# Patents, Innovation, and Competition in Pharmaceuticals: The Hatch-Waxman Act after 40 Years

**Authors:** C. Scott Hemphill, Bhaven N. Sampat
**Citation:** "Patents, Innovation, and Competition in Pharmaceuticals: The Hatch-Waxman Act after 40 Years," 39 * Journal of Economic Perspectives*, Spring 2025, at 27 (with Bhaven N. Sampat)
**Source:** https://www.aeaweb.org/articles/pdf/doi/10.1257/jep.20241423

*p. 2*
typically capture the great majority of the market very quickly (Buttorff, Xu, and Joyce 2020). Rapid market penetration is promoted by laws in nearly every state, which require or permit a pharmacist to fill a prescription with a generic substitute where available. Private and government payers further encourage substitution by lowering a patient's out-of-pocket cost when a generic product is used. Generic medicines saved $3.1 trillion for drug purchasers over the decade from 2014 through 2023, and $445 billion in 2023 alone (AAM 2024). These provisions also provide a means to handle frequent, contentious, and high-stakes disagreements between the branded and generic firms about whether particular patents are actually valid and infringed, and thereby block generic entry.

*p. 2*
Meanwhile, the "patent term restoration" provisions extend the patent term to allow branded drug makers to recoup part of the time spent during clinical trials and review by the Food and Drug Administration (FDA). Restoration aims to augment the incentive to innovate. This matters because newly patented drugs play an important role in helping people to live longer, healthier lives (Cutler, Deaton, and Lleras-Muney 2006;Buxbaum et al. 2020;Lichtenberg 2022). From early-stage research and development investment to government-mandated clinical trials, bringing a new branded drug to market is very costly. Recent estimates place the cost above $1.5 billion per approved drug (DiMasi, Grabowski, and Hansen 2016;Wouters, McKee, and Luyten 2020)-a figure that accounts for the fact that most drug development efforts are a failure. A half-century of cross-industry economic research furnishes empirical evidence that in the pharmaceutical industry, patents are a uniquely important source of appropriability (Taylor and Silbertson 1973;Mansfield 1986;Levin et al. 1987;Cohen, Nelson, and Walsh 2000;Mezzanoti and Simcoe 2023).

*p. 2*
The effectiveness of both sets of provisions is open to question. Most obviously, high prices for branded drugs are a subject of public and political attention. Observers have raised concerns that patent-holding firms have erected barriers to generic competition, thereby forestalling generic entry and creating static inefficiency. As for dynamic efficiency, the other side of the ledger, partial patent term restoration leaves intact a distortion whereby drugs with longer clinical trials receive shorter periods of exclusivity.

*p. 2*
Measured against its goal of establishing a pathway for generic entry, the Hatch-Waxman Act has been a success. At the same time, setting the right balance between dynamic and static efficiency remains a pressing question. The 40th anniversary of the law presents an opportunity to assess its successes and failures in promoting innovation and competition. In this essay, we begin by reviewing the Act's origins and key features. We then present evidence on how well the law has encouraged competition and rewarded innovation in pharmaceutical markets. On the competition side, we show how the Act creates incentives for branded firms to accumulate patents and generics to challenge them, and various strategies deployed by branded firms to delay generic entry. On the innovation side, we characterize the prevalence and extent of patent restoration. The net result of the Hatch-Waxman compromise is a convoluted and expensive approach to balancing innovation and competition-one that is unlikely to align social and private rewards to innovation. Finally, we consider various avenues for reform, ranging from tweaks of Hatch-Waxman to alternative policy approaches that hold promise in achieving these goals at lower cost.

## Background Background

*p. 3*
Before the Hatch-Waxman Act Before the Hatch-Waxman Act

*p. 3*
In evaluating the evidence that has accumulated in the four decades since the Hatch-Waxman Act was passed, it is useful to reprise how we got here and the problems to which the Act was addressed. To do so, we must harken back to the birth of the modern research-intensive pharmaceutical industry at the end of World War II.

*p. 3*
A government-led medical research effort during that war, enlisting universities and firms across the country, helped support the mass production of penicillin, malaria treatments, a range of vaccines, steroids, blood substitutes, and many other treatments that were crucial for the Allied victory (Gross and Sampat 2023). During the war, most medical research contracts prohibited researchers from obtaining patents covering the fruits of government-funded research (Sampat 2020;Gross and Sampat 2023). Instead, a guaranteed market and wartime subsidies made the effort worthwhile for participating firms (Temin 1980). Thanks to this arrangement, firms developed research capabilities, identified drug innovation as a source of high profits, and built research and development programs (Temin 1979;Temin 1980;Cockburn et al. 1999).

*p. 3*
After World War II, the US government left drug development to firms, confining its funding to basic medical research at universities through the National Institutes of Health (Sampat 2023). In a departure from wartime cooperation and knowledgesharing, "innovator" drug makers increasingly relied on patents as a vehicle to limit competition and provide profits (Temin 1980). In turn, patent protection increased the profitability of other investments such as advertising (Temin 1980). Fueled by new technological opportunities created by the war, and increased appropriability secured by patents, the 1950s marked what is sometimes called the golden era of pharmaceutical innovation.

*p. 3*
With the rise of heavy marketing of patent-protected drugs, some members of the medical community raised concerns about over-prescription of drugs with limited efficacy (Podolsky 2015). In response, Congress passed a 1962 law that became known as the Kefauver Act. The law required evidence of efficacy as a condition for regulatory approval by the Food and Drug Administration, a new responsibility for an agency that had been focused primarily on safety up to that point. Specifically, a requirement for "adequate and well-controlled investigations" was institutionalized in rules requiring clinical trials. Senator Estes Kefauver ( D-Tennessee), the driving force behind the law, would have gone further by prohibiting patents on molecular modifications that did not improve efficacy (Greene and Podolsky 2012). Kefauver also proposed compulsory licensing of all patented drugs after an exclusivity period of just three years, with royalties capped at 8 percent of sales. 1The generic drug industry emerged as the byproduct of an FDA decision to apply the new efficacy rules to pre-Kefauver Act drugs. A retroactive review called the Drug Efficacy Study Implementation applied Kefauver Act efficacy standards to all drugs approved between 1938 and 1962. For drugs deemed efficacious, manufacturers could continue to market under their pre-1962 regulatory approval. Drugs that failed the test could no longer be marketed. The question arose of how to treat firms wishing to sell copies of drugs deemed effective once they came off-patent.

*p. 4*
Requiring a new drug application supported by clinical trials for drugs that were off-patent and already deemed efficacious "seemed both economically wasteful and potentially unethical" (Greene 2014, p. 66). Accordingly, in 1969, the FDA created a shortcut-approval by means of an "abbreviated new drug application" (ANDA), which removed the need for clinical trials. This less expensive approach provided an entry pathway for producers of off-patent drugs, who would come to be known as generic producers.

*p. 4*
Many generic firms "cut their teeth" developing pre-1962 drugs whose patents had expired (Greene 2014). As this sector grew to be a source of low-cost drugs, other policies supported further growth. In the late 1970s, most states passed laws that (as mentioned earlier) encouraged substitution with generics (Song and Barthold 2018).

*p. 4*
By the late 1970s, as patents on post-1962 drugs began to expire, the nascent generic pharmaceutical industry started lobbying to extend the abbreviated new drug application process to these drugs as well (Greene 2014). Absent such a pathway, a generic drug maker was forced to file a new drug application as if the drug had just been invented. The predictable result was that many (reportedly, more than 100) post-1962 drugs had no generic competition even after their patents expired.

*p. 4*
The Kefauver Act also had major effects on innovator drug firms. The introduction of costly clinical trials increased the dependence of such firms on patents: without the promise of some exclusivity, firms were reluctant to undertake that expense. Patenting practice, then as now, was to file a patent application at the time of discovery, with patent issuance several years thereafter. As a result, a substantial part of the patent term was consumed by clinical trials and regulatory review. For example, if a patent was issued in 1970 but trials were not complete until 1975, only 12 years of patent term were left out of the statutory 17. (In 1995, the US patent term was changed from 17 years from the date of issuance to 20 years from the date of application filing. 2 ) Branded firms argued that shorter effective patent protection was responsible for a decline in new drug introductions. Lost innovation was arguably more important than the clinical benefits lost while waiting for government approval, itself the subject of a significant literature (Peltzman 1973).

## Main Features of the Hatch-Waxman Act Main Features of the Hatch-Waxman Act

*p. 5*
The Hatch-Waxman Act embodied a compromise. To promote competition, it created a pathway for accelerating approval of generic versions of post-1962 drugs. The 1984 Act set up the long-sought pathway, still used today, under which a generic drug maker can file an abbreviated new drug application relying on clinical data about the existing branded drug. To obtain FDA approval, a drug maker must demonstrate "bioequivalence," meaning that the generic product uses the same active ingredient and is absorbed by the body at the same rate and to the same extent as the brand-name drug. Bioequivalence is also typically necessary for pharmacist substitution under the state laws discussed above.

*p. 5*
To address the problem of patent term lost to clinical trials and regulatory review, Hatch-Waxman provided for patent extensions. The extension is available for drugs with a novel active ingredient. Firms marketing a drug can choose one patent per drug for extension, and the extension is added back on to the expiration date of the chosen patent. Restoration is partial: all of the regulatory review period, but only one-half of the testing phase. 3 Any time lost to lack of due diligence by the applicant is subtracted. Restoration of patent time is subject to various caps, put in place at the behest of generic firms and others who argued that patent terms were already too long. A patent can be extended a maximum of five years,foot_3 and the resulting term of that patent cannot exceed 14 years from drug approval.

*p. 5*
In addition to these provisions for generic drug entry and patent term restoration, Hatch-Waxman provided for new regulatory exclusivity. For example, "new chemical entities" are protected from generic competition for five years after approval. A new chemical entity is a drug product containing no active ingredient that has previously been approved. This five-year benefit-which Engelberg (1999) suggests was key to the Hatch-Waxman compromise-was meant to guarantee each such drug a minimum amount of exclusivity without regard to patent protection. This lower-bound protection is important where patent protection is short-lived (near expiration by the time of drug approval, despite restoration) or unavailable-for example, because the underlying innovation is already wellknown (Roin 2009).

## Generic Entry Prior to Patent Expiration Generic Entry Prior to Patent Expiration

*p. 5*
One obvious strategy for a generic firm considering entry is to seek approval of its abbreviated new drug application once patent expiration occurs. However, matters are not so simple. The patent might be invalid or easily invented around. Moreover, the branded firm may have multiple patents on the drug, which are applied for and expire at different times, resulting in a lengthy period of exclusivity if the generic firm waits until the last patent expires.

*p. 6*
Branded firms have a large incentive to use these patents (and other methods) to delay generic entry. After all, later entry extends the period of high prices, increasing branded profits while reducing the static welfare of purchasers. To illustrate, suppose a branded drug has $1 billion in annual sales, and generics achieve 80 percent generic penetration in the first year of entry, during which they sell at a 40 percent discount to the brand. 5 Under these and other assumptions, a one-year delay of generic entry represents a transfer of more than $300 million from purchasers to the branded firm (that is, $1 billion × 80% × 40% = $320 million). Even one month of delay can be worth tens of millions of dollars. 6 With these concerns in mind, the 1984 law provides a three-step process governing patenting and generic entry, including a means for generics to enter prior to patent expiration.

*p. 6*
Step One: Patent listing in the Orange Book. As part of its new drug application, a branded drug maker is obliged to report to the Food and Drug Administration certain patents that apply to its approved drug. Most obviously, the drug maker reports patents that claim the drug's active ingredient. In addition, "drug product" patents claiming the formulation or composition and "method" patents claiming a novel use of the product are also listed. The drug maker has a responsibility to submit and describe these patents. The FDA does not evaluate these claims, viewing its role as purely ministerial. The FDA lists these (unaudited) patents in the so-called Orange Book, so named because its first edition was published with an orange cover. The official name is "Approved Drug Products with Therapeutic Equivalence Evaluations," and as the longer title suggests, the document describes the set of generic drugs that are readily substitutable for a branded product. The Orange Book was first published in October 1980; patent listings were added in 1985.

*p. 6*
Step Two: Paragraph IV certification. Often, a generic drug maker believes that a patent listed in the Orange Book should not block generic entry because the 5 These assumptions are conservative. Conrad and Lutter (2019) report a generic-to-brand price ratio of 0.61 with one generic entrant; the ratio is lower with more entrants. Buttorff, Xu, and Joyce (2020) report a typical generic penetration rate of at least 95 percent. The calculation in text further assumes that quantity is unchanged after low-price generic entry occurs. In fact, price elasticity is typically very low. The small consumer response to a price change has multiple sources, including the separation between a physician's decision to prescribe a drug and the decision by the government and private insurers to pay for it. Gatwood et al. (2014), consistent with other studies, reports very small elasticitybetween 0 and 0.157. 6 For hypertension drugs sold between 2000and 2008, Branstetter, Chatterjee, and Higgins (2016) estimate welfare effects of patent challenges, finding consumers gain $42 billion from entry associated with challenges, compared to producer losses of $32.5 billion.

*p. 7*
patent was wrongly issued (invalid), not infringed by the proposed generic product, or unenforceable. Accordingly, the drug maker applies for approval notwithstanding the unexpired brand patent. In this case, the drug maker's application includes a "Paragraph IV" certification, so-named after the relevant clause of the Hatch-Waxman Act, that explains why the patent does not legally cover the generic product. 7 For a new chemical entity, an abbreviated new drug application containing a Paragraph IV certification-a so-called Paragraph IV challenge-can be filed as early as four years after approval. A generic drug maker that uses a Paragraph IV certification must notify the branded drug maker of its submission.

*p. 7*
Step Three: Litigation and automatic stay of FDA approval. A legal battle will often ensue at this stage. Upon notification, the brand company may sue the generic for patent infringement before the generic begins selling its product. If the suit is filed in a timely fashion, FDA approval of the generic firm's abbreviated new drug application is automatically blocked for up to 30 months while the suit is considered by a court. 8 If the generic wins the patent litigation sooner than that, the stay expires early. Note that this delay arises even if the patent is invalid (for example, because the invention is obvious) or not infringed (for example, because the generic drug product uses a different technology), and even if the patent should not have been listed in the Orange Book in the first place (for example, because it does not pertain to the drug in question).

*p. 7*
Generic drug makers face a potential collective action problem in this setting: no firm may be willing to pay the costs of litigating the validity of a brand-name patent if the result of a successful patent invalidation by one firm simply opens the door to free entry by all other generic firms, resulting in a crowded market with low margins. For this reason, the Hatch-Waxman Act created a special incentive for generic firms to bear the expense of a Paragraph IV challenge. The first generic drug maker to file such a challenge is eligible for a 180-day exclusivity period to market the generic drug before other generics may enter the market. In this way, the first filer can capture the lion's share of sales and establish a leading market position. Thus, generic exclusivity is a lucrative "bounty" worth hundreds of millions of dollars for a blockbuster drug. Generic exclusivity can also create a bottleneck for subsequent filers as they wait for the period to expire. Once generic entry occurs, the branded firm often raises its price, earning a higher margin on a minority of brand-loyal customers.

## Hatch-Waxman and Entry: Orange Book Patent Listings, Challenges, Hatch-Waxman and Entry: Orange Book Patent Listings, Challenges, and Generic Approval and Generic Approval

*p. 8*
The pharmaceutical industry has been traditionally classified as a "discrete product" industry, in which one or relatively few patents cover a single product (Levin et al. 1987;Cohen et al. 2000). The classic "discrete product" drug has one strong patent, covering its active ingredient.

*p. 8*
But in the decades since Hatch-Waxman, real-world drug patenting has diverged from this initial pattern. Here we extend the analysis in Hemphill and Sampat (2011), which reported growth in Orange Book patents over time. Our starting point, using data from the Food and Drug Administration, is the set of new molecular entities-drugs containing a novel active ingredient-approved between 1985 and 2020 (FDA 2023a). To gain a complete picture of patenting for these drugs requires linking various editions of the Orange Book. (Each edition of the Orange Book contains a snapshot of unexpired patents.) The NBER Orange Book dataset does this, relying upon digitized archival versions of the Orange Book between 1985 and 2009 and digital versions through 2016 (Durvasula et al. 2023). We update the Orange Book data to include all editions to 2023. The online Appendix provides details.

*p. 8*
The final dataset includes 826 drugs approved between 1985 and 2020, and their 4,446 patents.foot_6 Where the same patent protects more than one drug, it is counted more than once; technically, there are 4,446 drug-patent "pairs" (and 4,217 unique patents).

*p. 8*
Figure 1 shows the growth in patents over time. For the 1985-1987 approval cohort-that is, branded drugs approved between 1985 and 1987-the median number of patents per drug is one. The median increases over time to seven patents per drug by the 2012-2014 cohort. The mean also increases, from 2.4 patents per drug to more than eight in later years. At the 75th percentile, patenting rises from three patents per drug to ten.

## Secondary Patents, Evergreening, and Nominal Patent Term Secondary Patents, Evergreening, and Nominal Patent Term

*p. 8*
To understand the increase, it is important to distinguish two types of patent. A "primary" patent covers the active ingredient of a drug. "Secondary" patents cover ancillary aspects of the drug, such as chemical variants, alternative formulations, and methods of use. The growth in median and mean patents for new drugs is driven by an increase in secondary patenting. Within the 1985-1987 cohort, 54 percent have at least one secondary patent based on patent categorizations from IQVIA's Ark Patent Intelligence (see the online Appendix for details). By the mid-2010s, this is true of nearly all drugs, a trend sustained through the 2015-2017 cohort (the most recent for which we have patent categorization data).

*p. 9*
Applications for these secondary patents are disproportionately filed later: on average the filing date of primary patents is twelve years before brand approval, compared to seven years before brand approval for secondary patents. 10 Ten percent of drugs have at least one patent filed after drug approval: these are almost all secondary patents. Later applications result in later patent expirations, given that patents expire 20 years after application (or, before 1995, 17 years from issue).foot_8 Later-expiring secondary patents can thus lengthen (in the language of Hemphill and Sampat 2011) the "nominal" patent term: the time from FDA approval of a drug to the expiration date of its last expiring Orange Book patent.

*p. 9*
Figure 2 shows the increase in nominal term over time, from about 15 years (at the mean) in the 1985-1987 cohort, to 18 years in the 2000-2002 cohort. While the data from all Orange Book patents shows a slight decline in the most recent cohorts, the second panel (excluding late listed patents) shows a steady increase over time, suggesting the recent dip is due to censoring. 1 9 8 8 -1 9 9 0 1 9 9 1 -1 9 9 3 1 9 9 4 -1 9 9 6 1 9 9 7 -1 9 9 9 2 0 0 0 -2 0 0 2 2 0 0 3 -2 0 0 5 2 0 0 6 -2 0 0 8 2 0 0 9 -2 0 1 1 2 0 1 2 -2 0 1 4 2 0 1 5 -2 0 1 7 2 0 1 8 -2 0 2 0 Brand approval year 1 9 8 5 -1 9 8 7 1 9 8 8 -1 9 9 0 1 9 9 1 -1 9 9 3 1 9 9 4 -1 9 9 6 1 9 9 7 -1 9 9 9 2 0 0 0 -2 0 0 2 2 0 0 3 -2 0 0 5 2 0 0 6 -2 0 0 8 2 0 0 9 -2 0 1 1 2 0 1 2 -2 0 1 4 2 0 1 5 -2 0 1 7 2 0 1 8 -2 0 2 0 Brand approval year This extension of patent term through secondary patents is part of a broader set of entry-delaying efforts called "evergreening" by its critics, and "life-cycle management" by practitioners. Even absent nominal term extension, when both primary and secondary patents are listed on the Orange Book, the result can be a "thicket" of temporally overlapping claims, which a generic drug maker must either address or wait out to market a competing product.

*p. 10*
A further reason that secondary patents have been more controversial than pharmaceutical patents in general, along with potentially extending exclusivity, is that they are of lower average quality from a legal perspective, in the sense that they are less likely to meet traditional standards of patentability (novelty and nonobviousness) and thus less likely to be valid. Such patents are sometimes granted erroneously by a US Patent and Trademark Office operating under resource and information constraints (Lemley and Shapiro 2005;Lemley and Sampat 2012;Frakes and Wasserman 2023). In addition, some of the secondary patents listed on the Orange Book may not be infringed by a proposed generic product. As reported above, some secondary patents are filed even after drug approval, and thus less plausibly relevant to the drug as approved.

*p. 10*
These patterns suggest a concerning scenario in which secondary patents are used to create expense, uncertainty, and delay for generic drug makers, thus 1 9 8 8 -1 9 9 0 1 9 9 1 -1 9 9 3 1 9 9 4 -1 9 9 6 1 9 9 7 -1 9 9 9 2 0 0 0 -2 0 0 2 2 0 0 3 -2 0 0 5 2 0 0 6 -2 0 0 8 2 0 0 9 -2 0 1 1 2 0 1 2 -2 0 1 4 2 0 1 5 -2 0 1 7 2 0 1 8 -2 0 2 0 Brand approval year 1 9 8 5 -1 9 8 7 1 9 8 8 -1 9 9 0 1 9 9 1 -1 9 9 3 1 9 9 4 -1 9 9 6 1 9 9 7 -1 9 9 9 2 0 0 0 -2 0 0 2 2 0 0 3 -2 0 0 5 2 0 0 6 -2 0 0 8 2 0 0 9 -2 0 1 1 2 0 1 2 -2 0 1 4 2 0 1 5 -2 0 1 7 2 0 1 8 -2 0 2 0 Brand approval year deterring some entry that would otherwise be attempted. A branded drug maker has an incentive to apply for and to list as many patents as possible, and thereby force a sequence of Paragraph IV challenge, litigation, and 30-month stay. The resulting litigation, aside from delaying entry, also promotes a dynamic in which the generic firm has an incentive to settle the lawsuit with a promise to accept a later entry date in order to preserve its entitlement to the 180-day bounty (Hemphill 2006). The bounty, often a major source of profit for the generic drug maker, is placed at risk if the generic litigates to judgment instead of settling.

## Patent Challenges and Effective Market Life Patent Challenges and Effective Market Life

*p. 11*
Patent challenges in pharmaceutical markets have grown sharply over time. About 21 percent of drugs in the 1985-1987 approval cohort have one or more patents challenged by generics; this figure rises to 55 percent by 2000-2002. Nearly 80 percent of drugs approved between 2005 and 2010 face challenges, though the share declines in later years.

*p. 11*
These challenges can and increasingly do start early in the life of a drug. There is a spike in challenges at four years after approval-for new chemical entities, the earliest point they are allowed. For most commercially important drugs, drug makers not only know to expect a challenge, but also when to expect it.

*p. 11*
Some have argued that the rewards from generic challenge are too large: that the high benefit from a successful challenge (particularly the profits during the 180-day bounty period) or profitable settlement, and comparatively low cost, leads to prospecting by generics. Voet (2005) comments that generic drug makers "rely on the law of averages-if you place enough bets, you are sure to win a few of them." Several papers also suggest that challenges disproportionately target patents on blockbuster drugs (Grabowksi and Kyle 2007; Higgins and Graham 2009;Grabowski et al. 2017). This is concerning because, based on commonly cited estimates of the returns to research and development spending, blockbusters have an outsized role in helping branded drug makers cover the average cost of research and development (DiMasi, Grabowski, and Hansen 2016).

*p. 11*
Our research on these issues suggests a more nuanced story. While higher-sales drugs indeed are more likely to draw challenges, in part that is because those drugs also have more patents per drug, including a larger number of secondary, late-expiring patents driving a long nominal patent term (Hemphill and Sampat 2011). In previous work examing patents, challenges, and generic entry for drugs with first-time generic entry between 2001 and 2010, we found that even within drugs, secondary patents are disproportionately challenged (Hemphill and Sampat 2012). From this perspective, the Hatch-Waxman regime confers upon generic firms a high-powered incentive to give weak patents on important drugs a strong second look (Engelberg 1999;Bulow 2004).

*p. 11*
Generics are also more likely to be successful when challenging secondary patents. Hemphill and Sampat (2013) show that for patent challenges litigated to completion, branded firms win on primary patents 92 percent of the time, but on secondary patents only 32 percent of the time. While litigation outcomes are notoriously difficult to interpret when settlement is an option, for a smaller subset of cases adjudicated in an era where settlement was less common, the branded firm wins nearly all litigation resulting from challenges to primary patents, and generic challengers typically prevail on secondary patents, consistent with the legal understanding of secondary patents as weaker or lower quality.

*p. 12*
Successful challenges on secondary patents (including settlements with an early entry date) can thus offset evergreening. A drug's effective market life-the time from brand approval to first generic entry-is typically less than its nominal patent term. For drugs with first generic approval (specifically, of an application on the same active ingredients) between 2000 and 2023, the average effective market life (measured by the timing of generic approval) is 13.0 years, compared to a nominal patent term of 17.5 years.

*p. 12*
Figure 3 illustrates how these dynamics play out over the sales distribution for drugs with first generic approval between 2008 and 2023, using sales data from SSR Health (2024). Patent accumulation and nominal patent term increase with sales. Higher sales drugs also draw more challenges and have a shorter effective market life. Challenges to late-expiring secondary patents may not be the whole story. Hemphill and Sampat (2012) present evidence that for more lucrative drugs, patent challenges to primary patents are also more common. The incentive to challenge primary patents may be buttressed by a racing dynamic. If one generic firm would otherwise wait until the primary patent expires, it may be induced to challenge that patent by its expectation that others will do so, lest the firm miss out on exclusivity (Hemphill and Sampat 2012;Grabowski et al. 2017).

*p. 12*
Despite these complex interactions, effective market life has remained fairly stable over time, as shown in Figure 4, generally ranging between 12 and 14 years since 2000, though it dips slightly in the most recent cohort. The general stability of effective market life over time, reported by Hemphill and Sampat (2012), Danzon and Furuakwa (2011), and others, is notable, given the growth of patents per drug, secondary patenting, and nominal term in the four decades since Hatch-Waxman.

*p. 12*
As further evidence on the nature of patents that are most relevant, we examine the share of patents that potentially matter for the timing of generic approval. We classify a patent as "non-binding" for a drug if the first generic approval occurs before the expiration of the patent. For drugs with generic approval between 2000 and 2023, we calculate the share of patents that are non-binding for generic entry. We find that 65 percent of secondary patents are non-binding, compared to 24 percent of primary patents. This result is consistent with our previous work, and that of other scholars, showing the relative weakness of secondary patents.

## Paying for Delay, and Other Games Paying for Delay, and Other Games

*p. 12*
One way for a branded drug maker to hinder generic entry, and thereby reap the benefits of delay discussed above, is to pay the would-be generic entrant to postpone or abandon its efforts to enter the market. This strategy, a form of collusion, exploits the fact that monopoly profits are typically larger than total profits in a competitive duopoly (Gilbert and Newbery 1982). This strategy predates the modern pharmaceutical industry. In the nineteenth century, Bell Telephone filed a patent suit against Western Union, a competitor in telephony. Under their 1879 settlement, Bell paid Western Union many millions of dollars and Western Union agreed to stay out of the telephone business for more than a decade (Farrell andChicu 2018, citing Brock 1981). Branded drug makers use the same playbook, with the branded firm in the position of Bell, and a generic entrant playing the role of Western Union. For the monopolist, payment makes sense as a way to avoid duopoly. As for the entrant, the incentive to compete aggressively is neutralized by the payment.

*p. 13*
In the pharmaceutical context, the payment is made as part of a settlement of Paragraph IV patent litigation. The patentee pays an alleged infringer to abandon its bid for competitive (and allegedly infringing) entry. Paying off one generic drug maker might seem pointless, given the risk that others would enter anyway Note: This figure shows patents per drug, nominal patent term (time from brand approval to last expiring patent for the drug), whether there was a Paragraph IV patent challenge, and effective market life (time from brand approval to first generic approval for the drug's active ingredient(s)), by sales category, for the 292 drugs with first generic approval between 2008 and 2023. Sales categories are based on SSR Health data from the calendar year prior to first generic approval. All sales are adjusted to 2022 dollars. Drugs are grouped into quartiles of non-zero sales, where Q1 represents the lowest quartile (mean sales of $118 million) and Q4 represents the highest quartile (mean sales of $2.9 billion). The 118 drugs without SSR Health data in the year before generic entry are plotted separately, as they may represent either zerosales drugs or those outside SSR Health's sampling frame. or demand payments of their own. But as discussed earlier, "first filers" under the Hatch-Waxman regime are potentially eligible for 180 days of generic exclusivity, before other generics can come in. Moreover, later challengers are impeded due to a regulatory bottleneck that inhibits approval of later filers (Hemphill and Lemley 2011). Thus, later filers have a relatively low incentive to pursue early entry.

## Z e r o / N

*p. 14*
In an ordinary patent settlement, one would expect payment to flow from infringer to patentee, as compensation for expected damages from past infringement. Thus, an observed payment from the patentee to the alleged infringer is a notable anomaly. For this reason, pay-for-delay deals are sometimes called "reverse payment" settlements. An ordinary settlement reflects the strength of the underlying patent-that is, whatever limitation on entry the patentee could secure based on its probability of winning the infringement suit. By contrast, when the patentee instead makes an additional payment to sweeten the deal, the result is less competition than the patentee could expect by asserting the patent alone. The payment, in other words, is for additional generic delay, compared to what is legitimately achieved by the patent alone.

*p. 14*
Accordingly, an observed and otherwise unexplained payment is a basis for inferring that the branded firm is profiting at the expense of consumers (Shapiro 2003;Hemphill 2006;Edlin et al. 2013). This inference has been adopted by the US Supreme Court in an important antitrust case addressing reverse payment

## Median effective

*p. 14*
Median nominal 2 0 0 0 -2 0 0 2 2 0 0 3 -2 0 0 5 2 0 0 6 -2 0 0 8 2 0 0 9 -2 0 1 1 2 0 1 2 -2 0 1 4 2 0 1 5 -2 0 1 7 2 0 1 8 -2 0 2 0 2 0 2 1 -2 0 2 3 settlements (FTC v. Actavis, Inc., 570 US 136 [2013]). In the wake of this decision, lower courts have struggled with what (beyond cash) counts as a payment, such as overpayment for services provided by the generic firm, forgiveness of a debt owed by the generic (such as liability for patent infringement on other drugs), or early generic entry on other drugs or in other jurisdictions (Hemphill 2009). 12 Lower courts have also considered what (if any) role is played by the merits of the underlying patent litigation that gave rise to the settlement. Branded drug makers use other strategies to delay generic entry. One such strategy is to file a large set of "citizen petitions" shortly before generic approval, to which the Food and Drug Administration is legally required to respond, thereby delaying generic approval (Carrier and Minniti 2016). The stability of effective market life as measured by the timing of approval, shown in Figure 4, suggests that the aggregate effects of these tactics may be limited. 13 But for blockbuster drugs, even a few months of additional exclusivity can make them worthwhile strategies for branded drug makers to pursue.

*p. 15*
A further strategy called "product hopping" increases the impact of delayed entry (Carrier and Shadowen 2017). A branded firm can shift patients and doctors from a drug facing imminent generic competition to a new version with stronger or longer-lived exclusivity. If the "product hop" is accomplished before the generic version of the old drug is approved, there is no foothold for generic substitution to take place with a pre-existing base of customers.

## Hatch-Waxman and Innovation: Evaluation of Patent Term Hatch-Waxman and Innovation: Evaluation of Patent Term Restoration Restoration

*p. 15*
The other side of the Hatch-Waxman compromise is patent term restorationthat is, provisions to add back time lost to clinical trials and regulatory review, which drug companies and some academics argue have been eroding patent terms and thus blunting innovation incentives.

*p. 15*
Of the drugs in our dataset, 69 percent (571/826) have a patent extended under these provisions. The average extension is about three years (1076 days at the mean, 1022 at the median). Notably, for 30 percent (169/571) of the extensions, the five-year maximum extension (or two years for certain drugs) is binding. 14 For 12 One common form of compensation involves the brand's ability to compete with the generic entrant by deploying an "authorized generic" marketed under the brand's original drug approval. By agreeing not to launch an authorized generic, the branded firm confers a benefit on the generic (at its own expense), a sacrifice that can be used to secure delayed entry by the generic. 13 One caution is that we are measuring generic entry by approval, not actual generic launch. The strategies discussed above can delay entry past the point of generic approval, at least in some cases. Understanding the extent of this dynamic is an important topic for future research. See the online Appendix for more details on the distinction between approval and launch, and potential empirical approaches to measuring the latter at scale. 14 By our count, 83 extensions were capped at five years. For another 86 "pipeline" drugs (see footnote 4 above for an explanation), the extension was capped at two years.

*p. 16*
another 26 percent, the post-approval cap of 14 years binds. Thus, of the drugs that receive patent extensions at all, only a minority get the full benefit of the restoration formula, a striking demonstration of the incompleteness of term restoration. Even for non-capped drugs, only half of the time lost in clinical trials is recouped (plus all of the time spent in regulatory review of the new drug application). 15In the previous section, we emphasized that new drugs increasingly have numerous patents. In choosing which patent to extend, the drug maker faces a tradeoff (Eisenberg 2012). Secondary patents often expire later but, as discussed above, are more vulnerable to challenge. For drugs approved since 1985 with an extended patent, primary patents account for 73 percent of the extensions, compared to just 28 percent of all patents on these drugs. This provides a "revealed preference" measure of which patents firms view as being most important for their drugs (Hemphill and Sampat 2011). Lietzan and Acri (2020) show that patents with longer trials have shorter terms, even after extension. We update and extend their analysis for our sample of drugs. Figure 5 shows the relationship between clinical trial length and different measures of patent term or market exclusivity for drugs with at least one extended patent and generic entry between 2000 and 2023. There is a strong negative relationship between trial length and original years of patent term, before patent term restoration. The term extension shifts the level of term upwards and reduces its slope (in expected value), but it is still negative. This result is unsurprising given the term extension formula and its built-in caps. Nominal patent term, sustained by all patents on the drug (not just the extended one) is flat: additional patents break the relationship between clinical trial length and patent term. However, as we have argued above, nominal patent term does not tell the full story, because secondary patents are often successfully challenged, especially for important drugs. The relationship for effective patent life, the time to actual generic entry, and clinical trial length is once again negative.

*p. 16*
Bringing the facts together: Hatch-Waxman term extensions restore life lost to clinical trials, but do not fully reverse the negative relationship between clinical trial length and the term of the extended patent. Secondary patents on the same drugs appear to undo the relationship at the drug level. However, these secondary patents are challenged by generic firms through the Paragraph IV process, leaving only the extended (typically primary) patent standing, resulting in a negative relationship between trial length and effective market life. These results are consistent with the conclusion of Lietzan and Acri (2020) that incomplete restoration is likely to distort drug development away from the most difficult challenges, such as Alzheimer's disease and pancreatic cancer.

*p. 17*
Whether the resulting level of effective market life is too high or too low is not obvious, but the distortion is troubling. Expectations about effective market life (including assessment of strength of different patents) shape the choices that firms make about research and development. Shorter exclusivity periods for longer development times may lead firms to reduce investments in long-term research. Consistent with this idea, Budish, Roin, and Williams (2015) show that firms' investments in cancer research are distorted away from drugs that require long clinical trials.

## Proposals for Hatch-Waxman Reform Proposals for Hatch-Waxman Reform

*p. 17*
The 1984 Hatch-Waxman Act uses a combination of patents and regulation in seeking to promote both dynamic and static efficiency in pharmaceuticals. Based on Hemphill andSampat (2025b, c, d), andNARA (2023). See the online Appendix for details. Note: This chart shows binned scatterplots (20 bins each) of nominal patent term (time from brand approval to last expiring patent for the drug), effective market life (time from brand approval to first generic approval for the drug's active ingredient), and original and extended term of the patent with extension, plotted against clinical trial length, for 382 drugs with an extended patent and first generic approval between 2000 and 2023. Dashed lines show linear regression fits through the underlying data. In this section, we take stock and set out some ideas about improvements to the patent listing, challenge, and extension provisions.

## Competition Competition

*p. 18*
By creating a pathway to generic entry, Hatch-Waxman helped to fuel the growth of the generic drug industry. Yet our discussion also suggests problematic manipulation of the regime. Brand-name drug producers have a strong incentive to obtain and list secondary patents, even those of doubtful validity or relevance. The extra patents drive up the nominal patent term and potentially raise barriers to entry for generic firms. That said, the data suggest that patent challenges, incentivized by the 180-day exclusivity bounty for some generic entrants, seem to be effective in targeting and largely negating these patents. Settlements and other gaming interfere with the machinery in important specific cases, but overall the challenge process does appear to ratchet back the nominal term.

*p. 18*
But might other approaches to the same end be possible? After all, our current system of patent accumulation, listing, challenge, and costly litigation both wastes resources and results in a selective lengthening for some drugs but not others, with a final exclusivity period that is unlikely to correspond to social value.

*p. 18*
One avenue for reform is to do a better job in preventing secondary patents from being issued in the first place. These proposals are sometimes framed as responses to the proposition offered by Lemley (2001) that it might be "rationally ignorant" for the Patent Office to screen lightly in general, given that most patents do not matter, relying on litigation to invalidate patents that turn out to be important. Rather than rely on litigation to challenge patents after issuance, there could be greater up-front investment in screening by the Patent Office. Increasing the review time of pharmaceutical patents by the examiners can be highly cost-effective, through reducing litigation costs and speeding generic entry (Frakes and Wasserman 2023).

*p. 18*
Perhaps even more promising, pharmaceuticals offer a rare context in which the knowledge of which patents actually matter for competition is publicly revealed, thanks to the Orange Book listing, which narrows the inquiry (Hemphill and Sampat 2012). Even if it is a practical impossibility to do an intensive review of all of the more than 600,000 patent applications filed at the Patent Office each year, selective review of pharmaceutical patents does seem feasible. In the 2023 edition of the Orange Book, there are only 5,633 unexpired pharmaceutical patents overall, of which only 645 were newly added since 2022.

*p. 18*
To ensure validity of these listings, the Patent Office could provide an intensive second layer of review to pharmaceutical patents at the time of Orange Book listing (Hemphill andSampat 2011, 2022). For drug makers, one benefit of this approach is that it could effectively "gold plate" high-quality patents against further litigation (Lemley, Lichtman, and Sampat 2005), limiting the "prospecting" type challenges that worry some observers (Higgins and Graham 2009).

*p. 18*
Another approach is to alter the process for clearing out invalid patents (Hemphill and Sampat 2022). In 2011, Congress established the Patent Trial and Appeal Board to adjudicate patent validity at a lower cost compared to traditional litigation. However, this process is not well integrated with the Hatch-Waxman framework, limiting the impact of its procedures for pharmaceuticals (Rai et al. 2022). The main issue is that if a generic drug maker goes before the Board and wins, it does not secure the 180-day exclusivity. Changing the Hatch-Waxman Act to incorporate this new mechanism for patent challenges is another important potential reform.

*p. 19*
Validity of patents is one issue. Whether a patent actually covers the drug is another. As noted above, currently the Food and Drug Administration performs no review of whether a listed patent actually pertains to the drug. One novel response was undertaken by the Federal Trade Commission (2023Commission ( , 2024)), which challenged as improper the listing of a large set of patents. Going forward, the regulator should find a way to move beyond a purely ministerial role in administering patent listings (Eisenberg and Crane 2015;Hemphill and Sampat 2022;Contreras and Rai 2023). As Contreras and Rai (2023) note, private patent pools scrutinize each patent for essentiality to the standard, suggesting the feasibility of a less searching scrutiny of each Orange Book patent for compliance with regulatory requirements. At a minimum, the agency should require more precise specification of what exactly the listed patents cover and deny listing of those that do not apply to the approved product (Hemphill and Sampat 2023).

*p. 19*
A final avenue for reform would directly alter the branded firm's cost-benefit analysis in erroneously asserting a patent that is invalid or not infringed by the proposed generic product. Currently, a branded drug maker experiences no adverse consequences when it blocks generic entry by means of the automatic stay, using a patent that is later shown to be invalid or not infringed by the generic drug product. Imposing a financial obligation on the branded firm in these circumstances would cause it to internalize some of the social costs imposed by its erroneous assertion. 16The primary objective of current proposals to restrict secondary patents is to reduce drug expenditures. How big an impact might these kinds of changes have? To examine this, we focused on the 399 drugs from our dataset approved between 1985 and 2019 that had at least one unexpired patent listed in the 2019 Orange Book. Overall, these drugs accounted for $134 billion in sales in 2019, among drugs in SSR Health's drug sales database (SSR Health 2024). Of these drugs, 6 percent had a primary patent only in 2019, 40 percent a secondary patent only, and 54 percent had both types. The finding that nearly all drugs from our sample that are on-patent in 2019 have a secondary patent is consistent with our data above on the growing prevalence of secondary patenting, and consistent with the idea that secondary patents are protecting economically valuable drugs.

*p. 20*
But the majority of these on-patent drugs-60 percent-are protected in whole or part by primary patents. These drugs are associated with an even larger fraction of SSR Health-recorded sales: 76 percent. Through eliminating or curtailing secondary patents could meaningfully reduce drug expenditures, the incremental value of such proposals is limited by the continued relevance of strong primary patent protection. To look at this another way, patent challenges have already had an impact that these figures do not capture, because many important drugs have already gone generic by 2019 thanks to challenges on secondary patents.

## Innovation Innovation

*p. 20*
While curtailing secondary patents (through challenges or additional reforms like those above) can hasten generic entry, doing so comes at the cost of the exclusivity period enjoyed by the innovator. Given the heavy reliance of drug companies on patents for recouping their substantial research and development investments, this raises a concern about innovation incentives. For example, Higgins and Graham (2009) argue that patent challenges might be "tipping the balance" of Hatch-Waxman towards generics and away from innovators, sacrificing dynamic gains for static benefits. The strength of this argument depends on whether optimal patent term is closer to nominal patent term, including that sustained by secondary patents, or patent term sustained by primary patents alone. We do not know the magic number for an optimal fixed patent term. But whatever it may be, we are skeptical that the right way to get there is through accumulation of patents of questionable relevance or validity.

*p. 20*
We can say with more certainty that the term restoration provisions of the act need rethinking. The partial restoration of time lost in trials, including the five-year cap, means that drugs with longer development time get less time on the extended patent, which we have shown translates to a shorter period of effective market exclusivity at the drug level. This penalizes research with long development times. If the goal is uniformity, then complete restoration, or simply beginning patent term at approval, would be more appropriate. A version of this approach that may be palatable to branded and generic firms alike would be a one-and-done approach: allowing firms to list one patent, with time lost to trials completely restored, but only one.

*p. 20*
However, uniform exclusivity for all drugs may not be the optimal policy (Budish, Roin, and Williams 2015). An alternative is to index the reward of exclusivity to the social benefits of research, or to time-to-market (Roin 2013). International agreements around patent harmonization limit the feasibility of implementing nonuniform patent terms in general. But in pharmaceuticals, one can imagine how the Food and Drug Administration might administer new exclusivity modeled on the Hatch-Waxman approach for extending patent terms. This approach might serve as an additional innovation policy lever (on top of or even instead of patents), which might be more proactively deployed to shape both the rate and direction of pharmaceutical innovation in a more flexible and nuanced way than possible through patent policy alone (Eisenberg 2012). A system based mainly on nonpatent exclusivities may also avoid the costs of the Hatch-Waxman Act's approach to balancing dynamic and static efficiency-which relies on the complex interplay of patent restoration, listing, and litigation to determine effective exclusivity termspotentially reducing uncertainty for branded firms and generics alike.

## Conclusion Conclusion

*p. 21*
In the 40 years since Hatch-Waxman, a pattern has taken hold in pharmaceutical markets. Branded drug makers accumulate patents on drugs and list them on the Orange Book, including a primary patent on the active ingredient (typically the one that is chosen for extension) and additional peripheral secondary patents. Patent challenges can commence four years after the approval of a new chemical entity-or sooner for other drugs-triggering litigation and sometimes settlement. Patent accumulation and speed and intensity of challenges are greater for blockbuster drugs. At the end of the process, much of any extra nominal term generated beyond the primary patent is ratcheted back. Primary patents sustain most of the effective market life of drugs, which has generally oscillated between 12 and 14 years since Hatch-Waxman.

*p. 21*
The right amount of exclusivity to balance innovation and access was unknown in 1984 when Hatch-Waxman was enacted, and in our view it remains unclear today. Various idiosyncratic features of the process we have described, including of the term restoration process, might be exploited to collect better evidence on the causal impact of marginal changes to patent term on innovation and competition. At the same time, the ritualistic aspects of the "listing, challenge, litigation" process caution against using patent challenges or their outcomes as exogenous shocks to patents or patent term: these steps are largely anticipated by all parties involved.

*p. 21*
The administrative data created by Hatch-Waxman are a promising source for economic research. Orange Book patent listings provide information linking patents to their products-which is not systematically possible in other industriespotentially a boon for empirical research on innovation (Durvasula et al. 2023). The major caveat is that all Orange Book patents are not created equal, and for many research questions primary and secondary patents should be considered separately. Relatedly, both researchers and policymakers describing the growth of patenting and patent terms need to distinguish between nominal patent term and effective market life: just because a patent is listed on the Orange Book does not mean it is likely to impede generic entry.

*p. 21*
Throughout this essay, we have focused on drugs that are new molecular entities, the canonical Hatch-Waxman drugs around which the regime was designed. Some of the dynamics may be different for drugs that are line extensions, not new molecules but modifications to earlier drugs. The challenge and generic entry process may also be different for drugs that are more difficult to imitate, or where FDA regulations for entry are more cumbersome (on the breakdown of the Hatch-Waxman system for drug-device combinations, see Feldman et al. 2022;Reddy et al. 2023).

*p. 21*
Moreover, a large share of US drug expenditures is now for biologic drugs, as opposed to the "small molecule" drugs created by chemical synthesis that are governed by Hatch-Waxman. Biologic drugs are not subject to Hatch-Waxman's patent listing and challenge regime, nor its mechanism for generic entry. But even here, there are some similarities and some potential lessons from Hatch-Waxman. For biologics, as for other drugs, a major policy challenge remains implementing a system of listing patents that provides clarity to potential entrants, while at the same time ensuring that invalid or improperly listed patents do not impede generic entry.

*p. 22*
Much of our discussion has considered variations on the Hatch-Waxman approach, which presumes that limited patent term exclusivity is the best way to incentivize innovation, and generic competition the best way to promote lower prices and access. But this dichotomy may blind us to possibilities for more fundamental policy change. In the 40 years leading up to Hatch-Waxman, policymakers employed mechanisms for promoting innovation without sacrificing access, including direct government funding of applied research and active use of government procurement as an innovation incentive (important in World War II, but largely absent afterwards until the Covid-19 vaccine development effort). Prize-based mechanisms and formal advance market commitments are other options. There are also approaches beyond Hatch-Waxman to implement the tradeoff between innovation and access, such as compulsory licensing or conditioning legal protection on therapeutic efficacy (two features, abandoned along the way, of the original legislation that resulted in the 1962 Kefauver Act). The 40-year milestone also presents an opportunity to revisit these and other policy levers-beyond patent extensions and generic entry-to promote the dual goals of dynamic and static efficiency in pharmaceuticals. ■ Sampat's research was supported by a grant from the National Institute of Healthcare Management for a project titled "Can Improving Pharmaceutical Patent Quality Promote Competition and Reduce Drug Prices?" We thank Daniel Francis, Margaret Kyle, Lisa Larrimore Ouellette, Rachel Sachs, Michal Shur-Ofry, and workshop participants at the Brookings Institution, New York University, and the University of Pennsylvania for helpful comments, and Erika Lietzan for suggestions and data related to the patent term extension analyses.

## Footnotes

> Congressional Record. 1961. 87th Congress, 1st Session, Vol. 107, Part 5 (April 12): 5369.

> The change was made to comply with international obligations under the Trade Related Intellectual Property Rights (TRIPS) Agreement.

> In cases where the patent issues after the regulatory review period begins, only the portion of that period after issue is counted, and similarly for the testing phase. Patents issued after a drug's approval cannot be extended.

> "Pipeline" drugs-drugs that were in trials but not yet approved at the passage of Hatch-Waxman-had a cap of two years.

> Another option, in the case of patent claims on a method of use of the drug, is to avoid the patent by means of a "section viii" statement, which asserts that the claims do not cover any use for which the generic seeks approval.

> If the abbreviated new drug application is filed less than five years after approval of a new chemical entity (recall that this can occur as soon as four years after approval), then the expiration date of the stay is extended. Under the extension, the stay expires 7.5 years (five years plus 30 months) after the date of approval of the brand-name drug.

> In addition to patents, nearly all of these drugs (802/826) receive regulatory protection as new chemical entities.

> We use "effective" filing date (also called the priority date) for this calculation, based on data from the Google Patents Public Dataset (2023).

> Not all late applications result in late expirations. For a discussion of the complexities, seeLietzan and Acri (2020).

> These calculations are based on a compilation of patent extensions maintained by the Patent Office. While this compilation is generally accurate, it does not fully reflect certain changes made to the ( pre-extension) patent expiration date after an initial extension decision. Earlier editions of the compilation attempt to make relevant adjustments. In the online Appendix, we show that the results are robust to using this alternative data source.

> Other jurisdictions provide instructive analogies. For example, Australian law provides for the branded firm to make compensatory payments to the state and to competitors if it secures a temporary legal block of entry while lacking a reasonable basis for doing so (Therapeutic Goods Act of 1989, § 26D). Similarly, a 2021 judicial opinion of Israel's Supreme Court prescribed disgorgement, payable to the generic firm, of improperly acquired profits secured due to an improper delay of generic entry (Unipharm v.Sanofi, CivA 2167/16 [2021]).