The stock market is swooning over the latest jobs report, as are some of the media outlets and pundits. Other media outlets and pundits are saying either that the BLS reports are lies, or that the numbers underlying the jobs creation of 243,000 and drop to 8.3% are much more bleak. I don’t buy the first assertion, but there certainly is some validity in the latter, as the numbers leaving the work force is pretty stark, and the percentage of the population which is employed is essentially unchanged from the low point which it reached at the bottom of the recession.
The question that occurs to me, and one that has not been explained, is what caused the increase in hiring? What is the source of this sudden increase in confidence that caused employers to start adding employees?
It’s not consumer spending, that was dismal in December, so why the January hiring? Housing prices were down in December. The European crisis is showing some small signs of resolution now, but it was still in full explosion mode when this hiring occurred. American politics have not improved. There is much horn honking and bell ringing about the new jobs being added, and rejoicing about the economy improving based on those new jobs, but no explanation of what invoked the hiring.
Do we have a self licking ice cream cone economy? The economy is improving because employers are hiring, and employers are hiring because the economy is improving. Which means employers are hiring because employers are hiring?
I don’t think these numbers are lies, per se, but I don’t trust them either. And that’s nothing against the Obama administration; I didn’t trust them when the Bush administration was doing them either. They are based on the “household survey” and the “birth-death model,” and they are not only “seasonally adjusted” but the January numbers are “annually adjusted,” and this years January numbers are “adjusted for the 2010 census” as well.
So in most months they are “adjusted estimates” and this set of numbers are estimates that have been triple adjusted. Just how seriously should we take that kind of thing?
The government estimate of retail sales uses a “birth-death model” to estimate retail sales, one which became so widely discredited that it is no longer even reported. Or if it is, they no longer report it by that name.
For the “birth-death model” the government had a group of retail stores which they interviewed monthly, and retail sales estimates for the nation was reported based on that group of stores. Then they make some complex and unexplained “assumptions” about new stores and the death of existing stores, but they do not assume that any of the reporting stores are in the “death” category, and any of those stores which did not respond were assumed to have zero change in sales.
You do see the flaw in that, right? The government did not. Think Circuit City. They didn’t accommodate the concept that stores might not respond because they were out of business and their sales had dropped to zero. So that meant that the increase reported by other stores was bogus, it was not true increase in overall sales, it was business they inherited from stores which had closed. Meanwhile the defunct stores were being included as if they were still selling at their previous pace, so retail sales estimates were wildly inflated.
Is jobs reporting by the BLS suffering from similar flaws with its “surveys” and its “seasonal adjustments” and such? Well, I certainly don’t know the answer to that, and there is no way to really be sure whether it is or not.
In the case of retail sales, there is one certain measure to evaluate relative retail sales levels, and that is to look at sales tax collected by states. Those are real numbers and are available in pretty close to real time, as retailers pay those taxes monthly for small retailers and weekly for large ones. By comparing those tax collections on a month by month basis we can get a pretty accurate estimate of what retail sales are doing. We used to see those numbers published fairly regularly, but I haven’t seen them in quite a long time. Perhaps someone got tired of them being in conflict with government estimates?
For employment there is a similar measure, and that is income tax withholding submitted by employers. That is also real numbers and is submitted in real time, since the money is paid to the government within ten days of being withheld from employees paychecks. The increase in job creation should be pretty easy to verify within a couple of weeks if those numbers are made public. Somehow, though, we never see confirmation of the BLS “estimates” by showing that they are reflected in the employer income tax submissions.
I can’t help but wonder why not.
No comments:
Post a Comment