Much has been written about the jobs puzzle - the slower than usual recovery of employment as the economy has pushed out of the recession. In today's New York Times, Eduardo Porter's piece is titled "Estimates of Job Creation Versus the Facts." Mr. Porter states that economists will be watching today's jobs report to see if "the numbers that they were forecasting are as embarrassingly off target as they have been in recent months."
All indicators point to a strong economy and strong employment growth, but employment estimates from the BLS, the "facts" that Mr. Porter takes as given, have stubbornly refused to fall in line with past experience.
Ever hopeful, many economists are still expecting a burst of employment. The consensus forecast this week puts employment growth at 120,000 jobs in March. Goldman Sachs is betting on 160,000. And High Frequency Economics is forecasting 200,000, saying good weather and the end of a supermarket strike in California should give the jobs data a big boost.
Some "whisper numbers," which economists do not write down but which make their way into trading rooms along Wall Street, are even higher.
Here is an interesting conjecture, that if true, might solve the riddle. Brian Wesbury and Bill Mulvihill, economists at a Chicago Investment Bank, argue that the current BLS benchmark for employment is probably too low. Unlike other agencies, the BLS is slow to revise its benchmarks in light of contrary evidence. The BEA, for example, recently revised its January estimate for wage and salary income of private workers upward by 0.7%. The disconnect between income estimates and employment estimates is what has economists confused.
Wesbury and Mulvihill state that the BLS won't revisit its benchmark until February of 2005, and that employment figures may be off by as much as 100,000 per month. If this argument has force, employment will remain surprisingly low until next winter when the BLS restates the employment "facts." Can it be that simple?