CLARKWATCH: Follow news and updates regarding sanctions on Mayor Clark.

U.S. payroll firm challenges Statistics Canada with alternative employment data

Nov 16, 2017 | 11:30 AM

TORONTO — An inaugural private-sector Canadian jobs report calculates that the economy actually shed jobs last month, contrary to the sizable gain reported two weeks ago by Statistics Canada. 

The Canada National Employment Report from ADP — a U.S.-based payroll processor known for releasing jobs data ahead of the American government — estimates that 5,700 payroll jobs were lost in October.

That’s a big difference from the 35,300 jobs that Statistics Canada suggests were added in October, according to the Labour Force Survey of about 56,000 households released Nov. 3 — one of the major sources of data used by economists.

Ahu Yildirmaz, who co-heads ADP’s research arm, says it’s publishing the information as a public service and doesn’t charge a fee for the monthly report.

“Our goal is to provide insight, not only to our clients and prospects but also to government, to policy-makers, to business decision-makers,” Yildirmaz said in an interview.

She added that ADP’s reports are based on data collected from actual payrolls at 40,000 companies with 2.2 million employees in Canada, which provides a much bigger sample than the government’s surveys.

A statement from Statistics Canada said ADP’s report uses “completely different concepts and methodology” than the government’s employment data.

“Their occupational breakdown is based on various sources of data such as those derived from their clients but also from Statistics Canada’s Labour Force Survey and the Survey of Employment, Payroll and Hours.

“While the ADP report may meet the needs of some users, there are fundamental differences between the ADP report and the labour and employment data regularly produced by Statistics Canada in terms of concepts, level of geography and industry.”

Two of Canada’s most prominent private-sector economists — Doug Porter of BMO Financial Group and Avery Shenfeld of CIBC — say the additional data is welcome but ADP is unlikely to replicate the prominence it has in the United States.

The company’s U.S. payroll reports are a “very important addition the mix” but they come out about two days before the U.S. Labour Department’s flagship jobs reports, Porter said in an interview.

“I’m just a little bit concerned that this one (for Canada) might not get all that much attention over time because it comes out after the Labour Force Survey. In this case, almost two weeks later,” Porter said.

In addition, he noted that it may be unhelpful to have ADP’s information conflict with StatCan’s — as it appears to do in the case of ADP’s October numbers released Thursday.

“Which one would I believe more? I’m not sure,” Porter said

But Porter said that ADP’s reports may be useful for confirming or challenging big swings in government jobs estimates.

Shenfeld said in an email that StatCan’s flagship report — which is based on surveys of households rather on data collected from employers —  isn’t “that accurate as a measure of what’s happened in a given month.”

“Maybe the best thing about this new ADP data is that it will cause markets and economists to give StatCan’s payrolls data the respect it deserves, since it’s likely a more accurate reading on job creation in the short term,” Shenfeld said. 

Porter said the U.S. Department of Labor issues results from two surveys — of households and employers — at the same time each month, but Statistics Canada’s Survey of Employment, Payrolls and Hours comes out six to seven weeks later.

“StatCan will often point to that (SEPH) survey as being a better way to look at industry-by-industry employment. But it’s almost ignored it is so far behind,” Porter said.

However, given the new ADP Canadian National Employment report, it’s interesting to note that its estimate of Canadian jobs added in August (38,327) is closer to StatCan’s August payroll report (38,900 added) than StatCan’s Labour Force Survey for the same month (22,000).

Porter said that, no matter what the monthly surveys say, he’s more interested in longer-term trends.

“It just drives home the point that you really have to treat the monthly numbers with a grain of salt.”

David Paddon, The Canadian Press