Labor's suspicious statistics

| Friday, April 19, 2013, 8:57 p.m.

The Reuters report “Jobless claims drop, easing investor fear of soft labor market” (April 12 and cited a U.S. Bureau of Labor Statistics (BLS) report on the number of initial unemployment claims filed during the week ending April 6, 2013. It reported that the “seasonally adjusted” number of initial unemployment insurance claims filed fell by 42,000 to 346,000. No mention is made that the BLS reported that the actual number of claims filed rose by 37,025 to 353,933.

To report a seasonally adjusted figure of minus 42,000 claims when the actual number was plus 37,025 means that the statisticians were expecting a seasonal fall in claims of 79,000 (37,025 plus 42,000). That sounds to me like a totally unwarranted assumption.

What are the seasonal forces in April? I am not aware of any.

We are owed an explanation by the BLS of what the seasonal adjustments were that it made. The two figures, one reporting an unusual fall in the number of claims and the other an unusual rise in the number of claims, are not consistent with one another. Maybe only the methodology used for seasonal adjustments is at fault. Something is grievously wrong with the statistics.

It is a disservice by the BLS to the business community and to investors to downplay the actual number of claims filed. But journalists pay attention only to the BLS announcement of “seasonally adjusted” claims.

One only has to read the report beyond its first paragraph to learn how many claims were actually filed. Apparently, few journalists bother to do so.

What is CNBC's and Bloomberg's excuse for such shoddy reporting, which is watched by millions of investors?

Raymond L. Richman

The writer is professor emeritus of public and international affairs at the University of Pittsburgh.

Subscribe today! Click here for our subscription offers.


Show commenting policy