US DATA: Initial Claims Likely At A Healthy Level Last Week (1/2)

MNI estimates for initial jobless claims from state-level data suggest a further decline in the four-week average of seasonally adjusted claims, back towards levels that have been associated with historical labor market tightness. 

• MNI estimates initial jobless claims at a seasonally adjusted 218k in the week to Oct 11, down from 234k in the week to Oct 4 (revised slightly from an initial estimate of 233k).
• As should now be familiar under the government shutdown, this uses the state-level data released yesterday afternoon ET with pre-released seasonal factors.
• There is a wider error band to estimates this week. Massachusetts again hasn’t reported this week (for context, 5.1k claims in the previous week). However, whilst Hawaii has reported this time (1.2k in latest data), some larger states haven’t with Arizona (2.8k previous week), Nevada (2.7k previous week) and Tennessee (3.1k previous week) all missing.
• Note that when it comes to using last year’s seasonal pattern for estimates, Tennessee was likely impacted by Hurricanes Helene and Milton around this time of year, suggesting our estimate could err on the higher side.
• Taking 218k as a given, it would see a further encouraging decline in the four-week average to 223.5k. That’s down from a recent peak of 241k in the four-week average of the official series back in early September.
• Much of this prior increase looks to have been down to Texas fraud. Whilst there haven’t been revisions to that previous spike, Texas claims are at least back running at more typical levels (a non-seasonally adjusted 16.4k last week vs 32k in early September).
• As we like to reference for a comparison with past periods of labor market tightness, initial claims averaged 218k through 2019.
• Note also that when looking at the seasonality chart below, the uptick around this time last year should be viewed in light of those aforementioned hurricanes. The latest claims data look in keeping with pre-2024 trends more broadly.