From the New York Times Economix Blog by Floyd Russell:
Every year at this time, the Labor Department tells us how badly it blew the previous year’s job figures. In each of the last two years, it turned out that job growth was better than previously estimated. But not this year.
The revised numbers come because the government has access to better data, after a long delay. The earlier figures come from a survey of employers, with the numbers adjusted for the government’s estimate of how many jobs were created by new employers, and lost from failing companies that could not, of course, respond to the survey.
The revised figures come from state unemployment insurance premium figures. Employers must pay insurance premiums based on how many employees they have each month, and those numbers are deemed more reliable, even though it takes a long time to compile them.
The new numbers are called the “benchmark revisions,” and they are preliminarily announced in September, although they are not made final until the followed February.
Today’s announcement sounded positive if you did not study it, and some news reports reflected that misreading of it.
“The preliminary estimate of the benchmark revision indicates an upward adjustment to March 2013 total nonfarm employment of 345,000 (0.3 percent),” said the Labor Department’s announcement.
But then it explained that all of that gain, and more, came from changing definitions, not new jobs.
The monthly job figures exclude some workers, like the self-employed, and it counts household workers as self-employed. But now it has revised its definitions, and decided that “establishments that provide nonmedical, home-based services for the elderly and persons with disabilities” should be classified as health care companies. “Many of these establishments were previously classified in the private households industry,” it said.
That added 490,000 workers to the total reported number for last March. On an apples-to-apples basis, however, the result is to reduce job growth in the 12 months through last March by 124,000.
The old numbers indicated that job gains in those 12 months averaged 169,000 per month. This change will shave that figure to 159,000.
It is not a big change, but the direction is not encouraging.
Addendum: The Bureau of Labor Statistics says that it will adjust figures going back at least several years, so that comparisons will not be distorted.