Healthcare prices in the United States vary wildly — sometimes by a factor of ten — for the same procedure within the same city. A knee replacement at one hospital might cost $15,000. At another facility twenty minutes away, the same surgery with comparable outcomes might cost $75,000. For most of economic history, we'd call that an unsustainable market distortion. In healthcare, we've called it normal.

Reference pricing is a deceptively simple policy idea designed to address this. An insurer sets a benchmark payment — a reference price — for a given procedure and covers that amount in full. If a patient chooses a provider that charges more, they pay the difference out of pocket. The logic borrows from how we already shop for most things: there's a reasonable price, and if you want the premium option, you pay the extra.

The concept has moved from theory to practice, and the evidence is worth examining closely. California's public employee retirement system ran one of the most carefully studied experiments in reference pricing — and what happened next surprised health economists, hospitals, and patients alike.

In 2011, the California Public Employees' Retirement System — CalPERS — faced a problem familiar to any large employer. Healthcare costs for its 1.3 million members were climbing relentlessly, and joint replacement surgery was a significant driver. Prices across California hospitals ranged from roughly $15,000 to over $100,000 for the same hip or knee procedure, with no meaningful difference in quality outcomes.

CalPERS chose a targeted intervention. It set a reference price of $30,000 for joint replacement surgery. Members who selected a hospital charging at or below that threshold paid only their standard copay. Members who chose a pricier hospital paid the full difference between $30,000 and whatever that facility charged — potentially tens of thousands of dollars out of pocket.

The results were striking. Before the policy, roughly half of CalPERS joint replacement patients used hospitals priced above $30,000. Within a year, that share dropped dramatically as patients migrated to lower-cost facilities. Average payments fell by approximately 20 percent. Crucially, quality metrics — readmission rates, complication rates, patient satisfaction — did not decline. Patients weren't sacrificing outcomes. They were simply paying attention to price for the first time.

What made the approach notable wasn't just the savings — it was the design clarity. The reference price was set high enough that most hospitals in the state fell below it, ensuring genuine access to affordable options. And the policy targeted a procedure that was standardized, plannable, and measurable. Those conditions, as we'll see, matter enormously for whether reference pricing works at all.

Takeaway

A well-designed reference price doesn't restrict choice — it restructures incentives. When patients gain a financial reason to notice price variation, their collective decisions can move an entire market without any central authority dictating what hospitals should charge.

The most revealing part of the CalPERS experiment wasn't what patients did. It was what hospitals did in response.

When reference pricing took effect, hospitals charging well above $30,000 for joint replacements faced an immediate problem. CalPERS members — a significant patient population in California — were suddenly incentivized to go elsewhere. High-priced hospitals weren't competing on reputation or convenience alone anymore. They were competing on price, and they were losing volume. Several responded by cutting their joint replacement charges by 30 percent or more within the first two years, bringing themselves below the reference threshold to recapture patients.

This competitive dynamic is precisely what health economists have argued is missing from most healthcare transactions. Under traditional insurance, patients pay the same copay whether their hospital charges $20,000 or $80,000. Price variation is invisible and irrelevant at the point of care. Reference pricing breaks that insulation in a controlled way — patients retain coverage up to the benchmark, but enough price sensitivity enters the system to generate genuine competitive pressure on providers.

The broader implication is uncomfortable for the industry. Much of the price variation in American healthcare doesn't reflect differences in cost or quality. It reflects what the market will bear in the absence of competition. Reference pricing essentially calls that bluff. And the speed at which hospitals adjusted their charges suggests that many high-priced providers could deliver the same service for considerably less — they simply had never been given a reason to.

Takeaway

Price variation persists in healthcare partly because nothing forces it to shrink. When a policy creates even modest competitive pressure, the gap between what providers charge and what procedures actually cost becomes visible — and often closes faster than anyone expects.

Reference pricing works well for joint replacements. The question is whether it works for everything else. The honest answer is: probably not — and understanding why is as important as celebrating where it succeeds.

The conditions behind CalPERS' results are specific. Joint replacement is a shoppable service. It's planned in advance. Patients have weeks to compare options. The procedure is relatively standardized across providers, and quality is measurable through established outcome metrics. Enough facilities offer it that genuine competition exists. Not all healthcare looks like this.

Emergency services are the clearest exclusion. Someone arriving at an emergency department with chest pain isn't comparison shopping, nor should they absorb surprise costs based on which hospital an ambulance reaches first. Complex chronic conditions present similar challenges — ongoing cancer treatment involves multiple specialists, evolving care plans, and provider relationships that make price-based switching impractical and potentially harmful. Even among planned procedures, reference pricing requires careful calibration. Set the benchmark too low and you restrict access, particularly in rural areas with fewer providers. Set it too high and the policy loses its market-disciplining effect entirely.

Health policy researchers estimate that roughly 30 to 40 percent of healthcare spending involves services that could theoretically support reference pricing — lab tests, imaging, certain outpatient procedures, and planned surgeries. That's a meaningful share of the market, but it means the majority of healthcare spending involves situations where this tool has limited application. Reference pricing is a precision instrument. Treating it as a universal fix would stretch it past what the evidence supports.

Takeaway

The most useful policy tools are often the ones with clearly defined boundaries. Knowing exactly where reference pricing works — and admitting where it doesn't — is what separates evidence-based reform from ideology dressed up as economics.

Reference pricing represents something uncommon in health policy: an intervention with solid evidence, clear mechanisms, and honestly acknowledged limitations. It works by reintroducing a basic market force — price sensitivity — into a system that has largely operated without it.

The CalPERS experience shows that when conditions are right, reference pricing lowers costs without compromising quality and reshapes provider behavior in ways that benefit the broader market. But those conditions are specific, and stretching the model beyond its natural boundaries risks creating new problems.

The deeper lesson may be the most valuable. Healthcare markets aren't inherently broken — they're structurally insulated from the competitive pressures that discipline prices elsewhere. Reference pricing doesn't fix healthcare. It reveals what happens when you let a small piece of it actually function like a market.