It’s the Perspective, Stupid

Medicare Prime (MP) is my proposal for a health insurance system that would avoid many of the problems created by our current one. One of its proposed features is the classification of healthcare providers into two camps: in-network and out-of-network. The benefits of this bifurcation of providers deserve further explanation.

As explained in a previous post, providers charge patients the same amount of money for the same type of service regardless of the type of insurance they have or their income level. Only after the bill has been calculated are these charges discounted according to the patient’s insurance status. Public insurers, such as Medicare and Medicaid, get the largest discounts (61% and 67% on averagea, respectively), while private insurers receive only a 42% discount on averagea. Uninsured patients officially get no discount at all, but often default on their obligations through personal bankruptcy and effectively get a 47% discount on averagea.

Under MP, in-network providers would be paid the Medicare rate for services rendered to patients that have surpassed their MP deductibles.  In exchange, these providers agree not to charge patients any additional money for services paid by MP. In other words, they will not be allowed to balance bill patients who have satisfied their deductible.

Out-of-network providers, however, are a different matter. They will not be a party to any contract with the government and therefore will be free to balance bill their patients. Given this lack of a quid pro quo, one might believe that MP should not pay anything for services from out-of-network providers. But I would disagree with this sentiment. There are several good reasons that out-of-network providers should be paid a partial Medicare payment despite their freedom to balance bill patients.

The Case for a Two-Tier Healthcare System

Creation of a two-tier healthcare system might seem contrary to egalitarian ideals, but such a system effectively already exists in the US and there are several reasons why this makes economic sense. To see how this might be correct, we need to consider the uniqueness of the American healthcare system.

But first, a joke.

Two balloonists become lost while sailing through a wind-swept fog. Seeing a pedestrian below they call out, “where are we?”. The man below yells back, “you’re in a balloon”. One balloonist says to the other, “he must be a mathematical economist”. “How can you tell?” asks the other. To which the first balloonist replies, “his answer was perfectly accurate and completely irrelevant”.

“What does this joke have to do with our healthcare system?” you might wonder. One of the advantages of being a masked vigilante is the freedom to insult whole professions with impunity, even my own. It’s the new year and I haven’t insulted anybody in quite a while.

But the main point is to highlight a fact frequently cited by politicians and economists: the US spends much more money per capita on healthcare than any other country while suffering a high death rate. While mathematically correct, the insinuation that this is somehow a negative feature of our healthcare system is more a product of subjective perspective than some careful objective analysis. Like the pedestrian’s answer, it is perfectly accurate and completely irrelevant.

Public health is the science of protecting and improving the health of people and their communities. Its perspective is exclusively global as opposed to individual. So, when considering whether to approve payment for a new cancer drug, for example, a typical public health expert would compare the number of lives saved for each dollar spent to that of other cancer drugs and even treatments for other diseases.

Suppose the new cancer drug saves one year of life for every hundred thousand dollars spent. Is that too much to spend? The answer to that question depends more on one’s perspective than it does on the drug’s relative cost and effectiveness. Anti-biotics, for instance, are very effective fighting deadly infections and very cheap. They might save 1,000 life-years for every hundred thousand dollars spent. From a global perspective, it makes no sense to spend additional money on the new cancer drug as long as spending that same additional money on anti-biotics saves more life-years.

In economic terms, we would say that any expenditure on the new cancer drug is past its point of diminishing marginal returns. From a global perspective, expenditure on the new cancer drug would cost more lives than it saved. This is why denial of coverage for expensive drugs by single-payer healthcare systems is quite common. Since these systems rely exclusively on public money, they take an exclusively global perspective, the same perspective taken by most public health experts.

If you ever completed a high-school course on American literature, you should be well-acquainted with the characteristic that practically defines the American psyche, individualism. It should come as no surprise then that the US rejects the global perspective of public health more than any other wealthy country.

From an individual perspective, that hundred-thousand-dollar expenditure on the new cancer drug makes perfect sense. If you have that much money to spend or more likely, that much insurance coverage, you would spend it without hesitation. The fact that you could save more lives if you would only sacrifice yours is not persuasive to most Americans.

Suppose you were a beneficent ruler with absolute power. If an advisor proposed a policy that would cost two lives for every three lives it saved, would you implement it? If you took a global perspective and you valued all human lives equally, you probably would. How could you not accept a policy that lowered the death rate by 33%? That would be a great policy — unless you are one of the people sacrificed. Then it wouldn’t be so appealing. 

So, from a global perspective the fact that the US spends more money per capita on healthcare than any other country, while suffering one of the highest death rates is clearly a bad result. But from an individual perspective, it is indicative of the ability of many Americans to marshal vast resources to fight deadly diseases. The fact that we could save more lives while spending less money if we didn’t care exactly whose lives we saved is irrelevant.

Single-payer systems don’t allow different levels of healthcare intensity. It’s one-size-fits-all. There are not enough resources, i.e. doctors, hospitals, drugs, etc., to treat everyone at the same high-intensity level that many Americans can afford today. Any politician or economist who implies otherwise is either misinformed or lying.

The two-tier system of MP would provide a minimally-acceptable intensity of healthcare for those receiving taxpayer-financed health insurance while allowing those households that can afford it to purchase more intensive healthcare with their own money.

The Case for MP Payments to Out-of-Network Providers

To fully appreciate why out-of-network providers should be paid a partial MP payment, think back to why the ACA exchange market was created. The all-or-nothing nature of public insurance in the US forces private insurance premiums upward to cover the capital costs of treating publicly insured patients. Before the ACA, these high premiums did not discriminate between wealthy households and households that almost qualified for Medicaid. The elevated premiums were especially bad for people with pre-existing medical conditions.

The ACA tackles this problem by subsidizing the private insurance premiums of middle-income households. As we illustrated in our last post, subsidizing premiums in this way forces the equilibrium premium upward, partially off-setting the benefits of the subsidies and even making the situation worse for households with incomes greater than 400% of the FPL.

The ACA also restricts private insurers’ ability to charge higher premiums to people with pre-existing conditions and mandates that all citizens obtain health insurance. In the political debates that preceded its passage, the individual mandate was advertised as a fair quid pro quo for the premium restrictions imposed on insurers. That justification wasn’t false, but it wasn’t the main reason for the mandate either. The main reason was to stabilize the health insurance market. Without it, a shock to the system, in the form of an economic recession for example, would lead to its collapse.

MP would use a different tack to benefit middle-income households. Although the deductibles for middle-income households would not be zero, they would be much smaller than the deductibles for high-income households. When the patient uses an in-network provider, she and her private insurer would need to only cover the MP deductible. That would be a clear benefit to both the patient and her private insurer, if she had one.

But even when the patient uses an out-of-network provider, she would still benefit because a partial MP payment to the provider would lower her and her private insurer’s liability. Lower liability for the private insurer would lead to lower premiums for the patient.

For services provided by in-network providers, there would be a quid pro quo between MP and those same providers: full Medicare payments in exchange for no balance billing. For services provided by out-of-network providers, the quid pro quo would be between MP and private insurers: partial Medicare payments in exchange for not charging higher premiums to people with pre-existing medical conditions.

Unlike the ACA, under MP there would be no need for an individual mandate. As the primary insurer, MP protects both patient and private insurer. Rather than setup an inherently unstable health insurance market (like the ACA), MP would stabilize it.

In my next post, I’ll discuss how much MP would cost relative to our current system.


aSource: The Medical Expenditure Panel Survey, 2017

How to Lower Health Insurance Premiums

Continuing our outline of Medicare Prime (MP), an “optimal” public health insurance system based on economic principles and data-driven research, we explore one of its benefits for middle-income households, i.e. lower premiums.

Why Insurance Premiums Are So High

In an earlier post I explained how Medicare and Medicaid (M&M) affect the private insurance market. M&M competes with private insurers for healthcare providers, e.g. hospitals and doctors. Using its superior market power, M&M pays only the incremental operational costs of treating its enrollees. This forces providers to charge private insurers much higher rates than they otherwise would in order to cover their capital costs. On average, private insurers pay twice as much as M&M for the same services, excluding prescribed drugs.  These high rates are then passed along to households in the form of higher health insurance premiums.

Figure 16.1 illustrates this phenomenon. The supply curve for private insurance shifts upward. This is a decrease in supply. The result is an increase in the equilibrium price at a lower equilibrium quantity. Or in other words, public insurance partially crowds out private insurance and is one of the reasons the price of private insurance is so high.

Figure 16.1 The Market for Private Insurance

The Affordable Care Act (ACA) subsidies also affect premiums, but in a different way. Instead of decreasing the supply of private insurance, they increase the demand for it. Figure 16.2 shows the demand curve shifting upward by the amount of the subsidy. This causes the price (i.e. premium) to increase from P1 to P2. Households that qualify for the subsidy pay P2 minus the subsidy and increase their quantity from Q1 to Q2. Households that do not qualify for the subsidy pay P2 and decrease their quantity to Q3.  This is another example of crowding out and another reason why private insurance premiums are so high for households that make too much money to qualify for a premium subsidy.

Figure 16.2 Effect of ACA Subsidies on Private Insurance Market

Who Would Benefit from Medicare Prime?

In a previous post I stressed that for the rich and the poor, MP would be little different from the ACA. Households below 138% of the Federal Poverty Level (FPL) would still have 100% of their healthcare expenses covered by public insurance if they used only in-network providers. Affluent households would not directly benefit from this system. They would still need to pay their healthcare expenses through private insurance or out-of-pocket. Nothing new there.

The main difference in insurance coverage would be for households that fall in between, i.e. the middle-income households. These households earn too much money to have first-dollar coverage (i.e. zero deductibles) from MP. But that doesn’t mean they would not benefit from it.

A non-zero but relatively small MP deductible would mean that any potential private insurer would incur lower risk than would otherwise exist and would result in a lower private insurance premium. For many middle-income households, MP would be, in effect, a high-deductible, catastrophic coverage plan. But since MP is primary, that coverage protects both the household and its private insurer.

In the next post we will attempt to quantify the benefits of MP.

Medicare Prime

Our quest for an optimal health insurance system is starting to come into focus. It would have several features in common with Medicare For All (MFA), Obamacare, and The Public Option (TPO), but with a few critical differences.

Like MFA, it would be financed entirely through taxation, not premiums. It would recognize a minimally-acceptable level of healthcare as a human right. Involuntary exposure to large uninsured healthcare expenditures would be a thing of the past. Like Obamacare, it would recognize and even promote the vital contribution private insurance makes to financial security. And like TPO, it would give middle-income households a wider array of affordable health insurance choices.

What it would not do is cause unprecedented levels of moral hazard (like MFA), perpetual adverse selection (like Obamacare), or the crowding out of private insurance (like TPO). It wouldn’t rely on an individual mandate either. Or necessarily cost more in taxes than the current system.

I’ve referred to this optimal health insurance system as a free public option, but as an economist, my heart skips a beat every time I see the word “free” and it’s not really a public option anyway. The option is whether to buy private insurance to go along with the public insurance every household would be entitled to.

So, I need to come up with a better name. In my last post I explained that for this system to work, public insurance and private insurance need to change from substitutes into complements and that this could only be achieved by making public insurance primary and allowing some providers to balance bill.

By the title of this posting you might have guessed that the name I have selected is Medicare Prime. It would be a Federal program that uses Medicare payment levels to replace Medicaid, the Children’s Health Insurance Program (CHIP), and Obamacare, but would not replace Medicare Parts A, B, C and D. It would only affect people who don’t already qualify for traditional Medicare. In other words, people still in the prime of their lives. And more to the point, it would be primary. Hence the name.

Can Public and Private Health Insurance be Complimentary?

Two substitute goods each satisfy the same need while not enhancing the benefit of consuming the other. For example, Coca-Cola and Pepsi both satisfy the need for a cola-flavored soft drink. Most people either consume one or the other. Few people like to combine the two. And because they are substitutes, their suppliers are natural competitors.

In contrast, complements are normally consumed together — think milk and cereal or bread and mayonnaise. Consuming one enhances the benefit of consuming the other. The suppliers of complements cooperate rather than compete.

Under current government policy, public and private insurance are substitutes. Consequently, they compete. Competition between private insurers is beneficial to consumers, but competition between public and private insurers is not. It has contributed to high private insurance premiums and the coverage gap.

But I contend that this harmful competition is a result of bad government policies rather than the natural order of things. Public and private insurance could be complements if we only chose the right policies.

The Importance of Primacy

Some people are covered by more than one health insurer.  As profit-maximizing organizations, private insurers attempt to limit their financial liability by stipulating that all other insurers are primary.  In other words, only after coverage by the primary insurer is exhausted is the secondary insurer liable for any healthcare expenses.

When an individual is covered by two different private insurers, it is a matter for a court of law to determine which insurer is primary. Not so, however, when a patient is insured by a private insurer and a public insurer. Federal law clearly establishes that the private insurer is primary in such cases, except in rare instances. Consequently, if an individual is enrolled in Medicare because she is over age 65 and enrolled in a private insurance plan through her employer, the private plan is liable for almost all her medical bills. She has little incentive to pay a premium for the private insurance coverage and the private insurer has little market power to charge a premium high enough to cover its costs.

The Balance Billing Ban

The government prohibits balance billing by providers of healthcare to Medicare and Medicaid enrollees. When a healthcare provider accepts payment from Medicare or Medicaid, it must agree to forego the collection of any additional fees from the patient except normal cost-sharing like deductibles and copays. There is a partial exception to this prohibition in certain cases, but in general it holds. This guarantees that public insurance enrollees have access to healthcare, regardless of income.

Like insurance primacy, Medicare and Medicaid are following common health insurance industry practice by not allowing balance billing.  The typical managed care organization (MCO) contracts with a subset of available providers to form a network. These “in-network” providers agree to accept lower payments from the MCO with no balance billing in exchange for an increased volume of patients. They are also subject to practice constraints imposed by the MCO.

Since the MCO has no contract with “out-of-network” providers, it cannot stop them from balance billing MCO enrollees who choose to be treated by them. This is a strong incentive for enrollees to choose only in-network providers. Enrollees are further disincentivized from using out-of-network providers by bearing a greater share of costs.

They’re Substitutes Because They’re Not Substitutable … Huh?

Although private insurance primacy is a government policy meant to prevent the substitution of taxpayer funds for private funds, ironically this makes private and public insurance substitutes. It is commercially disadvantageous to offer private health insurance to individuals eligible for public insurance under the condition of private insurance primacy. The only exceptions are if the insurance is procured through a small employer (i.e. less than 20 employees) or for coverage of expenditures Medicare does not cover (i.e. Medi-gap insurance).

Also, the balance billing ban prevents public and private insurance from being complements which I contend is their preferred state. There is a reason that almost no one over the age of 65 has private insurance. And there won’t be any private insurance for people under 65 if either Medicare For All (MFA) or The Public Option (TPO) are adopted.

The way to make private and public insurance complementary is to stipulate that public insurance is primary and to allow balance billing by providers not in the public insurance network. For those long-acquainted with the health insurance industry, this first part will sound heretical while the second will sound somewhat familiar, if out of place.

A Variable Deductible

If public insurance was free and primary, wouldn’t providers always be paid by the government thus crowding out private insurance and forcing taxpayers to bear the full cost of the healthcare system, just like MFA?

Not if a household’s public insurance entitlement decreased as household income increased. For example, the ACA established a threshold for Medicaid eligibility at 138% of the Federal Poverty Level (FPL). This threshold applies only to those states that expanded Medicaid, however. For the other states, the threshold for Medicaid eligibility is somewhat lower. The free public option’s deductible could be set to zero for all families below this threshold and it would increase with household income above that initial level.

The result would be that the same households that qualified for Medicaid in the current system would have 100% public insurance coverage in the new system. There would be very little difference besides having a consistent threshold across the country instead of the highly variable and arbitrary one we have now.

In contrast, high-income households would have very high deductibles, perhaps a hundred thousand dollars or more. They would almost never be able to rely on public insurance to pay part of their healthcare bills. Instead, they would need to cover all their healthcare expenditures via private insurance or out-of-pocket. This too is little different from the current situation.

In-Network Providers

Wouldn’t balance billing allow providers to bankrupt patients without private insurance?

Not if the free public option was operated like an MCO.  Providers that are willing to accept the relatively low payment levels of the free public option in full (i.e. without balance billing) in exchange for a steady volume of patients would be in-network. Out-of-network providers would be paid only a fraction of the rate paid to in-network providers but would be allowed to balance bill.  

If you believe that this would create a two-tier healthcare system, one for the poor and another for the rest, then you would be right, but a little late. Such a system already exists.  It is an underappreciated fact by those not intimately familiar with the American healthcare system that there is a wide gamut of healthcare providers.

For example, physicians range from foreign-trained immigrants who barely scrape by to Ivy League graduates who cater only to the upper class. Almost all the high-end physicians avoid treating Medicaid patients and many of them avoid treating Medicare patients as well. There is no trade off to accepting low public insurance payments the way that there is for in-network MCO providers. This has led to an appalling lack of access to physicians for Medicaid enrollees in some states.

Unlike high-end physicians, high-end hospitals cannot afford to reject Medicaid enrollees entirely, but many actively minimize their exposure to this population, nevertheless. I’ve consulted on a Certificate of Need application where a for-profit hospital intentionally chose a location because it had very few Medicaid enrollees nearby.

This two-tier system is a major reason why I think MFA and TPO would be so disastrous for the Democratic Party if either were adopted. Single-payer systems are one-size-fits-all.  There is no room for variation in quality or intensity. Many people who were satisfied with their private health insurance will be apoplectic when they realize their healthcare choices have been so curtailed. Many of the high-end providers would simply exit the market. The resulting shortages would be met with outrage.

Balance billing by out-of-network providers would allow middle-income households to purchase relatively inexpensive private insurance that would afford them a choice from a wider array of providers, both in-network and out-of-network.

More about this next time.

How to Close the Coverage Gap

I’ve written several posts explaining the twin problems of health insurance: adverse selection and moral hazard.  Rather than solve these twin problems, the two main health insurance proposals, Medicare For All (MFA) and The Public Option (TPO), would exacerbate them.

MFA would solve adverse selection by providing health insurance to everyone without premiums or out-of-pocket expenditures, but would cause run away moral hazard, e.g. long wait times, wasted resources on ineffective treatments, high taxes, etc.). It would also crowd out private insurance, thus eliminating one’s choice between a relatively more expensive provider and a less expensive one.

With its reliance on cost-sharing, TPO would cause less moral hazard, but because it would charge a premium, there would still be adverse selection, i.e. the coverage gap and the need for an individual mandate. And like MFA, it would crowd out private insurance, albeit more slowly.

So how do we close the coverage gap (i.e. adverse selection) without wasting resources (i.e. moral hazard) and without eliminating the private insurance market (i.e. crowding out)?

In brief, these three goals can be achieved respectively by: (1) offering a “free” public option, (2) providing it in the form of a managed care organization (MCO), and (3) making private and public insurance complements rather than substitutes. And here’s the kicker: do all this without increasing the burden on taxpayers from what it is now.

You might be skeptical that a “free” public option is even feasible. [I put the word “free” in quotes because as an economist I am duty-bound to point out that nothing is truly free, even public health insurance that does not charge a premium. The cost of this new public insurance will be borne by taxpayers and private insurance consumers, just like the current one]. You might ask, how would it leave room for private insurance and not be ruinously expensive?

I believe this new system is feasible for a number of reasons. First, the current system is so expensive and inefficient that we have a lot to play with here. In 2017, government-financed third-party payments to providers of healthcare to the under-65 population totaled $460 billion. That was 40% of all healthcare expenditures for that age group.

Second, calling it a public option is a bit misleading. Since there is no premium, everyone qualifies for it by default. The public “option” would actually be an entitlement that depends on one’s household income. You have it whether you want it or not, whether you even know it exists or not. That is the only way to get universal health insurance. On this point, I agree with the proponents of MFA. The real option is whether to purchase private insurance or not.

Lastly, this new system is feasible because it and the current one have more similarities than differences. Like the current system, nearly 100% of healthcare expenditures for households with low annual incomes will be paid for by the government. And also like the current system, nearly 100% of healthcare expenditures for households with average annual incomes or higher will not be paid for by the government.

For the households in between, this new public insurance system would amount to a high deductible, catastrophic coverage insurance policy. The government would only pay for unusually high healthcare expenditures. The rest would be paid for by private insurance or the patient herself.

As I have explained in earlier posts, the ACA premium subsidies have increased the demand for private insurance (i.e. shifted the demand curve to the right). This has pushed premium rates higher than they would have been without the subsidies. Private premium rates are already elevated due to Medicare and Medicaid refusing to pay providers the capital costs of providing healthcare.  So, families without access to group insurance and that make too much income to qualify for a premium subsidy are forced to pay exceptionally high premiums.

As I will explain in future posts, this new system should lead to decreased private insurance premiums for these middle-income households and also stabilize the private insurance market. The ACA has setup an unstable system that is likely to collapse during a future recession.

Another question you might ask is, if this new system is feasible, has it been proposed before and, if not, why? I’m not aware of this kind of public health insurance being proposed before. As to why, my only guess is that it contradicts long-held conventions of the insurance industry. Some of it might even seem counterintuitive.

Why this new system is feasible and even efficient hinges on converting private and public insurance from substitutes into compliments. I’ll explain that in my next posting.

The Cost of Obamacare

Private health insurance alone can never provide protection for all Americans. People with pre-existing medical conditions (including old age) will not be able to afford the premiums charged to them and healthy, risk-neutral people will find the purchase of private insurance too much of an unfavorable bet.

To close this coverage gap, many government-financed solutions have been employed or proposed.  Obamacare, with its expansion of Medicaid and subsidized private health insurance, closed part of the gap, but has not achieved 100% coverage.  Several Public Options have been proposed to remedy this shortfall.

Medicare For All (MFA) has been proposed to not only replace Obamacare, but all private health insurance. One would imagine that MFA would be much more expensive than Obamacare, but it would at least close the coverage gap, something Obamacare failed to do completely.

I have written previously that the financial stability of the health insurance exchange created by Obamacare is highly suspect. The individual mandate is too weak, even when it is being enforced, to stop the collapse of the exchange when there is a shock to the market.

So, I predict Obamacare will have to be replaced eventually. This makes its cost and the costs of possible replacements or additional programs relevant to the discussion of how best to close the coverage gap.

So, what is the cost of this new program compared to other health insurance programs?

Table 2. Healthcare Expenditures by Coverage Type, 2017

*Adjusted for people covered by more than one insurer.

Table 2. shows my estimates based on the 2017 Medical Expenditure Panel Survey (MEPS). The number of enrollees is adjusted for people with more than one type of coverage. So, for example, if half of a person’s healthcare expenses were paid for by Medicare and the other half by Medicaid, that person counted as one half person enrolled in Medicare and one half in Medicaid.

The ACA “caused” via Medicaid eligibility expansion or the purchase of private insurance (i.e. premium subsidization and the individual mandate) 23 million previously uninsured people to be insured at an average cost of $3,180 per person.

The cost per person of the ACA compares favorably with that of Medicaid, Medicare and private insurance, but this is not because of some cost-saving feature of the ACA. It is because many of the previously uninsured are relatively young and healthy and have fewer healthcare needs than older and sicker people.

The real test of the efficiency of the ACA (from a taxpayer’s perspective) is found in the Average Discount column. Approximately half of the previously uninsured who are now insured because of the ACA are new Medicaid enrollees and the other half are privately insured. As you can see, Medicare and Medicaid enrollees receive the largest average discounts from billed changes at 65.6% and 66.6% respectively. Private enrollees receive the smallest discounts (48.2%). Consequently, the ACA population is in between (56%).

This 10% average discount difference between the ACA and Medicare amounts to $720 per person ($16 billion in total) that taxpayers and the newly insured are forced to pay for. This amounts to a large transfer of wealth from taxpayers and the previously uninsured to healthcare providers, especially hospitals. As I have written before, the biggest beneficiaries of Obamacare are healthcare providers, not the uninsured and definitely not taxpayers.

What’s Missing?

These costs only include expenditures by payers, i.e. insurers, and patients. They don’t include several other costs of providing health insurance. The MEPS doesn’t include the cost of premiums paid by enrollees nor the administrative costs incurred by insurers. I could have included an estimate of these extra costs, but I don’t think that would add anything to the accuracy of the comparisons.

In the long run, the premiums received by insurers equal the payments to healthcare providers and the costs of administration. These include costs such as claim validation and payments to investors. For private insurers, these costs typically range from 6% to 10%. For public insurers, however, they are much less.

But as I have written before, our accounting systems don’t accurately capture the costs of providing public insurance. Private insurers must attract voluntary capital investment. In other words, they have to pay for it. Governments don’t have to voluntarily attract capital. They can appropriate capital through involuntary taxation. But taxation is not free. All taxes come with an added cost, called deadweight loss.

So, it is at least plausible and, in my opinion, very likely that the administrative costs of private insurance are balanced out and possibly even eclipsed by the deadweight losses caused by public insurance. Otherwise, why stop at public health insurance? Why not have the government pay for all food consumed? A lack of food can be as deadly as a lack of healthcare. Why not clothes and housing and any other necessity?

There is nothing special about healthcare that makes it particularly well-suited for 100% public provision. National defense, police protection and roads are non-excludable. There exists a free-rider problem that prevents their efficient provision by private markets. Yet, healthcare is completely excludable.  If you don’t pay for it, you don’t get it. There is no free-rider problem there.

Public education creates a large positive externality. A literate populace is necessary for a well-functioning democracy and equality of opportunity is the hallmark of a just society. However, healthcare does not create more positive externalities than most other privately provided goods and services, except for protection from communicable diseases via vaccination.

This is the core of the difference between centrally-planned (aka socialist) and market-driven (aka capitalist) economies.  Who is more likely to efficiently and fairly invest our scarce capital: the government or private investors? The Lone Economist clearly prefers the latter in most instances.

A Brief History of Public Health Insurance in the US

Any discussion of the health insurance policy debate will benefit from some historical context. Prior to the 1960’s, the U.S. government had very little impact on the provision of health insurance. Virtually all health insurance was provided through private companies.

For a private health insurance market to thrive, two conditions must exist.  The first condition is that a high degree of uncertainty must exist for both parties, the insurer and the insured. If one side of the transaction has a better understanding of the amount of risk, then adverse selection will occur.

The second condition is just as necessary but gets much less attention in public policy debates. The insured must be risk averse. This means that they are willing to pay additional money for the certainty of paying only the average cost of care. For example, suppose the average cost of caring for a group of people is $1,000 per month per person. If they are sufficiently risk averse, each person will be happy to pay a $1,000 monthly premium plus a little extra for the certainty of not having to pay a lot more than $1,000 if they incur medical bills. The little extra is the normal profit the private insurance company needs to receive in order to stay in business.

These two conditions together make it difficult for a purely private health insurance system to cover 100% of the population. People who know they are more likely to incur large health bills will pay the premium for health insurance and people who are relatively healthy and are not particularly risk averse with decline. The insurer will raise premiums to avoid losses. The end result is that premiums for people with pre-existing conditions including the elderly will be too high to afford and many healthy, risk-neutral people will choose to be uninsured.

Medicare Part A and B and Medicaid

This was, in fact, the situation in the 1960’s when Medicare Parts A (hospitalization) and B (physician services) were created to provide insurance for the elderly, disabled and people with kidney failure. Medicaid was created to provide health insurance for the poor.

These programs initially compensated hospitals and physicians on a per diem and percentage of billed charges basis. Providers quickly learned that by increasing their charge rates and hospital lengths of stay they could increase their profits. Consequently, charge rates skyrocketed. This led to a dramatic increase in the income of physicians and hospitals. So many new hospitals were constructed to cash in on this government-financed bonanza that an enormous number of unused hospital beds existed and a process to limit hospital construction (i.e. Certificate of Need) was created.

In the 1980’s Medicare Parts A and B and Medicaid reformed their provider compensation designs. For example, Medicare Part A adopted a “prospective” payment system. Instead of paying 80% of billed charges for a hospital stay, Medicare paid a fixed amount of money for each stay. The amount was determined by the diagnostic and procedure codes assigned by the hospital. This eliminated the incentive to keep the patient in the hospital longer than medically necessary. The excess number of hospital beds declined steadily for the following 20 years.

To compensate physicians, Medicare Part B now uses the Resource-Based Relative Value System (RBRVS). Like the prospective payment system for hospitals, this is a code-based system which assigns a dollar value for each Current Procedure Technology (CPT) code assigned to an office visit or surgical procedure.

Although these code-based compensation systems solved many perverse incentive problems of the old charges-based system, they have their own incentive problems. Upcoding, the illegal practice of systematically assigning more expensive codes than justified, was attempted by several hospital systems and physicians. Tenet, a large for-profit hospital chain, was found guilty of gaming the prospective payment system by rapidly increasing its cost-to-charges ratios on a yearly basis. They were assessed a large fine by Medicare and the system was reformed to prevent this type of fraud from happening again. But perhaps the biggest perverse incentive created by the prospective payment system is to encourage the early discharge of patients so that they can be readmitted a day later, thus doubling the hospital’s payment.

One thing in common that the provider compensation systems of Medicare A and B and Medicaid have is that they are calibrated to pay only the marginal operational costs of providing care. In other words, they do not cover providers’ capital costs. To cover capitals costs, providers must charge much higher rates to private insurers.

This cross subsidization is a type of indirect tax discussed in my previous posting. There are some advantages to an indirect tax that will be discussed in a future post.

Although Medicare and Medicaid are enormously expensive, they would be even more so if both systems had to cover a proportional share of capital costs incurred by providers. By shifting coverage of capital costs onto private insurers, the deadweight loss of the avoided taxes is itself avoided. Additionally, indirect taxation creates the illusion that the tax burden of these public systems is smaller than it actually is, thus making it more politically palatable.

The Clinton Administration and the Managed Care Revolution

Medicare and Medicaid addressed much of the insurance coverage gap problem, but not all of it. By the start of the first Clinton Administration in 1993, 15% of Americans were still without health insurance. These were mainly people with preexisting medical conditions and healthy, risk-neutral people who chose not to purchase health insurance.

To bridge this coverage gap, the Clinton Administration proposed a system of subsidies that ultimately failed to win approval by Congress. The media portrayed this political debacle as a victory by vested interests, but that was only part of the reason. Besides a desire to fill the coverage gap, another major motivation for this legislation was to slow the rapid increase in the cost of healthcare.

Prior to the first Clinton Administration, annual healthcare costs had been rising by 8% or greater. (See Figure 6). The adoption of Health Maintenance Organizations (HMOs) on a wide basis in the 1990’s led to a significant slowdown in healthcare costs. There was a consensus that the market, through HMOs, was fixing the problem. Therefore, the need for a costly and divisive government program was unnecessary.

Figure 6.

HMO’s slowed the growth rate of healthcare expenditures by imposing incentives on providers to better control costs. There was a focus on primary care physicians acting as gate keepers in order to decrease the use of expensive specialists. Other incentives, like capitated fees, also contributed to slower growth.

The slowdown in healthcare cost growth was only temporary, however.  By 2001, HMO enrollment had peaked and started a long decline. In 2017 according to the Medical Expenditure Panel Survey, fewer than 19% of the US population were enrolled in a private HMO and those still enrolled are confronted with far fewer restrictions than in the past.

Although the Clinton Administration’s efforts to expand public insurance coverage were unsuccessful, I included this section to illustrate a point about the American healthcare consumer. HMOs and their cost controlling restrictions proved to be effective in controlling the growth rate of costs, but were nonetheless, very unpopular. Americans do not like restrictions on their choice of provider. American employers were forced to offer less restrictive HMOs and alternative plans, such as Preferred Provider Organizations (PPO’s), in order to compete for employees.

European-style, single-payer systems share many characteristics with HMO’s. Without cost controls, these systems would experience unconstrained growth and quickly swamp their respective healthcare systems. Promising to expand Medicare coverage to all age groups while eliminating private health insurance can only work by adopting strict cost controls, the very same cost controls emphatically rejected by most Americans in the 1990’s. I do not think the politicians promising Medicare-For-All or public options that will crowd out private insurance understand the trade-off they are proposing.

Medicare Parts C and D

Medicare is popular with the public in general, but being a form of socialized medicine, has always been unpopular in conservative political circles. So, when it became clear that the lack of prescription drug coverage was a political liability, the George W. Bush administration decided to create a market-oriented system to fill this coverage gap. The result was Medicare Part C (aka Medicare Advantage) and Medicare Part D.

Medicare Part C purchases private insurance on behalf of beneficiaries. The premium paid is the average expected healthcare costs that would be spent on the beneficiary had she been a Medicare Parts A and B (i.e. Pay-For-Performance (PFP)) enrollee. To induce private insurers to participate, an extra amount is paid to provide a normal profit.

Part C has proven to be very popular and now comprises one third of all Medicare enrollees. There is evidence that Medicare Advantage plans cherry-pick relatively healthy beneficiaries and that the premiums paid by Medicare over-compensate plans. Because Part C pays more than the average cost per member, it has accelerated the annual growth rate of Medicare.

Medicare Part D allows enrollees to purchase prescription drug coverage from private insurers. Although it subsidizes the program by compensating insurers for very expensive or low-income beneficiaries, unlike Medicare Parts A and B, it does not use its market power to pay lower prices for prescription drugs. Consequently, Medicare Part D beneficiaries pay as much for prescription drugs as privately insured people.

Ironically, the market-oriented approach to Medicare Parts C and D has contributed to the acceleration of the cost of the Medicare program and especially the amount of direct taxation to fund it. These programs are particularly expensive for tax-payers and Medicare enrollees. There are two groups that unequivocally benefitted from the design of Medicare Parts C and D, however, pharmaceutical manufacturers and private insurance companies.

The Affordable Care Act

After several decades of political stalemate, the Democrats seized the opportunity to address the insurance coverage gap when they took over the presidency and both houses of Congress in 2009. Instead of exclusively using the government’s market power to provide health insurance at the provider’s marginal cost, like Medicare Parts A and B, the Obama Administration chose a combination of program types. They expanded Medicaid eligibility (i.e. the providers’ marginal cost approach) and created a health insurance exchange, a Republican-style market-oriented approach like Medicare Part D.

By subsidizing the purchase of private insurance rather than providing public insurance at the marginal cost of care, Obamacare resulted in higher payments to hospitals and physicians at a higher cost to taxpayers. In addition, healthy, risk-neutral people were penalized if they did not choose to purchase private health insurance (i.e. the individual mandate).

Despite this mandate, the number of uninsured people in the US remains far above zero. Like Medicare Part D, the benefits of the program are mixed. Many people with pre-existing conditions are able to obtain health insurance, but some are not. Taxpayers shoulder the full cost of the program, including those who feel the mandate is fundamentally unfair. The only groups that unequivocally benefit from Obamacare are healthcare providers (i.e. hospitals and physicians) and private insurers.

The Next Phase

The current status of Obamacare is the Trump Administration no longer enforces the individual mandate and the percentage of the population uninsured is now increasing after dropping significantly following its passage. Candidates for the Democratic party presidential nomination have authored competing proposals that can be generally summarized into two different categories: Medicare-For-All and a Public Option.

Like most sweeping legislative proposals, many details are left for future negotiations. The projected costs are debatable, but there are some features that can be described. Medicare-For-All is a relatively simple plan to expand Medicare (Parts A, B and prescription drug) coverage to all age groups. This simplicity is, I suspect, a major reason for its popularity in public opinion surveys.

A major flaw in this proposal is that Medicare has few constraints on costs or utilization, a feature all single-payer systems must have. With its strict limitations on costs and coverage, Medicaid is a much more realistic model for a single-payer system in the US than Medicare. Although a more accurate and honest appellation, I doubt “Medicaid-For-All” would poll nearly as well.

The Public Option has been marketed as a more moderate approach to closing the coverage gap. The basic idea is for the government to sell insurance to anybody who wishes to purchase it. This would guarantee that there would be at least one supplier of high-quality insurance at a reasonable price (i.e. premium). Private insurance companies that charge unreasonably high premiums would be forced to lower their premiums in order to compete or exit the insurance market.

The biggest problem I see with the Public Option proposals is that their premiums can be set arbitrarily low while still providing high-quality insurance. The premium the government charges need not cover the cost of capital nor operational costs for that matter. No private insurance company has the financial resources to compete with the federal government. Over time they will not be able to recoup their cost of capital and will exit the market, leaving it entirely to the government. Therefore, the Public Option is simply Medicare-For-All in sheep’s clothing, a less transparent method of achieving essentially the same end. The only difference being a premium, that even if set very low and with an enforced individual mandate, will leave many people uninsured.

The only way to provide health insurance to the entire population is to provide a public option with a zero premium. But it should not crowd out private insurance, otherwise moral hazard and runaway costs would ensue.

This might sound like a contradiction. How can a private profit-maximizing insurer compete with free public health insurance if it can’t compete with low-priced public insurance? It can’t, obviously. But it doesn’t need to. The Public Option need only solve the coverage gap problem while being less desirable than private health insurance for those who have access to it. I’ll explain more in a future posting.

Who Pays for Medicare and Medicaid?

In our last post, we found that physicians receive higher payments from the privately insured and the uninsured for office visits than they receive from the publicly-insured (i.e. Medicare and Medicaid), but as long as the publicly-insured pay the variable costs of providing care, most physicians have a financial incentive to treat them. This incentive to treat all payer types causes physicians to prefer hospitals that treat all payer types. Therefore, even though the average hospital loses money from treating the publicly-insured in the long run and the uninsured in general, they have a financial incentive to treat all payer types.

Despite the losses incurred from treating the publicly-insured and the uninsured, do hospitals make positive economic profits in general? The answer is “yes”, with some qualifications. Figure 5. shows mean and median net income as a percent of expenses from the treatment of patients in 2017. In other words, this does not include government subsidies and other capital infusions. The chart also breaks the hospitals into three types: Private For-Profit, Private Not-For-Profit and Government-Owned.

Figure. 5 Short-term Hospital Net Income Percent of Expenses, 2017

Source: 2017 Health Care Cost Report Information System from the Centers for Medicare and Medicaid Services. Includes only hospitals that reported both revenues and expenses.

For all short-term hospitals in general, revenues received from the treatment of patients are about 1% less than expenses. But these loses are incurred primarily by the Private Not-For-Profit and Government-Owned hospitals. These hospitals are not required to cover their capital costs in the long run and the Government-Owned hospitals don’t even need to cover their operational costs in the long run.

The Private For-Profit hospitals earn a positive rate of return, 10% on average although the median For-Profit hospital only earns 3.3%. Remember that Private For-Profit hospitals must earn a positive rate of return to cover their capital costs. Something in the 3% range is quite sufficient.

The big picture here is that hospitals are not in economic crisis. The combination of profitable payments from private insurers and unprofitable payments from the other payers adequately covers both the operational and capital costs incurred by For-Profit hospitals on average. Not-For-Profit and Government-Owned hospitals charge lower prices for the treatment of patients and receive outside funding.

Private insurers are in effect covering the capital costs of hospitals. The government leverages its market power to force private insurers and their premium-paying customers to shoulder some of the cost of providing public insurance. This is a form of cross subsidization.

In a government-run single-payer system, the cost of health insurance is borne entirely by taxpayers. Remember that the cost of raising a dollar of tax revenue is greater than a dollar. Taxes affect economic decisions of consumers and suppliers. We call this the deadweight loss of taxation. Therefore, the cost of a single-payer system is undercounted if we rely strictly on the amount of money spent on purchasing healthcare.

This is the under-appreciated part of the US healthcare system. Countries which have single-payer systems, like the UK, must place very noticeable limits on the provision of healthcare in order to control costs. These limits take the form of long wait times to see a physician and denied coverage for some surgical procedures and prescription medications. Medicare (and to a lesser extend Medicaid) is practically devoid of these types of restrictions. The US can afford to be very generous with its single-payer system for the elderly because people who purchase private health insurance are subsidizing a substantial part of the costs of that system.

This is why the expansion of Medicare to all age groups (i.e. Medicare For All) would be financially ruinous. Without private insurance, Medicare would have to cover both the operational and capital costs of providing healthcare. The only way to limit an enormous increase in taxes would be to install limits like the UK. The US consumer has no tolerance for this type of limitation.

Is Treating Medicare Patients Profitable?

This post is the latest in a series that establishes what an optimal health insurance system would look like, how much it would cost and who would pay for it. Past posts analyzed the number of uninsured people in the US (it’s not as many as officially reported), the cost of a single-payer system (it’s larger than officially reported), and the future of Obamacare (it’s likely to fail).

Today’s topic answers the question: do healthcare providers make money from Medicare and Medicaid? The answer is a little complicated and depends on the meaning of “make money”.

Two of the earlier posts dealt with the market power of healthcare providers. For example, in 2017 the average percentage markup by hospitals was 360%. In other words, the average hospital’s undiscounted charges were 460% of its costs. That sounds like a huge markup, but insurance organizations only pay a fraction of the bills they receive due to negotiated discounts. And the uninsured frequently default on their financial obligations to providers.

In a later post we showed that for outpatient and emergency department visits and inpatient stays, private insurers pay a much higher percentage of billed charges than public insurers (i.e. Medicare and Medicaid) and the public insurers in turn pay a much higher percentage than the uninsured do on average.

All three of these healthcare services are provided by short term acute care hospitals. Table 1 combines the results from the 2016 Medical Expenditure Panel Survey (MEPS) for these three hospital-provided service types and includes the average markup percentage for all hospitals.

Table 1. Percent Charges and Price Markups, Short Term Hospitals

Sources: (A) 2016 Medical Expenditures Panel Survey from the Agency for Healthcare Research and Quality, (B) 2017 Health Care Cost Report Information System from the Centers for Medicare and Medicaid Services.

Private insurers pay on average a little less than 49% of billed charges. That means they get a discount of slightly more than 51%. This might sound like a lot, but it isn’t when we consider that hospitals markup their prices way above the cost of providing care and that public insurers get over a 78% discount on average.

The uninsured only pay 6.7% of billed charges. This occurs because the largest bills are for very expensive emergency services. Typically, the uninsured are the least wealthy of all hospital patients and frequently must default on their hospital bills.

When combined with the average hospital percentage markup of 360%, we see that hospitals make a very substantial profit from services provided to the privately insured despite the large negotiated discounts. A 125% markup is very high, much higher than almost any other commercial enterprise one could name.

The story is quite different, however, for the other payer types. The average hospital treats the uninsured at a high loss rate and the average percentage markup for treating the publicly insured is essentially zero.

That zero figure is no accident.  The average discount hospitals are required to give Medicare and Medicaid are calibrated to result in a zero markup. The payments made by CMS cover the operational costs of treating their beneficiaries, but nothing more. Remember that these are accounting figures. They don’t include the cost of capital all hospitals must employ. When the cost of capital is factored in, the average percentage markup for treating the publicly insured is well below zero.

Why then do hospitals continue to treat the uninsured and the publicly insured if they lose money by doing so? There is a short-run answer to this question and a long-run answer.

First, the short run. Capital costs tend to be fixed in the short run, but variable in the long run. So, in the short run the average hospital’s variable costs of treating Medicare and Medicaid beneficiaries are covered by the payments from CMS. Therefore, the CMS payment level (i.e. price) is just above the shutdown point. This is the minimum price at which the average hospital continues to treat publicly insured patients. At this price, it loses money; however, its losses would be even greater if it shutdown (i.e. stopped treating CMS beneficiaries) in the short run.

That explains why the average hospital continues to treat CMS beneficiaries in the short run, but what about the long run? And why would hospitals treat the uninsured even in the short run?

To understand why hospitals willingly treat patients that cause them to lose money, we need to understand the business model of a hospital. A business provides a service to a customer in exchange for money. The typical customer performs three different functions: 1) she chooses how much to buy and which business to purchase from, 2) she supplies the money to pay the price and 3) she receives the service.

In the hospital business, the patient performs only one of these tasks entirely (i.e. she receives he service) and performs one of them partially (i.e. she pays part of the price). The insurer pays for most of the service and the admitting physician chooses how much to buy and which business (i.e. hospital) to purchase from.

So, who is the hospital’s customer, the patient (who consumes), the insurer (who pays) or the doctor (who chooses)? The answer is the doctor. Hospitals market their services primarily to doctors. They don’t market their services to potential patients except in very limited circumstances.

A common way for hospitals to drum up new business is to purchase physician practices. This is an example of vertical integration. Hospitals grant doctors “privileges”. The more doctors with privileges a hospital has, the better off financially it will be. Therefore, hospitals are managed to keep their doctors (i.e. customers) satisfied.

The typical physician’s practice consists of a mixture of patient types. Some are privately insured, some are publicly insured and some have no insurance at all. If you remember, Figure 3 showed that the uninsured pay a higher percentage of billed charges for physician office visits than either of the other payer types. Therefore, physicians make a higher return from treating the uninsured than they do from treating the insured.  A hospital that only treated the privately insured would be of little use to most physicians. Consequently, if the hospital wants to attract physicians (its real customers), it must be willing to treat all types of patients, even ones that can’t pay all their bills.

To my knowledge, there is not a single short-term acute care hospital in the U.S. that does not treat Medicare and Medicaid beneficiaries. And if they treat Medicare and Medicaid patients, then by law they cannot refuse to care for uninsured emergent patients.

Back to the main question: do providers make money from treating Medicare and Medicaid beneficiaries? The answer is they do not directly make money from treating them, but they do indirectly make money from treating them. They make money (i.e. positive net revenues) from treating the privately insured and they must treat the publicly insured in order to treat the privately insured.

The same argument goes for treating the uninsured. By law and by professional custom, life-saving care cannot be denied to a patient on account of an inability to pay.

Next time we’ll see how all the price discrimination, unfunded mandates and cross-subsidization affects the bottom lines of providers and who really pays for Medicare and Medicaid.

Who Pays What and Why

It’s time to return to designing an optimal health insurance system. In previous posts we have specified the different types of healthcare that are covered by health insurance, defined and quantified the size of the uninsured population in the US, explained the costs of single-payer healthcare systems, and discussed the high degree of market power wielded by providers as demonstrated by high percentage price markups. Today we continue our discussion of market power and its importance to the healthcare market.

If the price of a hospital stay is greater than the price of an office visit, that is understandable because a hospital stay requires far more resources to supply.  Each hospital patient must be attended by a team of nurses 24 hours a day.  There is also the cost of maintaining the hospital building itself and all the monitoring and surgical equipment. Plus, there are the physician fees for diagnosing the illnesses suffered by the patients and performing surgery.

What is harder to understand is that the price a patient and her insurer pays for healthcare depends as much on who is paying for it as it does on the costs incurred by the provider. In economics, the practice of charging some customers more than others for the same services is called price discrimination.

Price discrimination can only occur when there is an imbalance between the relative market powers of the supplier and the buyer. When price discrimination occurs because of the relatively strong market power of the supplier (i.e. monopoly power), it is considered bad for consumers. Consequently, this form of price discrimination is outlawed by three different acts of Congress: the Sherman Antitrust, Clayton Antitrust, and Robinson-Patman.

When price discrimination occurs because of the relatively strong market power of the buyer (i.e. monopsony power), however, there is less concern for the impact this might have on consumers. Antitrust laws don’t address this form of market power.

In the provision of healthcare, price discrimination is quite common. Providers bill patients the same amount for the same services regardless of who is paying the bill. But after the bill is submitted, health insurers typically negotiate a discount, i.e. a decrease in the amount owed to the provider. Uninsured patients, of course, are legally obligated to pay the full undiscounted billed charges.

This billing custom does not apply to the sale of prescribed drugs from pharmacies, however. Those are “point of sale” transactions, so any discounts are already subtracted from the sale price when reported. We’ll explain this in greater detail later.

Figure 3 illustrates the percent of billed charges paid to providers by three payer types: private insurers, public insurers (i.e. Medicare and Medicaid) and the uninsured. Those who have read my earlier posts will recall that to be truly uninsured one needs to have no other source of payment than one’s own money.

Figure 3. Percent of Billed Healthcare Charges Paid by Payer and by Service Type, 2016

Data source: 2016 Medical Expenditures Panel Survey, Agency for Healthcare Research and Quality.

*There were no instances of an uninsured household purchasing Home Health services.

The percentages illustrated in Figure 3 might seem pretty random, but they actually demonstrate the relationship between prices and market power quite neatly. Remember that the more market power the seller has over the buyer the higher the price paid. The lack of a simple relationship in the graph is due to there being four different buyer types with different levels of market power. Public insurers (i.e. the government) and the emergent uninsured have more market power than the privately insured and the non-emergent uninsured. The heightened market power of the emergent uninsured is due to a law that requires all providers that receive Medicare and Medicaid payments to treat all emergent patients regardless of ability to pay. This is an example of an unfunded mandate. The significance of which will be discussed more in a later post. Figure 4 illustrates the market power ranking.

With the relative market power price theory in mind, let’s look at the different service types one at a time.

Office Visits: The uninsured pay more than the privately insured and the publicly insured pay the least. That is exactly as the theory would suggest. A single office visit is relatively inexpensive and non-emergent. That means that the doctor does not have to see the patient if he doesn’t want to. The doctor has the greater market power. The only way an uninsured patient gets a discount at all is if he defaults on his financial obligation.

Outpatient Visits: The discount structure is very similar to that of office visits except that private insurers pay a little more than the uninsured. Outpatient visits are also non-emergent; however, they are much more expensive than office visits and the uninsured are more likely to default.

Inpatient Stays. For this service type the uninsured pay the least. This implies the uninsured have much more market power than they do for outpatient and office visits. That is true because many inpatient stays are instigated by emergencies. The very large dollar amounts charged for inpatient stays also contributes to the low percentage of charges paid by the uninsured.

Emergency Department: The discount structure of emergency departments is very similar to that of inpatient stays for the same reasons.

Home Health: There are no figures for the uninsured payments for home health services. This is a service rarely purchased by the uninsured. As with all the other services mentioned, private insurers pay a higher price than public ones.

Pharmacy Sales. This is the only service for which public insurers pay a higher percent of charges than both the uninsured and the privately insured. This happens because, unlike all the other service types, Medicare does not pay pharmacies directly.  Rather, Medicare pays private insurers to pay pharmacies on its behalf.  That is why the percent of charges paid by private and public insurers is almost the same.

The uninsured pay a relatively low percentage of charges for pharmaceuticals because, although they have little market power, there is a high degree of substitutability in the sale of prescription drugs. Their demand for pharmaceutical products is relatively price elastic or sensitive to price changes. All their expenditures are “self-pay”. Only a small fraction of insured pharmacy purchases is borne by the insured party. Consequently, drug companies offer the uninsured discounts that they do not offer the insured.

Before I end this post with a brief summary, I should explain how I derived the percent of charges paid for pharmacy sales. As I noted above, pharmacy purchases are “point of sale” transactions. The undiscounted charges are not reported separately from the expenditures net of discounts and write-offs like the other service types. So, how did I derive the figures reported in Figure 3?

The MEPS, the source of data for Figure 3, provides a file of all the drugs purchased by the respondents to the survey. This includes the national drug code (NDC), quantity and the amount of money spent by the insurer and respondent, separately. From these data I calculated the average price per unit for each of the three payer types: private, public and uninsured.

I then found the highest price per unit for each drug (i.e. NDC). The undiscounted charges were the amount of money that would have been spent by each payer type if it had paid the highest price for each drug. The percentage of charges paid is actual expenditures divided by estimated undiscounted charges. I had complete data for 915 different drugs.


So, what’s the big picture here? In general, private insurers pay more than public insurers due to their relatively inferior market power. The uninsured pay the most except when there are exigent circumstances or their demand is price elastic. The next post will combine the post about percentage markups with this post about price discrimination to see who is really bearing most of the financial burden of our healthcare system and why Medicare can be offered with relatively few restrictions.