Statistics Canada will re-publish its flagship labor market report in two days after finding an error in the most recent edition, a move economists said may hurt its reputation.
The agency said yesterday it will re-release July’s employment data on Aug. 15 at 8:30 a.m. in Ottawa. The original report published Aug. 8 said Canada added 200 jobs during the month, compared with a consensus forecast of 20,000, and that 35,400 people left the labor force.
“It’s very unusual for them to restate and not just revise at the next release,” Blake Jespersen, managing director of foreign exchange in Toronto at the Bank of Montreal, said by e-mail. “The market is taking this as a sign that the number was understated.”
The monthly labor report is among the most watched economic indicators, often moving the currency, because it’s seen by traders as a barometer of the economy’s health and can signal changes in the level of interest rates. News of the botched release adds uncertainty to today’s C$3.4 billion ($3.1 billion) sale of federal government two-year bonds.
Canada’s dollar erased losses and bond yields rose after Statistics Canada’s notice was distributed at around 2 p.m. New York time, and the old report was removed from the agency’s website. Sylvie Michaud, director general at the agency’s education, labor and income branch, said by telephone today the human error was related to the start of a survey overhaul that is scheduled to be finished around January. The change “isn’t related” to new methodology used to complete the last census, she said.
The labor force survey “is the most important indicator followed by financial markets,” said Stefane Marion, chief economist and strategist at National Bank in Montreal, in an e-mail. “This tarnishes the image of Statistics Canada, which has been going out of its way recently to convince us about the reliability of the LFS survey. Is it a big mistake or a small mistake? Why do we need to wait three days to find out?”
“If it ends up that there was an error but it still leaves the message of a soft labor market, it will be inconsequential and soon forgotten,” he said. “If it involves a major revision to the last month, it will certainly be at least a temporary embarrassment.”
Statistics Canada said it “takes this matter very seriously and is immediately launching a review of the data verification processes in place,” according to its statement yesterday. “This does not affect other statistical programs.”
Derek Holt, Scotiabank’s vice-president of economics in Toronto, wondered why the statistics agency is waiting to release the revised figures, given the government bond sale today. “I don’t know why they’d wait until Friday,” Holt said in a research note.
David Tulk, chief Canada macro strategist at Toronto-Dominion Bank’s TD Securities unit, said he can’t remember a similar incident.
“It’s frustrating insofar as it introduces more volatility and the reputation of the survey inevitably takes a hit,” he said. “To StatsCan’s credit, they seemed to get out in front of it as soon as they could.”
To contact the editors responsible for this story: Paul Badertscher at firstname.lastname@example.org Chris Fournier