November 28, 2000 (The Editor’s Desk is updated each business day.)

Comparing hourly earnings measures

Earnings estimates produced using data from the employer costs for employee compensation program correspond fairly closely to the published average hourly earnings series produced by the Current Employment Statistics (CES) program, particularly among goods-producing workers.

Actual average hourly earnings and employer costs for employee compensation replicate estimates, March 1988- September 1999
[Chart data—TXT]

In a recent BLS analysis, long-term comparisons for the period 1988-99 were made between the actual CES average hourly earnings series and "replicate estimates" constructed with data from the employer costs for employee compensation program. For production workers in goods production (mining, manufacturing, and construction), the replicate estimate was on average 1.5 percent lower than the actual average hourly earnings for this group of workers—about $0.20 per hour lower.

For nonsupervisory workers in the rest of the private nonfarm economy, in contrast, the replicate estimate was on average 5.8 percent higher than the corresponding hourly earnings series—about $0.60 per hour higher.

This analysis uses data from the Current Employment Statistics and Employment Cost Trends programs. Additional information is available from "Replicate estimates of the average hourly earnings series," by Anthony J. Barkume and Michael K. Lettau, Monthly Labor Review, October 2000.

 

Happy 10th Birthday, TED!

The very first issue of The Editor's Desk (TED) was posted on September 28, 1998. TED was the first online-only publication of the Bureau of Labor Statistics. For 10 years, BLS has been committed to posting a new TED article each business day, for a total of over 2,400 articles so far.

Find out more about the story of TED