The gynecologist made several incisions and inserted the laparoscope. With the help of her surgical team of nurses, students and anesthesiologists, she removed the patient’s uterus, which had been bleeding uncontrollably for the past six months despite aggressive medical therapy. The price tag of the procedure? Around $6,000.

Meanwhile in a nearby hospital, another gynecologist removes another woman’s uterus, in a procedure no more complicated or time consuming than the first one. But at that hospital, the bill came to more than $17,000.

In my recent post on the growth of health-care expenditures, I overstated the role of physician pay in contributing to health care inflation (my bad). As I point out in that post, and as I will reiterate now: it’s prices that concern me, including hospital pricing practices.

Observers have long noted that hospital pricing practices are both opaque and inscrutable. The Princeton economist Uwe Reinhardt in fact has a nice series of blog posts on this topic which explain some of the quirky factors that have led hospital prices to be so economically unjustifiable. He summarizes hospital pricing as being “chaos behind a veil of secrecy.”

But behavioral economics yields an insight into an important phenomenon influencing hospital prices, a phenomenon I came across while reading a 2005 Lewin Group report on hospital pricing.  What hospitals charge for various services are strongly influenced by anchoring effects.

Anchoring effects are one of the better established phenomenon in the field of behavioral economics. Daniel Kahneman and Amos Tversky first uncovered anchoring effects in a series of clever experiments whereby, for instance, a person’s guess at the age Mahatma Gandhi died at changes depending on whether they are first asked whether Gandhi was older than 9 or younger than 140 when he died.  These extreme anchors—neither 9 nor 140 being credible answers to the question—nevertheless influenced people’s final guesses. Those asked whether Gandhi was older than 9 when he died anchored on that lower number, and  insufficiently adjusted their final response to provide a final estimate of around 50-years-old. By contrast, those anchored at the ridiculously high number of 140 guessed that Gandhi was closer to 67.

Neat little experiment. But hard to believe that anything significant could result from a little guessing game. Right?

Well back in the hospital, people setting the prices were playing a guessing game of their own when they were first asked to set prices for their services. As the Lewin report summarized: “Hospital charges have been set over several decades, long before facilities had a good sense of the costs of providing services.” In other words, when first asked by insurance companies to establish prices, they pretty much it up.

Now consider that most U.S. hospitals develop a fee schedule for 20,000 or more items—from the cost of Tylenol tablets to the price of getting a basic chemistry panel to the cost of an X-ray or a CAT scan or an MRI—and you can see the problem. No hospital can review the exact cost of all these items. So the people in charge of hospital prices simply adjust the charges every year, typically raising the price some set amount above the previous year. They might revisit, say, the injectable drug increases one year, or the surgery suite charges another year.  But if you want to know what any service costs this year, your best starting point is to see what the hospital charged for it last year. That starting point then—that anchor—dramatically influences today’s prices.

New procedures, being new, come under a bit more scrutiny. Nevertheless, their prices are still influenced by anchoring effects. According to the Lewin team: “These costs may be compared to other similar services provided by the hospital.  In other words, if a hysterectomy has traditionally been priced at around $17,000, then a new gynecologic procedure which takes the same amount of O.R. time and personnel will be priced similarly.

An arbitrary starting point—heck, a whole series of arbitrary starting points—have been handed down, anchored into practice with insufficient adjustment, this irrational marketplace enduring in large part because the prices are opaque to most customers who, besides which, are protected from most of the expense because they have insurance.

Hard to argue with one of the hospital administrators surveyed by the Lewin group who concluded that “there is no rationality to the charge master [jargon for: the hospital prices].”

With this kind of irrational pricing, co-existing with such dramatic differences across hospitals, it behooves us all to comparison shop whenever possible before undergoing any kind of non-urgent hospital procedure. If we don’t, we will be the ones exhibiting irrational behavior.

Peter Ubel is a physician and behavioral scientist who blogs at his self-titled site, Peter Ubel and can be reached on Twitter @PeterUbel.  He is the author of Critical Decisions: How You and Your Doctor Can Make the Right Medical Choices Together.