Americans, by and large, are not a bunch celebrated for their fondness of nuance and technical esoteric’s … so expect the methodology change associated with DOL’s calculation/reporting of Jobless Claims to fall on mostly apathetic ears.
The methodological update has been covered moderately well but, it’s important, so a quick high-level redux is worthwhile:
- You can find the DOL announcement HERE and the historical data & seasonal factors HERE
- The first week reflecting the adjustment is the week/data for 8/29 (this weeks data). Data in prior periods will not be revised, so the new and old seasonally adjusted (SA) will not be comparable.
- The adjustment formula is: SA Claims = NSA Claims * 100/Seasonal Factor.
- The adjustment represents a change from a “multiplicative” adjustment to an “additive” adjustment - effectively representing a shift from a percentage change adjustment to an absolute/levels adjustment. Use of percentage change and or absolute change doesn’t make a significant difference during “normal times” but materially impacts the reported SA during any macro shock period.
For example: NSA claims were 821K last week. The Seasonal Factor was 81.7 yielding (using the formula above) a reported SA Total of 1006K – an adjustment of +184K, roughly equal to a +22.5% adjustment. Using an additive approach, the adjustment would have been something in the range of +40-50K higher, not ~+200K higher.
So, the methodology adjustment provides a better reflection of reality but any improvement is mostly illusory relative to underlying conditions. Any perceived material, positive inflection in the labor market represents a false read and an outcropping of statistical distortion (of sorts).
Indeed, the Seasonal Adjustment Factor (highlighted yellow in table below) is the same for both this week and last week (SF = 81.7) so while the SA data are not comparable, the NSA data are and offer a clearer read on the underlying trend. On that score, NSA Claims rose +7.6K, from 825K to 833K, signaling no underlying sequential improvement on the separation side of the net payroll equation.
The larger macro narrative arc vis-à-vis the labor market remains largely unchanged:
- Labor market improvement remains ongoing but the pace is slowing and August may prove the last month of large-scale gains.
- With PPP funds exhausted, enhanced benefits dramatically reduced, Walmart/Best buy raising a cautionary flag on the prospects for further consumer strength, permanent job loss rising, Ford/American Airlines/SalesForce/MGM/etc announcing fresh layoffs and Consumer Confidence measures around the labor market now deteriorating, the derivative impacts are beginning to manifest and scope of structural damage beginning to take shape.
- Noise: Census Hiring to the tune of ~250K will positively impact the August NFP data while a temporal shift in school/teaching hiring may serve as a modest negative drag.
Of course, with estimates on NFP ranging from -100K to +2.4M with a Standard Deviation of +454K, no one really knows what’s going on other than we’re trending towards less good during what should (still) be the steepest part of the recovery curve and ~900K in new job loss 24-weeks into a "temporary" shock counts as laudable in our new Trending surreality.
.
