Seeing the ‘deep pocket’ effect when insurers pay for health care

Some Americans pay directly for their health care. Many others have it covered by their health insurance. Both groups should pay the same amount for the same services, but don’t. The “deep pocket” concept may be at work.

In their book “Overcharged,” Charles Silver and David Hyman argue that health care is expensive because it’s insured. Their thesis is that medical procedures and commodities would cost less if patients paid for them directly. Recent news stories support this argument:

• Sutter Health charges Aetna $85 for flu shots it advertises to the public at $25.
• The New Jersey School Employees’ Health Benefits Program paid some acupuncturists and physical therapists an average of more than $600 per visit in 2018. Pay out of pocket and the cost is generally less than $100 per visit.
• Blue Cross and Blue Shield of Minnesota paid a New York City physician practice $25,865.24 for an out-of-network throat swab that LabCorp prices at $653.

A more systematic comparison of prices paid in a retail milieu to prices paid by insurers is suggested by a 2019 report by Katherine Hempstead and Chapin White on the retail market for health care among the Amish and Old Order Mennonite communities in Lancaster and Lebanon counties in Pennsylvania. Members of these so-called Plain sects reject commercial and public insurance, and so pay cash for medical treatment.

Hempstead and White found that a local health system quotes prices to these customers in a way that is fully transparent; prices are for care bundles and nothing in its price booklet for these patients approaches the fanciful dollar amounts or nonsensical descriptions found in the same facilities’ chargemaster lists.

I compared prices from this health system’s booklet to the average allowed amounts for the same services reported by the Health Care Cost Institute’s Guroo tool for hospitals in the state of Pennsylvania. For all services for which insurers pay more than $5,000, the price quoted to the Plain communities is substantially less than the price negotiated by insurers, with most retail prices falling between 49% and 64% of insurers’ prices.

Put another way, the insurer-negotiated amounts are generally one-and-a-half times to twice the retail prices, as shown in the chart below.

While hospital outpatient prices have been increasing on average more than 3% each year, there was no increase in prices quoted to Plain consumers between 2018 and 2020.

Intrigued by this differential, I compared the publicly posted price list for Wooster Community Hospital in Ohio, another facility serving a large Amish population, to Guroo data, and got similar results.

These hospital systems have been forced by insurance-averse consumers to engage in yield management. This technique, developed by the travel industry, is used to sell “perishable” slots such as empty seats on airline flights or open hotel rooms at lower prices to price-sensitive customers. One insight the pricing differential reveals is that the hospitals perceive the Plain community as a customer segment analogous to leisure travelers and insurers as a segment analogous to business travelers who can afford to pay much more.

Plain patients, despite their apparently ascetic lifestyles, have average household incomes roughly similar to those of Americans in general. Insurance pools, meanwhile, are funded by the earnings of average American workers. So the distinction made between Plain households’ ability to pay and insurers’ ability to pay is, in the end, wholly artificial. This suggests that health care is indeed more expensive because it’s insured.

What mechanism is at work here? In “Overcharged,” Silver and Hyman point to the effect that third-party payment has on consumer behavior — indifference to prices the consumer doesn’t pay out of pocket — but as the yield management scenario demonstrates, it is the hospitals’ and doctors’ behavior that governs pricing. A better explanation may be deep-pocket bias, a phenomenon best known in the realms of civil liability and tort law.

Deep-pocket bias refers to the likelihood that jurors in a civil trial will levy higher damage awards against defendants with a greater ability to pay. It’s essentially a behavioral economics concept identified intuitively by lawyers and judges long before behavioral economics was a field. It first manifested itself in a rule of evidence prohibiting mention of insurance coverage in trials. Empirical research later confirmed that jurors do award more money when they know an insurer will be paying. It’s not exactly a cognitive bias: Insurers sit on large pots of money; they can, and do, write larger checks than individuals. But a jury’s task is to objectively determine the cost of restoring an injured plaintiff to health, and that cost doesn’t increase because a defendant is insured.

The irony is that a dynamic demonstrated by people who, as the joke goes, are “not smart enough to get out of jury duty” is exploited by providers and acceded to by a sophisticated insurance industry — an industry that, on the property and casualty side, vigorously pushes back on deep-pocket strategies. The result is a health care system that is paid the equivalent of last-minute, unrestricted airfares instead of the lower advance-purchase fares that middle-class American families can afford.

In documenting the dramatic price increases for cancer drugs following enactment of Medicare Part D, David Besanko and colleagues noted that insurance coverage permitted a pharmaceutical manufacturer to “capture more value than its product creates.” But we can feel confident that the prices paid by the Plain community reflect the value of medical treatments received by its members and to the community as a whole.

It is shocking that the purchasing power and expertise of insurers has not achieved similar price levels.

Jackson Williams is vice president for public policy at Dialysis Patient Citizens. He previously worked on health care quality and payment reform issues at the Centers for Medicare and Medicaid Services.