Wednesday, November 30, 2005

More thoughts on Election 2006

1. The Liberals greatest problem is voter fatigue. During the 2004 election I called it the fourth election problem. At the time I said: In recent decades with but a few exceptions, Canadians have tended to get fed up with the party in power, generally after two terms, but almost always by the time a party is seeking a fourth term. This applies to both federal and provincial governments.

This has no doubt shown up in Conservative focus groups, and explains why Stephen Harper talked incessantly yesterday about “change”, according to the CBC mentioning the word 'change' at least 50 times in his kick off speech.

2. It will be a long campaign and with some measure at least of voter dissatisfaction with the two leading alternatives, I expect some volatility in the polls. Judging by the first Ipsos-Reid survey, this has already started. SES Research announced today they will be doing an enhanced daily tracking poll. Their poll did catch the last second shifting to the Liberals in the last election and the losses suffered at the end by the NDP, but did not get all of it, making the election look much closer than it turned out to be. This time their samples look like they will be large enough to do some meaningful regional tracking.

3. The TV debates were announced tonight, and the change in rules suggests to me that they will work better overall, and that means they have greater potential to influence the outcome. The 2004 debates were a disaster. As I wrote at the time: Cross-talk and noise is incomprehensible to most viewers, and simply discredits the political process. It also means the debates are less influential as a consequence, and end up politically as a wash.

Most viewers will tune out a wall of sound but if debate is intelligible it is more meaningful.

Tuesday, November 29, 2005

And they are off...

A few observations as the campaign commences:

1. Most of the polls since the 2004 election campaign, once run through my number cruncher, have produced minority governments, usually Liberal although today’s quickie overnight Reid poll produces a weak Conservative minority.

2. It is likely the Liberals will close the gap in Quebec, as the election becomes more polarized around federalism and it becomes obvious that the NDP and Conservatives don’t have strong campaigns on the ground in the province. Right now the BQ is looking at 65-68 seats. However, the Liberals closed strongly in 2004 in Quebec under similar circumstances and could get back most if not all of their 21 2004 seats.

3. It is harder for the Conservatives to win a majority than the Liberals. Assuming the Liberals could win 20 seats in Quebec (they took 21 in 2004) they would need to win about 58% of the seats in the rest of Canada whereas the Conservatives must win 66% of those constituencies, in both cases to obtain a bare majority.

4. The Liberals face a Conservative party with two strengths this time and one glaring weakness: they are much better organized (remember, they had just barely established themselves as a party in 2004) and better financed. However, their Leader, Mr. Harper, is a liability. He comes across as sour and angry (whereas Mr. Layton and Mr. Martin generally come across as sunny and optimistic) and what is worse for the Conservatives, he is better known than in 2004. He also strikes me from time to time as having a death wish. Either that or some of the foolish things he says, such as linking the Liberals to organized crime (fyi, the Republicans in the United States might be better candidates for such charges) simply show a political ineptitude I have trouble believing on part of some one as obviously bright as Mr. Harper.

5. The NDP was the hard luck party last time round, losing a dozen or so seats by tiny margins. They appear set to conduct both a better campaign, and with a leader whose communication skills have grown exponentially.

6. Using a weighted average of the three most recent polls and applying the seat forecaster we get the following seat distribution: L – 123, C – 80, NDP – 38, BQ – 67.

Saturday, November 26, 2005

Two New Polls Can't Agree on B.C.

There are two new polls out today - from Ekos and Ipsos-Reid - with completely different media spins (see the poll links below to the Star and the Gazette). I just want to note the contradictory findings for B.C. with respect to the Liberals and Conservatives. Ekos reports that it finds: L-43, C-21, NDP-33. On the other hand Ipsos-Reid says it is the Conservatives who are surging in B.C., and the current party preferences are L-26, C-33, NDP-33. Only the NDP numbers create some resemblance between the two polls. Certainly the most intriguing province.

Added note: I estimate the Ekos poll size as about 105-110 while the Ipsos-Reid sample is 117. With sample sizes that small the possibility of large errors is in my view considerable. That is one reason that I like to aggregate and average. If we average these two polls, we get L - 34.5, C - 27, & NDP - 33.

Ralph Klein's Alberta

The most politically pertinent fact about Alberta in the federal election is a Conservative one-party state, and oh, by the way, it is ruled by King Ralph. He was in the news again this week quoted as saying the Liberals would probably win a minority: "If I were a betting man, I would say there would be another minority," he said. "It's a damn shame.” Why will the Liberals win? It is because Harper while a "bright, articulate individual, perhaps he's seen as too much on the right".

The conventional wisdom about the latest pearl from Ralph is that once again he is helping out his nemesis Paul Martin. However, this particular comment could paradoxically help the Conservatives. How? If the Conservatives were too have any chance, and their prospects are indeed not too bright, they would have to come from behind, and equally important the NDP would have to do well, stealing votes from the dreaded Grits. Ralph’s comment effectively says I like Harper but he hasn’t got a chance to win, so you socialists are safe in voting NDP. The most likely impact of his comment is to help the NDP a little by helping to suppress the strategic voting that causes soft NDP votes to melt away to the Liberals.

The Conservatives at this point are likely to lose again, but there is enough unhappiness about the status quo – a recent Decima poll reports that it is desire for change that is the Liberals’ greatest weakness – that one can’t be sure. The real problem for the Conservatives is that they and their program are viewed positively by relatively few Canadians.

As for Alberta and the polls, there is not much to say. The Conservatives are down a bit here too, 4.1 points below their 2004 finish, but not by enough to matter. The Liberals post Gomery are also down just below their 2004 performance. The gains in Alberta seem to have gone to the NDP which is 5.5 points ahead of their 2004 pace. Based on these numbers my forecast model appears to give the NDP one constituency, Edmonton Strathcona (currently represented by Rahim Jaffer). I do not think they can actually win here but it illustrates how well the NDP is doing generally. The only change I expect to see is that the Conservatives should win David Kilgour’s riding (he is not running again) of Edmonton Beaumont. The Liberals were running nicely ahead of their 2004 pace in Alberta earlier in the fall but the Gomery report has moved Alberta back close to the 2004 election result.

Thursday, November 24, 2005

Regional numbers – West to East, beginning on the west coast

Now that it is clear that an election campaign will commence next Monday I intend to begin laying out a series of regional assessments about where the parties stand going into the election. I want to continue the before and after Gomery comparisons but also look closely at current poll performances compared to the results of the 2004 election.

Overall the region that appears to have changed the most since the 2004 election is British Columbia. There have changes elsewhere but it is on the west coast where the potential exists for a large number of seat switches.

As in the case of every other region, the Liberals lost support in B.C. between their September-October average prior to the release of Gomery and in surveys conducted afterwards, losing an average of 3.4%. However, we also find that the Conservative party also lost ground between the two periods, dropping 1.5%. The NDP picked up 3.1% and the Greens gained 1.3%. This is representative of the bigger picture. Compared to 2004, if we consider post-Gomery polls we find that the Liberals average 36.1%, 7.6 points higher than in the 2004 election, the NDP averages 31.1%, 4.6 points higher than 2004, while the Conservatives at 24% are 12.1 points below their 2004 result. If this prevailed on election day, B.C. would elect 18 Liberals, just 8 Conservatives and 10 New Democrats.

However, there is a history of the Liberals not holding their pre-election support. In 2000 the Liberals averaged 38.5% in a group of polls in the weeks before the call of that election, but on the day of the election the party received just 27.5%, a loss of 11 points. Could it happen again? The Liberals under Paul Martin have been aggressively chasing votes on the west coast. I would expect some drop off for the Grits, and indeed some recovery on the part of the Conservatives. However, in 2000 the pattern across the West was identical to B.C. This year the Liberal support post-Gomery is below their 2004 performance elsewhere. In 2000 NDP support in B.C. in the pre-election polls essentially predicted its election performance, while the then Canadian Alliance increased its support strongly at the expense of the Liberals.

Recent developments in provincial politics do augur well for the NDP. They did better than expected in May 2005 provincial election (receiving 41.5% of the popular vote), and more recently, benefited from the mishandling of the teachers’ walkout by the Campbell government. The NDP’s weak performance in 2000 was linked in part to the unpopularity of then Clark/Miller/Dosanjh governments. Apart from purely federal considerations, the NDP benefits from the current provincial context.

I don’t have regional numbers on leadership but at the national level Harper’s negatives are high, 58% in the October Strategic Counsel poll. The same poll notes that his negatives were increasing in Ontario at that time. My guess is that the same was true of B.C.

When I look at other regions’ numbers it does not seem a stretch to me to think that the 2006 election could produce numbers similar to 2004. At the starting gate, British Columbia looks dramatically different.

Addendum: I forgot to note in this post that B.C. has a history of being the most volatile, politically diverse province in the country. Back in the fifties and sixties, B.C. was represented in the House of Commons by four different parties, Liberal, Progressive Conservative, the CCF-NDP and Social Credit. In different elections, the Liberals, PCs, NDP, and Reform have dominated, winning at least 70% of the constituencies in at least one election each. Its unpredictable character, unlike say Alberta, makes it always intriguing.

Wednesday, November 16, 2005

Election 2005-06

Rarely has there been an election campaign that spanned two calendar years but that appears to be what we are about to have in Canada, starting apparently on November 28. While the election could have taken place in February if the Liberals had acquiesced in the NDP’s proposal, instead a collision resulting in a non-confidence vote now appears inevitable.

So what is the political position of the parties at this moment? I have averaged all the recent polls conducted after the Gomery report was delivered, and compared them to an average of national polls from September and October. Gomery has weakened the Liberals, but they are still ahead. Although it is the Conservatives who have exploited the issue most energetically, they are not biggest beneficiary of Liberal losses. The Liberals averaged 4.3% less in the post Gomery polls compared to earlier in the fall, the NDP 2.1% more, the Conservatives 1% more and the Bloc gained 0.8% (a 1.7% gain in Quebec).

This has produced some spectacular results for the NDP in terms of potential seat gains. For example, an average of the three recent Ipsos-Reid, Léger Marketing and Ekos polls would give the NDP 36 seats. The SES poll out today would translate into 49 seats. Only the first Strategic Counsel poll for the Globe taken in the days just after the report was released would have given the Conservatives the greatest number of seats; otherwise the Liberals are ahead.

My view is that the preferences now are highly unstable, and that the campaign will matter a great deal. More to come….

Monday, November 07, 2005

Post Gomery Polls - Bad News for All?

The two post-Gomery polls from Ipsos-Reid and Strategic Counsel have bad news for everyone. I initially thought the report might not have much impact. I was wrong.

Clearly, after the impact of the spring hearings and the election that never was courtesy of the Stronach defection, public opinion drifted back to a default distribution that left the Liberals ahead. However, it must not be forgotten in these situations that the preferences at the margins are quite soft. In the new configuration that remains true.

So how is this bad news for all?

1. It is obviously bad for the Liberals who lost about 9 points from the three most recent pre-Gomery polls and currently are well below the 36.7% they received nationally in 2004.

2. The bad news for the Conservatives is that their support, while it recovered by about 2.5% from the pre-report polls, is barely above the 29.6% they received in 2004 a figure that is within the margin of error of the two polls even when combined.

3. The NDP on average is up 3.8% from the 2004 election and 2.8% from the pre-report polls. My seat estimate for the Ipsos-Reid polls suggests they would win 43 seats. The NDP would do best in exactly the scenario portrayed by the polls, a strong third place in the context of weakness for both of the leading two parties.

So isn’t this good news? The problem for the NDP is that it sets up a classic strategic voting paradigm. The Liberals can now stir up fears of a Harper victory with the prospect of attracting soft New Democrat voters more fearful of a Conservative win than upset by Liberal ethics.

4. Surely it is good news for the Bloc, at least. Again the answer is no. The Bloc has had trouble cracking the 50% popularity barrier in any significant way (since June they have averaged 51%). The Ipsos-Reid poll gives them 52% and Strategic Counsel 57% but most of the 13 point Liberal drop in the Ipsos-Reid poll in Quebec went to other federalist parties while in the Strategic Counsel poll the report did not change the results much from a mid-October poll.

When the election comes it will be the Liberals, who are organized in Quebec with strong candidates, not the Conservatives and the NDP so the Liberals, despite all the negative fallout from Gomery, could potentially equal their total of 21 seats won in 2004.

I would not expect the Gomery effect to last given that it has decayed on two previous occasions. However, the Conservatives, the Bloc and the NDP can now use the findings of Gomery in political communications such as TV ads during the election campaign. Perhaps the worst news out of these polls for the Liberals is that the exoneration of Martin in the report has been discounted by a public looking for someone to blame. Strategic Counsel reported that only 33% believe Martin’s contention that he was not involved in the scandal.

The colourful Chrétien sideshow grabbed media attention, but as the former PM will not be running in the next election, his effort to clear his name has no broader political significance.

Today it appears Layton may be ready to provoke an early election. An election campaign to watch during the holidays would be a great Christmas present for T.C.

Tuesday, November 01, 2005

Reaction to Gomery

The report is out and I suspect interest will be high among the public.

One key difference from the AG's report: Gomery made a low key statement that was little more than a restatement of elements of the report, and refused to take questions, unlike AG Sheila Fraser who held court and expressed shock and outrage. I have long suspected that her highly-charged emotional news conference triggered a sympathetic response among the general public, who regard the AG as an almost God-like figure (very wrongly in my view).

However, it does mean response to this report will be dictated by how the public reads the reaction of the key figures: Martin, Harper, Chrétien, Layton, etc. The initial response may then depend on prior partisan preferences without any authority figure such as Sheila Fraser to prompt moves in an alternate direction.