Senate Commerce Committee Issues Testimony From Northwestern University Professor
* * *
In contrast to other developed countries,
Therefore, despite many contentions to the contrary, a market-based system remains the best mechanism for providing the appropriate incentives for long term welfare maximization in the
Concerns about the appropriate role for markets in healthcare are perhaps most frequently discussed in the world of pharmaceuticals. These discussions are motivated by high and rising pharmaceutical prices. While many claim these high prices provide prima facie evidence of a market failure, in reality they are the result of the complex and delicate balancing of incentives that sits at the center of the
This delicate balance is necessary because market failures at the center of the innovative process for developing new drugs requires some degree of market intervention in the first place. This failure results from that fact that the scientific advancements generated by firms developing innovative pharmaceutical products are essentially a public good, i.e. the knowledge is effectively non-rival and non-excludable.1 Rational firms realize they will be unlikely to capture a sufficient amount of the value generated by the large, fixed, and sunk investments necessary to bring a product to market. This results in an economic phenomenon known as "hold up" whereby firms, absent some form of government intervention, are unwilling to make value creating investments in the first place.
* * *
1 The degree to which this is fully a public good depends on how much information can be gleaned from the actual product, the regulatory filings, and the published research. For example, small molecule products can be more easily reverse engineered and therefore absent intellectual property protections are relatively easier to copy. Biologic products, however, have a more complex production process and therefore copying the technology is easier than making the product de novo but harder than for a small molecule product.
* * *
To address this initial market failure, governments offer various forms of intellectual property protection. Through patents or other forms of market exclusivity, governments arm firms with time limited periods of enhanced market power that allow them to capture a larger portion of the value created by their innovative products. During this limited time period, higher prices than would otherwise exist curtail some access to valuable medicines. This reduced access is deliberately traded off for the development of new products in the future.2 These new products, however, provide access to patients for whom there would otherwise be no available treatments.
In this way, policies governing the development of pharmaceutical products involve trading off the static inefficiency of reduced access to products today in order to create the dynamic efficiency of the increased development of new products. To the extent the value created by the new products exceeds the welfare losses resulting from the high prices (and decreased quantity), the granting of these periods of market exclusivity is welfare enhancing. This could be true even if the prices today are quite high.
This tradeoff is a source of much of the controversy surrounding prescription drug prices because it involves some number of readily identifiable individuals who are unable to access existing and potentially life-saving medications.3 Unsurprisingly, this particular form of a lack of access garners large amounts of press and political attention. However, it is critical to remember a perhaps far greater access problem for patients suffering from conditions for which no treatment options exist at all.4 For these individuals, there is no price at which they can purchase a treatment. These patients will gain access in the future only as a result of the dynamic incentives created by intellectual property protection. As we consider the optimality of policies governing the pharmaceutical market, we must balance the oft-discussed need for access to existing products with the less-discussed lack of access from the absence of effective treatments.
* * *
2 In considering this tradeoff it is important to consider the role of health insurance in mitigating decreased quantity resulting from high prices. To the extent that insurance mitigates some of this quantity decline it is possible that the welfare loss are smaller than would be expected. See D. Lackdawalla and
3 Garthwaite, Craig, and
4 This is particularly true because the impact of high prices on quantity is far more complicated in a world of widely available health insurance. Those who are insured may not suffer as much decreased access as they would in a market without third party payment. However, those for whom drugs do not exist certainly will not access a treatment at any price.
* * *
A central parameter of this tradeoff of static and dynamic incentives is the relationship between the elevated prices paid for prescription drugs today and the incentives of innovative firms to develop new products in the future. Economic research has clearly documented a relationship between increased market size and investments in research and development.5 Therefore, to the extent high prices signal expected economic returns for the providers of the risk-based capital necessary for innovation then the prices could represent a welfare enhancing policy choice. However, if the revenue generated by high drug prices is instead captured by other parts of the value chain there are valid concerns that our current policies are not providing an optimal level of innovation to outweigh the welfare losses from the price related reduced access.
Determining the optimality of this tradeoff in today's market requires a more careful understanding of the pharmaceutical supply chain. In particular, it is important to understand how various firms capture a share of the value created by innovative pharmaceutical products. Figure 1 provides a broad overview of this supply chain and the flow of funds across firms at its various stages. Perhaps most important for today's hearing is the relationship between manufacturers, payers, and pharmacy benefit mangers (PBM), which is depicted in the figure's upper right corner.
While largely unknown to customers, PBMs are the private firms that effectively manage all aspects of insurance coverage for pharmaceuticals. Despite their relative lack of attention, these firms occupy a central role in nearly every facet of the pharmaceutical distribution and insurance market. At a high level, PBMs sign contracts with plan sponsors (e.g. risk bearing health insurers or employers) to undertake activities such as negotiating drug prices, establishing pharmacy networks, processing pharmaceutical claims, and developing drug formularies.
* * *
5
* * *
In return for these activities, PBMs earn revenue through a variety of means. These include, but are not limited to, direct per member per month (PMPM) fees paid by plan sponsors, the ability to keep a negotiated share of the rebate (i.e. the discount from the manufacturer that the PBM is able to negotiate), spread pricing (i.e. the difference between what a PBM is paid by a plan sponsor for a drug and what they pay to the pharmacy to fill the prescription), and various administrative fees from manufacturers.
The primary role of PBMs is to help manage the static inefficiency resulting from high prices. Historically, these firms emerged to implement some degree of managed care and negotiation to the pharmaceutical benefit offered by plan sponsors. In particular, they allowed relatively small insurers to pool together and negotiate as a group against manufacturers.6 By constructing formularies, PBMs negotiate lower prices and can increase access to products and potentially to insurance overall. Of course, such activities limit revenues to pharmaceutical manufacturers and have been shown to blunt incentives to develop new products.7 This demonstrates the importance of the role of PBMs in the tradeoff central to drug development.
It is important to note that if the construction of formularies represents the preferences of consumers for access to new products, a reduction in innovative activity is not necessarily a problem. After all, our goal is to maximize welfare not innovation. However, if the reduction of revenues to manufacturers comes instead from PBMs capturing an inappropriately large fraction of spending as profits - there could be concerns about whether the pharmaceutical market is operating in a way that maximizes welfare. In particular, concerns about whether the welfare losses from the lack of access today are sufficiently offset by incentives to develop new products in the future. These concerns are central to today's hearing investigating the PBM market.
* * *
6
7
* * *
Concerns about this possibility stem from features of the existing market. For example, the PBM market is dominated by three large firms - Caremark, Express Scripts and
Given these concerns, I will concentrate my testimony today on the relationships between plan sponsors (i.e. third party payers such as insurers and employers), PBMs, and manufacturers. In particular, I will focus on the degree to which features of these relationships may allow PBMs to capture more value than might be appropriate or whether negative features of the market instead reflect the actions and incentives of firms in other parts of the value chain.
A consistent point to consider throughout my testimony is that any analysis of this market is meaningfully hampered by a lack of information about numerous features of the contractual arrangements between the various types of firms. While it is easy to identify potential areas of concern, without more information about the nature of these arrangements it is difficult to truly understand the validity of such concerns. Therefore, Congressional action in this area should be initially focused on creating more insight for regulators into these areas. That said, I will also highlight several policy options that exist to more directly confront potentially undesirable features of the current pharmaceutical market without generating unintended consequences.
I. Pricing, Rebates, and Cost Sharing in the
In the
* * *
8 In earlier testimony, I discussed the potential benefits and concerns of this vertical integration. This testimony is available at: https://www.judiciary.senate.gov/download/garthwaite-testimony.
* * *
PBMs are able to secure the discounts based on their ability to shift customers across competing therapeutic substitutes. For example, if there are two brand-name statin medications that treat high cholesterol, the PBM can place the product from a manufacturer offering a lower net price on a more preferential tier of its formulary, thus lowering the out-of-pocket payments from an individual enrollee when they purchase the drug. This should result in this product selling higher quantity, albeit at a lower price. In extreme cases, a PBM could entirely exclude a product from its formulary if the manufacturer is unwilling to provide a sufficiently low net price (i.e., they are unwilling to pay the PBM a sufficiently large rebate). The use of exclusion lists has grown in recent years. Figure 4 shows the number of products that are excluded by the largest PBMs. It is this ability to credibly threaten to move volume across products that results in larger discounts from the list price.
The increased use of strict formularies and exclusion lists has contributed to a growing spread between the list and the net (i.e. post rebate) price. Figure 5 depicts these prices from 2014 - 2020 and documents a large spread between the publicly known and often discussed list prices and the actual prices received by manufacturers. This figure demonstrates that any discussion of list prices provides an, at best, incomplete picture of the returns to innovative manufacturers in this market.
The spread between list and net prices has resulted in a large amount of total rebates in the system.
Figure 6 shows that in 2016, pharmaceutical manufacturers paid total rebates of approximately
While the increasing magnitude of rebates in the system is often discussed in a negative light, it is not necessarily a problem. After all, higher rebates could simply reflect more sophisticated or effective bargaining by PBMs. The ultimate question is which parties in the supply chain capture the value of those rebates and what features of the market determines the ability of those firms to capture that amount of value. The split of the rebate between the PBM and the payer is dictated by a contract that is the result of a bilateral negotiation between those firms. The specifics of this contract depends on the relative bargaining power of the two parties. Figure 7 contains estimates of the existing contractual structure in the commercial market with respect to rebates over time based on whether plan sponsors are large or small employers. From 2014-2018, there has been a marked increase in employers with PBM contracts that entitle them to receive 100 percent of the rebates. By that year a majority of both types of employers were using such contracts.
Unsurprisingly, PBMs often point to increasing rebates as evidence of their effectiveness. It is not clear this is accurate. After all, a large rebate can come from a higher list price, a lower net price, or a combination of both. If rebates are only the result of higher list prices then the actual price paid in the market (and the return to manufacturers) has not necessarily changed. It is tempting to think that in that situation the high list prices have little economic effect. However, even in contracts where 100 percent of the rebate flows to the plan sponsor, higher list prices can negatively impact other market participants.
In particular, high list prices can have direct and economically meaningful impacts on consumer out-of-pocket payments. This relationship between cost sharing and list prices results from the desire to maintain the confidentiality of negotiated prices. Such confidentiality provides stronger incentives for larger discounts. For this reason, the size of rebates paid to each PBM is kept strictly confidential, up to and including onerous audit restrictions in the contracts that limit the ability of the payer to monitor the financial activities of the PBM.9 To maintain this confidentiality, consumers whose cost sharing for pharmaceutical products is tied to prices (either because of a deductible or percentage based coinsurance) make these cost sharing payments as a function of the list rather than the net price.10 Thus, any inefficiencies that create incentives for higher list prices (even if those are entirely offset by rebates) affect consumer out of pocket spending.11 In the presence of liquidity constraints, this cost sharing could meaningfully reduce access to drugs in ways that magnify the static inefficiency of high drug prices. For this reason, high cost-sharing is not simply a financial inconvenience for consumers. Recent evidence has shown that increased cost sharing for consumers results in the decreased use of prescription drugs and increased mortality.12
* * *
9 Weinberg, Neil, and
10 This is mainly an issue for consumers enrolled in certain high-deductible health plans, as well as Medicare beneficiaries.
11 While the number of consumers with this type of cost-sharing has grown, it should be noted that customers in the pharmaceutical market are largely shielded from list prices.
* * *
The importance of cost sharing for prescription drugs has grown over time. Consider the evidence in Figure 8, which contains the average annual out of pocket payment for Medicare patients purchasing insulin. According to these data, in 2018 nearly 30 percent of Medicare patients purchasing insulin were paying more than
Insulin is not the only place where we see high cost-sharing. Overall, prescription drugs enjoy far less insurance coverage than other parts of healthcare. Figure 9 shows that insured patients are exposed to only 3 percent of their hospital spending. In contrast, patients directly pay 15 percent of their prescription drug spending out of pocket.
Given the negative health and financial effects of high cost-sharing, it is at first puzzling why such high cost-sharing persists in the market. Cost sharing is intended to be a form of utilization management that attempts to overcome the potential moral hazard arising from patients that are fully insured for their pharmaceutical purchases. This moral hazard could occur both through the overconsumption of products where the price exceeds value or more often from purchasing products that have less expensive therapeutic substitutes. Both of these would be negative features of an insurance product that cost sharing was intended to mitigate.
The ability to use cost-sharing to move patients across products is a key tool that PBMs use to negotiate lower net prices from manufacturers. However, we increasingly see high cost-sharing on products that are unlikely to be overconsumed (e.g. insulin and oral oncology products) or in areas where there are no therapeutic substitutes. This suggests this high cost-sharing serves goals other than simply utilization management.
It is not obvious that cost-sharing at the levels we observe is an independent strategic choice PBMs undertake to maximize their profits. After all, if plan sponsors desired less onerous cost-sharing they certainly could instruct their PBMs to construct such a formulary. In fact, in recent testimony before
* * *
12 A. Chandra,
* * *
Instead of signing contracts that pass rebates to customers, plan sponsors increasingly demand higher rebates from PBMs - even when those rebates are not associated with lower prices.14 Such rebates come from surging list prices and contribute to higher cost sharing payments by patients. It appears that this is because the combination of large rebate payments and high cost-sharing for expensive products provides a mechanism for plan sponsors to offer lower premiums to healthy patients and higher expected costs to sick patients requiring expensive medications.
Consider the stylized example in Figure 10 where a consumer in the deductible period pays the full list price of the drug. This customer does not benefit from any of the negotiation efforts of the PBM. Both the PBM and plan sponsor, however, can profit from the consumer's purchase of a prescription drug because these firms still collect a rebate when one of their patients buys a pharmaceutical product. This is true even when the product is entirely paid for by the patient. A similar logic exists when a patient makes a very large cost sharing payment because of coinsurance. Sponsors are then able to use those extra rebate dollars to lower premiums or decrease the cost of employer provided healthcare. In this way, high cost sharing combined with large rebates reintroduces medical underwriting and unwinds the community rating of health insurance premiums.
This stylized example is not simply an academic exercise. In a recent
* * *
14 https://www.finance.senate.gov/imo/media/doc/Grassley-Wyden%20Insulin%20Report%20(FINAL%201).pdf
15 https://www.finance.senate.gov/imo/media/doc/Grassley-Wyden%20Insulin%20Report%20(FINAL%201).pdf
* * *
Understanding these dynamics is important in considering the causal role of PBMs with respect to increasing list prices and rebates. It suggests that much of the furor at PBMs over increasing list prices, rebates, and cost sharing may be aimed at the wrong target. If such contractual features are being dictated by PBM clients (i.e. plan sponsors) than regulators should more carefully consider the incentives of those plan sponsors when constructing policy in this area. Furthermore, as I discuss below, if the concern about high list prices is primarily motivated by the effect on cost sharing there are policies that can be considered which more directly address this cost sharing.
II. Lack of Transparency in Financial Relationships in the Value Chain
While a large portion of plan sponsors have signed contracts that allow them to collect all of the rebates associated with prescription drug purchases by their customers, there are still many contracts where the PBM receives a percentage of the rebate as compensation. In addition, PBMs collect other fees that I discuss below which are also a function of the list price. Some have proposed that this provides a perverse incentive for the PBM to prefer higher list priced products where there is a large rebate compared to even lower prices products with a smaller rebate.
The concern about PBMs being attracted to higher-priced drugs can be best demonstrated by a simple example. Consider a drug that currently has a list price of
Ultimately, the unanswered question is whether the
Strong competition is even less likely to emerge if payers are unaware of the full scope of surplus created by their prescriptions. As discussed above, many large firms hire sophisticated benefit consultants and increasingly demand fully transparent contracts that provide them a complete picture of all "rebate" dollars. In theory, this provides information about the surplus created by their prescriptions. That said, there are reasons to be concerned that despite these efforts at disclosure, payers remain unaware of all of the funds (particularly those not labeled as rebates) flowing between the PBM and the manufacturer. For example, in addition to rebates, PBMs also receive various administrative fees and other payments from manufacturers - fees that are often a function of the list price of a drug.
The PBM and the manufacturer determine which of these payments are classified as "rebates" (and therefore covered by the price transparency and rebate sharing requirements), and what is instead an "administrative fee" (that does not need to be disclosed or shared).16 These fees are not trivial - for some contracts they can account for 25-30% of the money moving between the manufacturer and the PBM.17 Furthermore, since these fees are often structured as a function of the list price there is little economic distinction between an "administrative fee" and a "rebate." Describing this system, the
* * *
16 Eickelberg, Henry C. 2015. "The Prescription Drug Supply Chain 'Black Box': How it Works and Why You Should Care."
17 Dross, David. 2017. "
18 https://www.finance.senate.gov/imo/media/doc/Grassley-Wyden%20Insulin%20Report%20(FINAL%201).pdf, page 81.
* * *
If we consider the simple example above, the situation for the payer could be even worse if, instead of offering a "rebate" of
To further complicate matters, sophisticated payers hoping to gather more information about the flow of funds between the PBM and manufacturers that results from their prescriptions often face meaningful restrictions on the ability to audit their PBM-payer contracts.20 These can include the exclusion of particular auditors that are deemed to hold views that are hostile to PBMs, requirements that audits be held at the headquarters of the PBM, unwillingness to provide contracts with manufacturers, restricted access to claims data, and strict limitations on the number of years that can be audited.21 While many of these restrictions can be cast as attempts to maintain rebate confidentiality, they also increase the amount of asymmetric information between PBMs and payers about the amount of available surplus. Such information asymmetries can affect the efficiency of bargaining between these two groups.
As a result of these concerns, some have proposed policies where PBMs are not allowed to have contracts in which they are compensated based on the size of the rebate or the list price of a product. While this would certainly eliminate any perverse incentives for large rebates, it would also diminish the incentives for PBMs to push for large discounts. If the primary motivation for such policies is an underlying concern about the competitiveness of the PBM market, eliminating the ability for firms to sign incentive compatible contracts could have meaningful unintended consequences 19 To the extent manufacturers have preferences about this labeling it is likely related to the intersection with cost sharing discussed above. Note that high cost sharing impacts manufacturer revenue by reducing demand for pharmaceutical products.
* * *
20 Weinberg, Neil, and
21
* * *
In a similar vein, the
It is perhaps not surprising that policies from both parties are coalescing on attempting to end rebates. Frustrated by rising drug prices, people are looking for a scapegoat and a system of shrouded prices by large firms fits a convenient narrative. That said, it would be extremely unwise to limit the ability of PBMs to negotiate large discounts. Instead of ending the current system of confidential rebates, I've proposed (along with
III. Congress Should Address Cost Sharing and Price Negotiations More Directly
While the optimality of the existing PBM market remains unclear, it is becoming apparent that
* * *
22
23 Garthwaite, Craig, and
* * *
As a starting point, there is a clear case for a reform to Medicare Part D's reinsurance program. Currently, this program blunts the incentives of firms to negotiate price discounts for the most expensive drugs and increases consumer cost sharing. Figure 12 shows the distribution of spending responsibilities under Part
Furthermore, for high priced products the private firms empowered to negotiate on behalf of Medicare are largely shielded by reinsurance from the costs of most price increases - limiting the ability of the market to lower these drug prices. Perhaps more concerning, PBMs operating in both the commercial and the Part D markets may face different incentives for rebates across these different markets and could use the confidential nature of rebates to unnecessarily increase government Part D spending. Initially, reinsurance was not a dominant feature of Part D. This has changed. Figure 13 shows the average national plan bid across Part D firms by its component parts - the direct subsidy from the government, the base premium from the enrollee, and the expected reinsurance payment. These data show that from 2007 to 2018, the reinsurance component of Part D spending has grown from a relatively minor part of the program (25% of the plan bid) to the dominant source of payments to firms under Part D (60% of the plan bid).
This level of reinsurance shields plans from the costs of the most expensive specialty drugs - a category of products that represents a growing share of overall prescription drug spending. While such a large amount of reinsurance may have been necessary to attract plans to the newly established Part D market, it is highly unlikely this remains true today. Part D is now an established market where firms have sufficient data to make reasonable projections about potential risk. Therefore, I propose that
Beyond changing the incentives to negotiate prices, it is clear we should find policy solutions to pass along more of the negotiated discounts to consumers. However, it is critical that any policy solution saves the proverbial baby while throwing out the bathwater by maintaining the ability of PBMs to effectively negotiate larger rebates with manufacturers. Therefore, I propose that PBMs be required to base cost-sharing payments on a number that more closely approximates the net price of the product even if it is not the exact net price associated with that purchase. For example, this number could be the average net price across PBMs for that product, the average net price for the therapeutic class, or the minimum price paid in the market, i.e., the Medicaid best price. Assuming PBMs have sufficient ability to modify their formularies, any of these options should still expose the patient to enough of the cost of the product to address moral hazard concerns while not exposing consumers to artificially high prices that unwind the generosity and efficiency of the insurance contract.
Some have complained that policies that pass along rebates to consumers at the point of sale would lead to higher premiums. This fact is almost certainly true. However, this is not necessarily a problem. Our current system of using cost sharing by patients requiring expensive products to lower the premiums paid by healthier patients subverts many popular policy goals regarding the treatment of pre-existing conditions in the health insurance market. In addition, these higher premiums would reflect, in part, a more complete insurance product. It is not immediately clear consumers are fully aware of the financial exposure they have to expensive medications, and therefore we should not think that increasing the completeness of insurance in this setting is clearly a negative outcome.
* * *
24 As I discuss below, very large consumer cost sharing (such as the 5 percent of spending that patients must pay under Part D) can decrease the efficiency of insurance.
* * *
IV. More Information is Needed Before Implementing New Policies Aimed at PBMs
The role of various entities in the supply chain is clearly complicated. Pharmaceuticals move through a relatively lengthy supply chain inhabited by private firms with differing incentives, information, and market power. Given their central role in both negotiating prices and establishing formularies, it is tempting to blame PBMs for every negative feature of the system we observe. And it is possible that such blame may ultimately be valid. However, it is also apparent that we simply lack the information necessary to determine the degree to which these aspects of the market are actually caused by the independent motivations of PBMs to maximize profits versus how much they reflect the incentives of other firms in the value chain. For example, as mentioned above PBMs have offered contracts where rebates are passed along to customers at the point of sale and plans sponsors have largely avoided those plans. This suggests a more complicated story is necessary to explain the current market dynamics.
Given the uncertainty in this area, it is incumbent on policymakers and regulators to gather more information before attempting to develop and implement solutions. Certainly, the recent
Senate Commerce Committee Issues Testimony From David A. Balto Law Offices Antitrust Attorney
NICB PARTNERS WITH MARYLAND GOVERNOR HOGAN ON POST-DISASTER FRAUD PROCLAMATION
Advisor News
- Debt is limiting American retirement savings
- How starting a conversation with a stranger is like dating
- Year-end planning errors retirees should avoid
- Social cultivation: How to talk with wealthy prospects
- How important is a written financial plan?
More Advisor NewsAnnuity News
Health/Employee Benefits News
- Private health insurers in Colorado will need to cover abortion care beginning in January
- Generations
- Affordable health insurance may just be a few clicks away for Texans
- Disruption hitting Medicare for 2025
- Choosing a health insurance plan
Choosing a health insurance plan
More Health/Employee Benefits NewsLife Insurance News