Statistics Canada’s November 2014 Labour Force Survey (LFS) release generated the typical morning headlines, like Canada lost 10,000 jobs in November (courtesy CBC News). The associated write-ups highlighted bank economists’ ‘cautious optimism’, as they typically have. While Statscan’s brief report twice mentioned significant self-employment gains, both month-to-month and year-over-year, most national news reports (like CBC News’) omitted any such mention. As the saying goes, you can only lead a horse to water.
The almost statistically insignificant change reported in the latest and last LFS release of the year nevertheless merits mention, if only for reflecting the year that was(n’t) in the Canadian labour market.
Employment class, type, wages
The November 2014 LFS estimated monthly employment changes fell well below the reported standard errors for the estimates (which aren’t the actual standard errors for the estimates, but that’s another topic). Table 2 of the release indicates the estimated number of employees fell 23,000 as the number of self-employed rose 12,300 over the preceding month; the standard errors for these estimated changes were 35,700 and 25,500, respectively. That is, the practically insignificant changes were relatively meaningless, hence the asterisk in the title.
That said, Table 2 also indicates self-employment accounted for 27 percent of total LFS-reported employment gains year-over-year. That’s up from the 20 percent share of year-over-year gains reported in the November 2013 LFS release. For context, the November 2013-2014 figures represent a year-over-year estimated rise in self-employment of 1.4%, double the year-over-year rise in actual (payroll) employment of just 0.7%.
The silver lining bank economists supposedly saw in the report was that the monthly job losses were mostly in part-time employment. As Table 1 of the November 2014 LFS release indicates, part-time employment fell an estimated 16,300 as full-time employment rose 5,700. (Here again the estimated changes were well below the standard errors for the estimates, 36,100 and 39,200, respectively.)
In a month when larger losses in actual (payroll) employment were tempered by moderate gains in self-employment, is an estimated drop in part-time employment a positive sign?
Also, since the LFS doesn’t collect self-employment earnings, in a month when relatively fewer survey respondents reported wage earnings than a year earlier, is an estimated improvement in average wages cause for optimism, let alone a meaningful measure?
The intuitive answer to both questions is ‘no’ – which is at odds with what has been and continues to be reported by most Canadian news outlets.