Technological Self-Sufficiency and the Role of Novelty Traps

If there is one thing that the COVID pandemic has taught us, it is the value of technological self-sufficiency. Technologically sophisticated countries are pulling themselves out of the coronavirus crisis. 1 They do so by developing, manufacturing, and distributing the inventions needed to test people for SARS-CoV-2, including its variants, and to trace, cure, and immunize the local population. 2 Nations that lack the appropriate resources are not faring as well. 3 In India, for 1.

See, e.g., Covid-19: The Global Crisis -in Data, FIN. TIMES (Oct. 18, 2020), https://ig.ft.com/coronavirus-global-data/ [https://perma.cc/4FUM-QFTZ] [hereinafter Covid-19: The Global Crisis].

See, e.g., Shawn Radcliffe, Here's Exactly Where We Are with Vaccines and Treatments for COVID-19, HEALTHLINE (Jan. 25, 2022), https://www.healthline.com/health-news/heres-ex-actly-where-were-at-with-vaccines-and-treatments-for-covid-19 [https://perma.cc/XYQ3-5VBT].

example, the disease raged on long after developed countries had stockpiled the materials needed for treatment and had immunized a significant portion of their residents. 4 Blaming their fate partly on a global intellectual property regime that blocks access to supplies at prices they can afford, low-income countries demanded that the World Trade Organization (WTO) waive their obligations under the Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS Agreement) to protect patents and trade secrets. 5 Waiving their obligations, the low-income countries claimed, would allow them to suspend their patent laws, manufacture the products they need to deal with the pandemic, and make those products available at cost or prices that do not include the payment of royalties to right holders. 6 Although many countries joined in this initiative, 7 it was never clear that a waiver would constitute an effective response. As the Max Planck Institute argued, shortages in supplies were largely caused by the insufficiency of worldwide production capacity. 8 That problem, the Institute suggested, did not stem from patents or other forms of intellectual property protection but rather from the paucity of manufacturing know-how and technical infrastructure in many low-income countries. 9 There is, however, a way in which intellectual property law was a core obstacle to universal delivery of the material needed to contain the coronavirus. As currently structured, the international intellectual [Vol. 24:3:441 property system makes it difficult for countries to acquire the capabilities necessary to develop modern technologies, to adapt them to local conditions, or even to simply manufacture and disseminate them. 10 In this Article, we identify for the first time an important source of that problem: the emergence of an absolute standard of novelty. 11 Under this standard, no country will award a utility patent to an invention that was disclosed in, or rendered obvious by, prior art available anywhere in the world-that is, if the elements of the invention were published, patented, or even simply practiced in any place, regardless of its local accessibility. 12 As we explain, that standard traps technologies that could improve social welfare in a no man's land where there are inadequate incentives to do the work needed to make, distribute, and use them. 13 Hence our label: the novelty trap. 14 For developed countries in the North, this high standard of novelty largely makes sense. As long as the people working in a field have the absorptive capacity to learn from publicly disclosed materials, one can assume that they will identify promising technologies, make any incremental changes needed to promote their domestic use, and develop distribution networks. 15 Moreover, it is likely that customers in these countries will be receptive to-indeed, eager for-advanced technologies, have the funds to buy them, and possess the complementary assets, such as electricity and refrigeration, needed to adopt them. 16 Thus, one can also assume that customer demand will provide an impetus to innovate. 17 Although incentives to commercialize may sometimes be desirable, 18 patent law generally focuses on inducing inventions that would not be created otherwise through the skills of ordinary industrialists and artisans in the field, subject to conventional market forces. 19 For developing countries in the Global South, however, these assumptions do not always hold. "State-of-the-art" patentable inventions, of the type prevalent in Northern countries, are often beyond the reach of local technologists. 20 While indigenous inventors may be in a position to make "low-end" advances ("good enough" technologies by the poor, for the poor, 21 such as appliances that work without a steady supply of electricity) truly effective use of foreign technology would take considerable effort. 22 What might be called "midlevel inventiveness" can be needed for diffusion-to build factories capable of manufacturing high-tech products, adapt inventions made in developed countries to local conditions, supply complementary assets, INNOVATION HANDBOOK] ("[T]he word diffusion is commonly used to describe the process by which individuals and firms in a society/economy adopt a new technology, or replace an older technology with a newer . . . it is also an intrinsic part of the innovation process, as learning, imitation, and feedback effects which arise during the spread of a new technology enhance the original innovation.").

[Vol. 24:3:441 educate and train the domestic market, and create systems to distribute products and build demand. 23 Unfortunately, it is precisely the type of advances necessary that are most likely caught by the novelty trap: because the basic technologies are already known, the incremental changes inventors would wish to make are not always patentable. Yet without adequate incentives to diffuse Northern technology, the activity necessary to do so will be inefficient, if not entirely absent. If that is the case, the social welfare gains associated with the use of foreign technology will not be realized. Moreover, without an incentive system geared to local inventive capacities, the ecosystem required to support entrepreneurship and risk-taking, human capital formation, as well as capital accumulation and investment, is likely to be inadequate. The result is that developing countries will tend to remain technologically dependent on developed ones, and the likelihood is that inventions of importance to the South will not be adapted or produced. Examples include COVID vaccines, which may be scarce and, as formulated for the North, can be difficult for nations in the South to make, store, and transport; 24 plants that grow in the arid environments characteristic of many low-income countries; 25 and agricultural technologies, such as harvesters, adapted to local conditions. 26 There are many actions a country that is behind the technological frontier might take to become technologically self-sufficient-a term that we use to encompass acquiring the capacity to absorb foreign technologies, produce these advances, and make effective use of them, with the objective of reaching the point where the country is inventing at the global knowledge frontier, at least in sectors 171 (2004) [hereinafter PUNJAB DEVELOPMENT REPORT] (describing the steps taken to achieve a "Green Revolution"). of local priority. 27 We see as one key to that endeavor the introduction of a second-tier regime of patent protection. A second-tier regime would be designed to break the novelty trap and release innovative opportunities for local inventors. In essence, we recommend that developing countries create, as a supplement to utility patents, a new and distinctive intellectual property right, which we term "diffusion patents." These rights would be loosely analogous to pre-TRIPS patents of importation and would incentivize the efforts necessary for a country to reach technological self-sufficiency. 28 To that end, we recommend three changes to the novelty, inventive step, and disclosure requirements of patent law. 29 First, and most important, we would define the landscape against which the novelty and inventiveness of a technology is determined to exclude art that is not patented (or perhaps simply not worked) locally. Second, we would modify the inventive-step inquiry by focusing on the skills of the ordinary local artisan. Third, we suggest adapting the disclosure requirements of patent law to respond to the capacities of domestic inventors. In addition, we consider various implementation issues, including narrowing the scope of protection, restricting the availability of the regime to inventors from low-income countries, and imposing price controls. 30 This Article proceeds as follows. Part II elaborates on the reasons that diffusion patents would mitigate technological dependency. Part III describes the novelty trap in greater depth and discusses the details and advantages of a diffusion patent regime. Part IV considers justifications, both empirical and theoretical, for the moves we suggest. Finally, Part V discusses the consistency of this approach with international intellectual property law.

See, e.g., Jan Fagerberg & Manuel M. Godinho, As the International Monetary Fund (IMF) has demonstrated, the pandemic revealed that alongside low-income countries, many middle-income countries also suffer from technological inequality. See IMF, A Fair Shot, Fiscal Monitor 28, 31 (Apr. 2021), https://www.imf.org/en/Publications/FM/Issues/2021/03/29/fiscal-monitor-april-2021 [https://perma.cc/MA46-46UV] (Click "Full Report"). Thus, although we focus our proposal on low-income countries, we acknowledge that it could be applied, at least partly, to middle-income countries.

[Vol. 24:3:441 II. TECHNOLOGICAL DEPENDENCY Although there are many reasons for growing global inequality in the technological sphere, an important factor stems from the decision to link intellectual property with trade and to incorporate obligations to protect intellectual property into the WTO regime. 31 To a considerable extent, the TRIPS Agreement's one-size-fits-all regime was intended to benefit the high-income, developed countries of the North. 32 The thinking was that commodification of intellectual property rights would facilitate global trade in information. 33 Furthermore, without WTO-wide commitments to protect intellectual property, opening Northern markets to goods produced cheaply in low-wage countries in the South would leave Northern inventors without a return on the knowledge inputs embedded in manufactured products. 34 That said, proponents of the TRIPS Agreement also made strong arguments that TRIPS would likewise help the South. 35 The Washington Consensus suggested that rigorous intellectual property protection would encourage foreign direct investment (FDI) and technology transfer, promote a local innovation culture, and enable the South to catch up and progress to the technological frontier. 36 Furthermore, the intellectual property obligations required by the WTO would provide Southern legislators with political cover and allow them to escape short-term concerns about higher prices; in the long term, developing countries would achieve the social welfare gains associated with scientific advancement. 37 More than twenty-five years have elapsed since the TRIPS Agreement came into effect, and the North does appear to have benefited from stronger protection. 38 However, it is also evident that the South has suffered. 39 The projected benefits to the Global South have not materialized. 40 World markets for the commodities sold in the South did not compensate for the higher costs associated with intellectual property-protected goods. 41 Nor did high levels of intellectual property protection induce significant FDI or give rise to world-class indigenous inventorship. 42 Instead, technological dependency has persisted, the technological divide endures, and, as evidenced by the uneven distribution of COVID-19 vaccines, TRIPS is contributing to increasing global inequality. 43 For example, Keith Maskus and Lei Yang have found that although the changes TRIPS required have improved performance in research-intensive industries, the impact is smaller in lower-income economies, with minimal increases in indigenous innovation. 44 Similarly, the United Nations Conference on Trade and Development (UNCTAD) has expressed concerns about a widening [Vol. 24:3:441 technological gap and associated it with increased inequality across a variety of dimensions. 45 As it turns out, reaching technological self-sufficiency is a far more complex task than acknowledged by the Washington Consensus. 46 Indeed, extensive literature has grown up around questions concerning the conditions and mechanisms that would improve the innovative environment in the South and equalize its access to modern technologies. 47 As we explain in more detail below, 48 among other things, commentators suggest that catching up to the global knowledge frontier requires skilled workers, firms, and customers, as well as an infrastructure capable of supporting research, development, and distribution. 49 These insights have led intellectual property scholars to focus on creating job opportunities that enable workers to "learn by doing"-to develop technological capabilities by using imported high-tech equipment and processes, and facilitating access to such materials. 50 In accordance with that view, efforts have been made to ensure that international intellectual property laws give countries the flexibility to require patent holders to "work" their inventions-manufacture patented products or use patented processes-locally. 51 At the same time, a strong emphasis has been placed on relaxing intellectual property obligations so that patented inventions are available in the South at reasonable prices. 52 The proposed COVID waiver stems from that view, and the one modification that has been made to the TRIPS Agreement-the addition of another provision on compulsory licensing-reflects a similar approach. 53 Like the waiver, it reduces the degree of exclusivity rights holders enjoy in order to promote access-in the case of the modification, by allowing one WTO country to produce pharmaceuticals for the benefit of a WTO member that lacks manufacturing capacity. 54 The scholarship of innovation economists suggests, however, that these steps are not nearly enough-workers who learn by doing become adept at imitation. 55 In the short term, countries can improve social welfare using that approach because their firms can exploit wage differences with higher-income countries and profit from selling cheap versions of advanced products on global markets. 56 But there is a limit to this strategy: over time, the workers with imitative skills demand higher pay, and countries with even lower wage scales develop the skills to compete in the same sectors. 57 Reducing the cost of high-tech materials through compulsory licensing or other mechanisms is often also ineffective because both private and commercial consumers may need sufficient knowledge capital to understand what technologies are available. Similarly, they may need that capacity to evaluate these technologies, implement them, and adapt them to local conditions and social practices. 58 Adaptation can also require fresh investments.

[Vol. 24:3:441 Zvi Griliches, notably, studied the speed at which hybrid corn diffused across the Midwest of the United States. 59 He found that the rate of diffusion depended on how fast suppliers could customize the seed for use in a particular geographic area. 60 Similarly, Kristine Bruland demonstrated that the development of the Norwegian textile industry turned on the training that British machinery suppliers provided. 61 Put differently, reducing the profits available from information-based products works at cross purposes with the objective of promoting investments in diffusion because lowering available returns makes it difficult to recoup the extra costs associated with activities such as customization and training.

A domestic cohort of midlevel inventors and entrepreneurs could help resolve both the inadequacy of the imitation model and knowledge-capital inefficiencies. Countries that are behind the technological frontier have employed a variety of strategies to support the activities needed to promote the development of local inventiveness. Some have relied on government subsidies, direction, and involvement (examples include Japan's Ministry for Trade and Industry (MITI), Korea's Industrial Technology Research Institute (ITRI), and China's state-owned enterprises); some have benefited from investments by family-or business-based cartels (Japan's zaibatsus and keiretsus; Korea's chaebols). 62 There are also examples of firms in low-income countries entering into co-development deals or other arrangements with foreign firms that enhance local technological capacity. 63 Significantly, in a comprehensive study of the countries that have successfully caught up technologically, Keun Lee notes that many countries have found it effective to "detour" into a supplementary regime of patent-like protection. 64 In addition to offering utility patents, these countries have a petit (or second-tier) patent system with standards that reflect local technological capacities and domestic priorities. 65 Once the catching-up phase is over, Lee suggests that this that affect the benefits received, those that affect the costs of adoption, those related to the industry or social environment, and those due to uncertainty and information problems.").

59. detour process (and others that he offers) can be eliminated, leaving only the utility patent regime found in the North. 66 It is not difficult to understand why such a detour strategy would be fruitful. Most obviously, it has the potential to create a mechanism for incentivizing the incremental innovation required to distribute technologies that improve social welfare. 67 And like intellectual property systems more generally, it avoids the hazards associated with depending on government selection processes and support. 68 Just as important, a second-tier patent regime possibly establishes a lower, easier-to-reach rung on the protective regime ladder. Thus, it enables people who are behind the technological frontier to enter the system while adapting foreign technology to local needs. That, in turn, arguably promotes entrepreneurship, risk-taking, and the accumulation of human capital. With the right choice of technological sectors, these inventors can eventually develop the capacity to innovate at world levels, acquire utility patents in foreign jurisdictions, and even leapfrog over existing technologies to compete successfully in global markets. 69 At a minimum, this strategy can help a country become technologically self-sufficient in critical areas, such as healthcare.

Lee does not, however, describe what a lower-tier patent regime should look like and how it should differ from the regimes common in the North. 70 He leaves that to lawyers. 71 Accordingly, it is to this task that we turn. One aspect is clear: to the extent that the goal of a lower-tier regime is to create incentives for midlevel inventors, the inventive step should be adjusted downward to recognize the types of incremental innovation that they have the potential to accomplish. 72 But lowering the step does not supply a complete specification of the regime. Protection also requires a determination of novelty, and that analysis entails a comparison between an advance and what is already known. Thus, the determination requires countries to take a position on what is considered knowable and what people in the field can glean from prior art.

There are many ways in which this landscape of prior art could be characterized, and for a very long time, there were marked [Vol. 24:3:441 differences in how countries approached that question. For example, fifteenth-century Venetians considered the landscape to include only inventions available in Venice; anything new brought to the city could be patented. 73 Until 1952, the United States included patents and publications from anywhere in the world, but inventions not codified in one of those ways were considered to be prior art only if they were practiced locally. 74 Similarly, there were differences in whether patent applications should constitute prior art, and whether elements inherent in a disclosure, but which were not specifically mentioned, were in the prior art. 75 However, in an effort at the turn of the last century to negotiate a substantive patent law treaty to bring greater harmonization to national laws, the World Intellectual Property Organization (WIPO) suggested the adoption of an absolute standard of novelty. 76 Under this approach, any invention published, patented, known, used, or sold anywhere in the world would be considered in the prior art. 77 WIPO's International Bureau justified the standard this way:

"It is a fundamental objective of the patent system that nothing be alienated from society which already belongs to it. Indeed, granting a patent on an invention already known would impose constraints on society in respect of the use of known information without offering any return or benefit." 78 Although WIPO's substantive patent law treaty initiative was never promulgated, the last two decades have witnessed a growing consensus on its view that the optimum standard is one of absolute novelty. Thus, multiple countries-including the United States-have largely adopted this approach. 79 But in the evolution of that rule, we believe that an important cost was overlooked: this standard creates a novelty trap. We describe the novelty trap and the problems it creates in the following sections as part of our discussion of how we would structure a lower-tier patent regime.

The previous sections suggested that technological dependency stems in part from a global patent system that fails to reflect the needs of the Global South and the potential for midlevel inventorship that characterizes its creative communities. 80 Introducing a second-tier patent regime would alleviate the problem and create an ecosystem supportive of local innovation. 81 The question, then, is how that regime should be structured. In our view, it is important to recognize that this regime must do more than simply lower the height of the utility regime's inventive step, although it must do that as well. Rather, it is crucial to understand that the international patent system has evolved to a point where developed countries can shower their inventions on the developing world and leave valuable technologies trapped by a legal framework that prevents anyone from undertaking the efforts needed to distribute, adapt, and improve upon them.

This phenomenon possibly applies also when an inventor in the North obtains a utility patent on its invention in the South but does not choose to engage in efforts to exploit it locally. As the Patent Cooperation Treaty (PCT) has made remote patenting easier, the problem of Northern patents obstructing exploitation efforts in the South is becoming more commonplace. 82 [https://perma.cc/83L7-TGLD] (choose "2b-Grant for PCT national phase entries" from the "Indicator" dropdown; choose "Total count by filing office" from the "Report type" dropdown; choose "1985" and "2020" from the Year range dropdowns; then select "Low-income" from the "Office" list and click "Add"; then select "Non-resident from the "Type" list and click "Add"; then click "Search"). In the next tier up (low middle-income), the numbers are 26 and 31,817. Id. (choose "2b-Grant for PCT national phase entries" from the "Indicator" dropdown; choose "Total count by filing office" from the "Report type" dropdown; choose "1985" and "2020" from the Year range dropdowns; then select "Lower middle-income" from the "Office" list and click "Add"; then select "Non-resident from the "Type" list and click "Add"; then click "Search").

[Vol. 24:3:441 be left with an array of inventions that are not suitable for local use. To be sure, there may be cases where the utility patent holder tries to exploit its inventions in the South. However, these efforts can easily fail. As students of innovation have shown, successful exploitation can require a local presence: 83 Feedback loops that lead to incremental changes based on user experience; networks and other social systems which help spread information, including know-how, among the users of technology and their customers. 84 Establishing a local presence can be difficult, and many foreign inventors do not try. 85 Nevertheless, once the invention is patented locally, anyone who does try to adapt the invention becomes a potential infringer.

Even when an invention is not covered by a domestic utility patent, there can be a problem under an absolute novelty standard. That is, under the WIPO prior art policy, the mere existence of an invention anywhere in the world can be enough to put that invention, and even the incremental improvements needed to make efficient use of it, in the public domain. 86 The basic invention will no longer be considered novel, and incremental improvements will often not be considered sufficiently inventive to merit independent protection. 87 Admittedly, that result-the public availability of existing inventions and marginal variations on them-was WIPO's objective in promoting 83.

an absolute novelty standard for utility patents. 88 As we saw, its negotiators thought that releasing an invention from legal constraints would ensure its public availability. 89 However, it is simplistic to equate the existence of an invention in the public domain with its effective availability to those who could benefit from it. Customization, training, feedback loops, and networking may be required. Moreover, potential consumers may be unable to discern from remote disclosures that inventions capable of improving their welfare are available. From the perspective of the South, the absolute novelty standard thus constitutes a trap out of which new technologies cannot always emerge.

As noted above, we propose a second-tier regime that includes three key features: a relative novelty standard, an inventive step measured by the capacities of domestic artisans, and a disclosure requirement geared to local absorptive capacity. 90 Properly designed, this regime-a diffusion patent system-would break the novelty trap, create the incentives needed, and avert the associated social losses. And, as explained earlier, it would also function as the type of detour that Keun Lee suggested would enable countries to begin the process of becoming technologically adept in their own right. 91 In addition to discussing these modifications, this section considers other issues that a state implementing the system would be required to consider. These include questions on the scope of protection, who should be allowed to obtain these patents, and at what pricing?

The key to breaking the novelty trap is to redefine the landscape of prior art so that only disclosures that are locally accessible are relevant to the determination of whether an invention is novel and thus worthy of patent protection. As Part II suggested, it can require intellectual labor to identify and import remote knowledge and engage in the efforts required to diffuse it domestically. 92 Thus, awarding a form of intellectual property protection is easily justified: the protection

Traditional Knowledge and Intellectual Property -Background Brief, WIPO, https://www.wipo.int/pressroom/en/briefs/tk_ip.html [https://perma.cc/NL6G-M9WT] (last visited Jan. 22, 2022).

[Vol. 24:3:441 not only recognizes the intellectual contributions necessary for diffusion, but also provides economic motivation to engage in that work.

Local accessibility could, of course, mean more than one thing. One approach would be to consider an invention inaccessible only when it is not patented, published, or practiced locally. For this purpose, a country could consider an invention patented if it is covered by a patent that has been awarded, if it is covered by a patent application that is pending, or if the priority period provided by the Paris Convention for obtaining patents in third countries has not lapsed. 93 These are the most difficult inventions for a local innovator to identify as potentially useful domestically. Moreover, because learning by doing or observing is not possible, they are the hardest for a local to duplicate, improve upon, and distribute. Thus, their diffusion requires considerable intellectual effort, effort that readily justifies a separate incentive system. Diffusion patents for such inventions also make sense because if an invention is not patented locally, there is no possibility of infringement. Moreover, if the invention is not locally patented, there may be no one else with a strong incentive to do what is necessary to diffuse it.

A system that awards patents on this basis would have another advantage. Knowing that the failure to patent would allow another to obtain a diffusion patent, the original innovator might itself choose to obtain a utility patent, publish, or practice the invention locally. In other words, the potential for diffusion patents might spur foreign inventors to engage in local diffusion-oriented activities. Diffusion by foreign right holders would not necessarily be as effective in promoting an ecosystem conducive to technological self-sufficiency as awarding local diffusion patents, but it would surely provide jobs and training, bring in foreign investment, and create opportunities for locals to interact with foreign innovators and learn from their entrepreneurial cultures.

At the same time, however, it must be recognized that under a regime that considers local patents and publications to be in the prior art, whether worked or not, a foreign innovator could block local inventors by patenting or publishing locally but not doing the work needed to make the invention effectively available. As noted above, the PCT has sharply reduced the cost of that tactic. 94 There are two ways a country could counter that problem while also dealing with foreign patent holders who try to exploit locally and fail. First, the country could define the prior art to include only inventions practiced locally, even if a local utility patent had been granted, and include in the patent regime a compulsory license system to deal with the problem of blocking patents-that is, the inability of the holder of the diffusion patent to exploit it without licensing the utility patent and the inability of the utility patent holder to sell the adapted invention without the authorization of the diffusion patentee. The possibility that a compulsory license might be granted would put more pressure on foreign inventors to engage in exploitation efforts. Moreover, the compulsory license would recognize the contribution made by the original inventor and provide compensation for that effort. Alternatively, a country could require the holder of a local utility patent to work it locally within a reasonable time period (under the Paris Convention, four years from the date of the patent application or three years from issuance). 95 If the patentee chose to work, the benefits of jobs, training, foreign investment, and learning would be realized. If that period lapsed without working, then a diffusion patent could be awarded. Again, compulsory licenses could be used to break blocking positions.

Novelty is a technical impediment to patenting because it requires the presence of all the elements of the invention in a single piece of prior art. 96 In contrast, the inventive step, or nonobviousness, requirement allows examiners to consider whether a person of ordinary skill in the relevant art could combine disclosures from multiple sources to make the leap to the claimed invention. Thus, this requirement creates a stronger impediment to patenting than novelty. 97 Changing the prior art landscape is one way to alter this requirement for diffusion patents because it means that the only art that can be considered is art that is published, practiced, or patented locally (or perhaps only art that is practiced locally). As argued, that would significantly reduce the source material from which elements can be drawn. 98 [Vol. 24:3:441 countries, the height of the inventive step has fluctuated over time. 99 In the United States, for example, it has gone from very high in the mid-1900s to extremely low toward the turn of the century, settling in 2007 in the middle. 100 The current standard looks to the person having ordinary skill in that art (PHOSITA), a person who is deemed to have the capacity to make predictable advances and repurpose known inventions in response to market and design pressures. 101 US examiners, therefore, ask whether that person could have reacted to market forces to make the advance in question based on the art in the person's field. They would also examine other art that a person in the field would consider in solving the problem faced (analogous art). 102 If not, the invention is considered patentable.

For diffusion patents, these questions should be modified to deal with local conditions. Thus, rather than measure the capacity of PHOSITA by global standards, 103 the inquiry should be limited to the inventive capabilities of domestic artisans and the art they would regard as within their field or analogous to it. As important, examiners in low-income countries should consider the effect of market forces in their own environment. As noted earlier, in the North, high-income customers spur demand. 104 also account for the absence of that force in the South. Structuring the regime in this way would permit a local inventor to acquire protection for adaptations of foreign inventions even when the changes are ones that the North might consider to be too modest to merit protection. For example, it would allow inventors who technologically adapt patented seeds to local growing conditions to obtain protection. 105 It would similarly encourage domestic inventors to modify pharmaceuticals developed elsewhere to meet the needs of local patients, who might find the originator's version unpalatable or who might not have the refrigeration necessary to store it. 106 It would also spur inventors to educate consumers and find ways, such as new manufacturing techniques, to make inventions available at more locally affordable prices. 107 The PHOSITA standard has another useful feature: it operates as a policy lever that promotes technological advancement. 108 As Dan Burk and Mark Lemley suggest, when a new technology emerges, the knowledge in that area will be minimal. 109 Because PHOSITA-a person having ordinary skill in the art-will, at that point, have only modest capacities, even small advances will be considered inventive. 110 Naturally, it will be relatively easy to acquire patent protection. The ease of obtaining patents will attract other innovators. Once that happens, more art is produced, thereby expanding the knowledge base. As the field matures, those practicing in it will become more sophisticated, which will raise the standard for obtaining protection. 111 Inventors will then realize they must be more ambitious to merit a patent.

[Vol. 24:3:441 That same insight can be applied geographically. If the skill of the ordinary artisan is determined based on local-rather than global-abilities, advances that may be considered uninventive if measured by international standards might nonetheless be found to merit patent protection when created locally. As the level of technological capacity within a country grows, a person with ordinary skill in the art will be deemed to know more. As a result, the inventive step will rise with it, and simple adaptations will no longer be enough; local inventors will be forced to become increasingly creative. The field will, in this way, move towards the knowledge frontier.

In a sense, the heart of the patent bargain lies in the disclosure requirement because it is this requirement that obliges the applicant, in exchange for a period of exclusivity, to provide enough information about the invention so that people can enjoy it after the patent has expired. 112 Here again, the capacity of PHOSITA matters because the question examiners ask is whether the disclosure is sufficient for a person of ordinary skill in the art to make and use the invention without undue experimentation. 113 The standard of the ordinary local inventor may have several useful consequences. Because a midlevel inventor is likely to require considerable direction to practice an invention, this standard will require a diffusion patent to contain more information than the North typically demands for utility patents. Codification of more information will then make it easier for subsequent innovators to identify valuable art and figure out how to make and use it. With more art available, local inventors will become more proficient, ambitious, and able to catch up. Thus, even if patent disclosures tend to have a limited impact on research and development in the North, as some claim, 114 diffusion patents could be of considerable benefit in the South if they are geared to teach artisans with low absorptive capacity. Given that patents are drafted in the language of the relevant patent office, these patents would have the added advantage of providing this information in the local language. 115

In addition to adopting altered standards for the novelty, inventive step, and disclosure requirements of patent law, a country implementing a diffusion patent regime would also need to decide on the scope of the rights awarded, the beneficiaries of the regime, and various pricing questions.

Two factors determine the scope of a patent right: the breadth of coverage and the duration of protection. 116 The disclosure requirement discussed above is also used in the determination of breadth because it limits the claims to the information that is revealed in the specification. 117 For those who may be concerned about adding another layer of protection to an intellectual property system that already inhibits access to inventions, this is an important feature. If properly enforced, 118 the limitation aspect of disclosure ensures that a country that awards diffusion patents receives knowledge commensurate with the added cost associated with awarding protection. Since these patents would be geared toward encouraging marginal advances, they would only cover a narrow range of products or processes. Thus, others would be free to find other adaptations and uses for the underlying invention.

The second layer of patenting could also be constrained by a limit on the duration of protection. The patent period should be long enough to allow the right holder to capture a return on investment, but not so long that it creates its own novelty trap and inhibits further innovation. As J.H. Reichman noted in his exhaustive study of petit patents, the term of protection under these systems is usually well below the twenty years required by the TRIPS Agreement for utility patents. 119 [Vol. 24:3:441

Because national treatment is a cornerstone of the international trade regime, 120 a country must arguably award a diffusion patent to anyone who invests in the work necessary to make the invention available to its population. However, that practice could defeat the goals of adopting a diffusion patent regime. To be sure, if foreigners were allowed to obtain diffusion patents, they would be more motivated to modify their inventions in ways that meet Southern needs. They might also increase jobs and learning opportunities. However, as long as foreigners are more technically proficient than domestic inventors, local innovators could easily be crowded out of the system. 121 As a result, domestic innovation would not receive direct encouragement. Countries may therefore wish to limit diffusion patents to their own innovators or to their own inventors plus those from similarly situated countries. That limitation would, however, arguably violate the nondiscrimination provisions of the TRIPS Agreement. 122 Part V takes up the question of the conformity of our proposal to international intellectual property law. 123 Countries intent on developing a diffusion patent system to supplement utility patents could develop a common patent system akin to the system that the European Union is establishing. It would allow an inventor to acquire a single patent that covers multiple countries. 124 That would create a larger market for the holders of diffusion patents and allow more countries to obtain the benefits of Northern technology. 125

See EUR. PAT. OFF., UNITARY PATENT GUIDE 10 (2017), https://documents.epo.org/pro-jects/babylon/eponet.nsf/0/C3ED1E790D5E75E0C125818000325A9B/$File/Unitary_Patent_guide_en.pdf [https://perma.cc/4HW6-PLHK]. A regional patent system would also have the advantage of reducing administrative costs; a single nation's patent office could examine applications on behalf of an entire region. Id. at 10-11.

Diffusion patents could incur inventive costs and reduce access implying pricing considerations. On the whole, it seems unlikely that prices would put these inventions out of local reach because, under the basic system, the holder of the right could only sell the invention in countries where the original inventor did not receive a patent. Thus, the price would depend on local demand and purchasing power. Admittedly, the right holder could base the price on the demand of only the highest-income local customers. 126 In that instance, a country could impose price controls or price caps on advances protected by diffusion patents. 127 A country that decides to award diffusion patents based on the failure of the utility patentee to practice will face another pricing issue. Because the utility patent would block the holder of the diffusion patent from exploiting her invention (and vice versa), compulsory licenses might be needed to break the blocking position, a scenario that is especially likely if the initial patent is overbroad. 128 The issuance of a license would, of course, require the party with the less valuable patent to pay royalties to the other-under TRIPS, "adequate remuneration in the circumstances of each case, taking into account the economic value of the authorization." 129 Still, access should not be a profound concern because adequacy under the circumstances would allow for consideration of the purchasing power of local customers.

The previous Part recommended the introduction of a second-tier of patent protection, a right that encourages the efforts needed to diffuse new technologies within low-income countries. 130 Designed as a detour, diffusion patents would also grow the South's creative ecosystem, promote technological self-sufficiency in priority areas, and enable these countries to catch up technologically to the 126.

See Brook K. Baker, Patents, Pricing, and Access to Essential Medicines in Developing Countries, 11 VIRTUAL MENTOR 527, 527 (2009) (noting that pharmaceutical companies maximize profit by selling medicines at prices only the rich and well-insured can afford). 127.

Cf. id. at 528-29 (comparing alternative strategies to correct this "market failure"). 128.

Cf. Correa, supra note 52, at 10-11 (exploring varying international approaches to this issue). The "refusal to deal" is a ground for granting a compulsory license in many national laws. Id. at 10-11.

129. TRIPS, supra note 5, art. 31. In addition, the holder of the utility patent "shall be entitled to a cross-licence on reasonable terms . . . ." Id.

[Vol. 24:3:441 developed world. This Part takes a closer look at the reasons we believe this system will achieve these goals. In part, our view is based on observations of how developing countries have structured their patent systems over time. In addition, we consider the economic literature in greater depth and discuss how it supports our recommendations.

As noted in Part II, Keun Lee's view, that a second-tier patent regime can facilitate catch-up, stemmed from his study of how other countries developed and, in particular, from a comparison among countries in Latin America and Asia that were originally situated similarly but later diverged sharply in their technological proficiency. 131 Admittedly, his observations do not establish causality. 132 Still, the number of examples of technologically advanced jurisdictions that had modified patent regimes during their development period is probative of the benefits of this approach. 133 The United States' 2011 America Invents Act (AIA) is a prime example of a national policy that took one approach to the requirements for patent protection when the country was behind the technological frontier and then adopted a more rigorous standard when it became innovative at world levels. 134 Prior to the AIA, information "known or used," "on sale or in public use," or "made. . . by another inventor" was regarded as in the prior art only if the knowledge, sale, use, or invention took place in the United States. 135 However, for patents applied for on or after March 16, 2013, the AIA regards information "in public use, on sale, or otherwise available to the public" as prior art no matter where the use, sale, or availability is located. 136 In addition, under pre-AIA law, information in patent applications was considered prior art only as of their US filing date (and only if the US patent was eventually granted or the application published). 137 Under current law, in similar circumstances, disclosures in US patent applications are effective as of their foreign filing date. 138 In other words, information disclosed in foreign patent applications can now become part of the prior art landscape. 139 The value of the older rules to the United States is clear. Foreign inventors could not block the issuance of US patents based on most of their foreign inventive activity. 140 An inventor working offshore could, in short, find that if it later wanted to exploit an invention it made first, it would be required to pay royalties to a later inventor who did its work in the United States. 141 To prevent that occurrence, someone who created knowledge abroad was under pressure to bring the advance to the United States quickly, by describing it in a publication, putting the knowledge in a patent claim, or exploiting the inventions in the United States in a manner that made their technological contributions locally available. 142 Although the AIA ended this practice there are ways in which the United States continues to limit the prior art landscape to meet policy objectives it regards as important for promoting innovation. 143 For example, the current rule on novelty removes from the prior art disclosures made by the inventor or by someone who derived the invention from the inventor, if they are made within a year of the inventor's patent application. 144 It also removes certain art made by another after a public disclosure by the inventor or by someone who derived the invention from the inventor. Furthermore, it removes certain art in patent applications that were filed after a public disclosure by the inventor or by someone who derived the invention from the inventor. 145 Finally, in some circumstances, if the art in two patent applications is owned by, or subject to assignment to, the same entity, the art disclosed in the first application cannot be used against [Vol. 24:3:441 the second application. 146 None of these new rules appears designed (or needed) to help the United States catch up technologically, but each is thought to improve the creative environment. These rules establish a grace period that allows inventors to share their nascent inventions with a community that can help improve them. 147 They also allow inventors to give potential users notice of the innovation and enable researchers in the same firm to collaborate with each other without fear that transfers of knowledge will invalidate patent rights. 148 Japan and the Asian Tigers-a group of countries that have caught up to the technological frontier in the last few decades-demonstrate the value of a patent regime geared toward the capacity of local midlevel inventors. 149 For example, in a study sponsored by the Organization for Economic Co-operation and Development (OECD), Nagesh Kumar showed that strong intellectual property rights adversely affected absorption of knowledge spillovers, while countries that started with "soft" regimes, which favor local inventors, prospered. 150 Before TRIPS, Japan explicitly designed its patent policy to favor domestic inventors and encourage absorption spillovers from foreign activities. 151 It promoted a patenting culture with a utility model and an industrial design system that allowed and motivated local inventors to modify inventions made elsewhere. Kumar found that Taiwan and Korea took a similar approach. 152 Thus, he states:

[T]he east Asian countries, viz, Japan, Korea, and Taiwan have absorbed a substantial amount of technological learning under weak intellectual property protection regimes during the early phases. These patent regimes facilitated the absorption of innovation and knowledge generated abroad by their indigenous firms. They have also encouraged incremental innovations on foreign inventions by domestic enterprises and developed a patent culture through utility models and design patents. As the local technological capabilities matured and the domestic industry sought stronger protection for guarding their inventions, the IPR [intellectual property rights] regime was strengthened…. 153 Countries have similarly varied how they implement the inventiveness requirement. As noted above, the rule in the United States has fluctuated, only recently settling on a version that considers PHOSITA to be a person with a modicum of creativity, spurred by exogenous forces. 154 Prior to the change, only references that included much more were considered capable of barring or invalidating a patent. 155 The landscape against which the inventive step is determined has also varied over time and remains less absolute than the novelty standard. Thus, although all art can be used to determine novelty, only art in the inventor's field and in analogous fields are used to decide on nonobviousness. 156 Furthermore, although elements inherent in a disclosure are considered for novelty purposes, they are not combinable when determining inventiveness. 157 The European Patent Convention goes a step further and excludes disclosures in patent applications from the inventiveness-but not the novelty-determination. 158 A 2004 study by the WIPO Standing Committee on Patents showed that, at that time, there were similar variations among all the countries that it surveyed. 159 The utility models and design systems referenced by Kumar do nearly the same thing: they furnish protection for incremental advances. 160 Germany, for example, has offered midlevel inventors of products (but not processes) a "utility model" system that requires no substantive examination. 161 It also uses a relative novelty standard [Vol. 24:3:441 somewhat similar to the one we propose. 162 Significantly, however, interest in the system has declined over the years as firms have become more inventive and turned their attention to utility patent protection. 163 Dan Prud'homme observed an analogous trajectory in Asian countries. 164 Focusing on the world's largest users of utility model patents, namely China, 165 Japan, South Korea, and Taiwan, he found that each country recalibrated its utility model regimes over time in ways that mirrored that country's stage of economic development. 166 Thus, the regimes began by offering weak protection, as measured by requirements for novelty, inventive step, and the use of an examination procedure. 167 That created a lower rung in the protective ladder and also gave other inventors leeway to use the material in the disclosure for their own purposes. 168 As local innovators attained the proficiency needed to use accumulated knowledge, the requirements for obtaining rights became more demanding, and the regime offered broader protection. 169 Many industrialized countries followed a similar path. 170 Some have now dropped their lower-tier systems entirely. 171 However, utility model regimes endure in some places, and a handful of industrialized countries still continue to experiment with them. 172 The European Union, for example, has considered adopting a utility model system as a way to encourage improvements to known advances and to spur technological development within small and medium-sized firms. 173 Although the justification for such patents is in some ways similar to ours, these systems have arguably tended to pay little attention to the prior art landscape. Thus, in defining the state of the art as "everything made available to the public by means of a written or oral description, by use, or in any other way, before the date of filing of the utility model application," 174 the European Union's system fails to break the novelty trap.

As suggested earlier, economists have extensively studied why technological progress has been slow in many low-income countries, such as those in Latin America, and how it is that other countries, including the Asian Tigers, have prospered. 175 That work, which identifies many key factors, also provides support for connecting the availability of lower entry points in the intellectual property regime to a country's propensity to become innovative. 176 We summarize that literature here and demonstrate its connection to our proposals. To begin, economists essentially see two deeply interconnected problems which the following Section discusses. One is the lack of innovative capacity; the other is the difficulty of diffusing Northern technology in less-developed countries.

The economics literature suggests that technological progress in developing countries is largely a matter of dependency-it depends on adopting ideas and technologies developed elsewhere. 177 Innovation is a [Vol. 24:3:441 function of how socioeconomic, political, geographical, and legal subsystems interact, and, importantly, it is contingent upon the existence of diverse local capabilities. 178 In particular, innovation requires absorptive capacity, which is defined as "a set of organizational routines and processes by which firms acquire, assimilate, transform, and exploit knowledge to produce a dynamic organizational capability." 179 Thus, the extent to which a country relies on foreign technologies depends as much on its absorptive capacity as on its actual development needs. 180 A 2008 World Bank Report corroborates this observation. 181 It shows that much of the technological progress in developing countries measured over two decades has been associated with an increase in the exposure of developing countries to foreign technologies. 182 Unfortunately, however, as the World Bank Report also demonstrates, the capacity of the population of these countries to absorb technology has improved much less. 183 Lacking the capacity to create new or adapt and improve existing technologies, average productivity and income have remained low. 184 Since these countries cannot generate innovations at the technological frontier, those with a degree of technical proficiency tend to rely instead on an explicit policy of copying foreign technologies. 185 Steven Schnaars, following Theodore Levitt's 1966 seminal article, Innovative Imitation, categorized several distinct types of copying. 186 These include counterfeits or product pirates, knockoffs or clones, design copies, creative adaptations, technological leapfrogging, and adaptation to another industry. 187 Counterfeits and knockoffs are duplicative imitations. Counterfeits are copies sold under the same brand name as the original, often of lower quality, thereby depriving the innovator of owed profits. 188 Knockoffs or clones are different. Because they are based on products that are not (or no longer) protected by intellectual property, they are legally sold at lower prices. Clones may even exceed the original in quality. 189 From a development economics perspective, however, an economic policy based on imitation-of any not convey a sustainable competitive advantage, even when the copying is legal. If the wage cost in the imitator's country is significantly lower than the originator's, the imitator will enjoy a competitive edge in price. However, that advantage tends to be of limited duration. Eventually, lower-wage countries step in and take the market. 190 However, adjacent to policies supporting copying are policies directed at technology adaptation. These are more beneficial for at least two reasons. First, an edge here is likely to be more durable. Second, adaptation of foreign technology makes it possible for locals to understand the benefits of these advances, use them, and progress technologically. Indeed, many developing countries that began with simple technologies (and low-intensity R&D), such as textiles, clothing, food processing, and wood products, were able to rely on such policies to advance to somewhat more complex industries, such as metals, petroleum refining, or metal products. 191 In this regard, the 2008 World Bank Report notes the importance of both imports and exports. 192 Importing foreign technology helps developing countries raise the quality of their own products and production methods. 193 Exports put locals in contact with those who can provide them with guidance on how to meet global demand. Exportation also exposes domestic producers to foreign competition. 194 As the World Bank explains, countries differ in the ways and the rapidity with which they improve their capacity to invent. 195 The World Bank characterizes some countries as "traditionalist slow learners." 196 [Vol. 24:3:441 These countries rely heavily on learning from imports. 197 Others are considered "passive FDI dependent;" 198 they develop competencies from higher-tech exports. 199 Finally, some countries are categorized as "active FDI-dependent;" they strive to develop a strong domestic technological base. 200 Presumably, the differences stem from cultural and educational factors as well as from differences in their ability to attract FDI and technological assistance. However, given enough time, all of these learning techniques can lead to significant improvements in technological competence. 201 For example, Bangladesh, which the World Bank considered a slow learner in 2008, has since developed a vibrant manufacturing sector supported by its growing technological capability. 202 Significantly, in the last two decades, it has set up systems to ensure that rural communities can access the internet and information, created a portal for training, and improved its system of higher education. 203 The countries that succeed in expanding their capabilities may, in the end, enjoy special advantages from entering foreign competition as latecomers. 204 Thus, even in the North, there is often a lag between the time a new technology is created and the time it is adopted. 205 First, firms are reluctant to invest in new technology because they cannot always recover the sunk costs associated with the technology they already possess. 206 Second, it can be difficult for them to retrain workers and switch old employees to a new system of doing business. 207 In comparison, countries with newly acquired capacity do not experience these problems: when they enter the field, they do so with the latest equipment and skills and can accordingly capture gains from using more efficient equipment and production methods. 208 As a result, numerous countries have experienced these latecomer advantages and have proven that they can compete effectively on the world stage. These countries include Japan, South Korea, and Taiwan, where rapid post-war growth was rooted in the successful acquisition, imitation, and copying of technologies initially developed in industrialized economies. 209

The question then becomes: Why do industries in developing countries, which could potentially enjoy latecomer advantages, not regularly leapfrog over industries in advanced economies? Here, as the World Bank report shows, the core problem is that capacity can require familiarity with foreign inventions, yet the diffusion of new technologies can be extremely slow, especially in developing countries. 210 The World Bank looked at 102 country-technology pairings and compared how long it took, from the time of invention, for adoption to reach specified levels. 211 Over time, the pace of diffusion improved everywhere, but, at all points, it was faster in high-income countries than in low-income countries. 212 Economists have developed a variety of models to identify and characterize technology diffusion. 213 Capacity is (ironically) a major [Vol. 24:3:441 factor influencing the willingness to take risks on new-to-the-country and market technologies, as well as basic technological literacy, and the skills needed to undertake the research necessary to understand, implement, and adapt modern technologies. 214 However, other factors are important as well, including the availability of financing, means for capital accumulation, the rate at which information is exchanged, and thus the presence of networks of learning. 215 Of considerable importance are the net gains that firms expect to enjoy should they implement new technology. 216 Since revenue depends on the willingness of consumers to purchase their product, the buyers' absorptive capacity and the resources they have for learning are important here as well, as does the suitability of the product to their needs. 217 Aiding this process is complex. Both FDI and licensing opportunities can play an important role because they help make imported technologies available, sometimes without added domestic investment. 218 Furthermore, they produce technology spillovers; local workers receive training and accumulate experience as they deal with investors, which are typically multinational foreign firms. 219 Foreign affiliations can also give local workers the opportunity to work in other locations where they can learn management skills. 220 Their growing familiarity with high-tech products and techniques for using them eases the domestic adoption of similar-sometimes adapted-technology. 221 Public-private partnerships and other sorts of affiliations can have a similar impact. 222 In some countries, migration has also proved to be significant. While the so-called brain drain can deplete the population of technically trained individuals, the diaspora can also constitute a resource. 223 It establishes connections between countries at different stages of development and creates access to networks of learning. 224 Moreover, some migrants return with the knowledge and skills needed to bring foreign technologies home. 225 In India, for example, Sam Pitroda, an Indian-American, founded the Center for Development of Telematics, which introduced rural telephone exchanges that were cheap to maintain and designed to operate without air conditioning. 226 This adaptation later diffused to other developing countries. 227 Trade can also be a significant factor in improving the rate of diffusion. As Diego Comin and Bart Hobijn have observed, it has a push and pull effect. 228 Pushing-making sales of foreign technology into developing countries-gives locals exposure to high-tech products and to their advantages. 229 In addition, these sellers have the incentives Zvi Griliches observed in connection with hybrid corn, which prompt them to help adapt the technology to local conditions. 230 Pulling happens because the domestic rivals of the firms that import the technology must adopt it themselves to stay competitive. 231 These intertwined considerations-capacity and diffusionprovide developing countries with a choice of strategies. Some nations have made technological progress by emphasizing the capacity aspect of the problem. China, for example, has directed considerable efforts to building the capabilities of its domestic science and technology sectors. 232 Bangladesh now has a major pharmaceutical industry and has reached the point where it could, for instance, probably be manufacturing COVID-19 vaccines. 233 Other countries have focused on diffusion. Thus, the Green Revolution in India is largely attributable to the improvements to agricultural productivity and performance made possible by the widespread adoption of technological advancements. 234 [Vol. 24:3:441 Either way, a lower-tier patent system can help. It would afford protection geared toward the capacities of local innovators (and potential innovators) in the South and would create incentives for them to provide the education and make the adaptations required for successful diffusion. As locals begin to work in sectors of greater technological complexity, they would learn to convert proof of concept to product, develop communication styles and management techniques, and develop the ability to assess the risks and benefits of entering particular fields. 235 Over time, the manufacturing skills and scientific and technological knowledge acquired would also have the potential to enable domestic entrepreneurs to introduce evermore substantive innovations. They may also discover niches where latecomers enjoy special advantages and target those industries. These steps would, in turn, nurture the creative community and produce the ecosystem that is key to technological self-sufficiency. But because so much of this is dependent on the availability of technology that already exists in the North, breaking the novelty trap is a crucial element if a second-tier regime is to be successful.

For the most part, this Article's proposals are compatible with international law. The TRIPS Agreement requires that "patents shall be available for any inventions, whether products or processes, in all fields of technology, provided that they are new, involve an inventive step and are capable of industrial application" and allows members to "implement more extensive protection in their law than is required by this Agreement, provided that such protection does not contravene the provisions of [the] Agreement." 236 Thus, adding a second-tier regime would not in and of itself constitute a violation of TRIPS commitments. Indeed, the Paris Convention, which is incorporated by reference into the TRIPS Agreement, specifically mentions "utility models." 237 Furthermore, other than to say that "inventive step" is synonymous with "non-obvious," 238 none of the relevant terms is defined by the TRIPS Agreement. States therefore enjoy substantial flexibility in how they structure their systems. 239 Finally, there is nothing in the TRIPS Agreement that deals with pricing or limits the ability of WTO members to impose price controls. 240 There are, however, three somewhat difficult issues. The first concerns the key recommendation for breaking the novelty trap, which is to exclude from the landscape of prior art inventions that are not locally patented-or even inventions that are not locally practiced. The TRIPS Agreement requires that "patents shall be available and patent rights enjoyable without discrimination as to the place of invention, the field of technology and whether products are imported or locally produced." 241 It further requires that "each Member shall accord to the nationals of other Members treatment no less favorable than that it accords to its nationals about the protection of intellectual property." 242 Arguably, these provisions mean that foreign patent holders must be permitted to enjoy the right to prevent local inventors from acquiring protection over the same, or incrementally altered-or as patent lawyers say, patentably indistinct-advances. That is, the reference in TRIPS to the enjoyment of patent rights could be understood to include the right to block others from obtaining protection on incremental improvements. 243 It is not, however, clear that TRIPS guarantees were meant to cover anything more than the right to exclude others from practicing a patented invention. 244 Specifically, there is little to indicate that the TRIPS language should be understood as requiring states to recognize a right by one inventor to bar others from obtaining protection on related contributions. As noted above, there was a time when the United States regarded foreign knowledge and patent applications differently from US knowledge and patent applications. 245 Its trading partners objected to the practice because they saw it as depriving foreign inventors of opportunities to exploit and further develop their [Vol. 24:3:441 inventions in the United States. 246 Significantly, however, they did not challenge this practice in the WTO. Given that precedent, it would be difficult to argue that a second-tier system that relies on a relative novelty standard violates the TRIPS Agreement.

The proposed diffusion patent system is also different from the one US trading partners objected to in that a foreign utility patent holder could avert the creation of a separate right by engaging in diffusion activities itself. Moreover, if a foreign inventor has obtained a local patent, but the country nonetheless awarded a diffusion patent, the former patent would most likely block the use of the latter. The holder of the patent on the original advance could then either agree to license it or earn whatever royalties are set in connection with a compulsory license. Notably, the TRIPS Agreement specifically contemplates the possibility of blocking patents and allows countries to issue compulsory licenses to deal with them. 247 To put this another way, TRIPS can arguably be understood as recognizing that patent rights do not extend to blocking improvements and that later inventors might diminish some of the opportunities otherwise available to earlier innovators.

But the compulsory license alternative suggests another problem. To a considerable extent, the diffusion patent system is designed to put pressure on the original utility patentee to exploit the invention itself. That is not very different from the way the local working requirement in the Paris Convention is meant to operate. 248 That provision has been the source of considerable controversy precisely because it limits the choices that patentees can make. 249 It has been repeatedly amended to make it more difficult to invoke. 250 And it currently includes waiting periods that are not a principal part of our recommendations. 251 In addition, there are questions as to whether the local working provision was voided sub silentio by the TRIPS Agreement's ban on discriminating between imported and locally produced inventions. 252 Nonetheless, we believe that the differences between a local working requirement and a diffusion patent are significant enough for the proposal to survive a TRIPS challenge. Among other things, the diffusion patent recognizes that simply working an invention is not sufficient to diffuse it or to build a local creative community. Rather, it requires intellectual contributions to adapt it to local conditions and to create markets for the product.

The third problem is more serious. Limiting diffusion patents to local inventors conflicts with the national treatment guarantee-that is, with the obligation to accord to nationals of every WTO member "treatment no less favourable" to that which the member accords to its own nationals. 253 To avoid a violation, the class of beneficiaries could be broadened. It could be structured to include anyone willing to operate locally. But as we suggested, that practice could easily crowd out locals. 254 The same benefits might be achieved for diffusing the invention, but the protection is likely to be less effective at fostering domestic creativity. Thus, it may do little to promote technological self-sufficiency.

Another approach is to broaden the class of potential beneficiaries to include nationals in similarly situated developing countries. Even better would be to create a unitary system that protected and encouraged diffusion activities in multiple countries in the South. That would increase the pool of mid-level inventors who would be interested in undertaking local diffusion. And the larger market would enhance the available incentives. More important for TRIPS, if the beneficiaries are defined by economic status, other countries could not claim de jure discrimination. [Vol. 24:3:441

Other countries might, however, argue that such a system discriminates de facto: they might claim that status is simply a proxy for nationality. 256 As the WTO panel decision in the EC-GI dispute made clear, de facto discrimination can also be actionable. 257 That case involved an EU Regulation that set out one procedure for acquiring geographical indications for foods produced in the European Union and another (more onerous) procedure for foods made elsewhere. 258 Although the Regulation did not expressly discriminate among WTO countries, the Panel held the distinctions between the place of production did not "occur as a random outcome in a particular case but as a feature of the design and structure of the system," and the system would "operate in practice" to discriminate "to the detriment" of the nationals of countries that are not in the European Union. 259 The Panel then held that to permit members to substitute the place of production for nationality would undermine an important feature of the WTO framework agreements: the guarantee that right holders would enjoy "effective equality of opportunities" to trade in WTO markets. 260 To be sure, there are differences between the Regulation at issue in EC-GI and a law that treats inventors from developing countries in a special way. However, substituting economic indicators for nationality has essentially the same effect. Indeed, limiting the diffusion patent regime to nationals of similarly situated countries is intended as a form of protectionism: its goal is to enhance domestic technological proficiency by sheltering locals from being crowded out by foreigners.

It is, however, worth noting that it is not clear that the national treatment obligation has ever been faithfully honored in the patenting realm. In a series of studies of patent office actions in the IP5 (the USPTO, the Japanese Patent Office, the European Patent Office, the National Intellectual Property Administration of the People's Republic of China, and Korean Intellectual Property Office), Beth Webster and her coauthors observed systematic bias against foreign inventors. 261 Although the authors found that the bias is reduced when inventors rely on local prosecution attorneys, it does not disappear. 262 The PCT can even exacerbate the effect because it too creates a bias in favor of local applicants. 263 Indeed, in Kumar's close study of patenting in the Tiger economies, he found that until Japan developed technological capacities on par with developed economies, patent applications by foreigners experienced longer pendency periods in the patent office relative to domestic applications (and, since the term begins on the filing date, therefore enjoyed shorter terms of protection). 264 The claims that were eventually issued also tended to be narrower than claims in the patents of domestic inventors. 265 Similarly, Korea tolerated lax enforcement (and multiple complaints from the United States) to facilitate duplicative imitation. 266 In his report to the OECD, Kumar concluded that the TRIPS Agreement should be revised to give developing countries more flexibility until they reach a specified per capita income. 267 A strong argument can be made that even without amending the Agreement, this flexibility is available. While it is true that the transition provisions included in TRIPS (some of which have been extended several times) do not apply to national treatment obligations, 268 WTO law should be construed to give developing countries room to maneuver with respect to issues involving development. For example, Simon Klopschinski, Christopher S. Gibson, and Henning Grosse Ruse-Khan have suggested that the national treatment obligation in TRIPS should be interpreted in light of its analog in the General Agreement on Tariffs and Trade (GATT). 269 The relevant GATT provision requires nondiscrimination among products of [Vol. 24:3:441 different national origins. Notably, it applies only to like products. 270 Accordingly, WTO members can distinguish among products that are different in ways that further their own national policies. TRIPS focuses on the nationality of right holders rather than products. 271 True, it does not include the comparator "like." 272 But interpolating a comparison into the Agreement that permits countries to distinguish among inventors of different nationalities is consistent with WTO goals and values. 273 TRIPS would then recognize that inventors from different countries are in very different positions from one another and that affording them effective equality of opportunities requires nations to make distinctions to enable their inventors to compete. Significantly, Thomas Cottier, a member of the Swiss delegation during the negotiation of the TRIPS Agreement, has argued that the objectives and principles articulated by the TRIPS Agreement, which explicitly refer to "the promotion of technological innovation and to the transfer and dissemination of technology in a manner conducive to social and economic welfare," requires that the Agreement be interpreted in a manner conducive to technological development. 274 A softer view of national treatment also accords with recommendations that UNCTAD made when WIPO first considered the idea of harmonizing IP laws more thoroughly. 275 In its view:

"[E]quality of treatment only makes sense when the parties involved are in a general way equal; when they are not, equality of treatment simply gives the stronger party unlimited freedom to utilize his power at the expense of the weaker party." 276 Writing in 1975, UNCTAD was concerned that harmonization would create a "reverse system of preferences" that would favor highly developed countries and impose heavy costs on other nations. 277 Subsequent events have shown UNCTAD to be prescient. 278 As noted in the introduction, technological inequality has persisted under TRIPS, and the sequelae of the COVID-19 pandemic demonstrate its tragic consequences. 279 UNCTAD advocated for an international system that allowed for national differences designed to promote technological development. 280 Although TRIPS may represent more harmonization than UNCTAD considered wise, 281 it is a minimum standards agreement, and the commitment to flexibility should permeate the interpretation of all its provisions, including the ones on nondiscrimination. 282 Furthermore, although the WTO does not generally make normative judgments concerning the regulatory purpose of challenged legislation, 283 a strong argument can be made that the catastrophic consequences of technological dependency require the WTO to take into account the need of all countries to grow their technological sectors. 284

The inadequacy of the global response to COVID-19 has exposed a deep flaw in the international intellectual property system. By raising costs and impeding technological development in the Global South, the TRIPS Agreement has left many countries incapable of developing, manufacturing, and distributing innovations that are crucial to survival. Changing intellectual property law will, of course, not cure the problem. Technological dependency has many root causes, including the absence of a trained workforce and an appreciation for the value of acquiring human capital; the paucity of venture capitalists, entrepreneurs, and lawyers; deficiencies in infrastructure, such as the absence of laboratories, roads, and refrigeration; governance problems; and inadequate institutions to support education, capital accumulation, and risk-taking. But it is important to remember that these factors are interdependent: a country needs a research base to create incentives to [Vol. 24:3:441 adopt a strong patent system. 285 It needs a patent system to attract venture capitalists, high-tech jobs to make it worth acquiring human capital, and seekers of education to promote the need for universities. Thus, while modifications to the intellectual property system are not a cure, reform can be an important step.

A fundamental question is how that step ought to be designed. In this Article, we discuss one way in which intellectual property law has interfered with development: the structure of patent law has trapped novel advances in a legal limbo that makes it difficult for less developed countries to enjoy high-tech products, reach the technological frontier, or even attain technological self-sufficiency in areas of critical importance, such as health care. Accordingly, we have offered various approaches to breaking that trap.

See, e.g., Walter Park & Juan Carlos Ginarte, Intellectual Property and Economic Growth, 15 CONTEMP. ECON. POL'Y 51, 52 (1997).