Resolving the Controversy between Payroll and Household Surveys

The Bureau of Labor Statistics weighs in on what should be a thoroughly resolved debate: Which is the more accurate and reliable measure of Job creations, the Payroll or Household Survey?

Economists Arnold Kling and Steve Antler both referenced a recent Bureau of Labor Statistics analysis (BLS) on the divergence between the payroll survey and the household survey of employment. The BLS observed:

“As part of its annual review of inter-censal population estimates, the U.S. Census Bureau determined that a downward adjustment should be made to the household survey population controls. This adjustment stemmed from revised estimates of net international migration for 2000 through 2003. In keeping with usual practice, the new controls were used in the survey starting with data for January 2004. Estimates for December 2003 and earlier months were not revised to reflect the new (lower) population controls.

…As a convenience to its data users, BLS created a research series that smoothes the level shifts in employment resulting from the January 2000, 2003, and 2004 population control adjustments.”

-Bureau of Labor Statistics

It is somewhat perplexing to see controversy still dogging this statistical oddity.

Federal Reserve Chairman Alan Greenspan has noted that the Payroll survey is much more reliable than the household survey:

“I wish I could say the household survey were the more accurate,” Alan Greenspan, the Fed chairman, said in his testimony at a House hearing on Feb. 11. “Everything we’ve looked at suggests that it’s the payroll data which are the series which you have to follow.”

That hasn’t stopped several economists – arguably with political agendas from claiming the divergence between the two surveys is understating the strength of the economy.

As if Fed Chair Greenspan didn’t resolve the issue in his recent statements, the BLS itself has now weighed in. As these following charts make clear, when the household survey is “modified to make it more ‘similar in concept and definition’ to the payroll survey,” the divergement all but disappears.

The BLS did this by subtracting from the Household Survey:

1) Total agriculture and related employment;
2) Self-employed, unpaid family and private household workers;
3) Workers absent without pay from their jobs.

BLS then added back in non-agriculture wage and salary multiple job holders.

The use of the broader standard (including farm and unpaid family workers) is what apparently created the divergement, as shown by the Green lines.

Using data “similar in concept and definition” to the Payroll Survey eliminates the phantom missing jobs, as seen in the 10-year chart below:

Household and Payroll Survey employment,
Seasonally Adjusted, 1994-2004


Source: Bureau of Labor Statistics, March 5, 2004

Note the shift moving the Green line (Household Survey, unadjusted) to its new position – the red line (Household Survey, adjusted).

The BLS notes (in 10 year chart above, and 3 year chart below):

“The household series presented here has been smoothed for population control revisions. The “adjusted” household series has been smoothed for population control revisions and adjusted to an employment concept more similar to the payroll survey. Shaded area indicates recession.”

When the same statistical approach is applied to a more recent chart, the discrepancy between the two surveys again disappears:

Household and Payroll Survey employment,
Seasonally Adjusted, March 2001 – February 2004


Source: Bureau of Labor Statistics, March 5, 2004

We believe the comments of the Federal Reserve Chairman Alan Greenspan, when combined with the March 5, 2004 Bureau of Labor Statistics release, has now once and for all, turned the Household/Payroll Survey discrepancy into a “non-issue.”

The argument that the Household Survey more accurately reflects Job creation has been thoroughly discredited. Employ it at your own risk.


Bureau of Labor Statistics

The Two Employment Surveys
Arnold Kling
The Library of Economics and Liberty, March 14, 2004

Reconciliation of household and payroll employment surveys
March 13, 2004

Explaining the Recent Divergence in Payroll and Household Employment Growth
Chinhui Juhn and Simon Potter
Federal Reserve Bank of New York, December 1999

Two Tales of American Jobs
Edmund L. Andrews
New York Times, February 22, 2004

Bureau of Labor Statistics report, March 5, 2004

BLS report, March 5, 2004

Employment Situation Explanatory Note

What's been said:

Discussions found on the web:
  1. The Skeptical Speculator commented on Apr 10

    Some light on the contradictions on US jobs data

    This post by Barry L. Ritholtz hopefully sheds some light on the contradictions on US jobs data.

  2. Prometheus 6 commented on Sep 4

    So I don’t catch everything

    Not right up front, anyway.

    We’ve already seen Type Two Tim, aka Timothy Kane of The Heritage Foundation,

Posted Under