Intended for healthcare professionals


Public referendum on drug prices in the US: will it bring relief?

BMJ 2016; 355 doi: (Published 31 October 2016) Cite this as: BMJ 2016;355:i5657
  1. Thomas J Hwang, researcher,
  2. Aaron S Kesselheim, associate professor of medicine
  1. Program on Regulation, Therapeutics, and Law (PORTAL), Division of Pharmacoepidemiology and Pharmacoeconomics, Department of Medicine, Brigham and Women’s Hospital and Harvard Medical School, Boston, MA 02120, USA
  1. Correspondence to: A S Kesselheim akesselheim{at}
  • Accepted 18 October 2016

Thomas Hwang and Aaron Kesselheim assess how California’s proposed new law on drug pricing could affect healthcare costs in the US and beyond

Concerns over the rising cost of new medicines have prompted calls for a more sustainable system of drug pricing.1 2 3 Public unease is particularly acute in the US, where brand name drug prices are the highest in the world and the costs are increasingly being passed on to patients.4 This has led to federal and state policy efforts seeking to tackle high drug prices.5 6 In November, one such policy initiative will be put to a vote: voters in California will decide whether preferential pricing provided to the US Department of Veterans Affairs should also be available to other public payers in the state.

Successful ballot measures in California have historically triggered national reforms (box 1). The outcomes of the California referendum are also likely to have global implications, because attempts to reduce drug prices in the US could have a ripple effect on the prices negotiated by other countries.7 We examine the potential savings from the proposed change.

Box 1: Importance of policy reforms in California

California has a long history as a laboratory for national policy reforms. It is the most populous US state (about 39 million people), accounting for roughly 12% of the US population. California also grants its voters the power to initiate or repeal legislation by referendum. Examples of influential health policy changes in California include:

  • Tobacco control program—In 1988, voters in California passed Proposition 99, which increased excise taxes on cigarettes to fund a public anti-tobacco use campaign, serving as a model for similar initiatives globally

  • First rebates for Medicaid programs—Attempts by California and several other states to obtain discounts on prescription drugs led to federal legislation in 1990 establishing the first statutory rebate for states’ Medicaid programs

  • Rejection of single payer healthcare—In 1994, voters defeated a ballot initiative proposing to establish a state single payer healthcare system; parallel legislation at the federal level similarly failed

  • Climate change—In 2005-06, California ratified targets for reductions in greenhouse gas emissions and became the first state to adopt a multisector cap-and-trade program for statewide emissions

How much does a drug cost?

Most developed countries have governmental bodies that negotiate drug prices for all, or nearly all, of their consumers.8 By contrast, health coverage in the US is fragmented, and public and private payers obtain varying discounts and rebates through their negotiating power and statutory protections (box 2). The net, or actual, price paid for a medicine by an insurer can therefore vary widely and, in many cases, is below the catalog or average wholesale price.9

Box 2: Percentage of average wholesale price paid for branded drugs by US payers1011121314

  • Uninsured and cash paying retail customers: >95%

  • Commercial insurers: 80-85%

  • Medicare (part D): 78-85%

  • Medicaid: 50-55%

  • Veterans Administration: 35-40%

Medicare, the public insurance for people 65 years and older, is statutorily prohibited from negotiating prices on behalf of the individual health plans that provide coverage for outpatient medications (under part D) and is reported to pay a similar net price on average to private insurers.10 11 12 By law, other public payers are entitled to reductions in drug prices. Medicaid, which provides means tested insurance for people on low incomes and is jointly funded by federal and state governments, is required to receive a rebate of at least 23.1% off the average manufacturer price for most branded drugs (table 1). The Veterans Administration (VA), which provides healthcare for military veterans and their dependents, is entitled to a rebate of at least 24%.

Table 1

Summary of statutory rebates for key public payers in the US

View this table:

Medicaid and the VA are also protected from price increases that exceed inflation, and state Medicaid programs can individually negotiate supplemental rebates. The resulting overall average net price for branded drugs is estimated to be 50-55% and 35-40% of average wholesale prices for Medicaid and the VA, respectively.13 14

Referendum on prescription drug costs

Californians are being asked to vote on the California Drug Price Relief Act (Proposition 61), which requires that the net price paid by most of California’s publicly funded insurance programs be no higher than the VA’s lowest price.15 This policy would apply to the fee-for-service component of Medi-Cal, the state’s Medicaid program, as well as programs in which the state does not purchase the drug directly but is the ultimate payer, such as the AIDS drug assistance program, public employees’ retirement system, and prison systems. The initiative is expected to cover roughly five million people, who accounted for net drug spending of $4.2bn (£3.4bn; €3.8bn) in 2014 (table 2).16

Table 2

Summary of state entities covered by California’s drug pricing act

View this table:

The initiative follows the recent outcry over the cost of sofosbuvir, an important new treatment for hepatitis C infection, which was listed at $84 000 for a 12 week course after it was first approved in 2013. Since a substantial proportion of patients with hepatitis C infection in the US are covered by public insurance, directors of Medicaid programs and other insurers warned the drug could bankrupt states.17 18 Proponents of the initiative argue that the burden of rising healthcare costs on taxpayers and public programs could be reduced if the imbalance in drug prices paid between government payers is rectified. Opponents—led by members of the drug industry, which has spent over $100m to campaign against the initiative19—believe that the measure will hinder investment in medical research and could interfere with patients’ access to necessary medications if manufacturers refuse to participate or decline to offer their products at the required lower prices. However, there is limited empirical evidence to assess these claims.

Potential savings

Drug price negotiations are confidential, and public payers, including Medi-Cal and VA, generally exempt their agreements with manufacturers from freedom of information laws. However, the VA releases its negotiated prices, and state Medicaid programs provide the formula for calculating statutory rebates. We examined differences in net prices paid by the VA and Medi-Cal for the top 200 brand name drugs by spending in the Medi-Cal fee-for-service program in 2014, which accounted for 71% of the program’s branded drug expenditures and 63% of total drug expenditures (fig 1). In brief, we calculated Medi-Cal’s net price by applying the 2014 average wholesale price and inflation changes since the drug’s launch to the statutory rebate formula.20 For each drug, we then compared Medi-Cal’s price to the VA’s net price. We projected possible savings assuming Medi-Cal had obtained the lower VA net price for the quantity of the applicable drug purchased by Medi-Cal in 2014.


Fig 1 Calculation of drug use and spending for Medi-Cal and Department of Veterans Administration (VA) in 2014

For these 200 branded drugs, we estimate that the net price paid by Medi-Cal after statutory rebates exceeded the negotiated VA price for 65 (33%) of the drugs by 30% on average. Achieving price parity between Medi-Cal and the VA for these 65 drugs would imply cost savings in 2014 of $100m-$125m. Most of these potential savings were associated with drugs to treat HIV and hepatitis C, as well as blood and blood component products (table 3). For example, the estimated net cost of a 12 week course of sofosbuvir was about $64 000 when paid for by Medi-Cal versus $45 000 when paid for by the VA—a difference of 42%.

Table 3

Potential savings from introducing California’s drug pricing initiative*

View this table:

Historical precedent and limitations

Our findings suggest that for most branded drugs, the prices paid by the Medicaid program in California are similar to or better than those negotiated by the VA. The VA’s negotiated price seems to be lower than Medicaid’s mainly for newer drugs (which have few competitors and are less exposed to inflation linked discounting) and protected drug classes with lower statutory rebates.

Our calculation is likely to be conservative because it excludes confidential final stage rebates— namely, Medi-Cal specific supplemental rebates (about 5% of average wholesale price) as well as additional rebates given to the VA (5-10% on average). The state’s other major payer, the public employees’ retirement system, may benefit from other cost savings, but both we and the retirement system21 could not estimate these since relevant drug level data were not publicly available.

However, the long term budgetary effect of the current ballot initiative is uncertain, and strategic pricing shifts could influence the likelihood of payers achieving these illustrative savings. The VA has historically had disproportionate pricing power relative to its size: it operates an integrated healthcare system and has a restrictive formulary whereas Medicaid and Medicare are generally required to cover all FDA approved drugs. Yet, since the VA’s preferential pricing agreements must be periodically renegotiated, any savings could be eroded if manufacturers raised their prices to the level currently being paid by other payers.

One precedent occurred in 1990, when the US Congress passed a short lived provision (ultimately revoked in 1992) tying Medicaid’s statutory rebates to the lowest price achieved by any payer, including the VA. Early anecdotal studies suggested that the provision was associated with increased prices for the VA and other payers, although a recent review concluded that a causal relation with the provision could not be credibly established, and that any real price increase was likely to be small (less than 4% on average across all branded drugs).22

Lessons for future policy

The California referendum shines a light on the challenges faced by public programs in paying for costly medicines. High drug prices in the US also have consequences for other health systems: payers in several countries, such as Japan and Canada, reference US prices directly or indirectly in their reimbursement decisions.23

In response to these challenges, several countries have developed coordinated negotiating strategies. For example, the European Union is trialing a joint procurement system, starting with pandemic vaccines and medical countermeasures, to harness the collective negotiating ability of individual countries.24 Importantly, the EU pilot provides participating manufacturers with volume guarantees and market share. These features could ensure that greater affordability for public payers is balanced with sufficient economic incentive for innovation. Another example is Canada, where the federal government partnered with provincial and territorial authorities (as part of the Pan-Canadian Pharmaceutical Alliance) to obtain greater savings on prescription drugs for all publicly funded drug programs. In the US, this model of pooled procurement has already been applied, in part, to federal purchasing of pediatric vaccines, although implementing a similar policy for a broader set of medical products will invariably face considerable resistance.

Neither the current system nor the ballot initiative adequately addresses questions around the fairness of a drug’s catalog price or the appropriateness of fixed statutory discounts based on this price. For example, some new medicines may provide value to patients and payers (through reduced risk of hospital admission, increased quality of life, or extended survival) even without the required rebate, whereas others may need larger rebates to be cost effective. Rigorous, evidence based reviews of drugs’ clinical effectiveness and value—such as those developed by the National Institute for Health and Care Excellence in the UK, the Institute for Quality and Efficiency in Health Care in Germany, and the Institute for Clinical and Economic Review, and the Drug Effectiveness Review Project in the US—could inform the development of policies that better link the price paid for a drug to its value to patients and the broader healthcare system.


For the first time in recent memory, California’s ballot will give the public a direct voice in the ongoing debate about the rising cost of new medicines. We have estimated that the state’s public payers could obtain additional savings if they were able to access the lower prices negotiated by the VA for a relatively small number of high priced drugs. In an era of tenuous public finances and constrained healthcare resources, any meaningful budgetary respite could contribute to the fiscal health and future sustainability of socially valuable public insurance programs.

However, the initiative is unlikely to cure rising healthcare costs or to resolve underlying social tensions around the ability of patients and payers to afford new medicines. Regardless of whether it passes, the distributional and health equity consequences of pharmaceutical payment systems require greater public and empirical scrutiny.

Key messages

  • Prescription drug prices vary widely in the US between payers (and compared with other countries) because of differences in negotiated rebates and discounts

  • A referendum in California will attempt to eliminate the gap in prices paid for drugs between public payers

  • This policy could have saved the state’s Medicaid roughly $125m of its branded drug spending in 2014

  • It is unlikely to solve the current fragmented system of purchasing or address questions about the fairness of drug prices more broadly


  • Contributors and sources: ASK is director of the Program on Regulation, Therapeutics, and Law (PORTAL) at Brigham and Women’s Hospital and has written extensively on prescription drug regulation and pricing. TJH’s research focuses on the development and use of new drugs and devices. Both authors contributed to the conception, design, and execution of the study, and approved the final draft for publication. TJH is guarantor.

  • Funding: ASK’s work is supported by the Laura and John Arnold Foundation, with additional support provided by the Engelberg Foundation and the Harvard Program in Therapeutic Science.

  • Competing interests: We have read and understood BMJ policy on declaration of interests and declare the following: TJH has received funding from Harvard University and the Interfaculty Initiative in Health Policy, and was employed by Blackstone and Bain Capital; and ASK has received unrelated research support from the FDA Office of Generic Drugs and Division of Health Communication.

  • Provenance and peer review: Not commissioned; externally peer reviewed.


View Abstract

Log in

Log in through your institution


* For online subscription