Health Affairs, 23, no. 1 (2004): 89-102 doi: 10.1377/hlthaff.23.1.89 © 2004 by Project HOPE	New Online   How Would Obama, McCain Cover The Uninsured?   Debating Cost Of Uninsured   Try Medicare-For-All   HA Blog Top 10

This Article Abstract Reprint (PDF) Submit a response to this article Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Services E-mail this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Archive Download to Citation Manager Reprints & Permissions Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (12) Citing Articles via Google Scholar Google Scholar Articles by Newhouse, J. P. Search for Related Content PubMed PubMed Citation Articles by Newhouse, J. P. Related Collections Medicare Pharmaceuticals Research And Technology

Medicare Challenge

How Much Should Medicare Pay For Drugs?

Abstract

 
Discussions of the Medicare drug benefit have focused more on what beneficiaries will pay than what pharmaceutical manufacturers will receive. A key choice is the degree to which Medicare must become involved in setting manufacturers’ prices. If prices must be set, Medicare could do so using average wholesale price, comparison with prices in other markets, cost, or rate-of-return regulation. Because all four methods have substantial drawbacks, Medicare should not initially attempt to set prices, but to prevent abuses in pricing, Congress should allow cost to be considered in coverage decisions.

Much of the public and press attention around the Medicare drug benefit has focused on the demand price, or the size of the deductible, coinsurance rate, and the infamous "doughnut," or region of no coverage. In contrast, I focus primarily on the supply price. The context is traditional Medicare, but similar issues also arise for Medicare managed care plans because they must provide the same benefits. Furthermore, the issues arise irrespective of whether the benefit is voluntary and whether the benefits come through private stand-alone drug plans (Part D) or are incorporated into Part B of traditional Medicare.

A drug’s supply price determines its profitability, as well as the overall cost to taxpayers. Taxpayer cost is affected in three ways: (1) The obvious direct effect—holding beneficiary payment constant, the higher the supply price, the more Medicare pays for each drug. (2) A higher supply price for a drug makes marketing efforts for that drug more profitable and thus can increase the quantity that is prescribed.² (3) Expectations about the future supply price are critical to investors’ decisions about how much capital to supply to the pharmaceutical and biotechnology industries to develop new drugs. The higher those expectations, the more capital will be supplied and the more drugs developed.

I first discuss four approaches to establishing supply prices that vary in the extent of price setting. I then review some economics of the pharmaceutical industry and their implications for pricing. I next sketch an economically efficient price, before turning to a discussion of four feasible alternatives for price setting and their drawbacks. I conclude with a discussion of price in coverage decisions as providing some protection against abusive pricing.

Four Approaches To The Supply Price Of Pharmaceuticals

Top Editor's Notes Four Approaches To The... The Economics Of Pharmaceuticals What Is The Right... Options In Setting Administered... Is There A Way... Excess Profits In... Notes

 
Manufacturers set a profit-maximizing price. In the least restrictive approach to supply prices, pharmaceutical manufacturers are free to charge a profit-maximizing price, but beneficiaries face a demand price with cost sharing. For drugs that face competition, especially when generic drugs are available, this approach can work well. By contrast, for drugs that face little or no competition, this approach can greatly raise supply prices. Because most drugs will be bought by beneficiaries who have exceeded an initial deductible, the coinsurance rate principally determines the manufacturer’s profit-maximizing price.³ To show the resulting leverage on the supply price, suppose the insurance has only a coinsurance rate c, a number between 0 and 100 percent that shows the percentage of the price paid by the patient. In that case, the manufacturer’s profit-maximizing price for a drug with no competitors relative to a market with no insurance increases by a factor (100/c). In the proposed Medicare legislation, c is variously 10, 20, and 50 percent, so the price of drugs facing no competition could be two to ten times higher than with no insurance. If there is a catastrophic ceiling beyond which insurance fully pays (c = 0) or if there is supplementary insurance that covers the cost sharing, the profit-maximizing price has no bound. Medicare’s use of average wholesale price (AWP) to reimburse certain drugs it now covers through Part B, together with the Part B 20 percent coinsurance, is a variant of this first method.

Medicare sets prices for each drug. The next two approaches involve varying degrees of price setting. Medicare could treat supply prices for drugs as it does supply prices for institutional and physician providers. Using this model, Medicare would set or negotiate a price for each drug, just as it sets prices for each diagnosis-related group (DRG) and Current Procedural Terminology (CPT) code. This pricing method is not under serious consideration for Medicare, although it has been used in other countries.

Medicare sets prices with PBMs. Alternatively, Medicare could set or negotiate a price with pharmacy benefit managers (PBMs) for all covered drugs, just as it sets a price for health plans for all covered medical services.⁴ In turn, the PBM would negotiate a supply price with manufacturers for specific drugs, as it now does for its commercial business—and as health plans now do with individual hospitals and physicians. If the health plan price analogy were followed exactly, Medicare would set a price, limit surcharges to beneficiaries, and ask PBMs to bear full risk, so that PBMs would have a strong incentive to negotiate for low prices. If the price Medicare set for PBMs was too low, PBMs would not participate, just as many health plans do not participate in Medicare+Choice. But unlike health plans, PBMs in commercial insurance rarely assume risk, because they have little or no control over prescription volume, so literally using the health plan model does not appear practical.

PBMs bid for Medicare business. The last approach, the one now being considered, also uses insurers or PBMs as an intermediary between the government and beneficiaries but does not formally involve price setting. Instead, PBMs bid for Medicare business, as they do now for commercial business. Current legislation envisions multiple PBMs or insurers in a local area, with beneficiaries choosing among them on the basis of their bid price (and benefits), with Medicare possibly sharing risk.⁵ In this arrangement PBMs have two related incentives to negotiate low supply prices with drug manufacturers. First, lower prices paid to manufacturers facilitate low bids, thereby attracting beneficiaries. Second, if the PBM can negotiate even lower prices than those anticipated in its bid, it keeps some or all of the savings.

Even though they typically do not bear risk in commercial insurance, PBMs attempt to restrain supply prices by increasing price competition within a therapeutic class of drugs. To induce such competition, a PBM has the beneficiary pay less for drugs on its formulary, thus bringing demand prices into the picture. Drugs on the formulary are those for which the PBM has secured a favorable supply price from the manufacturer. For those drugs, the consumer typically pays a relatively modest amount, perhaps $5 or $10 for a month’s supply, but for other drugs, the consumer pays a much higher amount.⁶ If a formulary is used, the determination of which drugs are in the same therapeutic class is a crucial implementation decision, as is updating that decision over time.⁷ Although the proposed Medicare legislation specifies a deductible and coinsurance, it permits actuarially equivalent tiered copayments. Such copayments would be much higher than one observes with commercial insurance, however.

I prefer the last approach because it promotes price competition among manufacturers better than the first but minimizes price setting relative to the other two approaches.⁸ Nonetheless, I doubt whether Medicare can fully escape price setting. The potential problem arises around a drug that is on patent, represents a major therapeutic advance, and has no close competitors.⁹ Erythropoietin when it first came on the market is an example. In a formulary, such a drug would likely be the single drug in its class and by definition would face no price competition. If Medicare were prepared to pay any price a manufacturer named for such a drug or reimburse the PBM for that price, there would clearly be little or no incentive for the manufacturer to hold back.

The Economics Of Pharmaceuticals

Top Editor's Notes Four Approaches To The... The Economics Of Pharmaceuticals What Is The Right... Options In Setting Administered... Is There A Way... Excess Profits In... Notes

 
Thinking about how Medicare ought to approach supply prices for drugs requires some understanding of the economics of the drug industry.¹⁰ There are three salient economic facts for this discussion: (1) The industry’s fixed costs are high relative to its marginal costs. Fixed, or at least sunk, costs stem in part from research and development (R&D).¹¹ Moreover, a portion of marketing and manufacturing expenditure is fixed in the short run, or at least subject to substantial economies of scale. Once production begins, the marginal cost of producing another pill is typically negligible. (2) New chemical entities (NCEs) that come to market will have product and process patents that give them a temporary legal monopoly. (3) Developing new drugs is risky; only about a fifth of drugs that enter phase I clinical trials ultimately obtain marketing approval.¹² In short, there are many "dry holes." These three facts have five implications.

Patent owner can exploit monopoly power. A drug on patent has some degree of monopoly power that the patent owner can be expected to exploit when pricing the drug; indeed, the intent of the patent is that the owner should exploit the monopoly. If the drug is an important therapeutic advance with no close competitors, the degree of monopoly power can be large, with corresponding profits for the manufacturer or patent owner. If an insurer must cover the drug and beneficiaries are not liable for any increase in its supply price, there is potential for an abusive supply price.

R&D costs must generate a sufficient return. R&D requires capital, and a manufacturer will not obtain or invest that capital unless it can, on a probabilistic basis, obtain a return at least equal to the capital’s next-best use.¹³ Recovering the capital costs together with a return implies that the difference between marginal revenue (the price received for the last pill sold) and marginal cost (the cost of that pill) must be substantial.

Supply price must cover development costs. Once the drug has been developed, the drug company can increase its profit by selling the drug to some markets at supply prices that, if they prevailed everywhere, would not recover fixed costs. Unless the company is engaged in charity, the supply price always must cover incremental manufacturing costs, but those are often negligible. In other words, the company can maximize profit by selling at different prices in different countries as long as drugs cannot be arbitraged between countries (that is, bought in low-price countries either by residents of high-price countries directly or through importers). Within a country, as well, the drug company may maximize profit by selling at varying prices to different groups of consumers.¹⁴ Thus, hospitals and large health maintenance organizations (HMOs) can buy many drugs at much lower prices than uninsured individuals can. If Medicare prices are linked to prices in other markets, however, drug companies will price differently in those markets, as I discuss below.

Regulators can "hold up" manufacturers for costs of development. Once a drug is developed, the manufacturer can be exposed to what is known in industrial organization as a holdup problem, because the manufacturer cannot write an enforceable contract with price regulators prior to development.¹⁵ Specifically, the manufacturer must make the R&D investment in advance of knowing the price that regulators will allow. Once the manufacturer has made the investment, however, a regulator only needs to set a price greater than the manufacturing and marketing costs to make production profitable; as a result, a regulator can "hold up" the manufacturer for the costs of development. If the manufacturer or investors, anticipating this situation, expect that the prices the various jurisdictions will ultimately set will not in the aggregate cover the cost of development plus a return to capital, the manufacturer will not develop the drug, even though the willingness to pay for the drug in the world might greatly exceed the drug’s development costs. A holdup is often politically attractive, because taxpayers and beneficiaries will pay less for the existing drug, whereas there are usually no identifiable people being harmed by the failure of a future drug to be developed.¹⁶ This situation is exacerbated by free riding by various countries and payers, each of which would prefer that someone else pay the development costs. An economically efficient outcome requires that regulators in the aggregate not hold up manufacturers, but the political incentives to do so might be overwhelming.

Manufacturers have powerful incentives to spend money on marketing. Because prices for drugs on patent are typically well above the negligible marginal production costs, manufacturers have powerful incentives to spend money on marketing.¹⁷ With low production costs, the revenue from any additional sale mostly drops to the bottom line. The resulting incentive to sell additional quantities of drugs invites abuse, a subject much in the news of late.¹⁸ Supply prices that are close to marginal cost would eliminate the financial incentives for abuse but would not sustain R&D.

One sometimes hears it said that the industry would have more money for R&D if it would cut down its marketing costs. This comment reflects misunderstanding of the economics of the industry. If a firm did so, it would be less profitable and would attract less capital for R&D or would have fewer internally generated funds to invest. Put another way, drug manufacturers are allocating resources to both development and marketing to maximize profits; unless firms have made mistakes, reallocating monies from marketing to research will lower profitability.

What Is The Right Supply Price?

Top Editor's Notes Four Approaches To The... The Economics Of Pharmaceuticals What Is The Right... Options In Setting Administered... Is There A Way... Excess Profits In... Notes

 
To frame the issues with respect to specific price-setting methods, I next sketch several factors that determine an economist’s optimal supply price. It will quickly become clear that there will never be nearly enough information to determine such a price with any degree of precision, which is why prices are generally better set through markets.

Economic efficiency. The economist’s optimal supply price would satisfy economic efficiency. Economic efficiency has a technical meaning, but it suffices here to say that it implies that the last dollar spent on drugs should buy as much benefit as that dollar would buy if it were used to buy any other good or service. This already makes clear the complexity of determining an economically efficient price, since valuing the benefit of the last dollar spent on drugs requires valuing how much health that dollar buys, and there is no consensus on how to do that.¹⁹

Furthermore, an efficient price requires an additional dollar spent on drugs to buy the same clinical benefit as an additional dollar spent on other medical inputs, something that is difficult to achieve within the silo-like structure of pricing in traditional Medicare.²⁰ In other words, optimality requires the correct Medicare budget for each kind of medical input, but both substitution among inputs and technological change make determining such budgets difficult.²¹

To complicate matters further, there are two types of efficiency, static and dynamic, whose implications for pricing point in different directions. Static efficiency takes the drugs developed at a point in time as given; dynamic efficiency considers how the development of new drugs will be affected by supply prices.

Static efficiency. Consider static efficiency first. Assuming that the production costs of the additional pills are minimal, both supply and demand prices should be relatively low so that people will not forgo using drugs that, because they require few resources to produce, do not much impinge on the production of other goods. The observed prices of generic drugs may be approximately efficient, since there is effective price competition, and the large R&D expense is not relevant.

Would an increase in some hypothetical supply price increase efficiency? Suppose the increase calls forth the production of additional pills. In addition to shifting some minimal resources to more pill production and away from other goods, the increase also might call forth additional marketing resources. Whether that effort is worth those resources depends upon how the mix of overuse, underuse, and misuse of the drug changes as the marketing effort changes.

Dynamic efficiency. Turning to dynamic efficiency, an optimal supply price should call forth the right amount of R&D, so it must be above—probably far above—the efficient static price. How far above? Far enough so that the last dollar spent on research buys the same benefit as the last dollar spent on other goods or services. Hence, to determine the optimal supply price, one must know the expected benefit from those drugs that would be developed if expected supply prices were incrementally higher. Not only is this unknowable with any precision, but the benefit could change over time as the basic science changes.

The patent system is an effort to approximate the dynamically efficient price. While the patent is in effect, price will be well above the cost of production, potentially allowing the manufacturer to recoup development costs, but once the patent expires, price typically falls to near production cost, as with generic drugs. Nonetheless, a patent is given for an arbitrary period, and how well the resulting prices approximate dynamically efficient prices is unknown.

Options In Setting Administered Prices

Top Editor's Notes Four Approaches To The... The Economics Of Pharmaceuticals What Is The Right... Options In Setting Administered... Is There A Way... Excess Profits In... Notes

 
Despite the virtual impossibility of knowing whether any administratively determined price for a drug is economically efficient, the possibility of abusive pricing in the case of a clinically important drug with no close competitors, together with modest or no coinsurance, means that Medicare might be tempted to adopt some form of the pharmaceutical industry’s nightmare—price setting—at least for such drugs.²² I now consider four ways in which Medicare might go about price setting. All four are in use in some form, but all have substantial drawbacks.²³

Average wholesale price. Although many do not know it, Medicare already does cover some outpatient drugs—namely, those that a patient usually does not self-administer. In fact, Medicare spent more than $8 billion on such drugs in 2002, more than 3 percent of total program spending.²⁴ Altogether it covers 450 drugs, many of which are for cancer chemotherapy, but it also covers erythropoietin, a drug that stimulates red blood cell production and is used primarily for patients on renal dialysis and with cancer and AIDS.²⁵ How Medicare determines supply prices for these drugs is a useful starting point for this discussion.

In the case of cancer chemotherapy, as well as erythropoietin for cancer and AIDS patients (but not erythropoietin for dialysis patients), physicians, who typically dispense the drugs, receive 95 percent of the AWP. Medicare pays 80 percent of this amount, and the physician collects the remaining 20 percent from the beneficiary as Part B coinsurance. The AWP, however, is not defined in statute, and in practice manufacturers have much discretion in determining it.²⁶ It is roughly analogous to a sticker price for an automobile, in that manufacturers name the AWP, but few physicians actually buy at the full AWP or even at 95 percent of it.

This method of setting the supply price has been criticized for two reasons. Because physicians typically purchase these drugs for less than the 95 percent that Medicare pays, it is argued that Medicare overpays.²⁷ Furthermore, because the dispensing physician keeps the difference between 95 percent of AWP and the purchase price, there is a financial incentive to overprescribe, although it is not known to what degree, if at all, physicians act on that incentive.²⁸

The AWP has the virtue that Medicare does not set a price, but whether it can serve as a long-term basis for supply prices in a general drug benefit is problematic. In effect, Medicare would be agreeing to pay a percentage of any price that a manufacturer chose to name, which leaves the incentives for abusive pricing in place.²⁹ An approach based on AWP is analogous to the initial design of Medicare, which was a passive insurer that, by using a usual, customary, and reasonable fee schedule to reimburse physicians, gave physicians substantial discretion in fee setting and partially relied on coinsurance to restrain prices. Ultimately this method was abandoned.

Comparable prices. Medicare could look to prices paid by other payers, even those paid in other countries, as a guide to setting administered prices. The Omnibus Budget Reconciliation Act (OBRA) of 1990 mandated that Medicaid use a variant of this method by requiring manufacturers to give Medicaid the lowest price at which they sold a drug. There is, however, no reason to expect that the methods used by other payers to set prices come close to an efficient supply price.

More practically, such a method is problematic for Medicare because of its large market share of many drugs: Medicare beneficiaries account for more than a third of U.S. drug spending.³⁰ Thus, if Medicare asks to receive the best price, manufacturers might simply raise prices to other payers, analogous to the increase in prices to the Departments of Veterans Affairs and Defense when legislation required that Medicaid obtain the lowest price.³¹ Similarly, if Medicare priced at current Canadian prices, manufacturers would likely insist on higher prices in Canada or withdraw their products from the Canadian market. This method would also be problematic for a new drug; there might not be a comparable price.

Cost. A third alternative is to emulate the cost-based methods Medicare uses to set supply prices for institutional providers. Here Medicare’s stated policy is to pay the marginal cost of the efficiently run provider.³² But in the case of drugs, such a policy does not appear to be feasible, because marginal cost is near zero for many drugs. Paying marginal cost thus means that other payers would have to cover almost all of the fixed costs. If they were unwilling to do so—and it seems especially improbable if the drug is mainly used by the elderly—manufacturers might well be unwilling to sell to Medicare at such a low price or, in the more relevant case of drugs not yet on the market, might not go forward with development.

As an alternative to marginal cost pricing, Medicare could in principle pay some fraction of average cost, but the basis for setting that fraction is not obvious.³³ It may seem natural to use Medicare’s share of the drug’s sales, analogous to how Medicare initially paid hospitals, but that figure may neither be economically efficient nor correspond to federal budgetary needs. For the latter reason, Medicare does not use cost as a basis to determine physician fees.

More fundamentally, using some fraction of average cost presumes that average cost can be ascertained, but in fact any such estimate would be highly arbitrary. Large drug companies have many products, and allocating joint costs across those products is arbitrary. For example, how would CEOs’ or research directors’ salaries be allocated by product?³⁴ Furthermore, companies must cover the R&D costs of their unsuccessful drugs or go out of business, so reimbursing only those costs that can be directly associated with drugs on the market seems pointless.

Just as there is precedent in Medicare for using the AWP to set prices, there is precedent for using a drug’s average cost. When erythropoietin first came on the market, Medicare based its reimbursement rate in part on an estimate by the Department of Health and Human Services (HHS) Office of Inspector General (OIG) of the manufacturer’s average cost, including development cost.³⁵ The average cost estimate included a 20 percent rate of return to the manufacturer (Amgen), but to calculate the supply price that would yield that rate of return, the OIG used a market size of the then roughly 100,000 dialysis patients. The current number of dialysis patients is more than 350,000, but subsequent changes in reimbursement have not been adjusted for this growth, implying that the manufacturer’s return on this drug has been well above that initially contemplated by the government.³⁶

Return on assets. Instead of setting prices for individual drugs, the British National Health Service (NHS) limits a manufacturer’s return on its entire book of business with the NHS. The limit is 20 percent of assets in the United Kingdom, approximately the same figure that the OIG used in the erythropoietin example. Subject to the 20 percent overall constraint, the manufacturer is allowed pricing freedom on individual drugs. Not surprisingly, drug manufacturers mark up the drugs in their portfolio that face less competition to a greater degree, emphasizing the problematic nature of pricing for important drugs that do not face competition.

As does attributing average cost at the level of a drug, this method faces a problem in determining assets. Standard U.S. accounting conventions for the pharmaceutical industry "expense" both R&D and marketing costs, meaning that such costs are not included in the value of the company’s assets.³⁷ Economically, however, this makes no sense. Both development and marketing costs create assets because they generate future returns.³⁸ That is, if development is successful, the drug can be expected to have sales and profits for many years. Similarly, marketing helps build a brand name, which should support future sales.

Because accounting methods do not include these expenditures in a company’s asset base, they are not an appropriate method for determining an economic return on assets and hence will not generate efficient prices. Interestingly, when the OIG determined a rate of return in the case of erythropoietin, it did not use the accounting definition of assets but rather calculated past investment in Amgen to measure its assets.³⁹ Unfortunately, it is not clear if it is generally feasible to obtain the data necessary to compute an economic return, given that the firm’s audited financial statements are based on the accounting conventions. Mergers and acquisitions would further complicate calculations. Finally, a proper economic calculation would account for assets outside the United States, as the example of a company based entirely outside the United States makes clear, but non-American firms use different accounting conventions, and it might be difficult to put their data on a comparable footing with U.S. data.

Even if meaningful numbers could be obtained, this method has drawbacks. It would artificially strengthen the existing trend toward consolidation in the pharmaceutical industry. Consider a start-up biotech company whose only drug is a blockbuster. Because the profit it could receive as a stand-alone company would be limited, it would want to merge with other companies that had on average less successful drugs. The price it received from being acquired by such a company, one less constrained by the overall limit on profit, could more than compensate it for its forgone profits. It might want to merge with a large company in any event, but it is not sound policy to force a merger simply to evade a pricing constraint.

Finally, this method offers the firm an incentive to add excessive capital. The incentive arises because for a firm at the constrained rate of profit, each additional dollar spent on assets relaxes the constraint.⁴⁰

Is There A Way Out?

Top Editor's Notes Four Approaches To The... The Economics Of Pharmaceuticals What Is The Right... Options In Setting Administered... Is There A Way... Excess Profits In... Notes

 
In general, formularies and tiered or reference pricing ought to achieve reasonable supply prices when there is competition among drugs, as is consistent with their widespread use in commercial insurance. But when there is no competition, the proposed Medicare drug benefit appears vulnerable to abusive supply pricing by manufacturers. Why does this problem appear not to arise with commercial insurance?

In the case of commercial insurance, the insurer can threaten to not cover a drug that it deems to be too expensive or not cost-effective. This threat is certainly credible if there are substitutes in a therapeutic class. Even if the drug has no close substitutes, the PBM retains some leverage because of its threat not to cover the drug. No coverage means that the manufacturer is in the position of the textbook monopolist, who, facing a market with no insurance and thus some price-sensitivity to demand, can in principle define a finite price that yields a maximum profit. As long as the price negotiated with the insurer gives the drug manufacturer at least as great a profit as it would earn from the uninsured market, the manufacturer should agree to the insurer’s price instead of having the drug excluded from coverage.⁴¹

Nonetheless, it is clearly difficult for a private insurer to exclude a drug from coverage, especially one that is a major therapeutic advance. Consumers with insurance naturally expect that the drug their physician prescribes or a satisfactory substitute will be covered. As a result, the manufacturer may be able to negotiate a price that earns much more than would be earned in an uninsured market. However difficult it is for a private insurer to exclude a drug because of cost, it is even more difficult for a public program such as Medicare to do so.

By law Medicare is to cover services that (1) are considered reasonable and necessary for the diagnosis and treatment of illness or injury or to improve the functioning of a malformed body member and (2) are not statutorily excluded.⁴² Because cost is not explicitly mentioned, it is not clear what weight it can carry in decisions to exclude a product or service. In 1989 HCFA (now the Centers for Medicare and Medicaid Services, or CMS) issued a Notice of Proposed Rule Making that proposed to account for cost in coverage decisions, but this rule was never made final because of pressure from the medical device industry.⁴³ If cost cannot be considered, Medicare’s bargaining power over price is reduced. So too is the bargaining power of any PBM that is acting as Medicare’s contractor.⁴⁴

The Clinton Health Security Act did propose that the HHS secretary could exclude a drug from coverage if its price were too high. Not surprisingly, this proposal attracted the wrath of the drug industry. Even if such a provision had been enacted into law, it is an open question whether Medicare’s threat to exclude a drug with major therapeutic benefits and no good alternatives would be credible.

Nonetheless, given that manufacturers with unique and important drugs could ask extortive prices and that all of the formal price-setting methods described above have serious drawbacks, it would seem that Congress should allow cost as a factor in coverage decisions. This could lead to a formal cost-effectiveness criterion in coverage decisions, but even the ability to consider cost might suffice to mitigate abusive pricing.

Excess Profits In Pharmaceuticals?

Top Editor's Notes Four Approaches To The... The Economics Of Pharmaceuticals What Is The Right... Options In Setting Administered... Is There A Way... Excess Profits In... Notes

 
The four methods of controlling price I have described all have their drawbacks, as any price-setting method must, simply because of limited information, not to mention possible distortions from political rent seeking. For these reasons, I favor minimizing the degree of price setting that a drug benefit would entail. Because the profit-maximizing price for a monopoly drug is divided by the coinsurance rate, I also favor the use of a formulary, with beneficiaries paying markedly higher prices for off-formulary drugs, instead of relying solely on coinsurance to restrain prices.

Many people favor a more dirigiste approach to pricing on the grounds that pharmaceutical industry profits are already excessive and that Medicare’s adopting a less restrictive approach could swell them even further. But the profit figures usually cited are based on the accounting conventions that exclude both R&D and marketing expenses from the asset base. As a result, the usual profit figures greatly overstate the industry’s economic profit rate. The U.S. Congress Office of Technology Assessment (OTA) estimated that the industry’s economic rate of return from 1976 to 1987 was only two to three percentage points above the return in other industries, not the four to six percentage points estimated using standard accounting conventions.⁴⁵ That the pharmaceutical industry would have a somewhat above-average profit rate is not surprising because of its extensive use of patent protection and the riskiness of its R&D investments. Although some degree of administered pricing for drugs might ultimately be desirable in Medicare, sharply lowering the pharmaceutical industry’s current rate of return does not appear to be a strong argument for that course.

Editor's Notes

Top Editor's Notes Four Approaches To The... The Economics Of Pharmaceuticals What Is The Right... Options In Setting Administered... Is There A Way... Excess Profits In... Notes

 
Joe Newhouse is the John D. MacArthur Professor of Health Policy and Management at Harvard University.

The author thanks the Alfred P. Sloan Foundation and the Agency for Healthcare Research and Quality (Grant no. P01-HS-10803) for support and Adam Block for able research assistance. Richard Frank, Tom McGuire, Bob Reischauer, and three anonymous reviewers gave him helpful reactions to an early draft.

Notes

Top Editor's Notes Four Approaches To The... The Economics Of Pharmaceuticals What Is The Right... Options In Setting Administered... Is There A Way... Excess Profits In... Notes

 

1. In reality, there are prices all along a supply chain from manufacturers to wholesalers to pharmacy benefit managers (PBMs) to retail drug stores and mail-order houses. In the interest of simplicity, I abstract from these various supply prices.
2. Given both overuse and underuse of certain drugs, one cannot say a priori whether greater use of a drug is good or bad.
3. The Congressional Budget Office projects that in 2005, 75 percent of beneficiaries will spend at least $500 on drugs, and they will account for 99 percent of total drug spending. Congressional Budget Office, Issues in Designing a Prescription Drug Benefit for Medicare (Washington: CBO, 2002). Current legislation has proposed deductibles in the $250–$275 range but also has a well-publicized "doughnut," or a range of spending with no coverage. Beneficiaries in the doughnut will face the full incremental price of drugs. In what follows, I focus on those beneficiaries who have exceeded the deductible but are not in the doughnut or, more precisely, who expect the marginal dollar of drug spending not to be in those regions. In the case of a fixed-dollar copayment (for example, $10 per month), the marginal coinsurance rate is zero.
4. As with health plan pricing, this price could in principle be risk-adjusted.
5. Because multiple PBMs invite efforts by each PBM to enroll good risks and avoid bad ones, I prefer that PBMs bid for temporary local monopolies, just as employers only contract with one PBM at a time. See H.A. Huskamp et al., "The Medicare Prescription Drug Benefit: How Will the Game Be Played?" Health Affairs (Mar/Apr 2000): 8–23.
6. For reasons explained by Haiden Huskamp and colleagues, I favor the consumer’s paying the entire incremental cost (reference pricing) rather than just a higher copayment for drugs not on the formulary. Ibid.
7. P. Kanavos and U.E. Reinhardt, "Reference Pricing for Drugs: Is It Compatible with U.S. Health Care?" Health Affairs (May/June 2003): 16–30.
8. Huskamp et al., "The Medicare Prescription Drug Benefit."
9. Similar cases can arise even if multiple drugs are available for a given clinical problem. Some patients, for example, may only respond to one of the drugs available or have intolerable side effects from all the others. For those patients, the manufacturer of the relevant drug would be in a monopoly position.
10. See F.M. Scherer, "The Pharmaceutical Industry," in Handbook of Health Economics, ed. A.J. Culyer and J.P. Newhouse (Amsterdam: North-Holland, 2000), 1297–1336.
11. Recent estimates place the average R&D cost to bring a drug to market around $800 million. J.S. DiMasi et al., "The Price of Innovation: New Estimates of Drug Development Costs," Journal of Health Economics (March 2003): 151–185. Although some believe that the $800 million figure is too high, it is still clear that there are substantial fixed costs.
12. The Office of Technology Assessment summarized two earlier studies that placed the probability of ultimate approval at 13 and 23 percent. U.S. Congress Office of Technology Assessment, Pharmaceutical R&D: Cost, Risks, and Rewards, Pub. no. OTA-H-522 (Washington: U.S. Government Printing Office, 1993), 52. Joseph DiMasi and colleagues ("The Price of Innovation") estimate a probability of 22 percent (calculated from the 31.4 percent probability of a drug’s entering a phase III trial, Table 1, and a 68.5 percent success rate in phase III, page 163).
13. By probabilistic basis, I mean that the expected revenues, accounting for the likelihood of success, will exceed the costs of development. Because successful drugs must cover the development costs of unsuccessful drugs, successful drugs will have to more than recover their costs of development for the firm to stay in business. Moreover, this must be true at each stage of the development process, since a manufacturer can abandon development at any time if it receives new information suggesting the drug will not recover investments that must yet be made to bring a drug to market. R.G. Frank, "New Estimates of Drug Development Costs," Journal of Health Economics (March 2003): 325–330.
14. R.G. Frank, "Prescription Drug Prices: Why Do Some Pay More than Others Do?" Health Affairs (Mar/Apr 2001): 115–128. Maintaining different prices within a country requires that there be no arbitrage across the consumer groups.
15. J. Tirole, The Theory of Industrial Organization (Cambridge, Mass.: MIT Press, 1988).
16. C.L. Schultze, The Public Use of Private Interest (Washington: Brookings Institution, 1977).
17. Any profit-maximizing manufacturer will expend marketing resources until the incremental net revenue from the sales resulting from the additional marketing expense equals that expense. R. Dorfman and P. Steiner, "Optimal Advertising and Optimal Quality," American Economic Review (December 1954): 826–836.
18. For example, TAP Pharmaceuticals in 2002 settled criminal and civil litigation brought by the federal government for marketing abuses with Lupron, a cancer drug, for $875 million. M. Peterson, "Two Drug Makers to Pay $875 Million to Settle Fraud Case," New York Times, 4 October 2001. See also U.S. vs. TAP Pharmaceutical Products, Inc., U.S. District Court for the District of Massachusetts, Eastern Division, Sentencing Memorandum of the United States, Criminal Action no. 01-CR-10354-WGY. In a similar case involving Zoladex, a competitor drug to Lupron, AstraZeneca pleaded guilty to a felony and paid a $355 million fine. M. Peterson, "AstraZeneca Pleads Guilty in Cancer Medicine Scheme," New York Times, 21 June 2003. Warner-Lambert, now part of Pfizer, is accused of paying physicians to promote the use of Neurontin for off-label uses, something that would be illegal for the company to do directly. M. Peterson, "Court Papers Suggest Scale of Drug’s Use," New York Times, 30 May 2003. Schering-Plough recently announced that it expected to be indicted over marketing abuses. M. Peterson, "Indictment Seen by Drug Maker over Marketing," New York Times, 31 May 2003.
19. Economists generally value the last unit of a good bought by the demand price paid for it, but this criterion is problematic in medical care for many reasons, including the role of the physician.
20. This consideration is often cited by pharmaceutical companies in debates over drug budgets. They argue that reducing dollars spent on drugs could be self-defeating by causing more than offsetting increases in other types of medical spending. There can be little doubt that their argument is sometimes correct. See S.B. Soumerai et al., "Effects of Medicaid Drug-Payment Limits on Admissions to Hospitals and Nursing Homes," New England Journal of Medicine 325, no. 15 (1991): 1072–1077[Abstract]; and S.B. Soumerai et al., "Effects of Limiting Medicaid Drug-Reimbursement Benefits on the Use of Psychotropic Agents and Acute Mental Health Services by Patients with Schizophrenia," New England Journal of Medicine 331, no. 10 (1994): 650–655.[Abstract/Free Full Text]
21. See, for example, the discussion of substitution among the marginal hospital day and various types of postacute care in J.P. Newhouse, "Medicare," in American Economic Policy in the 1990s, ed. J. Frankel and P. Orszag (Cambridge, Mass.: MIT Press, 2002). The problem is somewhat easier in the context of health plans, which can in principle trade off dollars for each type of medical input; for example, managed behavioral health care plans have substituted drugs for behavioral therapy to some degree. R.G. Frank et al., "Price Indexes for the Treatment of Depression," in Measuring the Prices of Medical Treatments, ed. J.E. Triplett (Washington: Brookings Institution, 1999), 72–117; and E.R. Berndt et al., "The Medical Treatment of Depression, 1991–1996: Productive Inefficiency, Expected Outcome Variations, and Price Indexes," Journal of Health Economics (May 2000): 373–396. But only a small minority of Medicare beneficiaries are in health plans.
22. If Medicare sets prices, there is a question of whether manufacturers would be allowed to charge additional uninsured amounts, the equivalent of balance billing. Although I doubt Medicare would permit this, Jacob Glazer and Tom McGuire argue that it is efficient to do so in the physician context. J. Glazer and T.G. McGuire, "Should Physicians Be Permitted to ‘Balance Bill’ Patients?" Journal of Health Economics (October 1993): 239–258.
23. Most of these options lend themselves both to take-it-or-leave-it pricing and to negotiation.
24. Medicare Payment Advisory Commission, Report to the Congress: Medicare Payment Policy (Washington: MedPAC, March 2003).
25. Ibid.; and J.K. Iglehart, "Medicare and Drug Pricing," New England Journal of Medicine 348, no. 16 (2003): 1590–1597.[Free Full Text]
26. Iglehart, "Medicare and Drug Pricing."
27. U.S. General Accounting Office, Medicare: Payments for Covered Outpatient Drugs Exceed Providers’ Cost, Pub. no. GAO-01-1118 (Washington: GAO, 2001). Physicians argue that these payments compensate them for Medicare’s failure to pay a fee for administration. MedPAC, Report to the Congress: Medicare Payment Policy.
28. R. Abelson, "Drug Sales Bring Huge Profits, and Scrutiny, to Cancer Doctors," New York Times, 26 January 2003. Although it potentially represents shifting among brands rather than increased overall prescribing, civil litigation arising out of the Lupron case alleges that the manufacturer (TAP) raised the AWP while keeping its actual sales price constant as an inducement to the prescribing physician to use Lupron rather than Zoladex, a competing drug. In Lupron Marketing and Sales Practices Litigation, MDL no. 1430, CA no. 01-CV-10861, U.S. District Court, District of Massachusetts.
29. An analogous method would be to pay the dispenser of the drug, such as a pharmacy, physician, or clinic, its acquisition cost. This would also invite high prices from manufacturers, since Medicare would reimburse the manufacturer’s price.
30. Calculated from J.A. Poisal and G.S. Chulis, "Medicare Beneficiaries and Drug Coverage," Health Affairs (Mar/Apr 2000): 248–256; and K.R. Levit et al., "Health Spending in 1998: Signals of Change," Health Affairs (Jan/Feb 2000): 124–132.
31. Although it is difficult to attribute causality to the legislation, the theoretical expectation of a price increase is very strong. F.S. Morton, "The Interaction between a Most-Favored-Customer Clause and Price Dispersion: An Empirical Examination of the Medicaid Rebate Rules of 1990," Journal of Economics and Management Strategy (Spring 1997): 151–174; and F.S. Morton, "The Strategic Response by Pharmaceutical Firms to the Medicaid Most-Favored-Customer Rules," RAND Journal of Economics (Summer 1997): 269–290. See also GAO, Changes in Drug Prices Paid by VA and DoD since Enactment of Rebate Provisions, Pub. no. GAO/HRD 91-139 (Washington: GAO, 1993); and CBO, How the Medicaid Rebate on Prescription Drugs Affects Pricing in the Pharmaceutical Industry (Washington: CBO, 1996).
32. See, for example, MedPAC, Report to the Congress: Selected Medicare Issues (Washington: MedPAC, June 2003), 23.
33. Economic theory suggests that markups over marginal cost that recover fixed cost be set differently by drug according to the distortions in behavior induced by the varying markups. F.P. Ramsey "A Contribution to the Theory of Taxation," Economic Journal 37, no. 1 (1927): 47–61. This is the rationale, for example, for paying less for discretionary admissions in the diagnostic-based risk-adjustment system used to pay health plans. The feasibility of adapting such a method to drug pricing, however, is doubtful.[CrossRef]
34. There is a similar problem for physician practice costs that cannot be directly allocated to specific HCFA Common Procedure Coding System (HCPCS) codes, such as rent or utilities, as well as in the DRG system, but it is likely an even more quantitatively important issue for drugs.
35. OTA, Recombinant Erythropoietin: Payment Options for Medicare, Pub. no. OTA-H.451 (Washington: U.S. Government Printing Office, 1990), Appendix E.
36. In addition to the unanticipated growth in the market, pricing might have also contributed to a higher return. From 1989 to 1991 Medicare paid $40 per 10,000 units of the drug if less than 10,000 units were given and otherwise paid $70. In setting this price, HCFA assumed that the average dose would be 5,000 units; in fact, it ranged from about 2,500 to 2,800 units. Nancy Ray, MedPAC, personal communication, 18 June 2003. The degree to which the manufacturer as opposed to dialysis providers profited is unclear. In response to concerns about underdosing, beginning in 1991 Medicare paid $11 per 1,000 units of the drug, a price that dropped to $10 in 1993. MedPAC, Report to the Congress: Selected Medicare Issues (Washington: MedPAC, June 1999). Relative to 1991, the real price by 2003 has dropped by about a third.
37. C.P. Stickney and R.L. Weil, Financial Accounting: An Introduction to Concepts, Methods, and Uses (Orlando, Fla.: Dryden Press, 2000), 437; OTA, Pharmaceutical R&D; and K.G. Palepu et al., Business Analysis and Evaluation Using Financial Statements (Cincinnati: South-Western College Publishing, 2000), chap. 7, 6. By standard U.S. accounting, I am referring to generally accepted accounting principles (GAAP).
38. The accounting treatment stems from uncertainty about future profit, but, as Clyde Stickney and Roman Weil point out, there is also uncertainty about many physical assets that accounting conventions do put on the balance sheet as assets. They find the rationale for the accounting treatment of pharmaceutical research to be "less than compelling." Stickney and Weil, Financial Accounting.
39. OTA, Recombinant Erythropoietin, Appendix E.
40. H. Averch and L.L. Johnson, "Behavior of the Firm under Regulatory Constraint," American Economic Review (December 1962): 1052–1069.
41. Because of the additional sales when the drug is covered, the negotiated price could be well below the profit-maximizing price in the uninsured market.
42. The language is from Section 1862(a)(1)(A) of the Social Security Act (42 U.S.C. 1395y), although the act is written in the negative; that is, no payment shall be made for services that do not satisfy the reasonable and necessary criterion.
43. S.B. Foote, "Why Medicare Cannot Promulgate a National Coverage Rule: A Case of Regula Mortis," Journal of Health Politics, Policy and Law (October 2002): 707–730.
44. A drug that is an important advance with no close competitors would be on any reasonable formulary. If a PBM had to cover the drug, it would have no ability to control price.
45. OTA, Pharmaceutical R&D, 101.

                      What's this?

This article has been cited by other articles:

				M. Fotaki, M. Roland, A. Boyd, R. McDonald, R. Scheaff, and L. Smith What benefits will choice bring to patients? Literature review and assessment of implications J Health Serv Res Policy, July 1, 2008; 13(3): 178 - 184. [Abstract] [Full Text] [PDF]

				C. White Why Did Medicare Spending Growth Slow Down? Health Aff., May 1, 2008; 27(3): 793 - 802. [Abstract] [Full Text] [PDF]

				R. G. Frank and J. P. Newhouse Should Drug Prices Be Negotiated Under Part D Of Medicare? And If So, How? Health Aff., January 1, 2008; 27(1): 33 - 43. [Abstract] [Full Text] [PDF]

				L. L. Berry and N. Bendapudi Health Care: A Fertile Field for Service Research Journal of Service Research, November 1, 2007; 10(2): 111 - 122. [Abstract] [PDF]

				A. Garcia-Altes, C. Borrell, L. Cote, A. Plaza, J. Benet, A. Guarga, and for the Montreal and Barcelona group on performanc Measuring the performance of urban healthcare services: results of an international experience J. Epidemiol. Community Health, September 1, 2007; 61(9): 791 - 796. [Abstract] [Full Text] [PDF]

				E. Kelley Health, spending and the effort to improve quality in OECD countries: a review of the data The Journal of the Royal Society for the Promotion of Health, March 1, 2007; 127(2): 64 - 71. [Abstract] [PDF]

				S. D. Ramsey How Should We Pay the Piper When He's Calling the Tune? On the Long-Term Affordability of Cancer Care in the United States J. Clin. Oncol., January 10, 2007; 25(2): 175 - 179. [Full Text] [PDF]

				E. T. kelley, I. Arispe, and J. Holmes Beyond the initial indicators: lessons from the OECD Health Care Quality Indicators Project and the US National Healthcare Quality Report Int. J. Qual. Health Care, September 1, 2006; 18(suppl_1): 45 - 51. [Abstract] [Full Text] [PDF]

				J. Hsu, M. Price, J. Huang, R. Brand, V. Fung, R. Hui, B. Fireman, J. P. Newhouse, and J. V. Selby Unintended consequences of caps on medicare drug benefits. N. Engl. J. Med., June 1, 2006; 354(22): 2349 - 2359. [Abstract] [Full Text] [PDF]

				N. E. Morden and S. D. Sullivan States' Control Of Prescription Drug Spending: A Heterogeneous Approach Health Aff., July 1, 2005; 24(4): 1032 - 1038. [Abstract] [Full Text] [PDF]

				M Weiner, T E Stump, C M Callahan, J N Lewis, and C J McDonald Pursuing integration of performance measures into electronic medical records: beta-adrenergic receptor antagonist medications Qual. Saf. Health Care, April 1, 2005; 14(2): 99 - 106. [Abstract] [Full Text] [PDF]

				N. K. Choudhry and A. S. Detsky A Perspective on US Drug Reimportation JAMA, January 19, 2005; 293(3): 358 - 362. [Full Text] [PDF]

				M. B. Rosenthal Doughnut-Hole Economics Health Aff., November 1, 2004; 23(6): 129 - 135. [Abstract] [Full Text] [PDF]

				E. A. McGlynn There Is No Perfect Health System Health Aff., May 1, 2004; 23(3): 100 - 102. [Abstract] [Full Text] [PDF]