[joshuahedlund]

The article claims that but I can't tell what evidence they have for that claim. One of their linked pdfs[0] is long but as far as I can tell involves working class insured folks, and the chart on p31 shows Drug Overdoses as the largest excess category by far for the 0-44 age group.. Also page 27 says "Note that recent period deaths may continue to shift from non-COVID to COVID as COD coding becomes more complete".

Edit: and the table on p43 seems to directly contradict their claim that the excess deaths are concentrated among the young "working" as opposed to non-working: Age 15-44 have almost the same percentages (18.9% vs 19.2%) for excess Q2 deaths; the insured cohort is the slightly lower one. "It's probably mostly drug overdoses and COVID" still seems like a reasonable null hypothesis to me.

[0] https://www.soa.org/4ac0fd/globalassets/assets/files/resourc...