Introduction and summary
Shortages of prescription generic drugs, which account for nearly 90 percent of prescriptions, are frequent and can last for several months or even years. A recent study shows that from 2017 to 2021, the U.S. Food and Drug Administration (FDA) received 731 manufacturer “supply chain issue reports,” which are meant to identify shortages that could affect the national supply of an important prescription drug.1 Of those, 113 drugs had a “meaningful” shortage, defined as a reduction of 33 percent or more in the quantity supplied within six months of the issuer report compared with the preceding three months. A substantial majority of these shortages were for drugs where generics are available, a majority lasted for more than a year, and the median age of the drug was 17 years.2
The consequences of these shortages for patient health can be severe: In 2018, for example, carcinogenic contamination of valsartan, a generic drug used to treat high blood pressure, forced an FDA recall of the drug. Manufacturers were unable to resolve the problem quickly, resulting in a serious shortage that lasted for years.3 Even though other blood pressure drugs were available, they were not perfect substitutes, and the health effects of the shortage were large. According to recent analysis, patients taking valsartan at the time of the shortage showed a statistically significant increase in cardiac events such as heart failure, heart attack, and stroke in the following six months.4
According to recent analysis, patients taking valsartan at the time of the shortage showed a statistically significant increase in cardiac events such as heart failure, heart attack, and stroke in the following six months.
The failure of prescription generic drug suppliers to avert and to respond to supply disruptions requires explanation. For many generic commodities such as copy paper or gasoline, significant, long-term supply disruption is not the norm. Producers of these commodities generally invest adequately in plants, equipment, and operations to prevent shortages, and buyers support those investments through the prices they pay. When shortages do occur, markets usually self-correct in a reasonably short time. Rising prices signal the shortage, inventories buffer the disruption, and there is often some spare capacity to make up for shortfalls. However, this is not true for generic drugs.
The structure of generic supply is part of the problem. The supply of individual generic drugs is characterized by a small number of firms selling products that are close substitutes, competing on price, and often earning little profit.
The structure of generic demand also contributes to shortages. A small number of group purchasing organizations (GPOs) are dominant wholesale buyers of generics, and their bargaining power exerts additional downward pressure on wholesale prices and manufacturing profits.
But this concentrated supply also makes it more likely that when output is interrupted at one firm—because of operational or quality problems or because of business reasons—the loss of supply will be significant. In addition, low profits mean that existing suppliers are unlikely to maintain inventories or invest in spare capacity that might absorb a supply shock.
These economic conditions together inhibit market self-correction and raise the probability of harm to patients whose health depends on a stable supply of generic drugs.
Policy interventions, however, can reduce shortages and avert patient harm. They consist of a coordinated set of public-private actions to address the market failures that put patients at risk, including:
- Identifying a set of priority generic drugs
- Building national inventories of these drugs
- Establishing large-scale nonprofit public-private partnerships to contract for reliable supply of these drugs
- Supporting domestic nonprofit public-private partnerships to manufacture priority generics when needed
- Limiting geographic risks to the supply of active pharmaceutical ingredients and final form drugs
An industrial policy along these lines would address a serious, persistent problem that otherwise would not be resolved.
Why generic supply is unreliable
Concentration and small market size are typical of generics. A recent economic study found that more than 40 percent of generics have a single producer. The median number of producers in a market for an individual drug is two and the 75th percentile number is five to six, with the number of competitors positively correlated with market size. Median annual sales were less than $1 million, 75th percentile revenue was $4 million, and average revenue was $12 million. The large ratio of mean to median revenue indicates that revenues for individual drugs are often highly concentrated in a small number of firms, even when there are several producers.5
Shortages are more frequent when markets are concentrated and revenue is low. A study of 2017–2021 drug shortages found that a substantial majority of meaningful shortages for all drugs occurred when there were four or fewer manufacturers or when market revenue was less than $5 million. Most shortages lasted for two or more years.6
Another recent study of 132 drugs with current supply chain issue reports—120 of which are generics—produced similar results. The report showed that more than 75 percent of current shortages are in highly concentrated markets, 56 percent are for drugs with a unit list price of less than $1, and 58 percent of the shortages have lasted two or more years.7
Recently, an FDA study of drug shortages from 2013 to 2017 concluded that drug shortages are often associated with drugs that have a “weak business case,” meaning either the drug or the facility in which it is made is insufficiently profitable to justify investments to improve manufacturing processes.8
When profits are low, existing suppliers have little incentive to invest in or to maintain spare capacity to meet market fluctuations. Also absent are incentives for suppliers to hold sometimes perishable buffer inventories, which could be used to meet spikes in demand or interruptions in production. This adds to the risk created by small numbers of producers.
Given the connection between the structure of supply and shortages, an important open question remains: Why do market prices and firm revenues fail to support stable supplies of essential generics when there is sustained need for them and there are few suppliers?
Economic theory suggests an explanation. When firms cannot differentiate their products and compete on price alone but the number of firms is small, they can sometimes have tacit agreements to hold the price above unit cost. When the market price is greater than unit cost, all firms in the market earn profits. If an individual firm were to cut its price, it could gain market share and profits in the short term. However, the firm will balance short-term profits gained against the loss of future profits when other firms retaliate by cutting prices. When the number of competitors is very small, price cuts may not be rational.
However, the reason for an individual firm to maintain a price above unit cost declines with the number of competitors. This is because total profits are divided among market participants. When there are more firms, the short-term profit that can be gained by cutting price and gaining market share may outweigh the loss of future profits. Hence, even small increases in the number of competitors can reduce the likelihood of tacit collusion, drive prices toward unit cost, and limit profits.9
A recent study of a cohort of 77 brand name drugs that lost patent protection from 2010 to 2013 shows a pattern of entry and price change consistent with this model.10 When a drug loses patent exclusivity, generic manufacturers have incentive to enter the market since existing prices are at monopoly levels. This incentive is amplified by a legal provision that gives six months of sales exclusivity to the first generic manufacturer to be approved by the FDA for marketing under an abbreviated new drug application (ANDA).
From 2010 to 2013, generic manufacturers responded to the profit incentives this created, filing an average of eight ANDAs for each of the 77 drugs in the cohort. The number of entrants was correlated with market size: The largest markets had greater entry and large price declines over 15 quarters, and the smallest markets had far fewer entrants and small or no price declines.11
These price and quantity outcomes are consistent with theoretical expectation: When firms are competing exclusively on price, small increases in the number of firms will push price toward unit cost.
Moreover, even when the number of firms in a particular generic market is small, the number of potential competitors may be high. For example, the FDA found that for each drug in its shortage sample, there were, on average, three firms with ANDAs that were not producing it.12 Generic markets such as these can be characterized as “contestable” because credible potential entrants and ease of entry limit incumbent pricing power.13 Since several large generic firms hold large portfolios of ANDAs, contestability may be characteristic of the markets for many generics even when the number of current producers is small.14
In short, small markets subject to intense competition keep profits and the number of producers low. Incentives for investment in inventories or spare productive capacity are therefore limited. The resulting structure of supply is vulnerable to economic and operational disruptions that contribute to shortages.
Effects of concentrated generic demand
The structure of generic demand also likely contributes to shortages. Economic theory suggests that buyer concentration confers bargaining power over price.15 Since wholesale buying of generic drugs is concentrated in the hands of relatively few GPOs, they have every incentive to use their bargaining power to exert downward pressure on wholesale prices.16
About 90 percent of drugs sold through pharmacies are purchased at wholesale by three GPOs that are joint ventures of large drug wholesalers and retail pharmacy chains.17 Similarly, four hospital GPOs make wholesale purchases of about 90 percent of hospital medical supplies.18 So if suppliers want drugs to have wide access to major pharmacy chains, they need to go through three GPOs. If suppliers want to have wide access to most hospitals, they need to go through the other four GPOs.19 Demand consolidated to this degree should confer significant bargaining power. A study from the U.S. Government Accountability Office reviews evidence that GPOs have bargaining power when purchasing drugs.20
Market response to generic shortage
When markets experience excess demand, they usually self-correct. As FDA economists have pointed out, the expectation is for excess demand to be followed by price increases, greater production from incumbent firms or entrants, and restoration of original quantities supplied. However, generic drug markets generally do not self-correct in this way.
FDA economists measured changes in prices and quantities following a shortage using data for a sample of drugs in shortage from 2013 to 2017. They found that price increases, quantity increases, and restoration of supply were not the norm. All three outcomes occurred together in only 2 percent of the sample. Moreover, only 18 percent of drugs experienced a sustained price increase, meaning an increase of 50 percent or more that lasted six months; only 42 percent experienced a significant production increase, or suppliers increasing output enough to restore at least 50 percent of the shortage; and only 30 percent had supply restored to pre-shortage levels within 12 months.21
These market responses to supply shocks are much different from those that occur in the retail market for gasoline, where brands exist but firms are all supplying the same commodity. Like markets for many individual generics, the U.S. gasoline market has a handful of producers that operate a relatively small number of refineries.22 Moreover, a single refinery shutdown, similar to a disruption at a generic drug manufacturer, can have a measurable impact on the flow of gasoline to a regional market. When that happens, prices rise, leading to sales from inventories. Those inventories are large, amounting to about a month of national gasoline consumption. Inventory drawdowns cushion the lost supply and help moderate price increases until production is restored. Shutdowns are usually temporary.23
The differences in market responses following shortages reflect differences in market structure. Supplying gasoline is a profitable business. Even though refiners are all selling the same thing, there is little danger that new competitors will enter the market to compete for share and drive down price. The last U.S. refinery with significant capacity was built in the 1970s, and there are limits to the expansion of existing refinery capacity.24 In addition, the market is large and profitable enough to support large inventories that can buffer shocks to supply or demand. Demand for gasoline is mostly from individual consumers, who lack bargaining power and can do nothing individually to push price toward marginal cost. Hence, the market for gasoline, while dominated by an oligopoly with all the associated inefficiencies, self-corrects in a way that markets for many generic drugs do not.
Policy recommendations to limit generic shortages
Public intervention is needed given the serious health consequences of prescription generic shortages and the observable failure of markets to prevent shortages or to self-correct when they occur.
Enact policies to subsidize demand
A potential strategy to consider is subsidized demand. If manufacturers can earn normal profits from maintaining production capacity and inventories that avert shortages, they are more likely to do so.
Policy experts have proposed a version of this strategy, which the U.S. Department of Health and Human Services (HHS) recently endorsed.25 The plan is to increase the Medicare hospital reimbursement rate when generics are purchased from designated reliable manufacturers. The intent is to divert demand to a subset of producers, raise the prices they receive, and thereby reduce shortages.
While this would affect hospital purchases, outcomes may be limited. Determining an effective reimbursement level is a challenging task. It requires estimating the prices needed to support capacity and inventories and ensuring that manufacturers are paid those prices. This requires either substantial information about manufacturer costs and profits or a willingness to guess at reimbursement levels and adjust them as production conditions change. Moreover, because the GPOs that purchase most hospital generics have bargaining power, they will capture some of the increased reimbursements and dilute the effect on manufacturers. And of course, this strategy would not extend to generics purchased outside of hospitals.
A larger-scale approach, using a wider set of demand and supply instruments, seems necessary. It would include five elements:
1. Identify priority generics
Efforts to prevent shortages need to be carefully directed given the large number of generic drugs produced and used. This requires analyzing factors such as the expected harm caused by a shortage, the expected number of patients affected, and the likelihood and duration of shortages to prioritize efforts.26 HHS’ Administration for Strategic Preparedness and Response has a priority ranking that considers many of these factors.27 There are proposals for refining the list, but the feasibility of constructing one is clear.28
2. Build national inventories of priority generics
Inventories can provide a short-term buffer that limits harm to patients when production flows are disrupted.29 However, generic manufacturers cannot be counted on to maintain them. Particularly for priority drugs, publicly supported and overseen inventories are needed in order to reduce harm and to deliver net economic benefits.30
These inventories could be managed using existing private infrastructure. Drug wholesalers hold large inventories of prescription drugs and already deliver drugs from their inventories to pharmacies and health care facilities across the country. The federal government could contract with these wholesalers to hold required buffer inventories, rotate them, and draw them down and sell them at acquisition cost during shortages.
An alternative would be to build on the infrastructure of the U.S. Department of Veterans Affairs (VA) Consolidated Mail Outpatient Pharmacies.31 These online mail-order pharmacies manage drug delivery for the VA and for military medical systems. Their experience with inventory management and distribution indicates that their role could be expanded efficiently.
3. Establish large-scale nonprofit public-private partnerships to contract for reliable supply of priority generics
Some chronic generic shortages could be addressed by coordinating existing large-scale purchases through nonprofit public-private partnerships (PPPs). For example, the VA medical system, hospitals, nonprofit medical insurers, and state Medicaid providers are large buyers of generics, and each has an interest in preventing adverse patient outcomes caused by shortages. They could collectively contract with manufacturers to provide priority drugs.
PPP contracts that are long term, with commitments to buy guaranteed minimum amounts and with payment levels conditioned on quality and delivery performance, would provide incentives for manufactures to invest in adequate capacity and operational conditions. Competitive auctions with multiple awards can determine prices for quantities that the partnership is able to estimate. Contract bid prices will reflect manufacturer estimates of the revenue needed to meet contract conditions. These contracts can also be used to source national inventories.
Civica Rx, a nonprofit established by hospitals, nonprofit insurers, and foundations, has implemented a version of nonprofit demand aggregation.32 It acts as a buying group for drugs that are often in shortage. Its hospital and Blue Cross Blue Shield members make volume commitments for five years. The buying group also provides some supply to the VA and to the Department of Defense.
4. Support domestic nonprofit PPP manufacture of priority generics when needed
If reasonable contracts with existing manufacturers are not possible, the federal government could help organize (and if necessary, subsidize) PPPs to manufacture priority drugs and sell them at cost. Members of the PPP could provide equity or debt to help finance needed plants and equipment and to guarantee a minimum level of demand. Other capital funding could come from guaranteed or federal loans. Privately provided equity and debt would incentivize PPP members to build and to operate efficiently. Given the need for close monitoring by both private and public partners, and the need for a resilient and secure supply chain, PPP-financed production facilities ought to be in the United States.
Civica Rx, in addition to acting as a buyers club, has taken initial steps toward nonprofit generic manufacturing: It is building a finished dosage form drug facility to produce generic insulin and other biosimilars and sell them at cost. The threat of entry from Civica Rx, along with changes to federal regulations and other forms of public pressure, has resulted in a remarkable decline in the price of insulin.33 Civica Rx has reportedly received $100 million as a subcontractor to Phlow Corp. to support plant construction.34
5. Limit geographic risks to the supply of active pharmaceutical ingredients and final form drugs
Geographic risk needs to be considered when designing policy to prevent shortages because supply chains for generics are spread around the globe.
The manufacture of drugs requires several discrete steps. Biochemical ingredients are used to create intermediate inputs. These inputs are combined into biologically active agents, called active pharmaceutical ingredients (APIs). APIs are then included in the consumable finished dosage form drug (FDF). APIs and FDFs are frequently produced by separate companies, most of which are outside the United States.35 The geographic location of drug manufacturers can provide competitive advantages such as low-cost labor supply and economies of scale and scope that have developed over time.
However, this opaque and often highly concentrated global supply chain has produced dangerous failures in manufacturing quality, leading to patient harm and shortages of essential generics. For example, the 2018 valsartan recall and the resulting shortage were prompted by the discovery of carcinogens in the medication. The source of the contamination was ultimately traced to the drug APIs, which had been manufactured in India and China.36 The FDA discovered that the contaminant was linked to a chemical byproduct created during the manufacturing process. The generic API manufacturers had switched the manufacturing process from the one used when the drug was under clinical trial.
Failure in the supply chain for heparin, an anticoagulant commonly used in surgeries and by patients undergoing dialysis, provides another example. China is the source of 80 percent of the world supply and 60 percent of the U.S. supply of crude heparin, which is obtained from the tissue of pigs and is the API of heparin. In 2007, when swine flu caused widespread destruction of herds, Chinese crude heparin was deliberately contaminated with another, cheaper chemical.37 Before the contamination was discovered, it led to 80 deaths and 800 serious adverse events in the United States and the subsequent heparin recall created severe shortages. Limits on the ability of the FDA to inspect manufacturing facilities continue to make it difficult to trace a given lot of heparin to a specific farm.
When priority finished form drugs, or their APIs, have quality or stability risk and are located offshore, there is compelling reason to move production onshore and increase oversight. Both demand and supply measures can be used to accomplish this.
Conclusion
Shortages of generic drugs—many critical to patient health—are frequent and long lasting. These shortages can cause significant patient harm.
The structure of demand and supply contributes to quality problems and production disruption for generics. Competitive conditions among producers and the market power of buyers discourage investment in capacity and inventories needed to sustain reliable sources of critical drugs, making generic supply quite vulnerable.
Policy intervention, including creating national buffer inventories of priority generics and encouraging large-scale public-private partnerships either to direct demand to reliable low-cost suppliers or to support the creation of reliable nonprofit production, can reduce shortages and help avoid patient harm.
Acknowledgments
The author would like to thank Professor Rena Conti for her help in understanding the structure and operation of the generic drug market; and Emily Gee, Andrea Ducas, and Natasha Murphy for their helpful comments on an earlier draft. The author is solely responsible for the content of this report.