In food, grocers are (roughly speaking, here) analogous to insurers and food is analogous to care utilization. Grocers want to maximize sales of food at the best price possible: they make more money by moving their product. Insurers want to minimize utilization: they make less money by "moving" their product.

In the case of food, the grocers' and buyers' incentives are aligned: people need food, and grocers want to provide it. In the case of medicine, this isn't the case: people need care, and the access providers don't want to provide it.

It makes sense to me that an insurer would never be a good analogy for a grocery store, but that problem seems specific to the insurance model, rather than to medicine itself. That's more what I meant to ask about. If we imagine that hospitals and patients work like grocery stores and their customers, a few obvious problems come up:

- Healthcare is too complicated for patients to make informed choices. - Emergency medicine is too urgent for patients to have much choice anyway. - Demand for care is unevenly and unfairly distributed, because of conditions people are born with.

So I guess my question is, how much of medicine (in terms of dollars) hits those problems? Cancer treatment seems to hit at least two out of three. But on the other hand, a lot of dentistry seems not to.

Of course you have incentives. Hospital administrators want to cut costs, doctors want to redirect patients to their private offices, suppliers want to increase profits (and will offer kickbacks to doctors and administrators), etc.

If you have humans, you have incentives. Regulation can change some, but not eliminate them.