Latest national poll median date: October 20
Projections reflect recent polling graciously made publicly available by pollsters and media organizations. I am not a pollster, and derive no income from this blog.

Friday, August 27, 2010

Harris-Decima: Tories Lead by 3

The Canadian Press reports on the latest Harris-Decima poll, showing a narrowing gap between the Tories and Liberals. In Atlantic Canada, this poll agrees with the latest Ipsos, Angus Reid and Léger that the Grits have a double-digit lead; now only EKOS shows a tight race there. The Liberals must also be encouraged by their 28% number in Québec. However, Ontario is still a dead heat. The NDP got another really low number in Alberta (6%), but their 30% in BC might give the Dippers hope that they've recovered from weak August polling there.

Minor changes in the projection give:

CON - 130
LIB - 91
BQ - 50
NDP - 37

The average Conservative national lead is 4.9%. That Léger poll increasingly looks like an outlier. Excluding it would make the gap roughly 4%.

Tuesday, August 24, 2010

Léger: Tories by 9 Nationally (3 weeks ago); Bloc by 13 in QC (last week)

Léger Marketing has posted not one, but two new polls this week. The older one is a national survey from early August, and it shows a much larger Conservative lead (9 points) than contemporaneous polls (3-6 points). This can almost entirely be explained by the 7-point Tory lead in Ontario (compared to 1-point in other polls at that time; the latest EKOS shows a 3-point Liberal lead there) and by a whopping 48% for the Conservatives in BC (against 19% only for both the Grits and the Dippers).

The newer poll is a provincial one, and shows run of the mill numbers for Québec. Only the Tory number is slightly higher than usual - Québec pollsters often produce such results, in anticipation of the common federalist ballot box bonus.

These polls do not cost the Liberals any seats in the projections, though it pushes them farther away from bigger gains in Ontario. The Tories do manage an extra seat at the expense of the NDP, which is now at its lowest level since mid-March.

CON - 131
LIB - 90
BQ - 51
NDP - 36

The average Conservative national lead is 5.5%, but expect this number to go back down as this poll may well be an outlier.

Thursday, August 19, 2010


As you may have noticed, I've created an "Economics" label so that you can find my posts on that topic more easily. I've also renamed the vague "Commentary" to "Opinion," and somewhat broadened the scope of that category. Please let me know if there are other tags that you would find helpful!

EKOS: Tories by 4.6

The newest EKOS poll suggests that things may have stabilized, with very similar national numbers for both weeks for the Grits and the Tories (the NDP did improve quite a bit on the Prairies and in the Atlantic provinces). The most intriguing development in this poll is the 3.3% Liberal lead in Ontario during the second week, a change from a roughly 1% Conservative lead during the first week and in previous polls. Are the Grits finally getting traction in that crucial battleground?

Like in other recent polls, the Grits managed to score a respectable 25% in Québec, so their recovery from the spring doldrums there seems real. On the flip side, after several strong polls in MB/SK, the Grits fall back into the teens in this one, to the profit of the NDP. However, the Dippers did dismally in the other Prairie province, which (barely) moved my Alberta projection to a Tory sweep. In BC, this poll confirms the recently observed widening of the Conservative lead over the NDP; are British Columbians already calming down from their HST anger?

The Liberal seat count continues its summer gradual upward trend:

CON - 130
LIB - 90
BQ - 51
NDP - 37

The average Conservative national lead did grow slightly, however, to 4.7%, mostly on the much reduced weight of the 7/28-8/3 EKOS polling that had the Tories up by just 1.2%.

Tuesday, August 17, 2010

How Rich Is Canada Relative to the U.S.?

According to Statistics Canada, Canada's real GDP per capita is catching up to America's: it went from 82% in 1999 to 92% in 2008. However, according to the IMF, during the same period, it went from 81% to 82% (other international organizations, like the OECD and the World Bank, show similar numbers). What's going on?

The reason why these estimates have come to differ so greatly is that Statistics Canada estimates that the purchasing power of each Canadian dollar was 90 U.S. cents in 2008, while the IMF reckoned that it was only 81 cents. In contrast, in 1999, both agencies pegged the CAD's real value at 0.84 USD. Of course, this begs the questions: why did the difference arise, and who's right?

Statistics Canada gives two reasons for the diverging figures, both of which suggest that its numbers better reflect the relative standing of the Canadian economy:

1. The International Comparison Program (ICP) is the main reference for establishing the true purchasing power of various currencies. It does so by collecting prices for a variety of goods in each participating country. The last round of the ICP was carried out in 2005, and determined that the CAD's purchasing power was about 0.82-0.83 USD. However, Statistics Canada used extra data and recomputed the figure by tweaking the basket of goods being compared, so that it better reflects North American consumption and investment patterns. The adjusted value was 0.87 USD.

2. After the 2005 benchmark, these statistics are adjusted every year to reflect differences in inflation between countries. For example, if inflation is higher in Canada than in the US, then the relative purchasing power of the CAD would decrease. Question: inflation of what? If Canada and the US did not trade at all, the answer would be easy: the inflation of all goods and services in each economy. But due to international trade, the G&S produced in a country do not correspond to the G&S consumed in that country. What basket of goods and services should be used depends on the purpose of the calculations:

- If one is trying to figure out economic (i.e. real GDP) growth for a given country, one should use production prices: the goal here is to compute the change in the volume of G&S produced. The relative prices of imports and exports are held constant.

- If one is instead interested in international comparisons of material quality of life, the consumption prices are the relevant ones. The relative prices of imports and exports follow changes in the market.

In Canada's case, this makes a big difference: as energy prices have risen considerably, a given volume of Canadian exports is now worth more imports than it used to. This has not been the case for the US. Thus, since 2005, while Canada's economic growth per person has roughly matched America's, Canadians' material quality of life has been increasingly more rapidly than Americans'. This is the second reason why international organizations, using the former measure, do not show Canada catching up, while Statistics Canada does.

(By the way, the second measure above - the trade-adjusted value of GDP - is sometimes termed Gross Domestic Income (GDI). Every year, GDP=GDI at that year's prices, by definition: your income is the value of what you produce. But because the previous year's GDP and GDI are not generally the same at current prices, real GDP and real GDI growth rates differ. So point 2 is simply saying that the international organizations' estimates reflect Canada's real GDP growth rather than its real GDI growth.)

To make a long story short: there is good reason to believe Statistics Canada's findings that Canadian economic standards are catching up to American ones (although I'm bit skeptical about the point 1 adjustment being so large)! This is consistent with the mood in the two countries, even during the 2003-2007 expansion: while Canada and the U.S. posted similar growth rates (and had similar population growth rates as well), Canadians were much more upbeat than Americans about the economy. After the next round of the IPC in 2011, international organizations will recognize part of the catch-up (the portion related to point 2), as the exercise will update the relative prices of all goods.

Interesting side note: although Statistics Canada puts the CAD's purchasing power at 0.90 USD, it is lower for private consumption goods and services (what you and I actually "feel") - only 0.84. The numbers vary wildly across categories of goods: food, alcohol and tobacco are a lot more expensive in Canada, while health care (even the private portion, like eye care and dentistry) and education are a lot cheaper. However, for government purchases and capital investments, the values are 1.01 and 0.98. The former reflects cheaper health care in Canada, while the latter may be partially due to Canada mainly using value-added taxes instead of sales taxes.

Thursday, August 12, 2010

Angus Reid: Conservatives Lead by 4

A new Angus Reid poll is now available. It confirms the tightening of the race between the Grits and Tories, and provides reassurance to the NDP after two bad polls earlier this week.

This poll has something for everyone. The Tories can be happy to still lead by 3% in all-important Ontario (and by 48% in Alberta). The Liberals can congratulate themselves over a whopping 47% lead in Atlantic Canada (with 65% of voting intentions!), and yet another strong result in MB/SK (32%, just 10 behind the Tories). The NDP can get excited about their 18% in Québec, and relax a little about MB/SK and BC, where they get OK numbers (24% and 27%).

The new aggregate projection is:

CON - 130
LIB - 88
BQ - 52
NDP - 38

The Tory average national lead is now at 4.4%. I'm already curious about what next week's EKOS will tell us - but that's a whole week from now! Will Nanos or Environics come save us?

Another statistic of note in the recent poll: Harper has fallen below Ignatieff in terms of momentum, and how! In previous Angus Reid polls, Ignatieff tended to have, by far, the worst momentum score. However, over the past month, Harper was at -26 (just 6% have a better opinion of him, 32% worse), while Ignatieff was at -12 (10-22). That BBQ tour has probably not generated much positive feeling for Ignatieff, but it can be argued that it has mitigated some negative ones. The census flap, however, has probably damaged Harper, though it remains to be seen whether the effect will last.

Wednesday, August 11, 2010

Republicans, Democrats and the Past 40 Years of the US Economy

Figures below refer to US median household income, in 2008 dollars, and come from the US Census Bureau.

How a Republican would see it:
1978: $45,625 one year after Carter takes office
1983: $42,910 (-6% in 5 years) two years after Carter leaves office
1989: $48,463 (+13% in 6 years) the year Reagan leaves office, and just before Bush Sr. betrays Reagan by raising taxes
2004: $50,535 (+4% in 15 years) three years after Clinton leaves office
2007: $52,163 (+3% in 3 years) just before the financial crisis caused by government intervention in the mortgage market

How a Democrat would see it:
1969: $43,557 the year Nixon takes office
1976: $43,649 (0% in 7 years) the last year of the Ford presidency
1979: $45,498 (+4% in 3 years) the 3rd year of the Carter presidency, just before the Fed got serious about inflation
1993: $45,839 (+1% in 14 years) the year Clinton takes office
2000: $52,500 (+15% in 7 years) the last year of the Clinton presidency
2008: $50,303 (-4% in 8 years) the last year of the Bush presidency

The truth, of course, is somewhere in the middle. Nixon/Ford were unlucky to face the Oil Crisis; the Fed's inflation fighting destroyed Carter's record and gave Reagan an artificially low base to start from; Bush Sr. had to raise taxes due to Reagan's deficits; Clinton probably benefited from Reagan deregulation; Bush Jr. cannot be held solely, or perhaps even mainly, responsible for the financial crisis, which had roots in the Reagan reforms and in the Fed's actions mitigating the bursting of the Clinton tech bubble.

Given this history, you can be sure that 25 years from now, people will still be arguing whether the sluggishness of the current recovery is Bush Jr.'s or Obama's fault.

Tuesday, August 10, 2010

Ipsos Reid: Tories Lead by 3; Harris-Decima: Tories Lead by 6 has reported on the latest Ipsos Reid poll, while Harris-Decima has its own new release. The two polls are not surprising and indeed virtually identical in Québec and Ontario: the Conservatives still lead by 1 in Ontario, while the Liberals seem to be on a slight up trend in Québec. Considering small sample sizes, Alberta and Atlantic results are also unsurprising. In Manitoba/Saskatchewan and BC however, the Liberals show unexpected strength in the Ipsos Reid poll (just 5 behind the Tories in both cases, and 13 ahead of the NDP in BC), while nothing of the sort appears in the Harris-Decima survey.

Both polls are very bad for the NDP: their regional numbers go from weak in Ontario (16%) and Québec (12%) to very weak in the Atlantic (20-21%) and Alberta (8-9%) to disastrous in MB/SK (14%) and BC (20%) - in the latter, they trail the Tories by 17-18%.

The Bloc and the NDP weaken to the profit of the Liberals:

CON - 131
LIB - 87
BQ - 52
NDP - 38

Still, the smaller parties are in a strong position. The Liberals are approaching respectability territory, though their inability to get ahead in Ontario means they're not competitive yet with the Tories.

The Tory average national lead is down to 4.6%.

Friday, August 6, 2010

Ontario: where Ignatieff Needs to Make Progress

I have done so in the past, but let me reiterate how important Ontario is for the Liberals to have any chance of defeating the Tories.

I applied a projection based solely on the most recent week of the latest EKOS poll, where the Tories only led the Grits by 1.2%, but had roughly the same lead (0.8%) in Ontario. The seat result? Not even close:

CON - 124
LIB - 95
BQ - 50
NDP - 39

This is the same seat count for the Tories as in the 2006 election, while the Liberals do 8 seats worse even though they lost by 6% that time. Why? In 2006, the Grits still carried Ontario by almost 5%.

I also made a projection adding 3% to the Liberals and subtracting 3% from the Tories in every region of the country. This gives the Grits a 4.8% national lead, but a dead heat on the projected seat count:

CON - 109
LIB - 109
BQ - 49
NDP - 41

Do the Liberals now need to win by that much just to get a seat plurality? My guess is 'not quite so': after all, the "Ontario gap," defined as the difference between the Tory lead nationally and the Tory lead in Ontario, is abnormally low in the latest EKOS poll. That figure stood at 0.4% in the survey, while in my current poll average, it is 3.5%. But even with an Ontario gap of 3.5%, the Liberals would need to win the national popular vote by 3-4% in order to tie the seat count.

Now, it is quite possible that due to random error, polls currently understate Liberal gains in Ontario and overstate them elsewhere. If that's the case, and the Ontario gap is roughly at the level of the last election (6%), then the required national popular vote lead for the Grits is even lower. But I'm pretty sure that unless a drastic relative regional change occurs, the Tories would win the seat count in the event of a tied popular vote at the next election.

Can Ignatieff deliver his home province? His job depends on it. (Although by returning to academia, he would collect a larger paycheck while barely having to do any work due to tenure...)

Thursday, August 5, 2010

EKOS: Tory Lead at 4.8 in Past Two Weeks, Just 1.2 in Most Recent Week

The most recent EKOS poll shows some movement toward the Liberals in the past week. Specifically, Liberal fortunes appear to have improved in Atlantic Canada and Québec in the most recent week. However, the Grits are still shown trailing the Tories by 1% in Ontario, which is a major impediment to substantial improvement in the projected seat count. Indeed, the Tories lose only one seat in the aggregate projection, although this did bring the Grits to their highest level in almost 4 months:

CON - 131
LIB - 84
BQ - 53
NDP - 40

The average Conservative national lead, however, dropped one full point to 5.1%.

Note: Once EKOS posts its own report on the poll, I may update the above figures: the CBC report only shows the number of decided voters, and not the total number of voters interviewed. I weigh polls based on the latter because many pollsters report only that. As a result, the current numbers currently slightly underweigh the most recent poll.

Update: Weights have been fixed, but that caused no change in the projection.

Tuesday, August 3, 2010

Conservatives and Statistics

A lot of people think that the Conservatives are getting rid of the mandatory census long form to deliberately produce unreliable data that cannot be used to attack their policies. I don't think the Tories are that devious - they are probably just incompetent with data, and therefore don't appreciate the value of accuracy.

For example, in the Stockwell Day crime flap today, we got several indications that the Tory cabinet sucks with data - they fail to use it correctly even when it supports their case:

- Day did not bother producing any data to back up his assertion that crime is actually increasing in Canada (despite a drop in reported crimes) because more crimes are going unreported to police.

- Later, his colleague Rob Nicholson came to the rescue with a complete non sequitur, stating that 34% of crime in Canada go unreported, without mentioning whether the proportion has gone up.

- Upon examining the actual 2004 Statistics Canada report, one realizes that 34% is actually the proportion of crime that is reported, and in fact 64% of crime is unreported (with 2% unknown). [By the way, the AP seems to have done the verification, while the CBC just reported Nicholson's statement. (Hopefully this will have been fixed by the time you click on the link.) Journalistic rigour, CBC?]

- That 34% reported crime figure is down from 37% in the previous report for 1999, which is itself down from 42% in 1993.

So the data suggests that indeed, an increasing proportion of crime is going unreported. But the Tory ministers either didn't bother with it, or stated the level of crime reporting both erroneously AND without referring to the time trend, which is required to make their case. You really wonder how these people made it through university.

By the way, the next report, covering 2009, is due next month, so we're debating these figures exactly at the time where we have the most outdated data. Are the Tories afraid that the rate of crime reporting has stopped decreasing? Again, my guess is that they were instead simply guided by anecdotal evidence, electoral interest and ideology.

Besides, in the 1999 report, StatCan suggests that part of the decrease in crime reporting may be driven by lower damage in property crimes (smaller proportion of $1,000+ cases and bigger proportion below $100) and higher insurance deductibles: many people won't bother reporting a theft if they wouldn't get reimbursed anyway. In fact, reporting of violent crime has increased from 31% to 33% between 1999 and 2004.

So even if one believes that crime incidence has increased, it's unclear that we should be worried about it: serious crime appears to be decreasing, so it may well be that the overall social harm from crime is also decreasing. And even if one agrees that crime is harming Canadians more than before, it's another logical leap to conclude that we should build more prisons. After all, Québec, with its relatively light approach to punishment, has much less violent crime than the rest of the country. (Of course, you could argue that the causation runs the other way, but the point is that it's not at all clear that more punishment leads to less crime. What is clear is that prisons cost money.)

But, even given supportive data, Conservative ministers are unable to craft a coherent argument for the simple proposition that crime incidence has increased. Is it any surprise that these people don't think much of weakening the census?