Courtesy of Mish.
Initial Reaction
The already wide discrepancies between the household survey and the establishment survey took a wild leap today.
The establishment survey showed a gain of 204,000 jobs but the household survey showed a drop in employment of -735,000. However, the BLS states that on unadjusted basis 448,000 of the drop in employment was due to temporary federal layoffs during the budget negotiations.
Because of the way the BLS calculates overall numbers, one cannot subtract the seasonally unadjusted number of -448,000 from the reported seasonally adjusted number of -735,000 and arrive at any kind of meaningful number.
Important Note: I will go through my normal calculations, but they are very distorted by the temporary layoffs.
How Federal Layoffs Distorted the Picture
The BLS reports “Estimates of the unemployed by reason, such as temporary layoff and job leavers, do not sum to the official seasonally adjusted measure of total unemployed because they are independently seasonally adjusted. Household survey data for federal workers are available only on a not seasonally adjusted basis. As a result, over-the-month changes in federal worker data series cannot be compared with seasonally adjusted over-the-month changes in total employed and unemployed.”
From the Commissioner’s Statement
Some federal workers who were not at work during the entire reference week in October were not classified as unemployed on temporary layoff.
Our review of the data indicates that most of these workers should have been classified as unemployed on temporary layoff. Such a misclassification is an example of nonsampling error and can occur when respondents misunderstand questions or interviewers record answers incorrectly. According to usual practice, the data from the household survey are accepted as recorded. To maintain data integrity, no ad hoc actions are taken to reassign survey responses.
If the federal workers who were recorded as employed but absent from work had been classified as unemployed on temporary layoff, the overall unemployment rate would have been slightly, but not substantively, higher than reported.
Revisions
This was the fourth straight month of revisions to the establishment survey. The first two revisions were significantly lower. Last month, the revision was slightly higher and I commented “Perhaps the BLS has numbers they are happy with now.” So much for that idea. Today, the BLS reports “August revised up by 45,000 (from +193,000 to +238,000), and the employment change for September revised up by 15,000 (from +148,000 to +163,000). …
