At What Price?

I want to muse on health care costs for a bit.

Let's start with an exercise. Most health services are paid for by payers falling into three broad categories: private health insurance, Medicare, and Medicaid. Together, these three account for nearly three-quarters of health consumption expenditures in this country.

But they pay very different prices for the services they buy. For a given health service, private health insurance will pay the highest price, Medicare a lower one, and Medicaid a lower price still. But let's imagine everyone had Medicare. More specifically, let's look at how much each of these three payment categories actually spent on health care in 2015 and then imagine what they would've spent if they had paid Medicare prices.

This is a little rough, but let's assume the price relativity across these three payers is: Medicare = 1.0, Medicaid = 0.67, and private health insurance = 1.6 (that last one seems plausible based on this AHIP data brief). In billions of dollars, we have:

	2015 spending, actual	2015 spending, if repriced to Medicare rates
Private insurance	$1,072	$670
Medicare	$646	$646
Medicaid	$545	$814
Total	$2,263	$2,130

So using my rough price relativity numbers, total spending on the patients these payers insure would be a little less than 6% lower if all services were reimbursed at Medicare rates (coincidentally, the average hospital margin on Medicare was almost -6% in 2014). The point of all this being the amount of revenue paid to health care providers is, loosely speaking, pretty close to what it would be if everyone currently in those three buckets was in Medicare. And that's because what you save by reducing private health insurance payments to Medicare rates, you mostly give up by increasing Medicaid payments.

So it's not completely out of line to think that insurance payments in the U.S. roughly average out to be in the neighborhood of what Medicare pays.

MedPAC, the body that advises Congress on Medicare payment policy, recently put up an interesting blog post that considered the gap between the prices Medicare and private health insurance plans pay hospitals in the U.S. In it they observed:

The literature and our analysis suggests that Medicare’s hospital payment rates are roughly 50 percent higher than hospital rates in other OECD countries (MedPAC 2014). Commercial insurers’ hospital rates in the United States are often over 100 percent above OECD rates. 

Just as Medicare rates are about 50 percent above OECD, U.S. hospitals’ input costs are also about 50 percent higher than input costs in other OECD countries after adjusting for the cost of living (International Federation of Health Plans 2013, Koechlin et al. 2010, Organisation for Economic Co-operation and Development 2012). 

That's an interesting point. Despite setting its own prices, Medicare still pays substantially more than insurers in other countries pay their hospitals. Which perhaps is obvious if we (1) accept the conclusion above that Medicare reimbursements are the average reimbursement level for American insurance, give or take a few percent, and (2) observe that the U.S. spends significantly more on health care per capita than the rest of the world.

So we (specifically Medicare) are paying hospital prices that are 50% higher than those in other countries, but the cost to hospitals of their inputs is also about 50% higher here. So what are those inputs?

Several years ago the Massachusetts Hospital Association produced a paper laying out the expenses that make up hospital costs. The data is over a decade old now and reflects the costs faced by Massachusetts hospitals, but I imagine the picture hasn't changed substantially and is more-or-less broadly applicable. It's the best source I know of that lays out the components of a hospital's costs and their relative contributions.

As you can see from the figure to the right, a full two-thirds of those input costs were labor costs: salaries, wages, and benefits.

(The story is even more tilted toward labor costs if you get out of the hospital world and go into the world of physician practices. The 2014 MGMA Cost Survey of physician practices found that total physician, total non-physician provider, and total support staff costs together comprised 85% of total practice costs!)

Previous explorations by MedPAC of cost differences in the U.S. vs. OECD countries have found that:

• Medicare pays physicians ~50% more than physician labor costs in other countries
• Nurses make 50% more in the U.S.
• Drugs and devices cost about 50% more here

Their conclusion in looking at all this is that Medicare prices are in line with costs in this country:

Given the similarity between these [price-to-cost] ratios, the relationship between payments and costs in Medicare appears similar to those in other OECD countries. In other words, Medicare rates and rates in OECD countries both reflect the relative input costs in those countries.

And if we accept my argument above that Medicare prices are more or less the American price for health care (with lower Medicaid prices roughly balancing out higher private health insurance prices), then more generally our prices--long known to be quite high by international standards--are in line with costs.

So then the question is: how much can we get those costs down? And how much should we?

In that MedPAC blog post, they're somewhat optimistic that putting pressure on providers (i.e., in terms of the prices we pay them) like hospitals will cause them to clamp down on those costs. But thinking just of hospitals, if those Massachusetts numbers are still roughly correct then drugs and devices, etc make up less than a quarter of hospital input costs. Meaning if hospitals could somehow be empowered and enticed to muscle their suppliers' prices down by 50% (which seems like a tall order), that would only close a third of the gap with the OECD nations.

Cutting costs in any serious way means substantial cuts to wages or jobs or both. (I'm not even considering the potential downstream labor impact of cutting prices in the device and medical supplies industries in half--I'm thinking here only of those jobs directly supported by hospitals).

As is often remarked upon, health care is a significant chunk of the national economy. Hospital systems are often the economic anchor of their region and it's not unusual for the largest non-governmental employer in a state or area to be a health care system. These are often solid middle class jobs, generally safe from outsourcing and the like (indeed, health care is sometimes considered one of the few "recession proof" industries--not entirely accurately but there's some truth to that).

We want to pay less for health care, but do we want to cut jobs and wages in that sector? (These seem to me to be flip sides of the same coin.) And if so, how much?

This is incidentally one of my major sources of discomfort with enthusiasm for single-payer health care: it hides this central question. Touting savings and lower prices while obscuring where those savings come from seems a bit disingenuous and shuffles past a conversation that desperately needs to be had.