Chief Economic Analyst (613-951-9162)

The historical data on output and employment come from P. Cross, “Alternative measures of business cycles in Canada, 1947-1992.” Canadian Economic Observer, Catalogue 11-010-XPB, February 1996.

The best known of the papers on labour hoarding is Walter Oi, “Labor as a Quasi-Fixed Factor.” Journal of Political Economy, Vol. 70, No. 6, Dec. 1962.

Firms also are slow to cut production early in a downturn, which is why inventories continue to accumulate early in most recessions.

The small differences that occur between labour productivity and output-per-employee usually happen at cyclical turning points, when changes in output-per-employee are dampened by the non-business sector, where by definition output cannot vary much from employment.

The time span shown in the tables does not necessarily accord with the peaks and troughs for the overall economy. Partly this reflects that the tables only show the business sector. As well, the data for each downturn are presented as long as any one of the 10 variables in the table are declining, so readers can fully understand the relationship between all the variables.

Overall, lower employment contributed more to the drop in labour input than a shorter workweek while the reverse was true of the economy -wide measures.

P. Cross, “Recent Trends in Output and Employment.” Canadian Economic Observer, Catalogue No. 11-010, March 2007.

While average hourly earnings in public administration are above the provincial average, the provincial sector was below average.

This analysis also sheds some light on the ongoing slowdown in output per employee in Canada after 2002. As discussed in our 2007 paper on output and employment, the gap between the two has narrowed since 2002. But the experience during the 2008-2009 recession suggests some of this slowdown in output per employee reflects a greater reliance on changes in the workweek to meet labour needs. During the peak of the resource boom, in 2006-2007, employers relied on a longer workweek to increase labour inputs, especially in areas like Alberta, where employees were in short supply (the workweek in Alberta rose nearly a full hour between 2002 and 2008, the most in Canada). Similarly, during the downturn in 2008-2009, employers relied more on cutting the workweek than on reducing employment when lowering labour inputs. This helps explain why employment was less affected than hours worked in Canada than was the case in the US, despite similar drops in real GDP in both countries. Source: The standard workweek in the Survey of Payroll employment, earnings and hours (CANSIM Table 281-0038).