Economic Damages

Evergreening and the transfer of patent value

The ongoing patent infringement suits between Apple and Samsung highlight the increasing importance of intellectual property rights to maintaining a competitive edge. In 2012, over 5,000 patent actions were filed before US Courts, an increase of over 29 per cent since 2011, representing compound annual growth of over 6 per cent since 1991.¹

Between 2007 and 2012, the median damages award in patent infringement disputes in the US was just under US$5 million, and in 2012 three damages awards exceeded US$1 billion.² With such large sums at issue, attention has focused on the appropriate methods of assessing the value of patents and the effect of their alleged infringement.

In recent years, there has been increased interest in the role of intellectual property rights in Asia. Initially, at least from a public perspective, this was perhaps limited to counterfeit consumer goods and internet piracy. The growth of Asian disposable incomes, however, has resulted in growing demand not only for branded goods and electronic media but also for improved services, such as health care, that depend on patented and branded products and processes.

Broadly, a patent acts to protect an innovation while a brand is designed to communicate differentiation between products. Patents and brands can act independently of one another or in combination to grow or retain a market position and, as such, both are considered assets.

Two of the most commonly used approaches for assessing the value of assets are the market-based approach and the income-based approach, typically expressed in the form of a discounted cash flow (DCF) model.³ A market-based approach estimates the value of an asset by reference to the prices at which similar or identical assets have been transacted in markets for which transparent price data is available.

Where reliable data is available about transactions involving the asset being valued, this generally provides the best evidence of value assuming that the transactions occurred close to the date of valuation and that expectations and market conditions have not changed materially between the date of transaction and the date of valuation.

Where there is no recent and reliable data on transactions involving the subject asset, a market-based approach requires the existence of sufficiently similar assets and the availability of information on the prices (and other terms) at which these similar assets were traded. A market-based approach is particularly difficult to implement in the context of a patent valuation. The prices at which patents are traded are rarely publically disclosed (unless the patent owner is willing to disclose the terms on which it licenses the innovation) and, by definition, no two patents are identical. However, a market-based approach is often useful in the valuation of other forms of intellectual property, such as trademarks, which are commonly licensed by the owner to third parties.

In addition, the value of patents covering functionally comparable innovations may vary considerably, for example, with the remaining life of the patent and its perceived legal enforceability. In relation to the latter, patents are rarely perfectly enforceable and there is often considerable uncertainty as to whether a patent (newly issued or otherwise) can be enforced. Therefore, the value of a patent typically increases markedly once it has been held to be valid in court proceedings and this uncertainty is reduced.⁴ Similarly, the value of a court-tested patent would be higher than the value of a patent covering a functionally similar innovation that has not been tested in court.

These complications make it unlikely that a valuer will be able to find reliable pricing information for relevant, closely comparable patents. A market-based approach may therefore be applicable only infrequently in patent valuation.

The primary alternative to a market-based valuation is an income-based valuation that seeks to identify the value of the incremental cash flows associated with a patent, for instance, by using a DCF model. A DCF model estimates the value of an asset by reference to the cash flows the asset is expected to generate from the date of valuation. A DCF model generally assumes that the patent holder has already exploited the patent to bring a commercially viable product to market and that the cash flows generated can be forecast to a tolerable degree of accuracy.

A third valuation approach, the 'option price' approach, assesses the value of a patent on the basis that its payoffs are similar to those of an option contract (in particular, a call option, which bestows the right, but not the obligation to purchase an asset at a future date at a known price). The option holder will choose to exercise the option if, and only if, the value of the underlying asset exceeds the strike price.

An option-based approach can be used to assess the value of a patent prior to the introduction of a commercial product. A patent grants its holder the exclusive right to develop and market a product or utilise a cost-saving process. The patent holder will choose to exploit the patent if, and only if, the expected cash flows from exploitation (the value of the underlying asset) exceed the costs of development (the option strike price). If this is not the case, the patent holder can delay development, at no incremental cost, until market conditions change and development is profitable (up until the expiry of the patent).⁵

Although distinct from the DCF method, the option price approach relies upon aspects of it. Indeed, in the context of patent valuation, it is often necessary to use a DCF model to estimate the value of cash flows that might be generated if the patent holder chooses to develop a product from the underlying patent.

When using a DCF method to value a patent, it is common to project the cash flows attributable to the patent up to the expiry of the patent. This approach has intuitive appeal; no commercial entity is likely to pay for a patent that has expired. However, there is a body of empirical evidence to suggest that a patent holder is able to derive value from the underlying innovation well after expiry of the original patent (and the loss of exclusivity).⁶

This chapter seeks to identify the circumstances in which a patent holder might be able to extract super-normal profits from the underlying innovation following the expiry of the patent, and assess whether it is appropriate to assign the value of those incremental profits to the original patent.

DCF valuations of patents

A patent holder can typically extract value from a patent in two ways, namely:

• it can choose to exploit the patent themselves; or
• it can choose to license the right to use the patent on exclusive or non-exclusive terms.

How the patent holder chooses to extract value from the patent will depend upon, inter alia, the nature of the innovation, the patent holder's appetite for risk, and whether the patent holder has sufficient commercial and technical expertise to exploit the patent successfully.

In theory, the value of the patent itself is independent from how the patent holder chooses to extract that value. Rather, the method of value extraction determines how the total value is allocated between the patent holder and the potential licensee. In one extreme, the patent holder might choose to bear all the risk associated with the development and launch of a viable product and therefore retains all of the value of the patent. Alternatively, the patent holder might choose to share the risks of commercial exploitation with the licensee, but also has to share the value of the patent.⁷

The value of any asset is a function of the amount, timing and risks of the cash flows that it is expected to generate. When valuing a stream of future cash flows, it is appropriate to discount the amount of future expected cash flows to a present value using an appropriate discount rate that reflects both the time value of money and the risks inherent in the projected cash flows. Ignoring timing and risk for the time being, the amount of the income stream is principally a function of the ease with which the patent can be exploited to develop a commercially viable product and the size of the potential market for that product.

In principle, DCF analysis can be applied to patents covering both products and processes. The future cash flows generated by a cost-saving process patent are equal to the net reduction in costs brought about by the innovation, less the costs to implement the process. For a product patent, the future cash flows are the incremental profits generated though the sale of products covered by the patent.

In the interest of simplicity, this chapter assumes that a patent can be associated with a clearly defined income stream, whether reduced costs or increased revenues. In reality, many products and processes use a suite of patents, and it may be difficult to clearly delineate the incremental value associated with any individual patent. For example, Gillette filed 22 patents in relation to its Sensor razor that were designed to be interlocking, making it impossible to duplicate the product and difficult to assign value to any specific patent.⁸

A further factor that will heavily influence the income generated from a patent is the extent to which the patent genuinely limits competitive entry into a market. Many patents generate only incremental improvements to an existing product, or can be bypassed using non-patented workarounds, severely restricting their value.

In cases where a patent effectively excludes competition from a market, the patent holder is commonly assumed to be able to exploit this monopoly position to earn super-normal profits for the patent's duration. Upon expiry of the patent, it is expected that competition will exploit the newly out-of-patent technology to enter the market and erode the incumbent's profit margins and market share. For this reason, when valuing a patent using DCF analysis, a valuator seldom forecasts cash flows beyond the expiry of the original patent.

However, there is an increasing body of literature to suggest that, in some circumstances, a patent holder might be able to enjoy monopoly profits for a period after the expiration of the patent.

Extensions of exclusivity

A patent holder can continue to earn monopoly profits for as long as it is able to prevent entry into a market, while maintaining prices above the level consistent with a competitive market.⁹ The prospect of continued monopoly profits encourages a patent holder to seek to extend the period of exclusivity beyond the expiry of the patent by deterring or preventing generic competition from entering the market.

Of the many strategies employed by patent holders to extend their exclusivity, one of the most commonly recognised is 'evergreening'. Patent evergreening refers to the practice of exploiting intellectual property law to extend the period of exclusivity. While examples of evergreening can be found across a range of industries, the practice is most commonly associated with the pharmaceuticals industry. Three major evergreening strategies employed in the pharmaceutical industry are:¹⁰

• stockpiling - the practice of obtaining multiple, separate patents covering different aspects of the same product. By staggering the expiry of these patents, the patent holder is able to extend the period of market exclusivity;
• line extensions - the efforts by pharmaceutical companies to gain additional periods of exclusivity by obtaining patents on modifications to the original drug, or its method of application; and
• franchise extension - the introduction of a new (or modified) and heavily marketed alternative to the original drug whose patent is about to expire. The intention under this strategy is to minimise the loss of market share and dissuade generic entry by switching customers away from the out-of-patent original.

Unsurprisingly, these strategies are contentious and often the subject of litigation. Empirical evidence shows that patents which do not cover a drug's active ingredient and serve primarily to extend the period of market exclusivity are more likely to be the subject of legal challenges.¹¹

These legal challenges have undermined the effectiveness of evergreening strategies. In the US market, drugs for which the first generic entry occurred between 2001 and 2010 enjoyed an average nominal patent term, defined as the time between brand approval and the expiry of the latest expiring patent, of about 16 years. However, as the result of challenges to the evergreening patents, the period of market exclusivity enjoyed by these drugs was only about 12 years.¹²

Advertising expenditure and brand building can help reinforce the evergreening strategies discussed above. In particular, a strong, well-recognised brand can assist pharmaceutical companies in the successful introduction of brand or franchise extensions, extending the market exclusivity period enjoyed by the underlying innovation.

In recent years, concerns over the continued ability of pharmaceutical companies to develop new 'blockbuster' drugs, coupled with increasing development costs, has focused attention on the role of branding in the pharmaceutical industry in extending market exclusivity and deterring the entry of generics.

Strong brands enable firms to differentiate their products from generic competitors, strengthening their competitive position. This is particularly true in countries with decentralised health-care systems, where drug purchasing and treatment decisions are made by individual doctors or patients.

The benefits of developing a strong brand presence to augment franchise extension and diversification strategies are perhaps best illustrated by the AstraZeneca's marketing of Prilosec. AstraZeneca introduced Prilosec, one of a new tranche of drugs known as proton pump inhibitors, in 1989. By 2001, sales of the drug had reached a peak of US$2.6 billion per annum. As early as 1995, AstraZeneca recognised the threat that generic competition might pose to its blockbuster drug and set about planning a mitigation strategy.¹³

Between 1997 and 2000, AstraZeneca spent between US$50 million and US$100 million per annum marketing Prilosec as the 'Purple Pill' solution for heartburn. In March 2001, AstraZeneca launched Nexium, a second generation proton pump inhibitor. Over the next six years, AstraZeneca spent up to US$240 million per year positioning Nexium as an improvement on, and natural successor to, Prilosec and 'today's purple pill'. Nexium sold for about US$5 per pill and was targeted at high-income customers.¹⁴

In November 2002, following a series of legal battles between AstraZeneca and generic manufacturers, the first generic substitutes for Prilosec entered the market at a discount to the pre-patent-expiry Prilosec price. AstraZeneca responded by introducing its own generic version of Prilosec for over-the-counter sales. The generic version retained the purple colour and brand identity of the original prescription drug, which AstraZeneca sold for US$0.71 per pill, significantly undercutting the generic competition.¹⁵

This strategy allowed AstraZeneca to capture the premium end of the market with Nexium while the generic version of Prilosec secured much of the low income, generic market. It is clear that the success of this strategy rested, in part, upon the success of the original Prilosec branding exercise. Notably, the opportunities to introduce extensions to a strong commercial brand are valuable, even if the firm has not yet decided to pursue such a strategy. In other words, the additional options provided by the brand are valuable in themselves.¹⁶

Even if they are eventually invalidated by the courts, evergreening strategies may oblige a potential generic entrant to engage in costly litigation proceedings before it is able to enter the market. The attractiveness of engaging in, and initiating challenges to, evergreening strategies depends, in part, upon the regulatory regime and legal framework within which the patents are contested.

In the US, the Hatch-Waxman Act grants a 180-day period of exclusivity to the first generic manufacturer to challenge a patent successfully. During this period, only the successful challenger and the original patent holder can market an authorised generic product. This encourages generic manufacturers to file patent challenges and to do so as quickly as possible after the patent is issued. In Europe, where no such provision exists, the incentives to challenge patents are accordingly reduced.¹⁷

Even if the patent holder does not actively engage in evergreening and the patent is not otherwise extended, there is often a lag between the expiry of the patent and the entry of generic competition. Regulatory frameworks play a significant role in determining the speed and ease with which a generic competitor can enter the market following the expiry of a patent.

In the US, the Hatch-Waxman Act lowered the barriers to generic entry by allowing potential generic entrants to demonstrate bioequivalence¹⁸ to the original drug, rather than having to replicate the costly product testing undertaken by the original patent holder. To accelerate generic entry, the legislation granted generic manufacturers the right to conduct the necessary bioequivalence tests and trials prior to the expiry of the patent, thus reducing the lag between the expiry of the patent and the entry of a generic competitor.¹⁹

The situation in Europe varies between countries. In general, bioequivalence tests can only be conducted once the patent holder's data relating to pre-clinical tests and clinical trials is no longer protected. In cases where the patent was filed long before market authorisation for the product was obtained, this data protection period can extend beyond the life of the patent. Therefore, under the European system, exclusivity can stem from either the patent protection or data protection.²⁰ In practice, instances where exclusivity is solely reliant upon data protection are relatively rare. Nevertheless, the existence of data protection rights can serve to frustrate the rapid entry of generic competition following the expiry of a patent.

The speed of generic entry is also influenced by the size and structure of national markets. Empirical research has shown that the probability of generic entry is positively correlated with the revenue earned from the patented drug, and the delay between patent expiry and generic entry is negatively correlated with revenue earned.²¹ Therefore, high revenue drugs are more likely to see generic entry, entry is likely to occur more rapidly following the expiry of the patent and entry first occurs in the largest national markets.

It is apparent that a combination of regulatory policy, market conditions, and the actions of patent holders can extend the period of market exclusivity beyond the expiry of the patent. As illustrated in table 1, these factors lead to significant international divergence in both the probability of entry and length of exclusivity across four national markets.

Table 1: Characteristics of generic entry

Source: P Danzon and M Furukawa, ‘Cross-national evidence on generic pharmaceuticals: pharmacy vs. physician-driven markets’ (2011). NBER working paper 17226.

The data in table 1 shows that generic entry is by no means universal. Even in the United States, only about one-third of drugs experienced generic entry. However, the proportion of molecules experiencing generic entry increased over time in each of the four markets shown. This perhaps reflects the increased regulatory tolerance of generic entry and the increasing rewards available to generic manufacturers.

Despite the increased prevalence of generic entry, the length of the exclusivity period enjoyed by patent holders has not declined uniformly across the four markets. Between 1998-2001 and 2006-2009, the UK and Germany saw declines in the average length of the exclusivity period, possibly reflecting changes in the European regulatory regime over that period as well as specific national policies to promote generic entry.

In contrast, in the US, the average length of the market exclusivity period remained relatively stable, increasing slightly from 11.2 years to 11.3 years. This is consistent with the findings of Hemphill and Sampat that legal challenges to the increased use of evergreening strategies have ensured that patent holders have been unable to significantly increase the period of market exclusivity enjoyed by their products.²²

Outside of Japan, there is relatively little data about the length of market exclusivity periods enjoyed by patented products in Asia. Foreign pharmaceutical firms have typically been reluctant to introduce their patented products to less developed markets at prices affordable in those markets. This is in part because, in order to make the products affordable to consumers in less developed markets, patent holders would have to sell the drugs at a significant discount to the prices charged in more developed markets. Patent holders fear that such sales would undermine the monopoly prices they are able to charge in 'developed' markets, and perhaps encourage unauthorised parallel imports.

As Asian markets develop, it might be expected that they will exhibit stronger protections for intellectual property, and an increased ability and willingness to pay prices comparable to those prevailing in the US or Europe. Singapore's emergence as a regional hub for intellectual property transactions and management is perhaps indicative that this process is already taking place within Southeast Asia. Such developments might increase the willingness of pharmaceutical companies to introduce their patented products to these markets.

In the meantime, certain developing economies have sought to use compulsory licences to overcome this reluctance. Under a compulsory licence, the local government authorises a company to exploit a particular patent without the consent of the patent holder in return for a royalty set by the government. This process may be conducted with or without input from the patent holder. The circumstances in which a government may grant a compulsory licence without breaching international agreements were expanded significantly under the Agreement on Trade-Related Aspects of Intellectual Property Rights and the Doha Declaration.²³

To date, political pressure, fear of retaliatory acts by the governments of developed countries, and the relative lack of manufacturing sophistication in developing counties have limited the instances of compulsory licensing. Between 1995 and 2011, there were 24 instances across 17 countries where compulsory licences were either publically contemplated by the local government or granted. The number of compulsory licences granted peaked in the 2003 to 2005 period and has declined markedly since 2006.²⁴

The majority of compulsory licences have concerned treatments for HIV/AIDS.²⁵ For example, in 2006 Thailand granted a compulsory licence permitting Thailand's Government Pharmaceutical Organisation to import a generic version of the antiretroviral drug Efavirenz from India, even though the drug was still under patent in Thailand. In return, Merck, the patent holder, was granted a royalty of 0.5 per cent of the total sales value of the medication imported or produced by Thailand.²⁶

It is clear that, where they have been granted, compulsory licences undercut the ability of the patent holder to extract monopoly level profits from a market before and after patent expiry.

Implications for patent valuation

Although in some markets the period of exclusivity following expiry of the patent is seen to have declined in recent years, it is apparent even for those products that are challenged with generic entry, patent holders often enjoy a monopolistic position after the expiry of a patent. Hudson (2000) found that, from 1984 to 1996, this period averaged about 2.6 years in the US and UK compared witjh 3.4 years in Germany.

This has led some commentators to suggest that when valuing a patent one should also consider the 'post-expiry patent value' (PEPV). In DCF analysis, this would involve projecting cash flows beyond the expiry of the patent and assigning the value of these cash flows to the patent. Hudson (2000) has suggested weighting these cash flows in proportion to the perceived likelihood of generic entry following the expiry of the patent. The PEPV is thus a function of:²⁷

• the probability of generic entry;
• the expected lag between patent expiry and the entry of generics;
• the expected profits if no generic entry occurs; and
• the expected profits following generic entry.

These factors in turn depend upon, for example, the geographic market covered by the patent, the demand for the patented product, and the expected behaviour of the patent holder. Such characteristics are, collectively, unique to each patented product and may vary considerably even across different products in the same geography and industry. It is therefore not possible to assess the PEPV without carefully considering how these factors relate to the specific product itself. In particular, it would not be appropriate to simply assume the existence of a PEPV without assessing the role played by each of these factors.

However, while Hudson's analysis elucidates the factors influencing the profits earned by the patent holder following the expiry of a patent, it is not clear that the value of these cash flows, where they exist, should be assigned to the patent itself.

Assigning the post-expiry cash flows to the patent implies that a commercial entity would be willing to pay a positive amount for an expired patent. This is not an intuitively appealing assumption. It can be safely assumed that no commercially rational entity will pay for a patent that grants it no enforceable intellectual property rights. On this basis, post-expiry cash flows should not be included in a patent valuation.

If the value of post-expiry cash flows is not embodied within the value of the patent, where might it lie? One possibility is that the value of post-expiry cash flows is attributable to the brand name under which the patented product is distributed. Thus, while a company may no longer be willing to pay for the original patent covering ibuprofen, it may be willing to pay for the Motrin, Advil or Nurofen brands under which the drug is sold. This view is consistent with the observation that pre-expiry advertising is one of the strategies used by firms to maintain their sales following generic entry.

Conclusions

Firms engage in a variety of strategies to extend the period over which their patented products enjoy market exclusivity. The success of these strategies depends, in part, upon the regulatory and legal framework with which intellectual property rights are enforced. However, empirical evidence suggests that, across jurisdictions, patent expiry is followed by a delay before generic entry occurs.

This delay allows the incumbent to continue to enjoy the benefits of a greater market share or higher prices than would be available in a competitive market, even after the expiry of the underlying patent. We have suggested that, rather than attributing the value of these post-expiry cash flows to the patent, they might be attributed to the brand covering the patented product.

The effective period of exclusivity enjoyed by patent owners can be extended considerably through the adroit use of branding and product differentiation. Empirical evidence has found a strong correlation between advertising expenditure and the firm's decision to launch a franchise extension.²⁸

Branding is playing an increasingly important role in pharmaceutical firms' business models. Between 1996 and 2005, the total advertising expenditure by pharmaceutical companies in the US increased by 160 per cent to US$29.9 billion per annum. The most rapid growth was seen in direct-to-consumer advertising which increased by 330 per cent over the period.²⁹ From 1990 to 2005, the returns earned by pharmaceutical companies on advertising expenditure were three times greater than the returns on R&D.³⁰

Branding strategies, and the returns they produce, are predicated upon the enforceability of the underlying, original patent. The patent grants the holder a period of exclusivity during which it can develop its market position and build a sustainable brand. In circumstances where the patent holder is denied this exclusivity period, it may be unable to successfully develop a brand around the patented product.

In those circumstances, it may be denied both the post-expiry cash flows attributed to the brand, and the opportunities for franchise extensions provided by a strong brand presence. Thus, when considering the damages arising from patent infringements, it might be appropriate to consider both the cash flows attributable to the patent itself and the extent to which the infringement has denied the patent holder the commercial opportunities associated with a strong brand.

Notes

1. PwC 2013 Patent Litigation Study.
2. PwC 2013 Patent Litigation Study.
3. Other valuation approaches, such as the cost-based approach, are not discussed in this article.
4. Teece, D, Sherry E F, Royalties, evolving patent rights, and the value of innovation. Res Policy 33(2):179-191.
5. Strictly speaking, holding a dormant patent is not free. In many jurisdictions maintenance, or renewal fees, must be paid to ensure that the patent remains in force. However, these costs are payable regardless of whether the patent is exploited or not and should therefore not factor in to the decision to exploit a patent, or leave it dormant.
6. See for example: Hudson, J (2000), 'Generic Take-up in the Pharmaceutical Market Following Patent Expiry: a Multi-country Study,' International Review of Law and Economics 20, 205-221.
7. I recognise that there may be instances where synergies between the operations of two entities may allow them to extract more value from a patent than either party could in isolation. A discussion of the different approaches to extracting value from a patent and their relative merits is beyond the scope of this article.
8. Rivette, K G, Kline, D, 2000a. Discovering New Value in Intellectual Property. Harvard Business Review, 78 (1), 54-67.
9. Assuming it does not incur significant costs in order to prevent entry (eg, marketing expenditure).
10. Dwivedi, G, Hallihosur, S, Rangan, L Evergreening: A deceptive device in patent rights (2010) Technology in Society, 32 (4), pp. 324-330.
11. Hemphill, C S, Sampat, B N. Evergreening, patent challenges, and effective market life in pharmaceuticals (2012) Journal of Health Economics, 31 (2), pp. 327-339.
12. Hemphill, C S, Sampat, B N. Evergreening, patent challenges, and effective market life in pharmaceuticals (2012) Journal of Health Economics, 31 (2), pp. 327-339.
13. Jain, D and Conley, G J, Patent Expiry and Pharmaceutical Market Opportunities at the Nexus of Pricing and Innovation Policy, Insead Working Paper.
14. Id.
15. Id.
16. Dias, S and Ryals, L. Options Theory and Options Thinking in Valuing Returns on Brand Investments and Brand Extensions (2002), Journal of Product and Brand Management 11(2), pp115-128.
17. Danzon, P and Furukawa, M, Cross-National Evidence on Generic Pharmaceuticals: Pharmacy vs. Physician-Driven Markets (2011), NBER Working Paper No. 17226.
18. Two drugs are considered 'bioequivalent' if they are chemically identical, allow approximately equal amounts of the active ingredient to reach the therapeutic site and have similar rates of therapeutic action.
19. Hudson, J. Generic take-up in the pharmaceutical market following patent expiry A multi-country study (2000) International Review of Law and Economics 20, pp 205-221.
20. Organisation for Economic Co-operation and Development (OECD). OECD policy roundtables: generic pharmaceuticals 2009. Paris: OECD Competition Committee; 2010; p143.
21. Hudson, J. Generic take-up in the pharmaceutical market following patent expiry A multi-country study (2000) International Review of Law and Economics 20, pp 205-221.
22. Hemphill, C S, Sampat, BN. Evergreening, patent challenges, and effective market life in pharmaceuticals (2012) Journal of Health Economics, 31 (2), pp. 327-339.
23. Beall R, Kuhn R (2012) Trends in Compulsory Licensing of Pharmaceuticals Since the Doha Declaration: A Database Analysis. PLoS Med 9(1): e1001154.
24. Id.
25. Id.
26. Steinbrook R. Thailand and the compulsory licensing of efavirenz. N Engl J Med 2007; 356:544-6.
27. Hudson, J. Generic take-up in the pharmaceutical market following patent expiry A multi-country study (2000) International Review of Law and Economics 20, pp 205-221.
28. Jain, D and Conley, G J, Patent Expiry and Pharmaceutical Market Opportunities at the Nexus of Pricing and Innovation Policy, Insead Working Paper.
29. Donohue J M, Cevasco M, Rosenthal MB (2007). A Decade of Direct-to-Customer advertising of prescription drugs. N Eng J Med 2007; 357:673-81.
30. Pattikawa, L (2007). Innovation in the Pharmaceutical Industry. 1st ed.