This week’s jobless claims release was significant in a couple of ways. For starters, the Department of Labor revised seasonal adjustment factors (more on that below) and second, the most recent print for the seasonally adjusted number hit one of the lowest levels on record. Claims came in at 166K which tied the number from two weeks ago (after revisions) for the second-lowest reading on record behind the week of 11/29/68. That week’s reading was only 4K lower.
As for continuing claims, this week’s reading was higher rising from 1.506 million to 1.523 million. Having come off last week’s low, that remains one of the strongest levels since 1970.
Pivoting back to initial claims, while adjusted claims are at multi-decade lows, non-seasonally adjusted claims are very healthy but not quite at a record. At current levels of 193.1K, the only year with a lower reading for the comparable week of the year was 1969 (170K) but at the seasonal lows in the few years prior to the pandemic, claims were even stronger than they are now. So while unadjusted claims are strong, the seasonally adjusted number was even more eyecatching this week.
As for why the seasonally adjusted number appears so much stronger, as shown in the second chart below, the current week of the year has historically been the second most consistent week of the year in seeing a week-over-week increase in claims. As a period with consistent seasonal headwinds, the seasonal adjustment would control for this by depressing the level of claims by a relatively strong seasonal factor. That usual increase was not seen this year as NSA claims were actually slightly lower falling from 196.8K to 193.1K. That means seasonal adjustment helped out what was already a strong number.
Additionally worth noting this week is the DOL changed its seasonal adjustment methodology. The DOL fully describes the change in this week’s release and links to further reading on the topic for those interested in going deeper into the weeds, but to briefly summarize: periods when the level of an economic indicator experiences historic shifts in levels as claims did in 2020 and 2021, multiplicative adjustment (which has been used in “normal times” like prior to the pandemic and from here forward) would tend to over or understate adjusted claims. Additive adjustment, on the other hand, is less affected by the actual level of the indicator meaning it would lead to more accurate approximations of seasonally adjusted numbers in times of sudden spikes in the series. This methodology has been used throughout the pandemic period. Now that claims have more or less returned to a more normal range of readings, the DOL sees it justifiable to switch the seasonal adjustment methodology back to multiplicative. While that does not mean too much for interpreting claims, that change in methodology has led to some large downward revisions. Click here to view Bespoke’s premium membership options.