There's No BLS Without 'BS'

by: TrimTabs

Last week the Bureau of Labor Statistics estimated that 171,000 seasonally adjusted jobs were added in October. That initial estimate was based upon a survey of several hundred thousand mostly large businesses and virtually all government employers, and then is seasonally adjusted.

How accurate are the BLS initial monthly job reports? They are mostly BS. For example, the BLS initially estimated that in August 96,000 seasonally adjusted jobs were added. Then this past Friday two months later, the BLS revised the initial August 96,000 to 192,000 new jobs. That means that the initial August number was revised by about 100%. Is that a joke or what? Why would anybody invest money based upon an initial BLS number?

This morning Jack Welch former GE honcho and now independent player talked about the disconnect between the BLS reporting an October jobs pick up to 171,000 jobs and how flat his businesses are and have been. Welch also pointed out that the BLS jobs number includes seasonal adjustments that change year to year and sometimes vary by hundreds of thousands, year over year. And no explanations are ever made as to what the heck is going on.

The hopelessly outdated survey methodology and seasonal adjustment nonsense means the BLS data is as reliable as what passes for economic date coming out of China.

But, the joke is even worse and the ultimate joke is really on us. Real time data is available as to how many people are working right and how much they are making based the withheld income and employment taxes sent to the Treasury daily by all employers. That has been the basis for TrimTabs job and wage and salary growth estimates since the late 1990's. And it works. The only problem is the BLS does not, or at least says they do not look at the real time data embedded in withholdings sent to the US Treasury.

Nor does the Bureau of Economic Analysis - which puts out National Income and GDP stuff. It appears government economists are happy with the way things are. Their methodology keeps lots of well paid economists gainfully employed. So what if the data they produce is garbage that is ultimately damaging the US economy? But hey, at least these government economists still have a job!

And then there is the truly garbage Census Bureau retail sales report. Do you know that right now anyone can buy aggregate real time sales data from Master Card, Visa, American Express, etc. at a combined annual cost of less than one government economist. Instead of real time data, the Census Bureau guesses at retail sales with a monthly survey of one out of every 166 retailers.

What follows is from the current Census Bureau Advance Monthly Sales. "The estimates are based on a subsample… Approximately 5,000 retail and food services firms (are surveyed) whose sales are then weighted and benchmarked to represent the complete universe of over three million retail and food services firms." So the Census Bureau spensd millions surveying 5,000 retailers monthly and then guess. What bullshit!

Yes, after several years and all the income tax returns are analyzed the jobs and income numbers are accurate. But they certainly do not start out that way. Not even close.

Would you rather make investment decisions based upon real time data as to how many people are working, what industry they are working in, how much are they making, what are they are spending upon?

Or would you rather make investment decisions based upon US government initial data releases? It would be nice if we had a choice.