Off-Target: The Pelosi Proposal to Lower Medicare Part D Drug Prices  


The purpose of paper is to show that the binding arbitration proposal championed by Speaker of the House Nancy Pelosi to lower Medicare Part D drug prices is off-target.

We believe that there three broad categories of drugs where binding arbitration based on some neutral party’s estimate of value is more appropriate than mano-e-mano direct government negotiations with heavy-handed bargaining chips sometimes used to reach agreements.

  1. High PMPY  large molecule biologic drugs where patented therapeutic equivalents would be rare.
  2. High PMPY cell therapy cancer drugs where outcomes are uncertain.
  3. High PMPY orphan drugs due to costs having to be amortized over patient population in the hundreds.

The problem is we estimate that only 19% of total 2017 Medicare Part D list price drug spend is suitable for Pelosi’s binding arbitration scheme.



Now that the Democrats are in control of the U.S. House of Representatives, they have taken the lead in Congressional efforts to lower prescription drug (Rx) prices with two proposals to modify Medicare Part D.  In addition to uncertainty as to whether either will be effective,  support for the two proposals within the Democratic Party is divided sharply on ideological grounds.

One proposal is a binding arbitration scheme using a neutral third-party arbitrator to evaluate a pharmaceutical manufacturer’s (Pharma)  price offer in comparison with some independent body’s estimates of the drug’s “value”.

The champion of this proposal is Speaker of the House Nancy Pelosi. She is backed by the centrist New Democratic Coalition house caucus.  The details of this scheme are being worked out behind closed doors headed by Pelosi’s aid on healthcare Wendell Primus.

STAT has reported that Primus is being advised by Harvard economist Richard Frank who co-authored a 2008 article proposing binding arbitration to lower Medicare Part D drug prices.

There are also reports that Pelosi’s drug price advisory group has met with Trump administration officials, including John O’Brien, the special adviser to HHS Secretary Alex Azar on Medicare drug pricing.

The other House proposal is championed by Representative Lloyd Doggett of Texas. Doggett has already introduced legislation — H.R.1046 Medicare Negotiation and Competitive Licensing Act  — that has been co-sponsored by 122 members of the Congressional Progressive Caucus.

The Doggett bill would allow the Department of Health and Homeland Security (HHS) to negotiate directly with drug manufacturers. Specifically, the Doggett bill copies the Medicaid formula specifying that Pharma give government its “best price” globally in return for coverage.  

In addition, the Doggett bill includes a new wrinkle to government negotiating schemes — a heavy-handed bargaining chip:

“…if drug companies refuse to negotiate in good faith, it would enable the Secretary to issue a competitive license to another company to produce the medication as a generic”

However, this bargaining chip has been interpreted as falling under  the “March-In Rights”  clause of the 1980 Bayh-Dole legislation specifying when government can or cannot suspend a drug maker’s exclusive patent.  As a result, the use of this bargaining chip is in doubt.

If the goal of modifying Medicare Part D is to lower senior co-pays, we believe that there is a simpler way to achieve this goal than the Doggett “best price” scheme.  We have tweeted about alternative and will provide details in a later paper.

The Origins of the Pelosi Binding Arbitration Proposal

As of April 2019, the Pelosi binding arbitration scheme is still in the talking stage. There are no white papers let alone a legislative bill.  What is known is that the idea for the scheme came from a 2008 article published in Health Affairs by Harvard economists Richard Frank and Joseph Newhouse.

The original Medicare Modernization Act of 2003 launching Medicare Part D specified that Rx drug benefits or seniors would be managed by private sector entities called pharmacy benefits managers (PBMs).  Frank and Newhouse’s 2008 Health Affairs article presented the case that Medicare Part D needed direct government involvement to augment, not replace, price negotiations managed by PBMs.

Unlike Medicaid, the Medicare Part D legislation specifically prohibited direct government negotiations with drug companies. A likely reason for going full “neoliberal” on Medicare Part D was that private sector PBMs had a successful track record of negotiating rebates from Pharma in return for coverage.  

Furthermore, there were plenty of studies showing problems with hard-wired formulas used by government to negotiate prices with Pharma  — the “best price” formula used in Medicaid, and the “buy and bill” formula of average selling price (ASP)  + 6% used in Medicare Part B.

PBMs had developed a wide-range of bargaining chips in addition to the binary “include / exclude” coverage choice that government-run plans settled on.

The problem that Frank and Newhouse saw was that power of PBMs to negotiate rebates for coverage worked only when PBMs could play one drug manufacturer against another.  

PBMs discovered in the late 1990s that they had negotiating power when a number of “me-too” brand drugs came on the market ( e.g. in the statin class – Pravachol, Zocor, Lipitor or in the ACE inhibitor class — Prinivil, Zestril, Accupril, etc).  

For instance, in therapeutic classes like statins or ACE inhibitors, PBMs discovered they could negotiate substantial rebates in neighborhood of 50% off brand list prices in return for coverage.  

 While PBMs were great at negotiating down list prices in what we have labeled “oligopolistic therapeutic classes”, Frank and Newhouse argued that PBMs were powerless to negotiate rebates in cases where there no therapeutic equivalents — what we have labeled “monopolistic therapeutic classes.”  

In a 2018 paper for the Brookings Institute, Frank and Kennedy School economist Richard Zeckhauser present another argument for government intervention based on design flaws in Medicare Part D that reduce PBMs incentives to bargain hard on price.

The first is the coverage gap — the so-called donut hole — where PBMs cost risk is only 5% in the case of brands.  The other is the so-called catastrophic coverage limit of $8,140 for 2019 after which senior co-pays are held to 5. At the same time PBM cost risk to limited to15% while the federal government picks up rest risk cost at 80% of total drug costs.  Below is a nice depiction by the Kaiser Family Foundation of the current distribution of cost risk in Medicare Part D.


Defining the Target of the Pelosi Proposal

To repeat, Politico’s sources say that Pelosi’s aid Wendell Primus is looking only at binding arbitration for “select group of high-cost drugs”.  

Frank and Newhouse offered two binding arbitration  variants:

“The first is a final-offer arbitration system in which both parties cases are heard and then offer their best and final prices. The arbitrator then evaluates the case and sets the drug price at one of the two final prices.”

“Another potential model would be tri-offer arbitration in which a third-party expert or “fact finder” offers a third price for arbitrator consideration. The tri-offer model may be desirable because neither the Center for Medicare and Medicaid Services nor drug manufacturers alone hold the final say in a drug’s price. The system encourages informed negotiations, and an arbitrator only steps in if the two parties cannot reach an agreement.”

By all accounts, the choice of independent  assessor of price based on value would go to ICER, the leading think tank on such matters.

Frank and Newhouse’s case for augmenting PBMs negotiating efforts is solid in cases where there are not Pharma with therapeutic equivalents to play against each other.  

The question is where exactly do those opportunities exist today and how much is the associated drug spend?

Frank and Newhouse were explicit in saying that in 2008 there were few therapeutic classes where there was both lack of negotiating leverage and, this is key, where the total spend was material, say in the $130 Million Medicare Part D drug spend. Their assessment was similar to a 2007 assessment by the Congressional Budget Office.

However,  they could see back in 2008 a pipeline of unique drugs with large drug spend that would justify augmenting PBMs with government-run binding arbitration. Quoting from their paper

We believe that it is premature to conclude that there are enough unique drugs to create a meaningful budget problem. Nonetheless, the likelihood of drug products better tailored to specific genetic defects raises the possibility of many more unique drugs in the future”

We see three broad categories of drugs where binding arbitration is more appropriate than mano-e-mano direct government negotiations with heavy-handed bargaining chips sometimes used to reach agreements.

1. High PMPY large molecule biologic drugs where patented therapeutic equivalents would be rare.

 In the case of large molecule biologics, it is impossible to produce molecule-for-molecule generics so that PBMs have no bargaining leverage.  The only possibility for negotiating leverage would come after the approval of therapeutically equivalent “biosimilars”.   Plus, we believe value become a factor in all cases where where PMPY costs exceeding $50,000.

Here is a table of the top high PMPY drug treatments covered by Medicare Part D in 2017 that we believe would be best suited for binding arbitration  based on an independent assessment of “value”:

Data comes from Medicare Part D Drug Spending Dashboard & Data.  It is based on list price, not net price after rebates.

2. High PMPY cell therapy cancer drugs where outcomes are uncertain.

Much of call for value-based pricing comes for cases in which the degree of success is more in question that the value of the of the outcome itself .  Neutral assessment of probabilities of success are appropriate in these cases.  Prime examples of such drugs are new drugs to treat melanoma  Keytruda and Opdivo whose costs exceed $150,000 a year and with remission rates only somewhat better than just chemotherapy. 

The problem here is that most of these new cell therapy treatments are covered under Medicare Part B, not Part D, as they are infused treatments occurring in physicians’ offices, out-patient clinics, and hospitals.

3. High PMPY orphan drugs due to having to be amortized over patient population in the hundreds.

The justification for binding arbitration as an augment to PBM negotiations is solid for so-called “orphan drugs.” 

The market for orphan drugs is less than several hundred patients so that amortized R&D and manufacturing costs alone exceed $100,000 per person. The cumulative total number of orphan drugs approved by the FDA has accelerated since 2012.



Here, it seems appropriate to treat orphan drugs as a “natural monopoly” with government rather than private PBMs setting the price on a cost-plus or value basis.  Here is a list of the top orphan drugs by PMPY drug spend:

While the number of orphan drugs is on the rise, the total drug spend is relatively low due to the small number of patients treated.  Subjecting orphan drugs to binding arbitration would have a minor impact on overall Medicare Part D drug spend.


An Estimate of the Impact of Pelosi’s Proposal

In terms of % of total Medicare Part D drug spend,  the question is how large is opportunity to lower drug spend using a binding arbitration scheme? 

To answer that question, we use data from the Medicare Part D Drug Spending Dashboard & Data.  It is based on list price, not net price after rebates:

“Drug spending metrics for Part D drugs are based on the gross drug cost, which represents total spending for the prescription claim, including Medicare, plan, and beneficiary payments. The Part D spending metrics do not reflect any manufacturers’ rebates or other price concessions as CMS is prohibited from publicly disclosing such information.”

For purposes of targeting opportunities, this list price data is better than net price data.

We first filter out drugs with multiple sources.  This are mostly generics where dispensing pharmacies have great success in negotiating purchase volume discount off wholesale acquisition costs.

Next, we filtered out drugs with PMPY drug spend > $50,000. We think this filter captures most of the drugs in the three categories better handled by binding arbitration  — unique biologics,  drugs with uncertain outcomes, and orphan drugs.

But, the filter is not perfect. In particular, there are two therapeutic classes  — Insulin and Hepatitis C Virus (HCV) — representing 11% of total drug spend that are arguably more amenable to direct negotiation than binding arbitration.  We chose to place the two therapeutic classes in the direct negotiation subset.

In the case the insulin, while biologics, they have been proved to be therapeutically similar.  Furthermore, the drugs have been around for decades, the outcomes are relatively certain and costs are less than $50,000 PMPY.

In the HCV drugs, while costs are greater than $50,000 PMPY, cure rates exceed 90%, and they have proved to be therapeutically similar by genotype.

In fact, because there are therapeutic equivalents in these two classes, PBMs have had recent success in negotiating rebates in these two classes.

The following is a listing of drugs in these two therapeutic classes we chose to place in the direct negotiation subset and rather than Pelosi binding arbitration subset.



After that, we filtered out drugs with total drug spend > $130 Million to represent the core opportunity for Doggett’s direct government negotiation bill.  Here are the top 36 candidates for direct negotiations in addition to the insulin and HCV therapeutic classes

The remaining drugs was single source drugs representing 61% of drug numbers — 1,786 –but only a paltry 9% of total drug spend.

The following table summarizes the division of drug spend by size of the opportunity targeted by the Pelosi proposal vs the Doggett bill.

We estimate that only 19% of total 2017 Medicare Part D list price drug spending is suitable for Pelosi’s binding arbitration scheme.



Another way to depict the relative targets of opportunity for the two proposals is to graph total revenue by rank.  A comparison of the graphs indicates a much more severe revenue rank power function for drugs best suited for binding arbitration than drugs best suited for direct government negotiations.





Published by Lawrence W. Abrams

Economist, Ph.D. WashU StL, B.A Amherst. I write at the intersection of economics, finance, accounting and high tech. I have been a critic of PBMs since 2003. I have received no remuneration writing these papers.

Exit mobile version