The CMS Report: What it Says

There was a lot of buzz yesterday about the report from the Centers for Medicare and Medicaid Services on the impact of health care reform.  Conservative spin:  the bill is an utter failure!  Liberal spin:  It shows that the bill is great, except for the cost control part, which we can totally fix later.  Tim Noah, in a category by himself, spins a conspiracy theory about a dark plot by CMS to protect its budget from cuts.  None of these is quite right.

First, the conspiracy theory:  it's true that CMS would have a smaller budget under the House and Senate bills.  And it is true that government bureaucrats will fight vicious turf wars to prevent their budgets from being slashed.  But the part of CMS's budget that is being cut is the part that doesn't give it much power; the biggest savings come from automatic cuts to provider payments.  As far as I can tell, HR 3962 actually increases the part of the budget that bureaucrats care about, which is the money to hire staff and expand their empire.  CMS doesn't lose power, influence, or important capabilities under health care reform. It gains them.

Second, what does the CMS report tell us about health care reform, and specifically HR 3962, the bill it analyzed?  For one thing, it tells us that the exchanges, the mandates, and the subsidies are mostly a complicated sideshow.  The real action is in Medicaid.  Expanding Medicaid to 133% of the poverty line accounts for most of the decrease in the number of uninsured.  The exchanges are expected to cover only 13 million new people, or less than 5% of the population.  We're not so much restructuring the health system as making one of its sectors much bigger.

Which accounts for another important finding:  the bill is not going to control national health care expenditures.  In fact, it's going to slightly increase them.  Under current law, the CMS has a projection of 20.8% of GDP going to health care in 2019.  The new bill will bring that up to 21.1%.  That's not particularly surprising, since they think the bill will cover 34 million new people.  Unless those new people weren't planning on actually consuming any health care--in which case, why are we bothering?--spending was bound to go up.

But the underlying message of the analysis was neither "We need to do a little better on cost control in the future" nor "The Democratic bill sucks".  It was that we don't have any very good way of controlling costs--at least not within the parameters of the current US political and economic system.  The only way we know to cut our health care spending is to stop consuming so much health care.  And no one seems to want to do that.

For a while, politicians have been promising that the Fantastic Four of health care reform would save us:  comparative effectiveness research, waste and fraud, prevention and wellness, and administrative streamlining.  The CMS report indicates that those savings are pretty much utterly trivial, $2.1 billion out of a trillion dollar bill.  When savings are the size of rounding errors, there's not much point in discussing them.

Substantial savings might be generated with a version of Britain's NICE, which could use CER to set global policies on coverage.  But merely making the information available has virtually no effect.  If we can't expect heartless private insurers to deploy CER in the name of cost control, how can we hope that our government will do so in the face of the inevitable backlash from angry voters who swear that their toe surgery changed their life?

That leaves either cutting benefits, or cutting payments to providers.  Except there's really no "or".  The CMS, which is the agency in charge of implementing those cost savings, pretty much says that cutting provider payments means cutting benefits

Virtually all of the Medicare cost savings that Democrats are using to pay for their plan come from one of three things:

  • Ending "overpayments" to Medicare Advantage providers
  • Adjusting (downwards) provider payments based on productivity growth in the broader economy
  • Demanding a 50% discount from pharmaceutical companies for seniors who are in Medicare Part D's "donut hole".

Medicare Advantage "overpayments" are, in fact, payments for extra benefits. They almost have to be, because of the way that the payment mechanism is set up.  The CMS report makes it very clear that cutting out the overpayments will result in a substantial loss of benefits by current enrollees.  Now, maybe we shouldn't be providing those benefits--but cutting them is not going to be popular. It's not clear how long the benefit reduction will survive the political process.

And the productivity adjustments are even worse, says the CMS:

It is important to note that the estimated savings shown in this memorandum for one category of Medicare proposals may be unrealistic. H.R. 3962 would introduce permanent annual productivity adjustments to price updates for institutional providers (such as acute care hospitals, skilled nursing facilities, and home health agencies), using a 10-year moving average of economy-wide productivity gains. While such payment update reductions would provide a strong incentive for institutional providers to maximize efficiency, it is doubtful that many could improve their own productivity to the degree achieved by the economy at large.8

Over time, a sustained reduction in payment updates, based on productivity expectations that are difficult to attain, would cause Medicare payment rates to grow more slowly than, and in a way that was unrelated to, the providers' costs of furnishing services to beneficiaries. Thus, providers for whom Medicare constitutes a substantive portion of their business could find it difficult to remain profitable and might end their participation in the program (possibly jeopardizing access to care for beneficiaries). While this policy could be monitored over time to avoid such an outcome, so doing would likely result in significantly smaller actual savings than shown here for these provisions.

Footnote 8 goes onto note that because of a phenomenon known to economists as Baumol's Cost Disease, medical productivity doesn't improve as fast as most of the rest of the economy--basically, activities that are very labor intensive don't tend to have massive productivity gains.  That's why it still takes just about as many teachers to teach 50,000 sixth graders as it did fifty years ago.  Similarly, it still takes one person to give you a sponge bath and administer your pills.

Since productivity growth in the health care sector will presumably continue to grow more slowly than the rest of the economy, we will end up in the uncomfortable position of letting Medicare go the way of Medicaid (which many providers refuse to take because the reimbursements are so stingy) or losing the cost savings, and ripping an even bigger hole in the budget than we've already got.

Perhaps the most worrying item is tucked into the CMS's "caveats and limitations of estimates" section, which is well worth reading.  They point out that they, like most other agencies, are assuming a sort of frictionless universe in which 34 million new people demanding more health services increases the supply of health services in order to meet that demand.  That is not, notes the CMS, a very realistic assumption:

In estimating the financial impacts of H.R. 3962, we assumed that the increased demand for health care services could be met without market disruptions. In practice, supply constraints might interfere with providing the services desired by the additional 34 million insured persons. Price reactions--that is, providers successfully negotiating higher fees in response to the greater demand--could result in higher total expenditures or in some of this demand being unsatisfied. Alternatively, providers might tend to accept more patients who have private insurance (with relatively attractive payment rates) and fewer Medicaid patients, exacerbating existing access problems for the latter group. Either outcome (or a combination of both) should be considered plausible and even probable.

The latter possibility is especially likely in the case of the higher volume of Medicaid services. Despite a provision to increase payment rates for primary care to Medicare levels, most Medicaid payments would still be well below average. Therefore, it is reasonable to expect that a significant portion of the increased demand for Medicaid would not be realized.

We have not attempted to model that impact or other plausible supply and price effects, such as supplier entry and exit or cost-shifting towards private payers. A specific estimate of these potential outcomes is impracticable at this time, given the uncertainty associated with both the magnitude of these effects and the interrelationships among these market dynamics. We may incorporate such factors in future estimates, should we determine that they can be estimated with a reasonable degree of confidence. For now, we believe that consideration should be given to the potential consequences of a significant increase in demand for health care meeting a relatively fixed supply of health care providers and services.

In other words, while we are nominally increasing the number of "the insured", it's not clear we're increasing their access to health care very much.  The supply of health care services is actually pretty inelastic, because it depends on relatively scarce labor.  There's already a nursing shortage, and doctors already don't want to become GPs because the pay is mediocre, the work is routine, and the hours aren't particularly compelling.  To some extent they can be replaced by nurse practitioners--but they are neither particularly cheap, nor in endless supply.  And there's a limit to how much of our healthcare costs we can fix by replacing current workers with less skilled labor.

When you increase the demand for something without increasing the supply, you either get price increases, or shortages.  Neither is what the authors are promising for their bills.

(Yes, yes, I know what you're about to say . . . end the AMA cartel's artificial restrictions on entry into the medical profession!  That's a different post, but here's the short version:  the constraint on the supply of doctors isn't the medical school slots, but the residency slots.  And we're already importing a substantial number of doctors to fill our family practice slots, because about a third of them go unfilled during the "match".  This does not suggest that there are hordes of eager potential doctors clamoring for a crack at family practice.  There's a lot of demand for specialist slots.  But creating more cardiac surgeons will not put much downward pressure on healthcare costs.)

But this is not an indictment of the bill's ability to control costs, as of the ability of any bill to control costs.  Controlling costs means consuming less health care.  There is no magic pot of money waiting to be painlessly seized from some undeserving wretch, preferably one that voters already hate.  The only way we are going to cut costs is by cutting someone's benefits.

Perhaps we'd be better off, in some metaphysical sense, if we did.  But no one wants to.  That's why politicians are speaking euphemistically about Medicare Advantage "overpayments" and frantically promising that no way, no how, will they damage anyone's Medicare benefits.  The CMS report says what Doug Elmendorf, the head of the CBO, hinted at in his letters to Congress:  cost control will be painful, and Congress will almost certainly undo it.

That means that whatever bill we pass will undoubtedly blast a giant hole in the budget; conservatives are right that this bill is effectively grossly fiscally irresponsible, no matter how "deficit neutral" its stated intentions are.  (But I would say that, wouldn't I?)  On the other hand, it also means that Medicare is probably going to blow a giant hole in the deficit anyway, with the able help of political figures like Michael Steele.  That's hardly reason to crow.