Wednesday, January 28, 2015

How well is the Harper government really doing?


Many political writers have concluded that there is growing support for the Harper government. However, there have only been three polls released in January, one of which from Ipsos reported a strong result for the Conservatives, leading to a claim the Conservatives were on the "cusp" of a majority (my own seat estimate from that poll, however, would actually see the Harperites winning 141 seats, enough for first place but nearly 30 short of a majority). The other two polls, from Ekos and Forum, would see the Liberals winning the most seats.  Averaging Ipsos with the two other January polls would give the Conservatives 132, Liberals 126, the NDP 71, Green 1 and Bloc 8.  A majority for anyone is a long ways off but the race between the Liberals and Conservatives for first place is appears to be close.

Many recent analyses have been comparing the late autumn/ winter to a summer when there were few surveys, some of which had exceptionally strong numbers for the Liberals (enhanced in part by spillover into the federal realm from Kathleen Wynne's victory in Ontario). If we look at quarterly polling averages for 2014 we find the autumn looks much like the spring (January polls are not included here):


Compared to four years ago, however, there is much stronger antagonism to the Harper government. Nanos Research has been conducting surveys annually since 2007 for the Institute for Research on Public Policy on the mood of Canada. Its most recent write-up reports finding "signs of life" for the Harper government. This document is useful because we can compare current circumstances to previous years, particularly 2010 and 2011, when the polls were conducted five or six months before and after the May 2, 2011 general election.

One question asked respondents to grade the performance of the Harper government (See survey details here on page 3). I have isolated those who thought the government's performance either "very good" or "very poor" and put the numbers into the chart to the right. My view is that the "somewhat" category should be seen as expressing ambivalence; they are therefore excluded from the chart (along with the "average" category).

While the Harper government continues to a have small group of enthusiastic supporters, beginning in 2012 and continuing in 2013 when the Senate scandal dominated headlines, a large percentage of Canadians came to judge the government's performance as "very poor". The number dipped a bit in 2014 but given the limited interest of most Canadians in politics the 30% who say the performance of the Harper government is "very poor" is a remarkably high number.

The last couple of months have produced events that should have strongly favoured the Harper government, including announcements of a balanced budget (now in doubt), imminent tax cuts, and a higher profile for national security issues following the events on Parliament Hill and the House of Commons vote the same month supporting the international campaign against ISIS. Even the drop in oil prices should have helped given that the Conservatives have the reputation of managing the 2008-2009 economic crisis well. The government has spent millions on TV ads to remind us all of their fiscal virtues, and terrify us about the prospect of marijuana being legalized. All the while the Conservative Party has offered up continuing reminders via radio, television and internet of why in their view Justin Trudeau should not be considered a suitable prime minister. Given all this shouldn't we be asking ourselves why the Conservatives aren't stronger?

While the polls have moved somewhat in Harper's favour in recent weeks, they are nowhere nearly good enough for the Conservatives to aspire to another majority. Ontario was the key to Harper's 2011 majority. There the Conservatives captured 44% of the vote and had an 18 point lead over the NDP and the Liberals. Recently an Angus Reid poll gave the Conservatives 43% of Ontarians in a poll of eligible voters (there was a slightly smaller share of 40% among the firm's estimate of likely voters). If accurate does the poll mean a repeat of Conservative majority looms? Here is a comparison of the 2011 election results in Ontario to a 2015 seat projection based on the Reid poll.


In a first-past-the-post electoral system an 18 point lead gave the Conservatives 73 Ontario seats in 2011 (83 if applied to the expanded House coming in 2015). The Tory performance was a mathematical sweet spot nearly impossible to repeat, as the projection based on the Reid poll illustrates. The 18 point advantage is gone; even with almost the same vote share the Reid survey tells us Conservatives would lose about 18 seats in Ontario alone. Some recent surveys have given the Conservatives a lead in Ontario, but nowhere nearly enough to repeat its 2011 performance there.

The importance of Ontario to Harper can be gauged by his recent decision to cave in after many months of refusing to meet with Ontario Premier Kathleen Wynne. What led to the change of heart? I have little doubt that this Abacus poll released January 2 (see page 7) played a role. It reported that a substantial majority of Ontario's citizens (57% to 23%) disapproved of his refusal to meet the recently re-elected premier of Canada's largest province

Utilizing an average of national polls from October through December I projected with my forecast model, a Liberal plurality of 135 to 129 for the Conservatives, 71 NDP, 1 Green and 2 BQ in the next House of Commons, again a narrow margin that could tilt either way.

Stephen Harper is no fool; better to keep the majority he has as long as possible. Even if the Conservatives fortunes have turned up (by no means certain) the current political context rules out a spring election. 

Liberal strength remains as far from a majority as the Conservatives. What the next election appears most likely to produce is a House of Commons we haven't seen since 1972, when Pierre Trudeau's Liberals emerged on election night with a one seat margin over Robert Stanfield's Progressive Conservatives.  The Liberals were able to govern because they secured support from the NDP, giving them a working parliamentary majority. A Liberal plus NDP majority has not happened since then (apart from the Trudeau/Chrétien Liberal majority governments).

However, with the near disappearance of the Bloc Québecois, polling over the past couple of years tells me via my seat projections that an election reflecting the polling would have produced the Liberal plus NDP majority of 1972 to 1974 in almost every case.

What could matter in the next parliament is inter-party cooperation between the Liberals and the NDP, something we have not witnessed in over forty years. Our Westminster-style House of Commons is too regionally fragmented to easily produce governing majorities given the current party system. It is time for the NDP and Liberals to develop a viable model of party cooperation - one that is in the interest of both parties.

The NDP in principle is committed to cooperative parliamentary behaviour, given that it would be an inevitable consequence of proportional representation, a key plank in their current platform. In running for leader of the Liberal party Justin Trudeau endorsed ranked voting, which, if implemented, would create an incentive for inter-party cooperation, although Liberal Party commitments with Mr. Trudeau as leader have so far been vague.

Some argue that recent conflicts between the Liberal and NDP caucuses make cooperation impossible. It seems more likely that parliamentary math would trump personal and partisan sensitivities, as would pressure from the parties' grassroots to oust the Harper government.

Saturday, January 10, 2015

The NDP leadership race is underway in Manitoba

CBC Manitoba has an infographic with photos of Manitoba MLAs arranged by their choice for leader. The count is 6 for Ashton, 7 for Oswald, 13 for Selinger with 5 undecided and 2 neutral.

Although caucus support likely will not perfectly reflect the distribution of votes at the convention come March 8, the distribution of support implies a competitive race. Based on the dynamics of the convention that selected Selinger, one would expect that the second choice of Oswald supporters would be Selinger and vice-versa but with the bitterness of the split between Selinger and Oswald it is unclear how this might play out. Given that he is already leader and premier Selinger's position looks weak. If this were 2009 with the same candidates one might expect the second choice of Oswald supporters to be Selinger and vice-versa, although with just three candidates a first ballot victory is a distinct possibility. Regardless, the party will likely emerge divided (at least behind the scenes).

While the NDP has a longer-term structural advantage in Manitoba politics conferred by the Progressive Conservative Party's failure to establish itself as a significant political force in the City of Winnipeg, the short run appears likely to bring some sort of defeat in the next election. While the Liberals don't seem set to repeat the experience 1988 performance of Sharon Carstairs (see details on that on pages 20 to 24 of this paper I wrote in 2008 on the Manitoba Liberals), they could easily draw enough disillusioned NDP votes to make a decisive difference. The PCs are currently running just four points ahead of their 2011 election performance. It is significant bleeding to the Liberals and Greens that is mainly responsible for the NDP's position in the polls. Will it continue? The NDP will certainly make an effort through attacks on the PCs to prompt tactical voting on the part of some Greens and Liberals to revive their fortunes.

Free Press columnist Dan Lett drew a comparison to 1988 in a recent column. However, the political context was quite different. Briefly stated, in 1988 an NDP government at the depths of the mid-term blues was forced into an early election that produced for them a disastrous outcome despite a leadership change as the campaign got underway. This description does not apply now.

The 2015 convention will produce a winner who, having achieved the status of victor, will benefit at least for a brief honeymoon period. Even if it it is the politically-damaged Selinger there remains a year to prepare for the April 2016 campaign. Politics in Manitoba is generally pragmatic as the continued presence of the anti-Selinger dissidents within the party caucus attests. While I would expect a PC victory in 2016 the ongoing strength and advantages of the Manitoba NDP render that outcome at least somewhat uncertain.


Wednesday, November 12, 2014

Selinger's challenge

It appears Manitoba Premier Greg Selinger will meet his leadership challenge head on and it now appears likely there will be a leadership convention next March. It is unprecedented among the provinces and the only federal precedents federally are among opposition PC leaders: Joe Clark's resignation and subsequent candidacy at the 1983 PC Leadership Convention is probably closest (John Diefenbaker also ran at the last minute at the 1967 leadership convention called to get rid of him). I have little doubt Selinger will face a similar problem to Joe Clark, who led on the first ballot but had little in the way of support on second and subsequent ballots.

Clark would go on to lose to Brian Mulroney. There is an ironic comparison to be made to Selinger here. In his second term Mulroney became easily the least successful Prime Minister ever - his coalition fractured east and west into the Reform Party and the Bloc Quebecois, the Meech Lake Accord failed, he brought in the 7% GST, there was a deep recession aggravated by tight money from the Bank of Canada, which worried excessively about inflation from the GST, etc., etc. Through it all Mulroney kept the support of his caucus and cabinet. It is clear that Selinger has failed a similar test even though his 'sins' are modest in comparison.

Instead of raising the PST Selinger should have chosen a larger deficit for a longer period. There would have been less political fallout, less internal dissent and the arguments for raising the PST to support infrastructure apply as well to taking on more debt to support infrastructure investment.

Manitoba easily has the fiscal room in terms of its debt obligations.  I derived the table below from the Department of Finance's fiscal reference tables. I divided net debt for the ten provinces plus Canada by population estimates from Statistics Canada.  The amount of debt owed by each person makes the situation of one jurisdiction comparable to the others. Manitoba has less debt per person than all provinces to its east, but more than the other western provinces who are resource rich.

Net Debt Per Capita
Canada $19,198.25
Newfoundland & Lab. $17,112.75
Prince Edward Island $14,492.46
Nova Scotia $15,659.54
New Brunswick $15,441.02
Quebec $22,151.22
Ontario $19,533.23
Manitoba $13,484.73
Saskatchewan $3,743.38
Alberta -$2,347.82
British Columbia $8,372.81

Josh Marshall of Talking Points Memo made an argument following the results of the recent U.S. midterm elections that I think applies to Canada and Manitoba. The whole thing is worth a read but here are the key points:
What is driving the politics of the country to a mammoth degree is that the vast majority of people in the country no longer have a rising standard of living....wage growth has basically flat-lined since the crisis....
Fundamentally, most people don't care particularly how astronomically wealthy people are living their lives. It is a distant reality on many levels. They care a great deal about their own economic circumstances. And if you are not doing any better than you were 5 years ago or a decade ago or - at least in the sense of the hypothetical median wage earner - 40 years ago, that's going to really have your attention and shape a great deal of your worldview and political outlook.
So, let me sign up with those who are saying that it was a mistake not to run more clearly on the President's (and the Democrats') economic record. Unemployment is back down to something like normal levels (under 6%); the deficit has fallen consistently and is now back to pre-crash levels judged as a percentage of GDP (which is the only meaningful way to judge it); the stock market has done incredibly well. Yes, totally.
But here's the thing: As long as most voters are still just treading water in their own economic lives, Republicans can say, "Oh yeah, they say the economy's doing great with all their fancy numbers. But that's not what I see!" To an extent that will just be another Republican paean to innumeracy. But it will resonate because rising employment is not leading to rising wages. And that's the core economic experience of wage earners who make up the overwhelming number of people in the country. In a critical sense, it is true.
Wage stagnation stemming from growing inequality is a phenomenon likely to drive our politics for quite a while. However, there could be some periods of wage growth. Bill McBride of the Calculated Risk blog thinks the U.S. economy is growing well enough that the United States might actually begin to see wage growth in the coming year, although the test will be if it exceeds inflation. If that happens, then with a six or more month lag we might see something similar in Canada. It has the potential to benefit incumbent governments still here in 2016 but it is likely that the Manitoba NDP government has simply been around too long to be re-elected, regardless of who leads it and the economic circumstances that will prevail when the next Manitoba election is held.

Thursday, November 06, 2014

The crisis in the Manitoba NDP: the politics of taxes

The current crisis in the Manitoba NDP has its roots in the decision of the Greg Selinger government to raise Manitoba's 7% provincial sales tax to 8% in its 2013 budget, making it effectively the same as that in Ontario (where it is blended with the federal GST into an HST) and some other provinces. The government was looking for revenue (see the 2013 budget speech page two) after over a billion dollars in flood protection infrastructure was identified in a report commissioned following the devastating Manitoba floods of 2011, which were, as the report put it of "a scope and severity never before experienced, in recorded history...". They decided on the sales tax increase despite promising in the 2011 election that they would not raise the PST. However, the government faced urgent revenue pressures if they were to fund the infrastructure required for flood protection. The alternative in a real sense was an infrastructure deficit, something other governments, especially conservative ones (think Mike Harris and the cemented in Eglinton subway and its role in Toronto's current transit crisis), have chosen in the past.
Manitoba Premier Greg Selinger

A sales tax increase regardless of the circumstances was inevitably going to be unpopular. Theresa Oswald, one of the dissidents who just resigned (and thought to be Selinger's most likely successor), saw the problem right away when she spoke to Free Press columnist Gord Sinclair a week after the 2013 budget. According to Sinclair, "Oswald said Gary Doer had a saying: 'You don't increase beer prices, and you don't hike the PST.'" What was Doer referring to? It is worth unpacking.

We live in an era where a large number of those in the labour force face stagnating real wages and incomes, in part as a product of the 2008 financial crisis, but more because over a long period incomes have become more unequal as many workers can't increase their incomes enough to keep up with prices. There are many reasons: union power is weaker while an international race to the bottom has depressed wages here and driven jobs abroad. On the other hand the power of the wealthy, particularly in the financial sector, has become stronger. For many struggling to make ends meet, taxes are an easy and tempting target. Hence the frequent efforts, particularly on the part of conservative governments, to promise tax cuts, especially highly visible sales taxes (think Harper and the GST), even while they raise invisible taxes.

However, the reality is that a community cannot increase its net income by cutting taxes (apart from the context of national macroeconomic stimulus policy and that does not apply here). Doer I think intuitively understood the dangers from hiking visible taxes like the PST in the context of growing income stagnation. Note he said beer prices, not single malt Scotch. But the reality remains that one can only increase private consumption by cutting public consumption and sometimes there are urgent public needs for expenditure. Flood protection is arguably in that category.

The trade-off often seems opaque, so unless the public is convinced that a vital highly valued service, say health care, is truly endangered, they are likely to recoil at this kind of tax increase. Infrastructure even for flood protection can seem like an abstraction to many and not nearly as meaningful as direct ongoing services such as health care and education.

PC Leader Brian Pallister
The trade-offs are real but conservative politicians, such as Manitoba PC leader Brian Pallister simply pretend they are not a problem. He met recently with the Winnipeg Free Press editorial board in a session the paper televised online.

Columnist Dan Lett wrote about it afterwards:
The problem for Pallister is that with an unprecedented opportunity to galvanize the support needed to form his own government, and bury the NDP once and for all, he continues to struggle on the most basic policy issues. ...
At times, we were reminded just how incomplete many of Pallister's plans really are. He was asked to explain, once again, how he would roll back the one-point hike in the PST now dedicated to infrastructure -- a central Tory pledge -- and still maintain current spending on roads, bridges and sewer/water projects.
Pallister said he can replace the $260 million generated by the PST increase via cost savings without cutting infrastructure spending, laying off staff or cutting core services.
When pressed, Pallister impatiently said that despite not having outlined how he will find hundreds of millions of dollars in savings without fewer people or services, it can be done. "I guess you'll just have to take me at my word on that." 
Pallister can't say because he understands all too well how unpopular a candid answer would be. Perhaps the most widespread dishonesty used by conservative politicians today is the pledge to reduce spending by only cutting waste and inefficiency. This was a canard used by Rob Ford, who promised in 2010 when he was elected Toronto's mayor "No cuts to services, guaranteed", but eight months later he was proposing hundreds of millions in such cuts. I have little doubt Mr. Pallister will be just as unsuccessful in this respect as Rob Ford, when, as now seems likely, he becomes Manitoba's next premier.

The 'lower taxes but no cuts to services' pledge is a delusion. It is a posture, however, that can pay political dividends. In the recent Toronto mayor's race, Rob's brother Doug received a third of the vote with a platform of large spending increases on subways combined with pledges to reduce taxes. It was pure fantasy, but did not get the political humiliation it deserved because the delusion of tax cuts for nothing has a powerful appeal in the context of stagnating incomes.

If you want to increase sales taxes it is possible, but an extremely difficult political sell. One successful example was provided by the Saskatchewan CCF albeit during a period of strong rising real incomes. In 1950 Provincial Treasurer Clarence Fines raised Saskatchewan's sales tax by one point in his March budget from 2 to 3 percent. He needed the extra revenue to  help finance Saskatchewan's hospitalization plan (implemented January 1, 1947) and argued that the sales tax increase was fairer than raising a per capita tax in the form of a health premium, which was introduced under the expectation that the federal government would have by this time have introduced a national health program first promised in 1945. It would have paid 60% of hospital costs and the health premium was expected to be a condition, but the federal hospitalization program would not see the light of day until 1957-58. Exemptions to the sales tax on food, drugs, cleaning supplies, second hand goods and other items, previously implemented by the government, meant Fines could argue that the sales tax was 'less regressive'.

The CCF had originally promised to eliminate the sales tax but even before taking office concluded they could not do without it in the short term. However, the sales tax was contentious. So it is understandable that Fines would complain in his speech: "This tax has been criticized more than any other provincial tax, but this criticism has been political and not based on sound reasoning." The difficulty remains today and is compounded by how such taxes interact with growing income inequality.

The split in the Selinger government's ranks has likely become more problematic politically than the tax itself or the breaking of a promise. To cite an example from history, the fall of the Diefenbaker government in 1963 was not about nuclear weapons, it was about disarray in cabinet. There is nothing more likely to sap confidence in an administration than division in the ranks. The only realistic way out now for Manitoba's New Democrats is for Greg Selinger to resign so that a leadership convention can resolve (perhaps only temporarily as the federal Liberals have shown) the internal conflicts. That said, the dissenters all voted for the sales tax and won't escape political responsibility unless the tax is rolled back, but then whoever governs were that to happen must deal with the consequences, which, as Mr. Pallister has so aptly illustrated, are not so easy to address.

The dissenters should have recognized that after 16 years in office the accumulated grievances that accrue to any administration would likely have produced defeat (the election is likely to take place in April 2016). I am quite skeptical of the claims of imminent electoral disaster. However, it is likely the dissenters think a change in leadership might bring yet another victory, which also seems unrealistic despite the weaknesses of the opposition parties on the sales tax issue. We have seen the inadequacies of Mr. Pallister in dealing with it; the same is also the case for the Liberal leader, Rana Bokhari, who has promised only to place revenues from the sales tax hike in a separate fund for infrastructure. The election remains a long way off, but the divisions within the NDP must be addressed. While the exact outcome is far from certain, five term governments in Canada (apart from Alberta) have been exceedingly rare.

There is an irony for the NDP finding itself in a political crisis based on the sales tax. The Manitoba sales tax was introduced in 1967 by the PC Duff Roblin government as the Education Revenue Tax (the Saskatchewan tax had a similar name at one time). When Roblin left to enter federal politics reaction within the Progressive Conservative Party to his record including the sales tax led in part the PCs to chose Walter Weir, a rural small 'c' conservative. Roblin's government was highly progressive building a modern state in the province, expanding provincial expenditures on education and funding considerable capital investments such as the Red River Floodway, which as most Manitobans know has saved billions in flood damage. The result of the PCs moving right (along with the provincial Liberals) was the election of Manitoba's first NDP government in 1969 under Ed Schreyer.


Sunday, October 19, 2014

TC Norris quoted on New Ekos poll on iPoltics

I was quoted in this new iPolitics post regarding a new Ekos poll, which has very unfavourable numbers for Stephen Harper:

Justin Trudeau’s Liberals are well out in front with 38.5 per cent of the vote. The Conservatives are at 26.4 per cent, with Thomas Mulcair’s NDP in a close third at 25 per cent. However, the NDP leader has the highest approval rating and the New Democrats have the most potential to grow, leading EKOS President Frank Graves to conclude the NDP could form the opposition, despite its current third place standing.
“The next Parliament, if it were to be created based on the poll that we have, would see a strong minority Liberal government with a sizeable NDP opposition with Mr. Harper reduced to a third party, rump status,” said Graves.
That assessment was backed up by veteran poll analyst Paul Barber, who blogs under the name TC Norris. He said if these numbers were to hold up through a campaign, and based on the regional breakdowns, the Conservatives would finish a distant third, 70 behind the Liberals and 27 behind the NDP.
Screen Shot 2014-10-19 at 1.24.27 PM

Friday, October 17, 2014

Forum polls and the Toronto Mayoralty

There has been a fair bit of skepticism voiced on social media this week about Forum polling in the Toronto mayoralty race.  Forum uses interactive voice recognition (IVR) polling as its method. Essentially a computer calls you and a recorded voice leads you through choices you enter on the telephone keypad - I have been polled by Forum a number of times over the years including during the current civic contest (When being quizzed by Forum I am reminded of the Woody Allen joke in Annie Hall: "I would never want to belong to any club that would have someone like me for a member.")

Forum's results are clearly different from others. Other polling companies including Ipsos-Reid, Nanos and Mainstreet Technologies have polled in recent weeks. Here is a table with results comparing the average of Forum's numbers with the others taken since mid-September.

Olivia Chow
Doug Ford
John Tory
Forum Polls
21.8
33.6
40.0
Other Pollsters
25.9
27.8
46.1


As one can see clearly Forum reports results that are clearly more favourable to Doug Ford than the others and less favourable for both John Tory and Olivia Chow.

What to make of this? One way would be to examine Forum's recent efforts in other elections. Most recently it conducted a poll during the New Brunswick election held September 22, 2014. Forum reported a tie in the popular vote but predicted that the PCs would win the most seats. By comparison, Corporate Research Associates, a Halifax based polling firm that uses conventional live telephone interviewer polling predicted that the Liberals would win the election by nine percentage points. The actual result produced a Liberal majority based on an eight point lead in the popular vote.

In Ontario there were many more polls. Out of seven pollsters who conducted a poll towards the end of the campaign Forum had a total error greater than all but one other pollster. Here is a table displaying the total error comparison:


One comment on Twitter suggested taking another Forum poll result with a truckload of salt.  As Sarah Palin would say: "You betcha".



Wednesday, September 17, 2014

Scottish Referendum: NO side likely win

The trend recently has been for the YES side to stop growing and for the NO side to maintain a consistent lead.  Two excellent analysts that I follow both conclude the NO side is likely to win. However, there has never been a Scottish referendum on independence so there remains some uncertainty.

Claire Durand in a post published today(and she may update once more) estimates the NO side lead at seven points. She has been following the youth vote closely and the most interesting part of her analysis is a chart that reports support for the YES among youth peaked about two weeks ago and started to decline.


Peter Kellner of the UK polling firm YouGov has an interesting analysis published two days ago. YouGov is the firm that reported the YES side ahead in a poll done for the September 7 Sunday Times. However, Kellner argues things changed during the week following when the NO campaigned counter-attacked:
Last week, following our shock poll, the No campaign fought back strongly, with Gordon Brown appealing to Labour voters flirting with independence. Better together was aided by retailers warning of higher prices and Scottish banks warning of moving their headquarters to London, should Yes win on Thursday. Our midweek poll caught a marked change of mood. Not only did we report a 3% swing back to No, which returned to the lead, albeit only narrowly. More significant was the six point rise in the number of Scots thinking Yes would be bad for their economy – and an eight point rise in the number fearing that their own finances would suffer.
It should not be surprising that a campaign that effectively raised the prospect of bad economic consequences if the YES were to win should be effective. The use of Gordon Brown as a communicator also helped.  For example, see this excerpt from a Brown speech just uploaded to Youtube today:



Polls close tomorrow night at 10 pm Scottish time, 5 pm Eastern time in Canada.  If there are good exit polls and the margin is more than a couple of points we will know the outcome early.