How The No Surprises Act worsened the problem of surprise bills

Next week, Sen. Bill Cassidy (R-LA) will convene a roundtable to discuss adjustments to the 2020 No Surprises Act, which has come under increased criticism for shifting costs away from patients while driving up overall costs without addressing the underlying problem.

The theoretical market failure

The No Surprises Act was meant to address a legitimate health care problem based on fundamental economic principles. While the claim that healthcare is incompatible with a free market and needs to be managed by government is a vast overstatement, that doesn’t mean certain interactions don’t suffer from true market failures. They’re just much rarer and narrower than critics argue.

For a market to function well, participants need information and agency over their decisions. Both can fail when decisions must be made instantaneously by third parties. This principal-agent problem has been deeply studied in economics and is relevant to government actors. These problems occur when a third party makes a decision on behalf of someone whose interests or preferences differ from its own. Economists try to develop ways to avoid that issue.

How it relates to health care

This is one of the primary complications in emergency health care. The principal, the person who needs care, may be incapacitated or unable to choose their own provider. Moreover, the insurance-based health care system means the principal has ceded much of their decision-making power. But in the case of hospitals and healthcare, the situation is even more complicated. Hospitals  also use third-party providers, such as anesthesiologists. The hospitals are therefore acting as an agent for the insurance company and the patient, without the incentives to choose providers based on quality and costs that are aligned with the preferences of the insurance company or the patient.

Prior to the No Surprises Act, if a hospital chose an anesthesiologist that was not covered by the patient’s insurance, the patient could be stuck with an enormous bill for a provider they had never agreed to.

As the problem grew larger, Congress decided to fix the issue and passed the No Surprises Act at the end of 2020. The No Surprises Act effectively transferred the financial onus away from the patient, but instead of transferring it to the hospital, which was ultimately the decisionmaker, the bill created a process in which an arbiter would choose between the practitioner (the anesthetist) and the insurance company. Both would submit their view of what the payment should be, and the arbiter would choose one or the other.

To give the legislators some credit, I don’t think anyone could have predicted the eventual way this process would fail. They assumed that arbiters would choose each side’s offer about half of the time, giving both sides incentives to make reasonable bids. Basically, the legislators assumed the arbiters would be absolutely objective and fair.

That assumption turned out to be incorrect. Arbiters consistently sided with the practitioners. Even when insurance companies raised their offers, it seemed to have no effect on the probability of winning. Consequently, practitioners raised their bids. And because the process became so profitable for them, they brought more claims into the process.

Consequently, the costs of out-of-network health care boomed. The system has not yet reached an equilibrium, but when the new system settles, providers will be much better off, and patients will not only still be paying for the original size of the Surprise Bills via their premium, but they’ll be paying higher prices for these services and they’ll be paying for more of them, too. All that money will come from the insured.

Proposals for reform

Currently, proposals to fix the process focus on setting a limit on how much the practitioners can win. That may reduce costs, but it is not a solution to the underlying problem. Below are two approaches that instead highlight the core misalignment in incentives and push actors to come to an agreement without government influence. Clearly these need some fine-tuning, but they are market-oriented solutions, not merely bandages on a wound.

Approach 1: Shift the onus from the payer to the hospital

One approach is to address the principal-agent problem directly by having Congress mandate hospitals incur the discretionary costs. This is an application of the Coase Theorem: because hospitals do not bear the cost of the decision, the market structure should be designed such that it does. Then the hospital can figure out a solution that will better balance costs and quality for the payer.

Congress could require hospitals that use third-party practitioners to pay the costs above the insurer’s in-network negotiated rates. If hospitals choose in-network practitioners, then the insurance company pays the in-network rate. If they choose an out-of-network practitioner, then hospitals would only receive the in-network payment from the insurance company and foot the rest of the bill. This approach would align the financial decision with the one incurring the cost.

A version of this was proposed in the discussions that led to the No Surprises Act. It was discarded because of practitioner and hospital resistance. However, their criticisms are largely addressed by the narrower proposal described. Furthermore, the concern of practitioners that hospitals have too much bargaining power would seem less important under this proposal.

Approach 2: Remove the arbiter and give the parties a reason to agree

A second approach would use the current arbitration system but shift the bargaining power away from practitioners. Right now, because of the structure of the health care system as it is, and the demonstrated bias in the arbiters, the practitioners hold an enormous amount of bargaining power and are collecting growing sums of health care dollars.

To remove the arbiters’ bias, they should be removed altogether as the decision makers. They should be replaced by mechanical rules that push the practitioners and payers to come to an agreement directly.

This can be accomplished by making failure to reach agreement hurt both sides, effectively making each side pay for the other’s proposal. Generally, the insurer submits a low proposal while the practitioner submits a high proposal because both want to be paid as much as possible. Under this approach, the insurer would pay the practitioner’s bid but not pay it out to the practitioner. The difference would go to a third party, like the government.

With these incentives, there is no real benefit to bidding something reasonable because the bid is completely unrelated to the payout. To make this system work, additional stipulations would be needed. First, the disagreement-tax outcome should only occur when the two sides can’t reach a direct agreement. If the parties reach agreement, then the payment will be market-generated and the failure case won’t be necessary.

Second, if the parties’ bids overlap — the insurer proposes more than the practitioner asked — the median of the two proposals would be the final payment. This is a well-known solution to an auction format known as the sealed-bid double auction under incomplete information devised in 1983 by economists Kalyan Chatterjee and William Samuelson.

The goal of this approach is to provide incentives to reach an agreement. The more each side could lose, the stronger the incentive for them to reach a deal.