Canada's economy has become an increasingly urban one, but its statistical analysis of that urbanization hasn't entirely kept pace. Now it looks like Statistics Canada is working on taking its granddaddy of economic measures, gross domestic product, to the cities.
The national statistical agency on Monday published what it called an "experimental" analysis of GDP by urban centre, breaking down the data across 33 census metropolitan areas (CMAs) from 2001 to 2009. Though Statscan isn't saying so, the study has the appearance of a test drive of a more detailed level of GDP analysis, which in its current form only digs as deep as the provincial level, on a quarterly basis.
The new paper, by Statscan economic analysts Mark Brown and Luke Rispoli, vividly illustrates why city-level GDP analysis is long overdue. It found that 72 per cent of Canada's economic activity occurs in the country's urban centres. More than half of Canada's GDP is produced in just six cities – Toronto, Montreal, Vancouver, Calgary, Edmonton and Ottawa-Gatineau. Toronto alone makes up an astonishing 19 per cent of Canada's economy.
"Toronto accounts for less than 1 per cent of Canada's land mass, but has an economy that is larger than that of every province except Ontario and Quebec," the paper said.
Despite Toronto's dominance, the shift of economic activity in Canada, from the East to the West was glaring at the city level. Statscan said that from 2001 to 2009, Calgary and Edmonton accounted for nearly as much GDP growth as Toronto – even though their combined population was less than half of Toronto's. Eight of the nine census metropolitan areas tracked by Statscan west of Manitoba increased their share of national GDP in the period.
Figures on a per capita basis (i.e. per person) also illustrate how urban GDP has shifted from manufacturing-driven centres (especially in Ontario) to resource-driven centres.
"Of the CMAs in the top 10 in terms of GDP per capita in 2001, Kitchener-Waterloo, Halifax, Windsor and Oshawa were no longer in the group by 2009, replaced by St. John's, Saskatoon, Victoria and Vancouver," the report said.
"Of the nine CMAs with 25 per cent or more of their output in manufacturing at the start of the period, six fell in rank, all of them in Ontario."
(Regina had the top GDP per capita in 2009, by the way, at $65,404, followed by Calgary ($61,246) and Edmonton ($59,941). Toronto was 7th in 2009, down from 3rd in 2001.)
Of course, there are problems with these data, not the least of which is their best-before date. Statscan's experimental analysis stops at 2009 – fully five years ago. The numbers tell us nothing at all about how Canada's urban economies have evolved in the post-crisis economic recovery, a period that has been marked with some pretty substantial upheaval. Economic numbers predating the recovery are, frankly, ancient history.
It's also problematic that the end point for the analysis is 2009 – a year that marked the depths of the Great Recession. Any numbers specific to 2009 would obviously reflect some abnormal economic conditions that would distort longer-term trends in the evolution of the country's urban economies. It was, to say the least, an awkward year on which to bookend the data.
Still, the "experiment" presents a positive direction for Statscan to take its GDP analysis, potentially marking a starting point for a new and valuable conversation on the nature of Canada's economy. The cities are, clearly, at the heart of the nation's economic activity. Canada's major urban cores account for three-quarters of the nation's jobs, and have generated more than 90 per cent of the new jobs created this year. Yet they also have a higher unemployment rate than the nation as a whole. Understanding what makes urban economies tick is critical to understanding Canada's economic, labour and demographic forces.
One of economists' biggest criticisms of Ottawa's cancellation of the mandatory long-form census is that it has reduced visibility of economic trends at the community level, where government policy and services can be most effectively directed. Increased granularity of the GDP data wouldn't replace everything that has been lost from the death of the long-form census (indeed, not even close), but it could certainly help fill some important gaps in understanding income, employment and productivity trends, and in a relatively timely manner.
Of course, as with the long-form census, there would need to be sufficient political will and budget for Statscan to take this from academic exercise to routine reality. But we've seen a glimpse of the possibilities and the importance of city-level GDP data in analyzing an increasingly urban national economy.
David Parkinson