By Arnold Kling
Tim Kane has written a timely paper on the behavior of the payroll survey, which is the source of data showing disappointing employment numbers. The key point is this:
The payroll survey double-counts any individual who changes jobs during the pay period in which the worker is on two payrolls. Such turnover overcounting would normally be irrelevant if (1) the turnover rate was stable over time and (2) pay periods were stable. But if turnover rates increase or pay periods expand from weekly to monthly, overcounting will inflate the payroll count of total employment.
The stronger the economy, the more likely it is that workers will change employers and be double-counted in this way. Thus, near the peak of a boom, the payroll survey will overstate employment by more than it overstates employment when the economy is operating below capacity.
As I read this analysis, it says that the some of the jobs “lost” since 2000 were not really lost–they were double-counted in the first place. I presume that the same holds true for aggregate hours worked, which is the measure that I use to construct my preferred measure of labor capacity utilization. Perhaps labor capacity utilization was overstated in 1999, so that the decline since then is not as pronounced as my data would have indicated. By the same token, measured productivity growth since 1999 would be overstated.
For Discussion. What policy implications, if any, does Kane’s analysis suggest?