More on the CMS Report

Robert Waldmann+

I paraphrase Kevin Drum

My post last night about the CMS report on healthcare reform … [was not] a model of clarity. For more, see Ezra Klein here and Jon Gruber here. Combined with … [Drum’s original] post and this Wonk Room post, you should get a pretty good idea of what’s up.

Kevin Drum’s orginal post is much clearer than mine and makes the point I tried to make in the last paragraph.

After the jump I have further thoughts.

I quote from the wonk room post

CMS doubts that the health care industry could “improve their own productivity to the degree achieved by the economy in large” and predicts that the productivity adjustments in the House bill could lead some Medicare providers “for whom Medicare constitutes a substantive portion of their business” to stop seeing Medicare patients.

First I note again, the unrealistic productivity estimates are used to calculate Medicare compensation for institutional care providers (hospitals, nursing homes and home health care agencies) not physicians in private practice. This is important, because physicians in private practice are much more likely to refuse Medicare patients.

Second I ask if CMS chief actuary Foster is familiar with economics 101. The clause “for whom Medicare constitutes a substantive portion of their business” makes negative sense. (I’m trusting Wonk Room’s Igor Volsky to have quoted without removing necessary context. I haven’t gone back to the pdf.)

As usual, I start with a very unrealistic economics 101 model. I assume that hospitals aim to maximize profits (totally false). In that case, hospitals will accept Medicare patients if the fee is below marginal cost. Marginal cost is increasing along the relevant range. Consider hospital A and hospital B (assume they have the same number of beds and same marginal cost as a function of patient population for simplicity). A does a lot of business with the CMS, B just a little. If A refuses Medicare patients, then the reduction in its patient population will be greater than that of hospital B if hospital B refuses Medicare patients. This means that the average marginal cost of treating the patients it turns away will be lower. This means that it can’t be profit maximizing for A to turn away medicare patients and for B to treat them.

Consider the extreme case of a hospital with only Medicare patients. If it refuses CMS rates it will have zero revenues and go bankrupt. Consider the extreme case of a hospital with no Medicare patients. Refusing to treat Medicare patients makes no difference at all.

OK so ospitals, including for profit hospitals, are not profit maximizing entities (I think no firms are). However, I don’t think that there are hospitals with many Medicare patients which can afford to refuse Medicare compensation. Note they still have to provide emergency care (by law) and they can not bill the Medicare patients rather than the CMS (If they accept CMS payment, then they can charge patients up to 15% in addition to what they get from the CMS with or without reform).

The logic of Foster’s argument is that hospitals which have lots of medicare patients will go bankrupt if the HR3692 is enacted and the limits are actually enforced. This logic is invalid, because the increase in insurance coverage implies more revenues for hospitals. One can’t consider money from the CMS alone when considering whether a hospital goes bankrupt.

The relative rates CMS vs other are relevant to the choice of refusing to deal with the CMS. I think this would be financial suicide for the average hospital even if the cuts in CMS rates are enacted and not waived. Only hospitals who have few Medicare patients could afford to refuse to treat Medicare patients.

Basically, I think Foster is confusing marginal and average costs. He decides that if you lose money in an accounting sense doing something, then you have higher profits if you don’t do it. That assumes that capital costs and overhead decline proportionally to patient population. That is a plainly false assumption and the inference based on it is nonsense.