Many mumble-mumble years ago, I worked at our local hospital lab as a lab assistant. The hospital itself has well over 500 beds, a Trauma 1 level emergency department, and the regional neonatal intensive care unit. The lab was the largest, most modern and best equipped hospital lab in the state. I worked the night shift and minimum staff was usually six or seven people, numerous more on the day and evening shift when outpatient services were utilized and the microbiology lab and pathology labs were staffed.
Since then, they've modernized further. The chemistry, hematology, urology, and coagulation departments were all unified into a robotic system. I got to visit again about two years ago in the afternoon. There were two Medical Technologists (plus another in the blood bank) and an assistant to prepare samples. Less than half the staff we had when I worked there.
The technology has gotten much more expensive but the labor has been reduced significantly.
Sure - so we know the costs have changed; and we can guess that they are now lower. But it's still very hard (and, perhaps ironically, too expensive) to work out exactly what the costs actually are.
Like all businesses, they will be 'calculated' based on educated guesswork and simplifying assumptions, some of which are inevitably wrong; and, like in all businesses, once a number is decided upon, it will be treated as gospel when used as an input into decision making and other calculations, with no estimate of error or hint that it might not be a hard fact.
All commerce is like this - lots of accountants doing highly accurate sums that are based on highly inaccurate initial data, and pretending that their results are meaningful.
It's astonishing that economies work at all. It's certainly amusing when economists are surprised by unexpected events - such events are (predictably) the norm in complex systems where the data about the system is fundamentally corrupt from the outset.
And of course the entire thing is even less accurate when demand is uncoupled from price, as is the case with health care. If you sell widgets, you know that widget prices are too high, because people stop buying widgets. But if you sell malaria tests, people don't stop getting malaria because your test kits are too expensive.
So prices tend to inflate, and in the absence of a signal from the market in the form of reduced demand, or any good understanding of the costs on which a reasonable price could potentially be based, they can inflate rapidly - and the more payers there are, the faster this can occur.