Procompetitive health care reform options for a divided Congress
High health care costs affect nearly all health care debates. High costs make it expensive to expand insurance coverage to new populations or cover new services. High spending in public programs strains government budgets, while high spending in the commercial market reduces wages.
In this piece, we highlight areas where the 118th Congress can make tangible progress in reducing health care costs by increasing competition. We specifically focus on a set of procompetitive policies that have attracted various levels of bipartisan interest in recent years. In each case, we describe the rationale for these proposals and discuss how they are likely to make markets work better, as well as how they would affect the federal budget.
This is not meant to be an exhaustive list of all policies that could attract bipartisan interest, but the list we offer here includes some of the most plausible options for passage in a divided Congress. While these policies are far from sufficient to address all inefficiencies in health care markets, they would make tangible progress in the direction of better functioning markets.
Transparency and antitrust enforcement
A large body of evidence finds that the merger of potential competitors within health care markets increases costs to consumers and is suggestive that it also reduces the quality of care. That is, when hospitals or physician groups in the same geographic market merge together (or one acquires the other) they are able to extract higher prices from private insurers and employers. Newer research suggests that similar effects also may arise from mergers that occur across different markets within the same state. Higher costs are at least partially passed on to individuals in the form of higher premiums and cost-sharing or to employees in the form or lower wages. A growing body of evidence now also suggests that vertical consolidation between hospitals and physician groups leads to increased costs (above and beyond the effects of any resulting horizontal consolidation in the physician specialties being acquired by hospitals) with unclear effects on quality.
Despite this evidence, many acquisitions that raise significant antitrust concerns go unchallenged by antitrust authorities or are allowed to proceed by the courts. This is the result of multiple factors. First, pre-merger notification to federal antitrust authorities is only required for transactions over $111.4 million in 2023 (adjusted annually for inflation). This means that many acquisitions – particularly of physician practices – go unnoticed until the merger has been finalized. These smaller mergers can still be consequential. Notably, institutional private equity firms often acquire large market shares through a series of small transactions that can be challenging to track. Evidence shows that falling beneath this threshold greatly increases the chances of a transaction going unchallenged. Moreover, limitations on available data can also make it difficult to determine which transactions raise antitrust concerns. In addition, resource limitations at federal antitrust agencies may impede their ability to bring cases with merit. Finally, antitrust authorities are not always successful in blocking mergers and acquisitions in court even when there is substantial evidence of net consumer harms.
Greater transparency and strengthened antitrust statutes could help reduce the amount of anticompetitive consolidation in health care.
To enhance transparency, legislation could require pre-merger notification to antitrust agencies if the cumulative value of acquisitions by a single parent company in a given market exceeds the reporting threshold, even if the most recent acquisition itself is of lesser value than the threshold (either in general, or specifically for those within health care markets). In conjunction, the threshold amount for reporting mergers and acquisitions could be reduced. This would directly increase the Federal Trade Commission’s (FTC) awareness of smaller mergers, particularly those that cumulatively have substantial impacts on market competition. In addition, the antitrust agencies could be tasked with systematically tracking market concentration levels in each health care market. Together, these two reforms would make it easier for regulators and to assess when more modestly-sized mergers present meaningful competitive concerns. Undertaking these new activities may necessitate additional funding.
Congress could further improve transparency about consolidation levels by making ownership structures clearer and improving existing transparency data that was implemented through rulemaking. Notably, recently proposed legislation would increase the transparency of private equity investments in health care. And the data required of insurers under the Transparency in Coverage rule could be enhanced to include information on the volume of each service that was utilized collectively by each plan’s enrollees.
To limit anticompetitive consolidation and conduct in health care markets, policymakers could also consider strengthening antitrust statutes. A recent testimony by Leemore Dafny, an antitrust expert, included some potential options. For instance, Congress could make it easier to challenge mergers by amending Section 7 of the Clayton Act to require regulators demonstrate that a merger “meaningfully” or “materially” lessens competition or “tends to create a monopoly” rather than “substantially” so.
The Clayton Act might be further modified to allow for a demonstration that a series of transactions by the same acquirer has such an effect. This could make it more feasible to challenge a series of small acquisitions that individually may not “meaningfully” lessen competition, but do so when taken together. Should policymakers consider this approach, it may be appropriate to target these changes to health care markets only.
In addition to strengthening antitrust statutes, appeals could be limited to the circuit in which the provider predominantly operates (rather the current allowance for forum shopping with appeals).
This suite of policies would all tend to reduce the rate of consolidation, which would translate to lower health care costs over time. That said, it is challenging to forecast how the CBO would score each policy. Generally speaking, policies that more directly impede anticompetitive consolidation (e.g., those that change FTC enforcement) would likely have larger estimated effects than those that focus purely on transparency.
Site neutral payments
Medicare typically pays more for a service performed at a hospital outpatient department (HOPD) than when that same service is delivered at a physician’s office. The same is true, albeit less so, for services performed at ambulatory surgery centers (ASCs). Because beneficiaries are typically liable for coinsurance equal to 20% of the Medicare price, these payment differentials also result in higher patient costs.
In addition to the direct costs, paying more for physician services at a HOPD encourages hospitals to buy physician practices. A growing body of evidence finds that hospital-physician vertical consolidation drives up costs both for Medicare and for commercial payers. Therefore, policies that limit site of service payment differentials are likely to also generate savings for the commercially-insured population.
In recent years, both Congress and the Trump Administration have taken steps to require site-neutral payments in limited circumstances. As a result, clinic visits at off-campus HOPDs and any service delivered at an off-campus HOPD that was established after November 2, 2015 are reimbursed based on the physician fee schedule (PFS) rather than the more generous outpatient prospective payment system (OPPS). However, these provisions did not address services other than clinic visits at grandfathered off-campus HOPDs, services at on-campus HOPDs, or services performed at other types of facilities such as ASCs, freestanding emergency departments, and cancer hospitals.
Requiring site-neutral payments more broadly would further reduce Medicare spending, beneficiary costs, and incentives for hospitals to purchase physician practices. Legislation could require payment based on the PFS for all off-campus HOPD services – along the lines of a proposal from Representative Victoria Spartz, which has reportedly garnered support from other offices. Congress could require the same for a subset of services at on-campus HOPDs, ASCs, and other facilities that can be appropriately performed in a physician’s office. Additionally, payments to HOPDs could be reduced to the ASC payment levels for certain services. In their June 2022 Report to Congress, MedPAC identified the set of services where such site-neutral payments would be appropriate. Similar policies have been proposed by both the Obama and Trump administrations in their budget submissions.
MedPAC estimates that aligning payments across sites of care for the list of services identified would have saved Medicare $6.6 billion and beneficiaries an additional $1.7 billion if in place in 2019, even before accounting for the potential dynamic effects on vertical consolidation. And the Congressional Budget Office estimated that paying PFS rates for all off-campus HOPDs and for certain services at on-campus HOPDs would reduce deficits by about $140 billion over 10 years.
 Beneficiaries without supplemental coverage directly pay the higher coinsurance. For those with supplemental coverage, the costs may be borne through higher Medigap premiums, their former employer, or Medicaid. These higher prices also result in higher Medicare Part B premium payments.
Part B drug payment
Medicare pays providers directly for certain drugs administered by clinicians through Part B based on its average sales price (ASP) – a measure of the typical post-rebate transaction price for the product across nearly all U.S. sales – plus 6% (or, effectively 4.3% after accounting for sequestration under the Budget Control Act). Medicare additionally pays a separate fee for the administration of the drug. Patients are typically responsible for cost-sharing equal to 20% of the total Medicare payment.
This purchasing setup dampens price competition for drugs covered under Part B relative to Part D. While providers earn more money if they can acquire drugs for lower prices, their ability to negotiate with drug manufacturers is likely lower than large pharmacy benefit managers (PBMs) or insurers. In addition to provider purchasers generally being smaller than insurers or PBMs, tools that generate price competition like formularies, step therapy, and prior authorization are generally not present in Part B.
In addition, the Part B drug payment system creates a misalignment of incentives between the purchaser (providers) and patients by incentivizing providers to administer more expensive drugs when there are competing treatment options. This incentive arises because the provider’s add-on payment (4.3% of the ASP) is larger for more expensive drugs. By contrast, in Medicare Advantage (where insurers can attempt to influence enrollee choices among competing treatments), enrollees are significantly more likely to receive the low-cost physician-administered drug in situations where there are multiple similarly-effective drug options than Traditional Medicare enrollees.
Finally, this system likely incentivizes vertical integration between hospitals and physicians. Because hospitals and health systems likely have greater market power to negotiate lower drug acquisition costs than independent physician groups, Part B drug reimbursement will tend to generate larger margins for hospital-owned practices and therefore creates an incentive for hospitals to purchase physician practices. The 340B program also helps many hospitals obtain deeper drug price discounts, as discussed in a separate section of this report.
One method to improve price competition among drugs would involve shifting the direct drug purchasing decisions from providers to payers or third-party vendors. In Medicare, legislation could set up a vendor model similar to those proposed (but ultimately retracted) by the Obama and Trump Administrations. Under such a model, large vendors with significant purchasing power would buy drugs and compete to provide them to physicians. Alternatively, Congress could shift the coverage of some drugs from the Part B benefit into Part D. One study suggested that doing so could reduce spending on affected drugs by 7 to 18%. Such savings should be weighed against the difficulties imposed on patients by the tools such as formularies, step therapy, and prior authorization that are largely generating the savings.
Congress could also reduce providers’ financial incentive to administer more costly drugs when there are therapeutic substitutes available. Both the Obama and Trump Administrations proposed doing so by shifting at least part of the 6% of ASP add-on into a flat dollar amount, which could also be done through legislation. Alternatively, Medicare could generate price competition in these cases by paying the same amount for drugs that are substitutes for each other. This gives providers a strong incentive to choose the cheaper option. MedPAC has suggested doing so when biologic drugs have nearly-identical biosimilar competitors.
Congress embraced a similar goal in the IRA by temporarily increasing the add-on payment for biosimilars to 8% of the reference biologic’s price (instead of 6%) as long as the biosimilar had a lower ASP than the reference product. This gives both products an incentive to compete on price while giving a leg up to the new competitor. Given that providers are generally averse to switching patients who are well managed by the originator biologic drug, it is not unreasonable to preference biosimilars in such a way. Congress could extend this policy beyond its 2027 expiration date.
Medicare could generate greater savings by extending this type of policy to classes with multiple clinically comparable branded products. This type of policy would come with additional considerations, like accommodations for cases where products may not be interchangeable for some patients.
The 340B program requires drug makers to give certain qualified health care providers large discounts (roughly 25-50 percent) on drugs they purchase. The set of “covered entities” that receive these prices includes disproportionate share hospitals, along with several other provider types like federally qualified health centers or hemophilia centers.
Over time, the program has grown. From 2000 to 2020, the number of covered entities has increased from 8,100 to 50,000. By 2021, 340B covered entities purchased $43.9 billion in covered drugs. In large part, this is because of a 2010 expansion that allowed covered entities to contract with an unrestricted number of pharmacies to dispense discounted drugs.
While the program mandates that drug makers give entities large discounts on drugs, it does not require that those discounts are passed on to those purchasing the drugs (private insurers or public programs like Medicare). Nor does it require that the hospital use those funds in any specific way. A hospital can, for example, purchase a physician-administered drug at a large discount but sell it to Medicare or private insurers at full market prices, keeping the difference as profit.
This creates problematic incentives. First, a 340B hospital has an incentive to acquire private physician practices since it can purchase drugs cheaper than the practice can, increasing consolidation in those fields. Second, these discounts could mean hospitals earn high margins on products that are more expensive for insurers, encouraging them to prescribe costlier medicines.
One paper found that hospitals have responded to these incentives by dispensing more drugs, acquiring more physicians in some drug-intensive specialties, and treating fewer patients on Medicaid. However, it did not find evidence that hospitals provided more or better care to low-income patients. On average, 340B hospitals also have higher Part B spending per beneficiary, which is consistent with (but not proof of) them altering their prescribing behavior due to the program.
As CBO notes, “a policy change that applied drug discounts under the 340b program on a patient-level basis—that is, to patients with certain characteristics rather than to all patients at certain sites of care—might reduce hospitals’ and physicians’ incentives to consolidate.” While we are not aware of a formal score for this kind of proposal, we expect cost savings to be small in the budget window, partly because of uncertainty and because it would take some time for this policy to play out. Still, this reform improves incentives, more effectively subsidizes providers treating many low-income individuals, and would potentially fit well within legislation that reformed other aspects of the program.
Alternatively, policymakers could reduce Medicare payment rates to 340B hospitals to reflect their lower acquisition cost. The Centers for Medicare and Medicaid Services estimated that reducing Part B drug reimbursement for 340B drugs from 106% of ASP to 77.5% of ASP (with certain exceptions) would reduce Medicare spending by $1.6 billion annually. HHS’ recent efforts to do just that was recently overturned in court because they were implemented without a requisite survey of hospitals’ drug acquisition costs. While the agency could potentially address that, Congress could pass legislation along these lines without first completing such a study.
Improving and focusing competition in Medicare Advantage
Roughly half of Medicare beneficiaries now receive coverage through a private Medicare Advantage (MA) plan. The MA program is supposed to harness health plan competition to reduce costs, increase consumer choice, and allow for ongoing innovation in plan design (in turn, ideally improving patient outcomes).
However, in practice, competitive pressures are often weak and much of the competition that does occur is focused on making a health plan’s enrollees appear sicker (and, therefore, more costly to care for) than similar enrollees in the Traditional Medicare (TM) program. Doing so generates higher risk-adjusted payments from the government to insurers, allowing them to offer more generous plans relative to those offered by other insurers. A poor bidding structure and complex choice environment also dampen MA health plan competition. Nearly half of all MA beneficiaries nationally are enrolled in plans offered by just two insurers – United Healthcare and Humana.
Despite evidence that MA plans do produce cost efficiencies, reflecting lower utilization, they are still paid more than the costs for a similar beneficiary in TM. And evidence suggests that only about half or less of these extra payments are passed through to enrollees.
Currently, MA plans are paid a risk adjusted, monthly capitated amount per enrollee tied to benchmarks based on local TM per beneficiary spending. MA plans bid how much they project it will cost them to cover the standard Medicare Part A and B benefits for an average-risk beneficiary and receive a portion of any difference between the benchmark and bid, which can be used to offer extra benefits or reduce premiums paid by enrollees.
There are numerous policy options to improve competition in MA. While much attention has focused on competitive bidding or premium support proposals, we focus here on smaller changes to the program.
First, rather than the government collecting about 35% of the difference between plan bids and benchmarks, more of that difference could be paid to plans to fund extra benefits or reduce premiums. Limiting what is effectively a tax on lower bids would encourage greater price competition. On its own, this proposal would likely increase federal Medicare spending and the generosity of MA benefits, but it could be combined with a reduction to MA benchmarks. Empirical evidence suggests that the reduction to benchmarks would also enhance MA plan competition, causing plans to price more aggressively. Federal savings or costs from this proposal will depend on the magnitude of reduction to benchmarks.
Others have argued that a clearer choice infrastructure could improve MA plan competition, for instance by more clearly communicating premium reductions below the standard Part B premium to potential enrollees.
Other policy reform proposals, including some from MedPAC, aim to limit the ability of MA plans to game the risk adjustment system in an effort to increase payments from the government. Lessening this incentive would discourage insurers from focusing as many resources on coding intensity and, instead, encourage more competition over prices and quality. CBO estimates that increasing the adjustments for coding intensity such that MA plan payments better reflect the actual expected costs of enrollees would reduce deficits by about $50 billion over the 10-year budget window (budgetary savings estimates would likely be substantially higher today).
The current administration also recently expanded risk adjustment data validation (RADV) audits to reduce excess or inappropriate payments made to plans. However, there are arguments to allocate more funding through legislation to expand this further by auditing more contracts and potentially increasing penalties, which is likely to have a large return-on-investment for the federal government.
Some health care markets are already relatively consolidated. In these markets, dominant firms can use contracting terms to impede potential competitors and raise costs. These contract clauses can be used by both providers and insurers. Specifically, they include:
- Anti-tiering clauses—Requirements that the dominant provider is not on a worse insurance “tier” than any other provider. This impedes health plans’ ability to incentivize enrollees to pick a different, lower-cost provider.
- Anti-steering clauses—Clauses that disallow the insurer from “steering” (or incentivizing) enrollees to see other providers.
- All or nothing contracts—Provisions that the insurer must include every provider affiliated with the dominant system in their network or none (e.g., a dominant hospital could require all affiliated outpatient facilities or physician groups to be included in network if the insurer wants to include the hospital). This could allow the dominant provider in one part of the market to extend its power to other parts of the market.
- Most favored nations clauses—Requirements that providers guarantee a dominant insurance plan the most favorable pricing among insurers. This protects the position of the most dominant insurer in a market.
These types of contracts are not necessarily problematic in a competitive health care market. For example, an insurer could choose to not contract with providers who demand all or nothing clauses, or demand pricing concessions in exchange for such an agreement. However, if a provider or insurer market is already relatively consolidated, they can help providers suppress competition. For that reason, efforts to restrict this type of contracting practice would primarily target markets that are already consolidated and interfere minimally with otherwise competitive markets.
Eliminating these contracting provisions would generally increase competition in many health care markets and lower premiums. However, it is important to recognize limitations of such a policy. Because these policies do not directly change market concentration, providers and insurers will try to use leverage to seek higher prices through other mechanisms.
The bipartisan Lower Health Care Cost Act included provisions that would have banned the use of these contracting terms. That bill was reported out of the Senate HELP Committee in 2019. The CBO estimated that these provisions would modestly reduce premiums in the private insurance market and, in turn, increase federal revenues by $1.1 billion over the budget window. The CBO recently indicated that new evidence would lead them to increase their estimated cost savings.
More recently, Representative Spartz introduced HR 8135, which would direct the Federal Trade Commission and the Department of Justice to evaluate the effects of these clauses and the agencies’ ability to enforce antitrust laws with respect to their use.
While legislative interest has arguably been strongest in provider and insurer markets, Congress should consider whether similar contracting provisions impede competition in the drug market. For example, biologic drug makers have been accused of using “rebate traps,” in which they condition large rebates for a biologic drug on the exclusion (or worse formulary placement) of a competing biosimilar. Because providers are generally unwilling to switch well-managed patients to a biosimilar, this can be a powerful exclusionary contracting practice. Policymakers might consider whether contracts should, for example, be prohibited from referencing the formulary placement of a competitor in these cases. The use of other contracting provisions, like most-favored nation clauses by Pharmacy Benefit Managers (PBMs), are worth similar consideration. While we are aware of less legislation on these topics, contracting reforms along these lines are a natural extension of the competitive concerns noted in the provider and insurer markets (indeed, the largest PBMs are owned by insurance companies).
Reforms to the pharmacy benefit manager market
Pharmacy Benefit Managers (PBMs) work on behalf of insurers to develop formularies, negotiate prices with drug makers, and contract with pharmacies. Policymakers have paid increasing attention to these intermediaries, arguing that their behavior may increase drug costs. While this represents one of the most active areas of policy interest in Congress, it is important to recognize where Congress may be able to lower drug spending and where such outcomes are unlikely.
Concerns about PBMs are largely based on a few key facts. First, the PBM market has become relatively consolidated over time, with three firms—Caremark, Express Scripts, and OptumRx—controlling 80 percent of the market. (The major PBMs are also now vertically integrated with insurance companies.) In addition, PBMs are also alleged to facilitate pricing structures that increase costs for some consumers. Notably, they may encourage drug makers to set very high list prices while selling to insurers or PBMs at much lower “net” prices. Because consumer out-of-pocket liability can be a percent of the list price, this increases costs to sicker enrollees taking expensive medications. They may also affect competition within the pharmacy market by steering patients to use pharmacies that they own. Using selective contracting can lower costs to consumers but also may drive independent pharmacies out of business. Finally, PBM contracts are relatively opaque, which makes it harder to assess the harms or benefits of their behavior.
There is a strong argument for increasing transparency into the role of PBMs. Absent a better understanding of current contracting behavior, it is hard for plan sponsors or insurers to make informed decisions and for Congress to craft effective policy in this area. There are a number of policy interventions aimed at improving transparency in this area.
The Federal Trade Commission has recently announced an investigation into PBM markets that may help further illuminate PBM contracting practices and their role in drug spending. Recent requests from the House Oversight Committee may also produce similar information.
Legislative proposals like the Pharmacy Benefit Manager Transparency Act of 2022 would increase transparency into PBM payments and restrict some types of contracting. For example, PBMs would be required to disclose drug rebates, costs, prices, fees, and other information, while disallowing contracting practices like “spread pricing”. The Lower Health Care Costs Act of 2019 included many of the same transparency provisions. The CBO concluded that additional transparency would allow plan sponsors to better evaluate contracts and trigger more competition between PBMs. The result was a reduction in commercial market spending, and a $1.7 billion decrease in the deficit over ten years.
While we view increasing transparency as reasonable policy goal, it is important to recognize where other prominent policies may not lower costs in this area. Efforts to constrain the form of payment to PBMs—like bans on spread pricing—may have minimal effects on costs because they do not fundamentally change the market power of these firms. PBMs can secure payments via many avenues (this often includes administrative fees) and can likely transition rather seamlessly away from spread pricing if needed.
Some policymakers have been critical of PBMs’ use of step therapy (or “fail first” policies) and prior authorization. However, efforts to restrict these types of utilization management are likely to increase costs. A recent study of the Medicare Part D market, for example, shows that prior authorization reduced spending on affected drugs by over 25%. This is an important consideration for policymakers regarding the use of these tools.
Policymakers have also long been critical of PBM’s potential role in the growth of rebates which drive a wedge between “net” prices paid by insurers and higher “list” prices that can be the basis of individuals’ out-of-pocket spending. This has triggered recent proposals to pass rebates through to the point of sale, including the Trump Administrations’ “rebate rule” proposal. While efforts to eliminate this type of pricing may be well-motivated and reduce out-of-pocket costs for consumers, doing so will likely increase costs overall.
 This is a strategy where PBMs charge an insurance plan more than they pay a pharmacy for a drug and keep the difference.
Making the No Surprises Act more pro-competitive
The No Surprises Act, which took effect in 2022, prohibits most forms of surprise out-of-network billing. While not the primary intent, the bill may discourage consolidation in some specialties, like emergency medicine and anesthesiology, where large staffing companies appear to have aggressively leveraged surprise billing prior to the law.
However, some elements of the law almost certainly work in the opposite direction. In particular, arbitrators are told to consider a host of factors when determining the appropriate out-of-network payment in surprise billing disputes. The “additional circumstances” to consider in these cases include the prior contracted rates between the two parties. These rates are likely to be highest for the most consolidated providers. For example, published data show that prior contracted rates were substantially higher for the two largest emergency physician staffing companies – TeamHealth and Envision – than competing provider groups. To the extent that prior contracted rates influence arbitration decisions, this creates a financial incentive for smaller medical groups to sell to the larger companies with higher prior contracted rates (and for doctors to go work for these larger companies), effectively entrenching the previous status quo.
While the Congressional Budget Office (CBO) anticipated that arbitration decisions, on average, would follow close to median in-network payment rates (termed the “qualifying payment amount”), it is unclear if that is true in practice or if other factors, like prior contracted rates, play an important role in decisions. Removing prior contracted rates as a consideration in arbitration under the No Surprises Act would almost certainly be scored by CBO as reducing deficits through its downward effects on commercial insurance premiums. The precise magnitude of this effect is uncertain and would depends substantially on how often it is currently factoring into arbitration determinations.
Along similar lines, Congress could also remove “the market share” held by each party in an arbitration dispute as a factor for arbitrators to consider, although it is unclear whether this factor is expected to increase or decrease awards for companies with higher market shares locally.
Disclosures: The Brookings Institution is financed through the support of a diverse array of foundations, corporations, governments, individuals, as well as an endowment. A list of donors can be found in our annual reports published online here. The findings, interpretations, and conclusions in this report are solely those of its author(s) and are not influenced by any donation.