Healthcare in America: don't expect prices to fall

November 15, 2023

health insurance healthcare costs cartoon

I recently finished the book "An American Sickness" by Elizabeth Rosenthal. In it, Rosenthal covers an array of incentives driving up healthcare costs in the US. Here I intend to distill the key lesson I took away from the book: the health insurance sector, with an assist from hospital systems, is the linchpin of abnormally high healthcare costs in the US.

Health insurance in the US adds a layer of obfuscation between a patient (the customer) and his or her healthcare provider (the business). Patients, as customers, purchase healthcare from their healthcare providers. Note that I'm using the term "healthcare provider" broadly - I would include a pharmaceutical company in the definition as the product they deliver provides healthcare to its intended customer. Patients have a personal relationship with some healthcare providers such as primary care doctors and pharmacists. With others, such as medical-device manufacturers, patients have no personal relationship. Yet a patient has a business relationship with each and every provider that serves them, and health insurance obscures this fact.

Generally speaking, patients operate under the assumption that their healthcare bills will be invoiced to their insurance provider. Patients then pay for the healthcare indirectly through some combination of monthly premiums, deductibles, co-pays, co-insurance, depending on their insurance plan. This is straightforward. The reason I lay it out in these terms is to highlight the fact that patients are customers that are a layer removed from the businesses charging them. In this relationship, the patient generally discovers the cost of services when they are charged for them. If health insurance didn't exist then, holding all else constant, patients would demand to know the cost of non-emergency services before they take place. Because patients don't demand to know costs ahead of time, providers are incentivized to overcharge.

But wait, don't health insurance companies negotiate with healthcare providers to bring costs down? Yes, but crucially, they have a weak incentive to do so. Before I get into the reasons why this is the case, let's break down the high-level business model of health insurance. For the sake of explanation, let's imagine a generic health insurance company called XYZ-health. XYZ-health brings in revenue from its customers' by charging monthly premiums. In exchange, it's obligated to cover a portion of its customers' healthcare costs. The portion of healthcare costs a health insurance company is obligated to cover is analogous to the company's cost of goods sold[1]. The difference between premiums and XYZ-health's healthcare-costs is equal to its operating costs and profit. Bear in mind that the way I'm framing this is a useful simplification - I'm using the term "operating costs" to encompass all non-healthcare costs.

As a corporation, XYZ-health has an imperative to maximize profit. It can do so in four ways: increase premiums, reduce its healthcare costs, reduce its operating costs, or grow its customer base.

To reduce its healthcare costs, XYZ-health has two productive options[2]:

  1. Promote healthy lifestyles amongst its customers to decrease the incidence of necessary medical interventions
  2. Negotiate prices with healthcare providers

The problem with these approaches are that if XYZ-health reduces its healthcare costs too much, it must also reduce its premiums in-tandem. This is due to the Affordable Care Act (ACA), which stipulates that 80-85% of a health insurance company's premiums must go towards healthcare costs. This percentage is called the medical loss ratio (MLR)[3]. So if XYZ-health is an extremely efficient business, it will operate at the minimum allowable 80% MLR. It is legally prohibited from being more efficient than this. If its healthcare costs drop from this point, it must reduce its premiums as well, and therefore it is disincentivized from reducing the cost of its customers' healthcare past a certain point.

Another way of stating this problem is that health insurance companies have a legal upper limit constraining their profit margins. If, in theory, a health insurance company created an algorithm to run the whole business, its only operating-cost would trend towards the cost of electricity in the limit. And yet, it would be prohibited by law from earning more than a 20% gross profit margin. In contrast, tech and financial companies commonly earn in excess of 50% gross margins. And indeed, before the ACA, it was not uncommon for health insurance companies to operate well above a 20% gross margin (i.e below an 80% MLR). As Rosenthal points out, Blue Cross Blue Shield of Texas had an MLR below 65% in 2010.

So, in our example, XYZ-health can only increase its profit to a limit by reducing its healthcare- and operating-costs. What's more, the health insurance sector in general is highly concentrated, so for most existing companies there is a limit to which they can grow their customer base before they run into antitrust concerns. A comprehensive report from the American Medical Association claims that in 2019, 74% of health insurance markets were highly concentrated[4], and recommends tighter state and federal antitrust scrutiny of the sector.

This leaves XYZ-health, and the sector broadly, with one reliable strategy to increase profit in the long-run: to increase premiums. But, as we just learned, a health insurance company must still spend at least 80% of those premiums on healthcare. Thus, health insurance companies have a weak incentive to negotiate prices with healthcare providers: while XYZ-health cannot increase its gross profit margin past 20%, it can increase its total profit indefinitely as long as the costs of its customers' healthcare continues to grow.

I said at the beginning of this essay that the linchpin of abnormally high US healthcare costs is health insurance, with an assist from hospital systems. Just like the health insurance sector, hospital systems are highly concentrated. According to The Healthcare Cost Institute, 77% of US metros have highly concentrated inpatient hospital markets[4]. But hospital systems are more than just inpatient healthcare businesses. They would be better described as healthcare conglomerates. It's common for hospital systems to own primary care and specialist practices, diagnostic labs, imaging centers, outpatient surgery centers, physical therapy centers, and much more.

Abstractly, the natural consequence of a hospital system's market dominance within a given metro is that insurance companies have the weaker hand in price negotiations. But hospital systems also take concrete steps to strip health insurance companies of their negotiating power. Rosenthal highlights the hospital system Sutter Health, which favors buying remote maternity wards in Northern California that are the only viable option for a specific population of expectant mothers. As a result, health insurance companies must have the wards in their network in order to ensure adequate access to care for their policy holders in the region. So Sutter Health negotiates "all-or-nothing" deals, requiring insurers that include any Sutter Health facility in their network to include all Sutter Health facilities in their network. Point being, when a hospital system corners the market on a specific medical intervention in a given region then a health insurance company risks stranding its customers without in-network care if it refuses to agree to the hospital system's pricing structures.

But the moral of the story is that at the end of the day, the math works out fine for the health insurance company. Higher costs one year mean higher premiums the next, which ultimately is the only sustainable way to increase profit in the long run. This is a problem of bad incentives. By mandating a maximum efficiency level in a highly concentrated for-profit sector, the federal government ensured that rationally-acting health insurance companies will accept higher healthcare costs over time.


Notes:

  1. COGS is not a financial metric used in analyzing a health insurance company's financial performance, but it's a useful concept for conceptualizing the health insurance at a high-level and comparing it to other industries.
  2. Other options are 1) Cover a smaller percent of medical costs or a more narrow variety of medical care and 2) make the healthcare system more difficult to deal with, which makes its customers likely to purchase less healthcare. Both of these options are counterproductive, as they would drive rational customers into the arms of competitors.
  3. For health insurance companies that sell large-group coverage (51+ people), the minimum required MLR is 85%. For all others it is 80%.
  4. "Highly concentrated" is defined as a score of ≥2500 on the Herfindahl–Hirschman index (HHI). HHI is calculated by summing the square of the market share of each market participant. A market with a full monopoly scores 10000. A market with two companies with 50% market share scores 5000. Four companies each with 25% market share scores 2500.