U.S. Employment Report Even Weaker Than It Seems

The U.S. employment report yesterday was regarded by the markets as a bullish one, because the decrease in payroll employment was somewhat smaller than expected. However, considering how the Bureau of Labor Statistics continue to insist not only on continuing to impute jobs from the "Birth/Death" model, but in fact imputes more and more of them, even as any honest observer would expect fewer and fewer (in fact at this point none at all, or even negative) jobs created in new businesses relative to the number destroyed in business failures, it should be clear that job destruction is accelerating.


In February, the assumed seasonally adjusted monthly addition from the Birth/Death model (The raw monthly number is seasonally unadjusted. I've calculated the adjusted number by taking the total number added during the latest 12 months and dividing it by 12) was 53,000, in July it was 71,000. This means that the payroll number is in fact getting more and more distorted by this flawed model.

If you add the 71,000 imaginary jobs added by the model to the 76,000 private sector jobs lost according to the official number, then we're talking about at least 147,000 fewer private sector jobs.

This weakness was also confirmed by the rise in the unemployment rate from 5.5% to 5.7% and the decline in aggregate hours worked of a full 0.4%, a decline confirmed by the even sharper increase in the household survey of part-time unemployment. All of this suggests the opposite interpretation of the report compared to the one the financial markets decided for: namely that it indicates an accelerating and not decelerating downturn.