
Here’s where Statistics Canada found 42,000 new jobs
Good news, Canada: The country actually added 42,000 jobs last month, which is 41,800 more than we thought. Statistics Canada released its revised July job numbers today, which suggest the economy is on a much firmer footing than originally believed.
The agency also offered up a few more details about the “human error” that led to the foul-up. (Details are still scant, and the agency has promised to publicly release a more detailed internal review of the errors in the next few weeks.)
As I wrote about in this post yesterday, StatsCan periodically updates both its surveys and the data it uses to transform its sample population into data that represent the entire country. I mentioned that the agency was undergoing what it calls a “rebasing,” in which it updates its base numbers to reflect the most recent census. It was set to upgrade to the 2011 census next year. (Since 2011, the survey has been based on the 2006 census.)
Today, Statistics Canada said the errors were related to a broader redesign of the Labour Force Survey (LFS) that happens every 10 years. As part of that update, StatsCan says it was undergoing a scheduled upgrade to its processing systems in July, and that one program was left out of the update. That meant a number of people (rougly 60,000, judging by the change in its data) were counted as not being in the labour force, when, in fact, they were all employed. This is one reason that today’s update didn’t change the seven per cent national unemployment rate, even though we added more than 40,000 new jobs. This bigger redesign does indeed involve updating the Labour Force Survey to reflect the new 2011 census, but there’s apparently a lot more to it than that.
Here are a few more details about the survey redesign, including the upgrades that were due to take place this year. This comes from StatsCan’s 2014-2015 Report on Plans and Priorities it submitted to the Treasury Board earlier this year:
The LFS redesign will be conducted in two phases. First, the sample design and allocation will be modified to reflect changes in population and employment conditions. The estimates will also be rebased using the 2011 Census population estimates. This first stage of the redesign will be completed in 2014-15, and will include introduction of an Internet response option for survey participants. The second phase will focus on updating the processing systems of the LFS by adopting common corporate business processes and systems. This phase of the redesign will be completed in 2016-17.
Specifically,
2014-15:
* Revise the sample allocation and some methodological aspects for the redesign of 2014 (e.g., smart targets).
* Rebase the population estimates based on the 2011 Census, and reproduce the various products for stakeholders (time series, public-use microdata files).
* Offer an Internet response option to respondents to the LFS, following an initial contact with them.
The move to allow survey participants to answer questions over the Internet, rather than in person or by phone, seems to be a big part of the changes going on. StatsCan writes that it hopes the shift to an Internet-based survey will help to improve response rates.
As for the new numbers, the biggest changes seem to be concentrated in a few areas: Ontario added roughly 25,000 new jobs, compared to last week’s survey, while Quebec added 15,000. Public sector jobs rose by 21,000 and education jobs jumped by 14,000.
Statistics Canada has removed all the old, incorrect data from its website, which makes it hard to spot all the changes. But here is a summary of some of them, which I’ve culled from various news reports that came out in the wake of last week’s erroneous jobs report:
National
Labour Force:
Old: fell by 35,500
New: rose by 24,700
Change: an increase of 60,100 people working or looking for work
Full-time employment:
Old: fell by 59,700
New: fell by 18,100
Change: an increase of 41,600 jobs
Part-time employment:
Old: rose by 60,000
New: rose by 60,000
Change: none
Self-employment:
Old: fell by 29,000
New: fell by 37,000
Change: a decrease of 8,000 jobs
Private sector employment:
Old: rose by 26,300
New: rose by 54,600
Change: an increase of 28,300 jobs
Public sector employment:
Old: rose by 3,200
New: rose by 24,200
Change: an increase of 21,000 jobs
Industry
Construction:
Old: fell by 42,200
New: fell by 38,800
Change: an increase of 3,400 jobs
Education:
Old: rose by 32,100
New: rose by 46,100
Change: an increase of 14,000 jobs
Manufacturing:
Old: rose by 11,500
New: rose by 12,500
Change: an increase of 1,000 jobs
Service sector:
Old: rose by 34,300
New: rose by 68,000
Change: an increase of 33,700 jobs
Health care and social assistance:
Old: fell by 28,500
New: fell by 25,800
Change: an increase of 2,700 jobs
Business and building support services:
Old: fell by 10,600
New: fell by 14,000
Change: a decrease of 3,400 jobs
Transportation and warehousing:
Old: rose by 12,000
New: rose by 4,700
Change: a decrease of 7,300 jobs
Accommodation and food service
Old: rose by 14,100
New: rose by 16,900
Change: an increase of 2,800 jobs
Information, culture and recreation:
Old: rose by 15,100
New: rose by 17,100
Change: an increase of 2,000 jobs
By province:
Newfoundland and Labrador:
Old: rose by 6,800
New: rose by 4,700
Change: a decrease of 2,100 jobs
Ontario:
Old: rose by 15,100
New: rose by 40,000
Change: an increase of 24,900 jobs
Quebec:
Old: fell by 13,400
New: rose by 1,900
Change: an increase of 15,300 jobs
New Brunswick:
Old: fell by 3,100
New: fell by 3,400
Change: a decrease of 300 jobs
Nova Scotia:
Old: fell by 3,400
New: fell by 3,900
Change: a decrease of 500 jobs