•  About Me 
  •  FAQs 
  •  Mail Me 
  •  Links 
  •  Archives 
January 4, 2013

Stagnant BC Jobs Subject to Revision

Although Statistics Canada's job numbers for BC in December showed a small gain of 600, statistically this was equal to zero; this is better than the job losses experienced in October and November. The government's news bureau spun those data with a news release headlined: "B.C. economy remains strong and steady at close of 2012". No economist would call 0.8% job growth strong (BC's December 2012/December 2011 seasonally adjusted figure), but stagnant performance does qualify as steady.

The government news release claimed: "With 19,400 job gains since December 2011, B.C. ranks fourth compared to other provinces, behind Quebec (+138,000 jobs), Ontario (+100,300 jobs) and Alberta (+31,800 jobs)." One of the problems with the government's figures is it determines rank using absolute numbers rather than percentage changes, thereby giving BC an edge over provinces with smaller populations. BC ranks eighth with only Nova Scotia and New Brunswick doing worse when December 2012 is compared to December 2011 in terms of percentage job growth by province. The government's claim of fourth takes little PEI with a population of only 120,600 and says it's December to December job growth of 1.4% is not as good as BC's 0.8% because applied to PEI's smaller base it yielded a gain of only 1,000 jobs. What nonsense! No wonder the Liberals have issues with respect to trust and credibility.

The government's news release also said "Since August 2011, B.C. has added 42,400 jobs". That's a slight improvement over previous attempts to attribute job changes since August 2011 to the BC Jobs Plan but it begs the question of why anyone would refer to that strange 16 month period. Earlier government news releases claimed the Liberals' jobs plan started in August 2011 even though it wasn't announced until September 22, 2011. By using August 2011 as a starting point, the government is able to capture a 27,900 jump in jobs in September 2011 that some economists believe was largely a measurement error. If September 2011 is used as a base (15 months still being an unorthodox reference), the job gain drops from 42,400 to just 14,500. If that was all there was to it, most people might not know who to believe but Statistics Canada has a surprise in store next month.

As part of its release of the December employment numbers, Statistics Canada included the following note:

"Seasonally adjusted estimates from the LFS will be revised using the latest seasonal factors, going back three years (January 2010 onwards). The revised estimates will be available on CANSIM (tables 282-0087 to 282-0094, 282-0100, 282-0116 and 282-0117) on February 1, 2013."
The reason Statistics Canada periodically revises its seasonally adjusted data is that the adjustment process uses a moving average over dates for which samples have not yet been made. If that weren't done we would not get timely data, but a consequence of the technique is the necessity to revise the adjusted data from time to time. In a few weeks we'll see whether the unusual bump in September 2011 remains in the data. The seasonally unadjusted job numbers for September 2011 showed an increase of 7,300 rather than the seasonally adjusted 27,900. Upon revision the mysterious difference of 20,600 might disappear as easily as it was created.

It will be fascinating to see if the Liberals revise their job claims after the February 1st seasonally adjusted Labour Force Survey data are available. With strained credibility with previous claims, don't be surprised if the Liberals use outdated numbers as well as bizarre reference periods.