My hope, perhaps naive, is that 95% of what happens in this Rube Goldberg machine basically amounts to the government paying private insurance companies to pay doctors, which is inefficient but ultimately straightforward.
I was going to push back against this, but when I googled "what percent of American health care is administrative costs" I got back numbers ranging from 15-34%, which is pretty darn high. Much higher than the last number I heard, which was something like 8%. The factor of two variance is kind of disturbing in itself.
All this means is that for insurance companies to maximize their administrative costs they push up the overall price - from their perspective making 10% of $10M is way better than 10% of $5M.
Drives me up the wall that the same people who can correctly identify when nominal nonprofits like rich people's charity organizations universities or even hospitals are pissing away money on officer salaries, perks, facilities, etc. because "nonprofit" think that somehow being a 10% or X% profit capped organization can't do the same thing with whatever expenses they aren't capped on.
In my country, social security and universal healthcare administrative cost are 8-11% depending on the year.
Compare that to Swisslife, my previous private healthcare insurance: for every 100€ I gave them, 66-72€ were paid to insure in need while I had them. Of course the mandatory dividends were pretty high before COVID, and reinvestment/marketing/sales has to take a bite of that, so perhaps only half of the 30% they take from customers are admin costs, but still, I'd rather give more to my country national healthcare if I could.
Probably the variance you have could be caused by non-admin costs being counted as admin costs.
Unfortunately there's a fraud problem, NOT by patients but by various entities claiming for unnecessary or never-performed procedures. https://pubmed.ncbi.nlm.nih.gov/25068880/