A few mentions of increased demand for temporary/contract labor to let businesses stay flexible and not have to commit to full-time workers.

Seeing that trend in ASA Staffing Index (up YoY since 9/21), but not in temp help services employment

Beige Book, Jan 2026, Minneapolis Fed

credit to @orcma.bsky.social for pointing out in absolute terms, retail does actually explain a large part of the difference, so maybe next month's methodology chg will help more than i originally thought

bsky.app/profile/dani...

Good call, I had been focusing on the % chgs by sector, not the # chgs. (If you'll excuse the quick & dirty chart) retail does explain a large part of this. So maybe the methodology chg to exclude internet sales will help much more to close the gap.

No, unfortunately not in this data

One possibly relevant note: In next month's report on Feb 11, applications "associated with internet sales" will be excluded from high-propensity & corporation applications (retroactively + fwd). Could help shrink the gap, but the sectoral distribution suggests to me that isn't the whole story

4/4

FWIW, WBA (business applications with planned wages) seem more plausibly likely to predict actual employment growth from new businesses so this is relevant from a jobs growth POV.

3/

There's not an obvious sector explanation (say, somebody who sets up an etsy store as a side gig). Almost every sector has seen business applications with planned wages underperform business applications in total.

2/

I don't view that as a bias issue, but instead a methodology one. I don't love the approach w/ OER but I'm also not convinced I have a strictly better solution either.

Also credit to OP for asking this q respectfully. I don't expect the average person to have a good understanding of how the BLS works and why we trust it so I don't mind getting the q, but people insisting it's wrong/biased w/o knowing what they're talking about is grating.