context of a political film, informed the argument on ss406 and 407. All these 9 'errors' that I now address are not put in the context of the evidence of Professor Carter and the Claimant's case, but by reference to the IPCC report and the evidence of Dr Stott.

If you noticed the quotation marks around 'error' then you are more observant than all of the journalists I listed above. Burton is not saying that there are errors, he is just referring to the things that Downes alleged were errors. Burton puts quote marks around 'error' 17 more times in his judgement. Notice also the emphasised part -- Burton is not even trying to decide whether they are errors or not. This too seems to have escaped the journalists' attention.

Al Gore deserves the Nobel, not the nonsense he has to put up with from the media.

UPDATE: There is a posting about this controversy now available on RealClimate. An excerpt:

A number of discussions of the 9 points have already been posted (particularly at New Scientist and Michael Tobis's wiki), and it is clear that the purported 'errors' are nothing of the sort. The (unofficial) transcript of the movie should be referred to if you have any doubts about this. It is however unsurprising that the usual climate change contrarians and critics would want to exploit this confusion for perhaps non-scientific reasons.

Ontario election, throne speech and possible federal campaign

I don't have much to add to what has been said about the Ontario election. I underestimated the extent of the Liberal victory a little, but my forecasting model worked fine, predicting 69 Liberal seats when you input the actual vote shares.

The only small surprise was the 8% Green share where I had guessed it at 6%, and it was a huge gain from 2003. I strongly suspect the schools issue was at the root of it. It will be interesting to compare to the next federal election. In 2004 and 2oo6 the Green shares were 4.4 and 4.7% respectively. The environment is now permanently a key issue politically - climate change is going to ensure its importance.

The new Ipsos poll gives the Conservatives a 12 point lead over the Liberals 40 to 28. That big a lead should normally be expected to produce a majority. However, my estimate only gives the Conservatives 145 seats based this poll, well short of the 155 they would need for a bare overall majority.

The Conservative vote is exceptionally inefficiently distributed. Just one illustration: the poll has the Conservatives at 63% in Alberta. This would give them about 600,000 more votes than the Liberals in just this one province, far more than they would need to win every seat.

The poll numbers are bad for the Liberals, but are likely to reinforce their determination to avoid an early election. This story in the Star today tells us what to look for:

Apart from a few pro-election hawks, most Liberals are saying they need more time to get the party's act together.

This is why you will be hearing virtually the same words from Liberals in the days leading up to Tuesday's throne speech. They have come up with the "script" and it sounds like this:

"We will take into account all the aspects of the throne speech with only one thing in mind – the interests of Canadians," Dion told reporters in Toronto on Friday.

Doesn't sound like someone who is preparing to hit the hustings.

Wednesday, October 10, 2007

Tory/Tories trouble deeper than thought?.

There are indications the rout of the Conservatives may be greater than the final polls anticipate.

Democraticspace is hedging his final prediction here saying:

All of this is to say that even when there appears to be a consensus in the polls — in this case, showing a strong Liberal majority — there is the potential that it could break differently.

That’s why we provide ranges. Here are the best and worst case scenarios for each party as we see it:

LIBERAL 56-72 41.0-44.0%
PC 25-38 31.1-33.3%
NDP 9-15 16.7-17.9%
GREEN 0 5.9-6.7%

OTHERS 0 1.6-1.8%

So, when you sit back and watch the results come in tonight, keep in mind that nothing in ever certain. We’ve done our best to anticipate the range of possible outcomes, but in the end, as always, it all depends on you, the voters.

And the electionprediction.org site has increased the Liberal total to 67. TC also has heard that internal Tory polling is showing them to be in worse shape than the public polls. They may be headed to the mid-twenties with dire consequences even for their already limited number of incumbents.

We will soon see.

Tuesday, October 09, 2007

Ontario Election - How Large the Liberal Win?

I have been reviewing the polls, averaging them and making some intuitive guesses. Here is my final view of the outcome.

In the last week, the PCs have continued their steady decline. However, the PC vote is older and therefore turns out at a higher rate. In past elections polls have tended to underestimate the Tories. Nevertheless, the news for the Tories is quite bad - leader John Tory is headed towards a significant defeat in Don Valley West - TC estimates by around 8% give or take a few points.

Although the Liberals will win, the very latest polls do indicate a slight dip for them while the NDP is climbing a bit. Both parties are the opposite of the Tories with a disproportionate number of younger voters, and polls have in the past overstated their support.

However, it is the Greens who tend to be the most overestimated of all for a similar reason - a youthful alienated core of support. This tendency for polls to find too many Greens has been acknowledged for the first time to my knowledge by SES principal Nik Nanos who said in their firm's final pre-election release today, "support for the Green Party is usually over-reported in polls compared to actual ballot box". They actually offered a Green adjusted version of their final poll. It will be interesting to compare this to the actual vote shares.

The Conservative campaign was an unmitigated disaster - not only did they latch onto an issue they could have determined ahead of time would be unpopular in their promise on religious schools, but they put too much faith (no pun intended) in their "broken promises" theme forgetting that they offered no basis themselves for persuading the electorate that they, unlike the Liberals, could be trusted to keep promises. John Tory then shifted ground on religious schools in the middle of the campaign to show he too can bend a promise. Incredible! I also thought there was nothing at all in their platform to appeal to voters in a positive sense and I think this latter point should not be underestimated. Their biggest promise seemed to be to cut the health care premium financed by a vague promise of finding efficiencies, something I think many including myself reacted to with a healthy degree of skepticism.

Parties must give voters a reason to support them on election day and this the Tories failed to do. By contrast the NDP offered a clear bundle of commitments - eliminating the health care premium by raising income taxes, rolling back tuition fees and a higher minimum wage.

I have adjusted the final poll averages to reflect my views above and made a few adjustments to the calculations of my model. Here are the numbers:

Liberal PC N.D.P. GP Other Total
Seats 65 30 12 0 0 107
Votes 41.3 32.5 19.2 5.9 1.1 100

One final word: Danny Williams won a big majority tonight but he had a vote share (69.56%) that could have given him every seat. The difference in TC's view: strategic voting.

Tuesday, October 02, 2007

Tory's Meltdown

There are now two polls out giving the Liberals a ten point lead - an SES came out today in addition to the Ipsos poll I wrote about in my last post.

Tory's reversal does not fix his problem - see this Star story about the first day following his turnabout. However, the flip flop will be helpful to a few incumbents in already safe ridings who were no doubt being tormented about the issue by their strongest traditional supporters. The whole affair simply makes him look ridiculous.

Right now we are looking at a repeat of 2003 although I sense at least some limited opportunity here for the NDP to get into the 12 to 15 seat range.