Alternately, May 2013 SEPH:The ‘unclassified’ jobs recovery
Chart 1: Net change in payroll employment by industry classification
Source: CANSIM Table 281-0025 – Employment (SEPH), seasonally adjusted, for all employees for selected industries classified using the North American Industry Classification System (NAICS), monthly (persons), Statistics Canada
This is an update to a couple of prior posts (here, and more recently, here) on the Canadian economy’s strong ‘unclassified’ job creation record since The Great Recession, at least according to Statistics Canada’s Survey of Employment Payroll and Hours (SEPH). Unlike the Labour Force Survey (LFS) which asks individual household members about their work and tries to guesstimate which industry they work in, SEPH collects the industry information directly from businesses. At least that’s how it’s supposed to work in theory.
Evidently, it’s not been working over the last five years. Since May 2008, nearly 200K of the 470K jobs Canadian businesses supposedly created, 42%, have no industrial classification. As the figures from Chart 1 illustrate, classified jobs have increased less than 2% as unclassified jobs doubled over the last 5 years.
And that’s after Statistics Canada started tinkering with the data following an inquiry by a federal MP who read about the SEPH issue on this site (see end note on the 28K job revision for September 2012).
As previously discussed, neither wages nor hours for unclassified jobs are included in Statscan’s SEPH tabulations, likely biasing the averages for both reported figures upwards. Also noted was the fact that future revisions to correct the non-classified jobs boom will make it impossible to discern which jobs were the result of real economic activity and which were simply reclassifications in a given month.
What had not previously been mentioned, but is worth noting given the newly minted Employment Minister’s oversimplified synopsis of the labour market situation: Without knowing which industries over the last five years created more than 2 in 5 Canadian jobs, let alone the occupation class, rate of pay nor hours associated with said jobs, the Minister is basically left to guess whether the 58% of jobs that were classified are representative of the job market, or not. In many ways, the failings of the SEPH mirror those of the National Household Survey (NHS).
In an effort to help the Employment Minister: While nothing is known about these unclassified jobs, their distribution by province is provided in the available (for the time being) data tables. Ontario and Quebec accounted for nearly 130K of the 200K unclassified jobs created over the last five years. Most disturbing is that Ontario appears to have created 85K unclassified jobs to only 8.5K classified jobs, a 10:1 unclassified-to-classified job ratio. We literally have no idea which industries have been driving job creation in Canada’s largest province since The Great Recession. British Columbia created 21.4K unclassified jobs, but only 6.6K classified jobs. Except for Newfoundland, each of the Maritime provinces lost classified jobs relative to unclassified jobs created. Collectively, PEI, Nova Scotia and New Brunswick created 9.8K unclassified jobs as they lost 21.8K classified jobs over the same five year period. According to the May 2013 SEPH release, only the provinces of Newfoundland, Manitoba and Alberta created more classified than unclassified jobs since May 2008.