A Closer Look at Friday's Non-Farm Payrolls

Having laid out a plausible, if somewhat crude, case for a miss ahead of the non-farm payrolls report on Friday, needless to say we were taken aback by a much better than expected +244K result. So much so, we went straight to the source for a closer look at the data - the Bureau of Labor Statistics' http://www.bls.gov/news.release/empsit.nr0.htm. Here, we observe there are two sources the government uses to derive its monthly employment results - Household data (for unemployment) and Establishment data (for net job creation). The former is in our opinion more worthwhile, as it reflects a direct survey of households while the latter mainly incorporates assessment from businesses (more on that below).

Turning to the Households results - http://www.bls.gov/news.release/empsit.a.htm - we note Civilian labor force practically unchanged (153.4K), the number of unemployed up by 200K (13.7K), and unemployment rate therefore up to 9.0% from 8.8%. So far so good. The rub is that participation rate (labor force / civilian population) is UNCHANGED at 64.2% - higher denominator is the only way it would be MATHEMATICALLY possible to have both higher unemployment rate AND job creation of as much as 244K. In fact, according to the Household survey, the number of employed in April actually FELL 190K. This is how unemployment rate rose to 9.0% while civilian labor force remained unchanged. So... where in the world did +244K come from?

Here we turn to the Establishment data. Unbeknownst to the public, these series operate under the assumptions of a statistical adjustment called the "Birth/Death Model" - http://www.bls.gov/web/empsit/cesbd.htm - or monthly estimates of creation/unwinding of all businesses and government agencies. While we are not privy to the statistical tools the government uses in making these monthly adjustments, we do note it implies a hard-coded model to "reduce a primary source of non-sampling error which is the inability of the sample to capture, on a timely basis, employment growth generated by new business formations." We are hardly conspiracy theorists, suffice to say the government's ability over assessment of net business generation should not be taken at face value to be 100% foolproof, particularly when every other economic indicator has argued in favor of a far softer jobs report. As expected, the bottom of the last link shows net creation of 175K (far beyond recent months!) jobs based solely on the statistical adjustment - the bulk of 244K payrolls "creation." This very discrepancy between rising unemployment, unchanged participation rate/labor force, and a very strong print in new jobs helped temper (along with the rumor-mill of Greece leaving EMU) the early risk-on sentiment later in the US session on Friday.

Turning to the early Asia session economic data on Sunday/Monday, Bank of Japan will releases its outdated policy meeting minutes that preceded the most recent meeting. In Australia, ANZ job ads will foreshadow the official monthly labor data on tap for 21:30ET.

By GG - AshrafLaidi.com Staff