From the lies, damned lies, and statistics department, Tyler Durden at zerohedge.com:
One of the main reasons why the BLS has been massively overestimating job creation ever since great financial crisis, is due to the well-known birth-death adjustment, aka the CES Net Birth/Death Model, which quantitatively is shown on the chart below, has resulted in the “addition” of some 5.3 million jobs, that don’t actually exist, but are merely modeled by the BLS which continues to assume the same new business creation/destruction dynamics that existed before the crisis.

The is a big problem with this core assumption, which has follow through effects not only for domestic fiscal policy, but also monetary policy (and explains why despite a 5.1% unemployment, there is zero wage growth, thus keeping the Fed pushing the ZIRP accelerator pedal years later), for the simple reason that as of this moment it is dead wrong.
Here is what Gallup CEO, Jim Clifton, wrote several months ago looking at the trends in new business creation and destruction in the US.
We are behind in starting new firms per capita, and this is our single most serious economic problem. Yet it seems like a secret. You never see it mentioned in the media, nor hear from a politician that, for the first time in 35 years, American business deaths now outnumber business births.
The U.S. Census Bureau reports that the total number of new business startups and business closures per year — the birth and death rates of American companies — have crossed for the first time since the measurement began. I am referring to employer businesses, those with one or more employees, the real engines of economic growth. Four hundred thousand new businesses are being born annually nationwide, while 470,000 per year are dying.
As Clifton adds “you may not have seen this graph before” and for good reason: it destroys the most sacred assumptions held by the BLS’ cubicled actuaries and various tenured economists locked up in their ivory towers: namely that the number of US business startups outnumbers the number of failures. This is no longer true!

Here is what the above chart shows: until 2008, startups outpaced business failures by about 100,000 per year. But in the past six years, that number suddenly turned upside down. There has been an underground earthquake. As you read this, we are at minus 70,000 in terms of business survival. The data are very slow coming out of the U.S. Department of Census, via the Small Business Administration, so it lags real time by two years.
Gallup adds that business startups outpaced business failures by about 100,000 per year until 2008. But in the past six years, that number suddenly reversed, and the net number of U.S. startups versus closures is minus 70,000.
To continue reading: How Student Debt Broke The “Birth/Death Adjustment”