What’s Happening to Wages?
One thing I find a bit annoying over at Statscan is that detailed tables related to a release sometimes appear on the web site a few days after headline numbers are released in the Daily.
I was phoned by a couple of reporters seeking comment following the release of wage data (from the Survey of Employment, Earnings and Hours) in the Daily on August 26. The detailed Tables were not published untilÂ September 2.
The headline number from theÂ Daily was that average weekly wages had jumped 4.0% in the year June, 2009 to June, 2010, the largest increase since February, 2008. Parts of the media seem to have seen this as evidence of wages getting out of line, the end of recession etc. etc.
I pointed out that it could partly be an increased hours story as opposed to an hourly wage story, and that the Labour Force survey showed Â an increase of hourly wages of just 1,7% for the same period. (The Daily includes a Box also warning readers that several factors can drive average weekly earnings.)
It turns out when I looked at the Tables posted on September 2 that fixed weight average hourly earnings of all workers (hourly and salaried) rose by less -than 4% Â – in fact by 3.7% – over the year -showing that at least part of the 4% gain was indeed driven by increased hours as opposed to increased hourly wages, and also by sectoral shifts in employment.
What is especially interesting (now that we can see the Tables) is that average weekly earnings of salaried employees rose by 5.5% over the year. This seems to be partly a time worked story, since the standard work week rose from 36.8 to 37.Â 1 hours. Meanwhile, average hourly earnings of hourly paid workers increased by only 2.7% – well below that 4% headline number.
That still leaves a significant discrepancy between the LFS and the SEPH data. There are lots of goods reasons why there will be differences in any given month which will disappear when averaged over time.
MyÂ take is that we are not seeing aÂ significant increase in real hourly wages, and that Statscan might be a little more careful in how they headline the Daily release.