Monday, May 28, 2012

The Harris-Decima poll on Mulcair's views

Polls frequently measure something other than what they appear to be estimating.  Take the example of the Harris-Decima poll for CP headlined:New poll suggests Canadians split over NDP Leader Tom Mulcair's energy views.

At first glance one might think that Canadians are assessing Mulcair's views on how the Alberta oil sands are causing the Canadian dollar to be overvalued, costing Canada manufacturing jobs, as a consequence of the failure of the federal government to enforce existing environmental legislation.

The problem with polling on this topic is that it is complex, and it is highly likely that relatively few understand it. As political scientist John Zaller noted in his invaluable work, The Nature and Origins of Mass Opinion (on page 76):
To state the matter more generally, most people aren't really sure what their opinions are on most political matters.... They're not sure because there are few occasions, outside of an interview situation, where they are called upon to formulate and express political opinions. So, when confronted by rapid-fire questions in a public opinion survey, they make up attitude reports as best they can as they go along.
The HD survey asked two questions; the first sought to measure awareness:

"Recently, NDP leader Thomas Mulcair has made comments regarding the impact of the oil sands on the Canadian economy. Have you read, seen or heard something about these comments?" In addition, Mulcair's views were described this way: “Mr. Mulcair suggested that the oil sands raise the value of the Canadian dollar, which is hurting the economy in other parts of the country as buyers can no longer afford to buy as much from Canadian manufacturers.”

Note that the question does not provide the environmental context of Mulcair's position, something he repeats whenever he talks about this issue.  See this exchange in the House of Commons May 17 where, for example, Mulcair says:
Mr. Speaker, 500,000 good-paying manufacturing jobs have been lost because we are not enforcing legislation. We are not enforcing the navigable waters act. We are not enforcing the migratory birds act. We are not enforcing the Fisheries Act. We are allowing these companies to use the air, the soil and the water as an unlimited free dumping ground. Their model for development is Nigeria instead of Norway. We know what we want: it is sustainable development to protect future generations. 
It should not be forgotten that many people are reluctant to acknowledge their own ignorance of an important issue, so the poll almost certainly overstates Canadians' actual awareness of his views. But the poll reported that just 44 per cent of Canadians even claimed to be aware of his views. Logically, only those who affirmed awareness of his views should have been asked the following question (although it appears all were asked): "Would you say you strongly agree, agree, disagree, or strongly disagree with Mr. Mulcair's remarks"?

The poll perhaps not surprisingly does not really give a picture of what Canadians think about this relatively complex argument.  Instead it taps into pre-existing attitudes to the oil sands and Mulcair. What matters most are the words in the question. The words almost certainly prompted respondents to articulate how they felt either about Mulcair and the NDP, or the oil sands.

In its release, HD notes the following:
Agreement with Mulcair’s comments mirrors partisan ideology fairly well with the majority of NDP supporters agreeing with him (55%) along with 64% of BQ supporters with Liberal supporters fairly evenly split (48% agree vs. 44% disagree). Green party supporters were actually the second most likely to disagree with Mulcair’s assessment (56% disagree).
Hardly a surprise.

What strikes TC about the views of Green supporters is while it might seem surprisingly at odds with the strong environmental content of Mulcair's views, it is easily explained.  It is likely that many Greens are low information protest voters who simply don't know what Mulcair's views are.  The HD questions don't include Mulcair's environmental views, and media reporting on this score has been lamentably deficient.

The pundit class and the media have largely echoed the government line put succinctly by Heritage Minister James Moore in the exchange linked to above:  "The NDP approach is to attack the west, divide Canadians and attack parts of this country that he has never even been to."

Not to pick on him but, for example, pundits like Bruce Anderson speculated: "Mulcair’s oil-sands musing risks halting NDP momentum."  Anderson suggests that "some" think Mulcair "wants to deliberately split the country, rallying Ontarians and Quebeckers to put the NDP over the top by blaming the West’s success for Central Canada’s misfortunes."

The regional conflict paradigm, the government's narrative, has been the dominant narrative about this story in the media and among pundits. The actual story is more complex and could have been summarized clearly and fairly.  It wasn't.
However, because the debate is complex, and not really of immediate relevance to people's lives (although the long-run implications of how the tar sands are developed do matter), the whole exchange has been mostly an elite "inside the Queensway" discourse. While many media pundits were sure it would harm the NDP, it has in fact appeared to have had little impact one way or another. Two polls have been conducted and released since the contretemps emerged. The NDP leads in one, and both rank Mulcair as the most popular leader.

Mulcair and the NDP have not been harmed by this debate, and Mulcair, to his credit, confounded media and pundit expectations by sticking to his principles.  However, he does need to communicate in terms that are more easily accessible. A few understand what "internalizing the costs" of the tar sands means but most of the Canadian public does not.


Sunday, May 27, 2012

Recent research at Harvard makes it clear the Harper EI changes make no sense

Employment insurance is there to provide income support to workers who have lost their jobs. Ideally, if the labour market is to work efficiently and well, then workers generally should be able to find comparable employment to what they had before.

The Harper government's changes are headed in the completely wrong direction.  I think Tom Walkom has it right in the Star in his column headlined "EI changes driven by contempt and ideology".

Moreover. there is recent research that contradicts the Harper approach to unemployment insurance.

As a kind of add-on at the end of a fascinating column in the Financial Times called "Lunch with the FT" (featuring a lunch discussion between FT columnist Martin Wolf and Paul Krugman) there is the following:
Young economists: The empiricists strike back
The empirical work of young economists, as championed by Paul Krugman, is influencing policy debates. Here are some names to watch, say FT leader writers Martin Sandbu and Ferdinando Giugliano:....

Raj Chetty, 32, is one of the youngest-ever tenured professors at Harvard’s economics department. He is best-known for going against the conventional theory that unemployment benefits reduce the incentive to look for work. Chetty’s research shows benefits can be helpful as they prevent people from rushing to take an unsuitable job.
The full Chetty study alluded to in the column can be found here:

I think someone should tell the Harper government that perhaps that practical, empirical research says they are headed in the wrong direction in this area as in so many others.

Sunday, May 13, 2012

Alberta election results

For most what was striking about the Alberta election outcome was that fact that there was almost no hint of what was coming in the polls. The polling error was significant and TC can't really account for why it happened. There are good discussions of what might have happened here and here. However, it was not without some precedent in recent history.

It should not be forgotten that there was significant polling error at the time of the last federal election. It was most important in Ontario but significant elsewhere (of course we know now that voter interference may have been involved).  The polls had the average lead in Ontario for the Conservatives as 9 per cent while on election day their actual lead was 18 per cent, enough on its own to make the difference between majority and minority.

In Alberta comparing the results of the 2008 election with 2012 reveals that the key region in the election was Calgary.  The PCs significantly overperformed in the city relative to their overall provincial performance.

As one can see in the table below the PCs lost 8.7% province-wide but actually gained a little vote share in Calgary.

Change 2012 minus 2008 PC Liberal NDP Wildrose Green Other
Alberta -8.7% -16.5% 1.3% 27.5% -4.2% 0.5%
Calgary 0.3% -22.0% 0.6% 26.7% -4.2% -1.3%
Edmonton -2.4% -17.2% 3.9% 16.8% -2.8% 1.7%
Lethbridge/Red Deer -8.0% -22.5% 8.4% 24.5% -4.2% 1.8%
Rest of Alberta -18.7% -11.4% 0.0% 33.9% -4.9% 1.1%

This does suggest to TC that the endorsement from Peter Lougheed mattered in Calgary as did the comments of Calgary Mayor Nenshi attacking the extremist comments of two Wildrose candidates.  So too did the climate change comments of Wildrose leader Danielle Smith.

Interestingly, the Alberta outcome has led to some despair among more right-wing Canadian conservatives such as Gerry Nicholls, a former VP of the National Citizens' coalition. For him the result in Alberta meant 'Canadian conservatism is dead':
Time of death: April 23, when Alberta’s conservative-leaning Wildrose Party, after being swept up high on the winds of the polls, came crashing down to Earth with a disappointing thud. What made this event the equivalent of an ideological house crushing is not so much the result of the vote, but rather how that result is being interpreted. Experts are blaming the Wildrose loss on its conservative agenda. They say Wildrose was just too radical to win.

For instance, in its electoral postmortem, the Toronto Star gloatingly pointed out that Albertans didn’t share “Wildrose’s enthusiasm for rehashed ‘firewall’ policies, privatized health and charter schools.” Of course, you would expect those on the political left to make such an argument. But surprisingly Wildrose leader Danielle Smith is also saying much the same thing. As she put it, “We have some soul-searching to do as a party. Our members have now seen that some of our policies were rejected by Albertans, quite frankly .… We will be revisiting some of those. You can’t run a government if you don’t get sanction from the people.”
He is right that what really matters is that the Wildrose leader herself acknowledges that her party is too extreme for the Alberta electorate.

Nicholls goes on to offer an alternative explanation but perhaps he is on to something.  Despite that great triumph a year ago for his former Citizens' Coalition colleague Mr. Harper, there are signs that Canadians are beginning to tire of this kind of politics.

One more footnote on Alberta, it provided yet another example of the continuing carnage being experienced by Canadian Liberals.  The Alberta party lost 16.5 % from one election to the next.  The focus was on PC/Wildrose but note also that in contrast to the Liberals, the NDP made a modest advance in a context where there was an incentive for voters on the left to vote PC to block Wildrose.