This morning’s revelation that payrolls shrank by only 20,000 (below the -80,000 consensus) was just the spark Wall Street bulls were looking for. The unemployment rate fell to 5.0%, defying predictions that it would in fact rise to 5.2%.
But things may be even better, says Larry MacDonald at Seeking Alpha. According to MacDonald, the labour Department’s jobs reports are flawed, and exaggerate the number of job losses because:
- Changes in employment in the service sector are not fully captured.
- Heavy seasonal adjustments often obscure trends in employment
- Employment growth at smaller copmanies and growth in self-employment are not fully captured
- The company “birth/death” adjustments are nothing more than educated guesses
- Only 60% of the survey for the latest month is complete
In a nutshell, the survey methodologies used by the Bureau of labour Statistics were designed when the US was still a manufacturing economy. Survey participation is heavily skewed to government entities, utilities, and manufacturing despite the fact that the US economy is now a service economy, and this is where all the growth is.
A better way to look at the jobs picture is by examining withheld income taxes. Investment research firm TrimTabs uses this method, and its analysis reveals a different story:
TrimTabs Investment Research reported today that the U.S. economy added 9,000 jobs in April… Biderman (TrimTabs’ CEO) thinks TrimTabs’ approach is more accurate. It estimates job losses from inflows of withholding taxes from the paychecks of all salaried U.S. workers. “Daily withheld income taxes rose 3.78% this April compared with April 2007,” said Biderman. “Growth greater than 3.7% year-over-year means the U.S. economy is adding jobs.” TrimTabs Online Job Postings Index also showed a significant increase in the availability of jobs.
It sounds almost too good to be true, and might be. But the market and financial media react to BLS jobs reports as though they were the gospel truth when, in fact, they are just surveys that are only slightly more accurate than (for instance) political polling information.