The Lone Economist Does the Urban Institute Study

I was all set to discuss the estimated cost of Medicare Prime (MP), my alternative proposal that would achieve many of the same goals as Medicare For All (MFA), but at a fraction of the cost to taxpayers. Then I looked at the detailed report of the $32 trillion 10-year cost estimate of MFA by the Urban Institute and Commonwealth Fund that has received so much press coverage lately.

Before I give my unsparing assessment, let me explain the predicament I believe the authors of this study found themselves in. The Urban Institute and Commonwealth Fund are associated with progressive causes and consequently also the Democratic Party. Two major presidential candidates for the Democratic nomination publicly proposed MFA, a policy so radical that it is difficult for people who know a lot about the healthcare industry and economics to take seriously.  But unlike me, a masked vigilante with a persecution complex, they can’t publicly point a finger at these candidates and burst out laughing. They have revenue streams to maintain, future employment opportunities to protect.   

So how does one convey to an impassioned base and credulous media just how over-the-top this proposal is without committing career suicide? Their strategy appears to have been to produce a comparative study of MFA and several alternative proposals rather than an unbiased estimate of the cost of MFA. Without the constraint of statistical unbiasedness, they were able to make ridiculously unrealistic assumptions about the cost savings of MFA. Yet, even with these assumptions, they still predicted a gigantic increase in federal expenditures.  It’s like giving a poor student an easy test and when they still fail, there is little doubt about the fairness of the process.

If you believe this to be a negative assessment, then you are mistaken. I am honestly very favorably impressed with this report. It is a thorough and transparent analysis of the relative costs and benefits of these proposals. To their credit, the authors chose the least bad of several bad options.

If the media interpreted their findings more literally than intended, that is not their fault. Henri Theil famously said, “Models are to be used, not believed”. The usefulness of their model is that it provides estimates of eight different health insurance reform proposals that rely on the same data and set of explicit assumptions. Consequently, the estimated costs are good relative measures, even if they are lousy absolute ones.

So, what evidence do I have that this was their strategy? It’s not from any direct communication with the authors. If I am right, they couldn’t give an honest answer to a direct question. I did correspond with the lead author, but that was about some technical issues concerning the data they used.

I’m basing my theory on the truly indefensible assumptions the authors adopt and on a statement the lead author gave to a reporter, “We felt we were making pretty optimistic, aggressive assumptions”. If you were trying for an unbiased estimate, that is not how you would do it.

But strange as it might sound, if the authors had made fewer outlandish assumptions, their cost estimate would have been even more biased. Let me explain.

What is “cost”?

You might think the answer to this question is obvious, but it isn’t. I’ll explain by way of parable because it succinctly conveys a complicated concept and gives me another opportunity to insult politicians.

A low-level politician who aspires for higher office drives a car for 20 years until it simply doesn’t run anymore. The most a junk yard will pay for it is $500, so he parks it in front of his house with a for-sale sign that says “$20,000”. After several weeks with no offers, he receives a notice from the city ordering him to remove the unsightly piece of junk from his front yard. Seeing an opportunity, he gives the car to a homeless person. The next day a headline appears in the newspaper “Politician Gives Homeless Person $20,000 Car!”

If you find this story absurd, then you know how I feel about healthcare accounting practices (and some newspaper reporters). Every year, almost all hospitals and many physicians essentially do what the fictional politician in my story does. They provide uncompensated healthcare services to poor people because they are professionally and often legally required to do so and then wildly exaggerate the dollar value of this care.

As I have detailed in a previous post, healthcare costs are initially reported as “undiscounted charges”. The average hospital’s markup from costs to charges is over three hundred percent. Some hospitals markup over 1,000 percent. Neither private nor public insurance providers pay the undiscounted charges, but uninsured patients are supposed to even though they rarely can. Nevertheless, hospitals that spend, for example, $500,000 treating indigent patients often claim that they provided charitable care worth several million dollars.

So, what was the cost of the car given to the homeless person? The $500 the junkyard offered to pay for it or the $20,000 the politician was asking for it? I think it is obvious that the $500 figure is more accurate.

Accounting vs. Economics

Economics is the science of how societies use scarce resources to produce valuable goods and services and distribute them among different people. Accounting is the science of measuring, processing, and communicating economic data. Although the two are intricately linked, economics is mainly concerned with theories while accounting is firmly rooted in the real world.

Consequently, “cost” can mean one thing to an economist and something different to an accountant. To an economist, “cost” is the foregone benefit of consuming a good or service. We call it “opportunity cost”. To an accountant, “cost” is the number of dollars that change hands in a transaction.

Normally the difference between these two definitions is small. But even when it isn’t, it’s considered unavoidable. Economists must use the data they have, not the data they wish they had, is the old saying.

I contend that this bromide should not and need not be followed when debating healthcare policy. Undiscounted charges are a poor proxy for the opportunity cost of healthcare. Almost nobody pays them. Why we continue to use them for measuring uncompensated healthcare costs eludes me.

The Urban Institute Assumptions

The dichotomy between accounting and economics plays a key role in the Urban Institute’s choice of assumptions. I’m not going to nitpick every assumption they made, but there are two, in particular, that deserve some discussion.

The study assumes that MFA can be achieved by lowering the price of healthcare and increasing the quantity supplied of healthcare at the same time. Figure 19.1 illustrates these two assumptions. The supply curve would have to shift to the right, leading to a lower equilibrium price and a higher equilibrium quantity.

Figure 19.1 Increase in Supply Leads to Higher Quantity and Lower Price

Specifically, the authors assume that MFA would pay all physicians the Medicare rate for physician services and the Medicare rate plus 15% for all hospital services. To put this into perspective, remember that private insurers pay approximately 50% more than Medicare for a physician office visit and 100% more for a hospital stay. So, the authors assume that physicians will accept a 33% cut in the average price of office visits by patients under age 65 and hospitals will accept a 43% cut in the average price of hospital stays from the same patients and their reaction will be to increase the number of patients they treat by more than 20%.

Separately, both assumptions are absurd on their face. Yet, had the authors not assumed supply would respond this way, their estimate would have been even less believable.

Figure 19.2 illustrates what would happen if we didn’t assume MFA would cause an increase in the supply of healthcare. By imposing a price ceiling well below the equilibrium price, the quantity supplied would decrease while the quantity demanded would increase. The result would be excess demand, i.e. a shortage.

Figure 19.2 Price Ceiling Causes Excess Demand

The combination of a decrease in price and a decrease in quantity supplied would result in a large decrease in the number of dollars spent on healthcare. Therefore, using the accounting definition of cost, MFA would decrease the total cost of healthcare and the increase in federal expenditures would be much less than $32 trillion.

The opportunity cost of MFA, a much more accurate measure of the economic cost, would be far greater than 32$ trillion, but that fact would be lost in the arcane definitions of “cost”. So, rather than make their underestimate even greater than it already was, the authors chose the lesser of two evils.

So, let’s see. I’ve insulted politicians, hospital accountants and the media, but hopefully not the authors of the Urban Institute study. Not bad for one post. Next time I’ll talk about the relative cost of Medicare Prime.

Published by TheLoneEconomist

I am a PhD economist who studies just about anything and proudly specializes in nothing.

2 thoughts on “The Lone Economist Does the Urban Institute Study

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: