The BLS, as part of the NFP report, has issued its preliminary estimate of the benchmark revision, which confirms that the BLS is really just BS. According to the report, for the period ended March 2010, the BLS has overestimated jobs by 366,000 (0.3%), or just over 30K jobs per month. While not as bad as the prior benchmark revision of almost one million for the period ending March 2009, this continue to be a blow to both the credibility and the data tracking capability of the US Bureau of Truth. By industry, the biggest hit was to the trade, transportation and utilities industry (-144K), Manufacturing (-114K) and Leisure and Hospitality (-91K). Luckily, losses in these critical sectors were offset by even more bankers than had been previously expected: Professional and business services ended up being revised higher by 14K.
From the Excerpt:
Preliminary Estimates of Benchmark Revisions to the Establishment Survey
In accordance with usual practice, the Bureau of Labor Statistics is announcing its preliminary estimates of the upcoming annual bench-mark revision to the establishment survey employment series. The final benchmark revision will be issued on February 4, 2011, with the publication of the January 2011 Employment Situation news release. Each year, the Current Employment Statistics (CES) survey employment estimates are benchmarked to comprehensive counts of employment for the month of March derived from state unemployment insurance tax records that nearly all employers are required to file. For national CES employment series, the average of the absolute values of the annual benchmark revisions over the last 10 years is 0.3 percent at the total nonfarm level. The preliminary estimate of the benchmark revision indicates a downward adjustment to March 2010 total nonfarm employment of 366,000 (-0.3 percent).
Table B shows the March 2010 preliminary benchmark revisions by major industry sector. As is typically the case, many of the individual industry series show larger percentage revisions than the total nonfarm series, primarily because statistical sampling error is greater at more detailed levels than at a total level.