Today's horrendous Initial Claims number was so bad that not even CNBC tried to spin it. In fact, as John Lohman points out, it was the second biggest miss to consensus in history! Of course, seeing how this is consensus and the BLS does not reveal any unknown information, we wonder just how difficult is it to factor in any special factors in determining project numbers, and if the answer is "very" then why do we need economists in the first place (that's rhetorical). And while Liesman is in Europe on some assignment, here is Goldman explaining why the historic miss was based entirely on special factors.
Charting historical misses to consensus.
And heeeeeere's Goldman:
1. Initial jobless claims rose to 474k in the last week of April, from 431k in the previous week. The Department of Labor (DOL) cited three main factors behind the increase. First, plant shutdowns in the auto sector caused temporary layoffs. Second, Oregon reformatted its extended benefits program which "also pulled in some new claimants" (according to Bloomberg). Third, initial claims in New York rose due to seasonal adjustment problems. According to the DOL, the change in Oregon accounted for only 1,000 or so of the increase, and the impact of auto shutdowns was "very small". The major factor behind the increase was "administrative", reflecting seasonal claims in New York that were all filed in a single week. This accounted for about 25,000 of the increase last week, according to the DOL.
2. Even with these distortions accounted for, the result was still poor, and suggests some slowing of employment growth. This is more likely to have implications for the May employment report than for tomorrow's report, which is based on a survey conducted in the week of April 10-16, well before the numbers reported today.
3. Nonfarm productivity grew at a 1.6% annual rate in Q1, better than we or the consensus expected. Nonfarm output significantly outpaced economy-wide GDP growth (3.1% versus 1.8%) while employee hours grew at a similar rate to the past two quarters (1.4%). Unit labor costs were up slightly more than expected, and 1.2% on a year-over-year basis, although in historical terms these remain extremely muted.