Intellectual Property Rights (IPRs) have become ubiquitous in the current debate and have emerged as the key issue of global innovation policy. The ‘Trade Related Aspects of Intellectual Property Rights’ (TRIPS) Agreement, signed in 1994 as a founding element of the World Trade Organisation, represents the most important attempt to establish a global harmonisation of Intellectual Property protection. The aim of this article is to re-examine critically what has become the common wisdom around IPRs, TRIPS and their effects. We argue that supporters of IPRs in western corporations and governments as well as detractors in global movements and developing countries have overestimated their importance in the process of generation and diffusion of knowledge and innovation. On the basis of some key learned lessons on the nature of innovation and technological change, we assess four theses about TRIPS and its impact on the global generation and distribution of knowledge. Finally, the policy implications concerning international organisations and technological transfer are discussed.
Intellectual Property Rights (IPRs) have emerged as the key issue of global innovation policy: through the ‘Trade Related Aspects of Intellectual Property Rights' (TRIPS) Agreement, there is an attempt to impose worldwide a western system of IPRs. Western governments and large corporations claim that strong IPRs are needed to maintain investment in innovation. This position is contrasted by new political and social movements, which assert that muscular IPRs enforcement hampers economic growth and welfare in developing countries. The article argues that both positions overemphasise what IPRs can actually do to promote or obstruct innovation. IPRs per se do not allow companies to appropriate the returns from their innovations unless they are matched to a wide-ranging strategy that includes continuous learning and dynamic innovation. There are substantial cross-industry differences in the role played by IPRs: while patents are quite significant in pharmaceuticals and copyright is important in the audiovisual industry, the majority of sectors are not seriously affected by either strong or weak IPR regimes. In order to catch up, developing countries should put specific policies in place to nurture their absorptive capacity through the creation of appropriate infrastructure and human resources. Competence building is not hampered by IPRs. Developing countries should concentrate on active learning policies to acquire the knowledge of the most industrialised nations. Western nations would better protect their well-being by focusing on promoting new knowledge and creativity rather than by impeding new entrants from accessing the know-how they have already generated. Intellectual Property Rights (IPRs) have become ubiquitous in the economic debate: the front pages of newspapers continually report major controversies among corporations, governments and advocacy groups. News such as the copyright issue of the Google Books project and the power of the Big Pharma’s patents over key drugs and vaccines have generated growing alarms and heated disputes. Some books, including Naomi Klein’s No Logo and Vandana Shiva’s Patents: Myth and Reality have become best-sellers. Science fiction has been quick to report these concerns, as shown by Michael Crichton’s Next. National parliaments, the European Union (EU) and the North Atlantic Free Trade Association (NAFTA) are repeatedly addressing the issue. A brand new Pirate party, whose main political goal is to get free access to software and copyrighted products, has even managed to elect its own deputies at the European Parliament. Above all, IPRs have become one of the core businesses of the World Trade Organisation (WTO). In a word, IPRs have emerged as the key issue of global innovation policy.
The ‘Trade Related Aspects of Intellectual Property Rights’ (TRIPS) Agreement, signed in 1994, is a founding element of the WTO. TRIPS constitutes the most important attempt to establish a global harmonisation of Intellectual Property (IP) protection and enforcement, creating international standards for the protection of patents, copyrights, trademarks and design. It also provides a dispute settlement schema and establishes enforcement procedures at the intergovernmental level. Not surprisingly, TRIPS has been highly debated by political scientists (May, 2002a; Ryan, 1998; Sell, 2003) and economists (Maskus, 2000; Maskus and Reichman, 2005, among others). But TRIPS has also been debated outside academe and has been vigorously opposed by nongovernmental organisations and global movements (Drahos and Mayne, 2002). The aim of this article is to re-examine critically what has become the common wisdom around IPRs, TRIPS and their effects. We argue that there has been an overestimation of the importance of IPRs in the process of generation and diffusion of knowledge and innovation. For both developed and developing countries, the key issue should be an active innovation and learning policy rather than the protection through IPRs of the already available knowledge. The debate has instead been concentrated on IPRs as such rather than on knowledge generation and diffusion because some key learned lessons on the nature of innovation and technological change have not been duly taken into account. On the basis of these learned lessons, we will assert four theses about TRIPS. This article will mainly focus on patents, while it will deal less with copyright and other IPRs (for a comprehensive collection of essays on this issue see Hess and Ostrom, 2006; see also Macmillan, 2006 for a focus on copyrights). In the next section we provide the rationale behind the establishment of the IPRs systems across modern societies. Section 2 presents four learned lessons that emerge from the literature of the economics of innovation and technological change. In Section 3 we present four main theses on the globalisation of IPRs and their effects on the global generation and distribution of knowledge. The final section discusses policy implications.
The modern patent system, based on the objective assessment of inventions, was introduced by the Venetian Republic in 1474 (May, 2002b). The two requirements indicated by the Venetian Republic – the usefulness and novelty of the invention – are still in vigour today in all states. As in a Faustian bargain, the inventor and the government undertake a long-term pact: the inventor commits him/herself to disclose all information of his/her invention, while the government guarantees that it will provide legal protection to give exclusive rights on the economic returns of the invention (for a history of IPRs, see May and Sell, 2006).IPRs have evolved substantially over the centuries but the Faustian bargain has remained unchanged. By providing intellectual property rights, the government assures the inventor the right to exclude others from using the outcome of his/her creative activities without his/her authorisation. Thus the government gives the inventor a legal monopoly to exploit his/her invention and capture the economic benefits for a limited period of time. Legislation is far from uniform: for copyright the disclosure is complete by the moment you publish a book or a film, while inventions generally have to pass a merit exam before being granted a patent. As happens with many deals, this one is rarely fully implemented and the inventor often tries to hide as much as possible about his/her invention, while the government is not in a position to assure full appropriation of the returns of the invention. Through this deal, the government manages to disclose information on the already generated knowledge, and perhaps more importantly it provides an incentive to individuals to invest their time and resources in creative activities. Creative activities are in fact time consuming and costly while it is always uncertain if they will produce something that will generate economic returns. Once the inventor has discovered a new device or a musician has written a new symphony, it becomes easy for others to exploit their outcomes at very low costs. Without legal protection, inventors and authors are not in a position fully to exploit their works and appropriate the economic returns. As a result, in the absence of public regulation there would be an underinvestment in creative activities that would be below a socially desirable level. An IPR regime can be defined as the written and customary rules that apply within a specific political community. In some countries, the government enforces strong protection of IPRs and the holders are guaranteed that any infringements will be persecuted by the law and compensation will be obtained. These are the strong IPRs regimes. In other countries, the IPRs regime is much weaker and there is much less public interest in enforcing IPRs. Policing violation is much more relaxed and courts are slow and/or permissive towards infringement.
Scholars in the field of innovation have largely explored the mechanism that lies at the heart of the creation and diffusion of knowledge and innovation, as well as the mechanism of technological change. In this section we will briefly outline four learned lessons derived from this body of literature that are relevant to assessment of the current IPRs controversy.
Lesson 1: Knowledge is not information. Successful knowledge transfer is not only a matter of transferring information, but it requires learning through acquiring a wide range of competences, skills and tacit knowledge. There is a basic distinction between information and knowledge. Information is a good that is costly to produce but which by the moment it becomes public can be appropriated and transmitted at very low costs. Since all the costs are on the shoulders of the producers of information and there is no cost on the users, the lack of institutional protection would likely lead to an underinvestment in these activities (Arrow, 1962). But can the outcomes of the creative and innovative activities for which intellectual property is requested be considered information? The Schumpeterian tradition argues that creative and innovative activities are the product of human knowledge, which cannot be transferred to potential users unless they are willing to invest effort and time in learning. Knowledge is therefore rather different from information since no user will be in a position to gain economic advantages from it without an active learning effort and creative adaptive processes (Nelson and Winter, 1982, p. 60; Pavitt, 1987). But this is not the full story. There is another important aspect that makes the transfer of knowledge difficult to achieve: not all knowledge can be properly codified. An important component of knowledge is tacit (Polanyi, 1967) and not even those who have generated it are able to articulate it properly in manuals, blueprints, patents or other ‘codes’. A good cook is not necessarily able to transfer all his/her knowledge in a book of recipes. We are dealing with tacit knowledge when ‘we know more than what we can tell’ (Polanyi, 1967, p. 4). In order to transfer tacit knowledge to an apprentice, the coded component should be complemented with experimentation and training. A key characteristic of IPRs is that they can protect the codified knowledge but not the tacit knowledge. The picture is therefore more complex than it may appear in the first instance. On the one hand, the producers of knowledge have a wider battery of instruments to profit from it, ranging from selling the codified component through the IPRs system to transferring it through direct contact (for example through teaching programmes, technical cooperation and so on). On the other hand, those willing to acquire knowledge should also invest their resources not just to buy IPRs, but also to get the infrastructures and the skills that make it possible actually to use knowledge for economic purposes.
Lesson 2: Without imitating it is impossible to learn and innovate. The development of emerging economies is associated with creative imitation and absorption. Pablo Picasso once stated that ‘good artists copy, great artists steal’. This also applies to knowledge generation: innovation cannot be created in a vacuum but rather is bred in an environment of creative imitation. In the 19th century, Germany and the United States benefited from the knowledge developed in the United Kingdom. In the second half of the 20th century, all countries that successfully managed to catch up, including Japan, South Korea, Singapore and Taiwan, undertook a process of continuous adoption and imitation of technologies developed abroad. Today, China, India and Brazil, among others, are acquiring the knowledge developed in the Organisation for Economic Cooperation and Development (OECD) countries. Every ‘emerging economy’ at some point in its history has relied on the adoption of foreign technologies.2 Technology transfer is a multidimensional phenomenon and to be effective it should pay attention to the features of the host countries, including their stage of development, economic and industrial base, characteristics of the institutions and last and not least the IPRs regime. There is not a single channel that guarantees successful technology and knowledge transfer. On the contrary, each knowledge domain requires activating a variety of interconnected channels, including: (1) foreign direct investments, since the branches of multinational corporations in host countries often provide the most straightforward way to assimilate production methods of other countries; (2) joint ventures and strategic alliances, which allow companies of different countries to combine their skills, resources and expertise; (3) technology licensing, which includes not just the acquisition of IPRs but also technical assistance and training; and (4) technology embodied in imports, especially in the case of capital goods and equipment.
Lesson 3: Knowledge is not about plug and play. To adopt foreign technology is a costly activity requiring a big deliberate domestic effort. The availability of scientific knowledge and technology from abroad is only part of the story, but it is not as such sufficient to foster development driven by technological change. The other part of the story is the endogenous effort that catching-up countries should be willing to undertake. In order to make sense and exploit the spectrum of knowledge, competences and technologies coming from abroad, each country needs to develop an ‘absorptive capacity’, that is, the endogenous capacity to learn from these opportunities and to exploit them economically (Cohen and Levinthal, 1989, 1990). This reminds us that learning is not a fully passive process. Countries need to monitor the advances occurring abroad, to create internal competences able to adopt foreign technologies and put in place specific measures to encourage their diffusion in the whole economy. Creating such an environment requires a comprehensive effort ranging from public policies, education and human resources, a reliable legislative system and institutions, incentives and trade policies, sometimes referred to as developmental state. In the second half of the 20th century, western technology was equally available to Latin American and to East Asian countries, but only the latter countries have managed to acquire it to a point where they are able to compete on a par, mostly because of the active learning policies implemented (Wong, 2004; Woo-Cumings, 1999). National R&D investment to absorb foreign technology has been a crucial enabling factor for US economic growth during the 1900–1946 period (Mowery and Rosenberg, 1989) and for Japanese reconstruction during the postwar period (Morishima, 1982). This is often associated with the activities carried out by both business and the public sectors. Firms in catching-up countries need to invest in R&D, capital goods, equipment and infrastructures to develop the appropriate knowledge and capabilities to adopt foreign knowledge (Freeman, 1987; Hou and San, 1993; Mowery and Oxley, 1997; Westphal et al., 1985). In turn governments need to implement dedicated policies to foster public R&D and education as well as to support trade and foreign direct investment (Bell and Pavitt, 1997). Countries that have succeeded in catching up relied on a combination of devices to acquire foreign technology with policies aimed at building competences and skills internally.
Lesson 4: The ways to profit from innovation are infinite. IPRs are only one of the several tools in firms’ competition and are effective in a few industries only. Innovation is one of the most heterogeneous economic activities. A hairpin and a song, a jet engine and a statistical method, a drug and a machine tool could all be innovations. The heterogeneous nature of the phenomenon is equally reflected in the sources and methods employed to appropriate the returns provided by innovation activities. The methods to guarantee returns from innovations vary considerably across industries, markets and countries, and also evolve over time. The most effective way to appropriate the returns from innovation is by combining a battery of different strategies (Teece, 1986). The methods used by firms to appropriate returns from their innovations can be broadly divided into two large categories. The first category comprises the legal methods associated with IPRs. The second category is represented by other economic methods, which include industrial secrecy, lead time, differentiation and market distribution. The relative importance of the two categories varies considerably across technologies and industries. But a wealth of empirical studies has consistently shown that IPRs alone cannot guarantee the full appropriation of innovations.4 On the grounds of in-depth statistical surveys carried out at the firm level, it emerges that patents, the most important and controversial component in the IPRs family, are very important only in the pharmaceutical industry, and, to a lesser extent, in the chemical industry. But patents are not a key competitive factor in other high-tech industries such as electronics, telecommunication, motor vehicles and machinery. In some high-tech industries such as aerospace and nuclear energy, patents are not important at all since industrial secrecy is much more relevant. Profit-seeking firms implement other strategies to exploit innovation including lead time, moving down the learning curve, sales and service efforts and secrecy. In the computer industry firms are increasingly relying on new forms of IPRs such as the GNU General Public Licence (GPL) for the development of so-called open source software. Contrary to the traditional IPRs logic, the GPL allows anyone to use and modify the software. As a result, we have witnessed over the last decade the emergence of new business models in this industry that allow firms to profit from relying on new strategies based on combinations of ‘traditional’ and new forms of IPRs (Chesbrough, 2003; Lerner and Tirole, 2002; Tapscott and Williams, 2006). Other significant differences emerge according to the individual characteristics of new technologies and, as expected, patents are better able to protect product rather than process innovations. This does not necessarily imply that firms consider patents useless; it would be difficult otherwise to explain why firms bother to file, apply and pay for several hundred thousand patents every year. But even when they have a substantial patent portfolio, firms declare that they are not able to profit from them unless they combine legal protection with other economic instruments of appropriation. While there is abundant evidence on the manufacturing industry and on the patent system, there is less evidence on the appropriability system in the service industry and on the copyright (for a significant exception, see Blind et al., 2003). But the available evidence suggests that manufacturing and services on the one hand, and patents and copyrights on the other, have several similarities. A few industries rely strongly on copyright and are damaged by copyright infringement. Cartoon films, for example, appear to rely on strong copyright systems as much as the pharmaceutical industry relies on patents. But copyright enforcement alone cannot guarantee the full appropriation of returns unless it is combined with other economic instruments. Similarly, in the industrial design sector, firms do not rely so much on the effectiveness of design registration as a means to protect their new products (Filippetti, 2009). The innovator’s and the imitator’s perspectives are quite symmetrical. Strong regimes of IPRs make imitation slightly more costly and slower but not impossible since there is often a possibility to ‘invent around’, namely a set of activities through which a competitor can produce an effective functional substitute for the product protected by IPRs without infringing its legal rights. By contrast, weak regimes of IPRs make imitation cheaper and faster. But prospective imitators need to acquire a wide range of expertise that can be achieved through a substantial investment in building the knowledge base requested.
This chapter starts with a simple proposition: Intellectual property right (IPR) regimes are not established independently of what is generally perceived to be in the "national interest." That is nearly axiomatic, at least in countries that have reached a minimal level of political development. Defining the country's national interest usually falls to elected officials and policymakers. Although it is unlikely that every single government decision has the national interest as its governing criterion, the choice of major legal and developmental institutions should have (and, I suspect, normally has) this criterion at its core. Although policymakers do not conduct complex calculations when arriving at such decisions, such processes imply an exercise in social welfare maximization. As argued in this chapter, there are few grounds in terms of either national or global welfare (other than to minimize the prospects of economic conflict or retaliation) for all countries to abide by uniform IPR regimes; there are even fewer grounds for claiming that IPR regimes should be uniformly tight. At the most general level, there is certainly no reason for countries to share social welfare functions or preference orderings that would justify uniformity in levels of protection. The fact that enforcement costs and budgetary constraints are different makes the convergence of IPR systems still less meaningful, at least from a social welfare standpoint. IPR regimes should accommodate major structural shifts in the economy, particularly the progressive maturation of a production and innovation base in the country. Yet, as suggested below, changes in IPR systems should not necessarily be unidirectional or patterned after an individual model, namely that of countries at the property rights legislative and enforcement frontier (where moving the frontier forward implies increased levels of protection). Developed countries' regimes are responding to intense technological competition by attempting to maximize the returns from their technological assets on a global scale. Even in such countries, changes in IPR systems are far from consensual, and it is sometimes argued that protection may have reached excessive levels, to the detriment of diffusion and technological innovation. The purpose here is to discuss the relative merits of a differentiated versus a homogeneous IPR system. This chapter argues for IPR regimes that are differentiated according to the level of technological and productive competence, so as to support a country's ability to absorb, adapt, and generate technology. There is little in economic theory to support convergence of IPR systems on a cross-country basis, particularly if convergence means an increase in the level of protection in developing and industrializing countries. This is true either from an individual country standpoint or from a global welfare perspective. Furthermore, countries with large research endowments do not believe it is to their benefit to loosen their IPR standards, nor do industrializing countries see benefits (except for lesser threat of retaliation) in equating their legislation and enforcement practices with those of the country that is at the forefront of the movement toward tighter regimes, namely, the United States. This chapter suggests that it is unlikely that an industrializing country would be serving its own interest by copying the legislation and enforcement practices of a developed country. Both may want to follow or have a set of minimum standards as guidelines,2 but the definition of such standards should not be a mere reflection of the developed country's perception of what constitutes the minimum: rather it should be patterned after what is. In fact, in 1979 the World Intellectual Property Organization (WIPO) introduced a model patent law that was followed by a number of developing countries and constituted a minimum standard for patent-related IPRs. The international conventions, on the other hand, do not necessarily provide for uniformity in cross-country treatment of IPRs. The most well known The Paris Convention—signed by 99 countries (as of 1989), has as a major feature, equality of treatment between domestic and foreign patentees. The specifics of the systems are generally left for individual countries to decide practiced domestically. The next section discusses some of the economic reasons for differentiation in IPR regimes. It argues that differences in individual country characteristics, stage of development, and budgetary constraints make differentiated regimes superior from a domestic welfare perspective. This proposition holds as long as a country's insertion in the global economy is not affected by choice of the IPR regime. The following section examines the case when this assumption does not hold. It suggests that differentiated regimes may entail substantial costs for developing countries in terms of trade, investment, and technology transfer flows, given that most are relatively open economies. The last section concludes with a discussion of IPRs from a global welfare perspective.
There are substantial intercountry differences in IPR regimes on all key dimensions: the level and scope of protection, the mechanisms used, the strictness with which legislation is enforced, and the way violations are penalized. In the case of patents, their duration varies considerably (5 to 20 years), shorter terms being more commonly found in industrializing countries. With regard to the scope of patents, most countries have exclusionary rules, except probably for the United States and Canada (where only scientific principles, abstract theorems, and atomic weapons go unprotected).3 Working requirements and compulsory licensing are also widespread. The latter is permitted if patents go unworked for one to four years, with shorter periods in Latin American countries, moderate (around three years) in Australia and Southeast Asia, and longer in Europe. Again, no working requirements are present in the United States or Canada. Although there is greater agreement regarding what comes under copyright and trademark protection (the former generally comprises works of literary and artistic expression; the latter focuses on any symbol or message that serves to identify and confer reputation to a product or a firm), there still remain differences in the length of protection and, more important, in enforcement practices (including penalty levels for violators).
A WIPO review of major cross-country differences, as summarized by Estache (1990:15), reveals that for patents, ''Exemptions of coverage are more frequently found in less developed countries ... 49 countries exclude from patent protection pharmaceutical products, 45 exclude animal varieties, 44 exclude methods for treatment of human and animal body, 44 exclude plant varieties, 42 exclude biological processes for producing animal and plant varieties, 35 exclude food products, 32 exclude computer programs, 22 exclude chemical products, 14 exclude nuclear inventions, 10 exclude pharmaceutical processes, 9 exclude food processes, 9 exclude microorganisms and 7 exclude substances obtained by microbiological processes." In enforcement are also a critical barrier for upholding the rights of firms relying on trade secrets in countries that recognize the legitimacy of this mechanism (which many do not, on the basis that the lack of disclosure disqualifies proprietary and closely held information from being protected by the IPR regime). The differences described above are at the core of major conflicts in international and bilateral forums. Much of the discussion has centered on patents and the protectability of a group of products comprising, until recently, foodstuff, chemicals (including fertilizers and insecticides), pharmaceuticals, and their manufacturing processes. A number of developing countries have argued against protecting these product groups due to their perceived importance in fulfilling the "basic needs" of the population, and on the grounds that countries should not become hostage to firms from developed countries by granting legal rights (patents) that allow for monopolistic pricing practices. The affected firms, however, view the exclusion of certain areas from patentability as attempts to free ride and subtract from their profits. More recently, discussion on the scope of protection has been extended to new areas, such as integrated circuits, computer software, and biotechnological products and processes, including not only microorganisms but bioengineered (transgenic) plants and animals as well. In biotechnology, positions differ not only between developing and industrializing countries, but among the latter as well. How are those substantial intercountry differences explained? They are certainly inconsistent with the concept that a uniform system is somehow economically superior, unless countries were acting against their self-interest. Yet it is unlikely that most countries act most of the time against their interests. It is true that certain types of legislation, policies, and bureaucratic procedures or practices often generate groups of "special interests" that help perpetuate or shape existing institutional and policy arrangements that are not in the country's interest. This phenomenon of "capture" is well recognized in the literature and is probably present to some extent in most countries [e.g., see McConnell (1966)].5 Yet why specifically IPR regimes that do not conform to the paradigmatic case (as defined by developed countries) are the ones resulting from bureaucratic or legislative failure is far from clear. Most countries, for example, with the exception of the United States and Japan, do not allow patenting of transgenic animals (except microorganisms), although the appellate committee of the European Patent Office (EPO) was reviewing an EPO decision to grant patents to such animals in late 1990. See Barreto de Castro (1991:3). The theory of capture can be considered closely related to the economic theory of regulation, the central thesis of which is that "regulation is acquired by the industry and is designed and operated primarily for its benefits." See Stigler, 1971. What could be argued is that required changes in legal institutions or policy regimes only happen slowly, because of entrenched interests. Further, it could be argued that in the case of countries that are not so politically developed, where debate is thin and acts of government are opaque, changes occur arbitrarily and without strict correspondence to the national interest. Nonetheless, from such considerations it does not follow that such countries should pursue or pattern their legislation after a single IPR model. Many countries may in fact have IPR legislation on the books and engage in practices that are outdated, reflecting the needs and capabilities of an earlier stage of development; a case could be built that they should migrate toward other models of protection. Yet economic reasoning does not seem to suggest the superiority of a uniform model. In fact it even raises questions about the need for protection in the first place and points to its adverse welfare implications, particularly for developing countries.
Although the economic rationale of patents and other property rights instruments is disputable on theoretical grounds, countries with the most productive innovation systems have fairly comprehensive IPR regimes. In some areas (such as chemicals, pharmaceuticals, or scientific instruments), where innovations are costly to generate but not as hard to imitate, a strong patent regime (or other form of protection) is probably quite necessary for firms to be rewarded for their innovative efforts. Yet the need for patents or alternative forms of protection is most clearly justified in the presence of domestic research capabilities. The presumption in this case would be that the missing link to innovation was a structure of incentives strong enough to mobilize this potential. What if the latter is missing, and the country lacks even the basic endowments to progressively build a production base in the relevant area (as in the case of very poor countries)? Then the presence of a patent or other system of protection would have little effect either way. The country would still pay for the results of research undertaken elsewhere in the form of imports. It is indeed suggestive that many poor countries abide by fairly tight IPR systems (when one considers that it See Primo Braga (1990:32) for an excellent and succinct review). He notes that "overall, economic theory has raised more questions about welfare implications of intellectual property than it has answered. The theory of intellectual property protection is fragmented and provides no robust answer to the questions of appropriate or optimal level of protection under various sets of real world circumstances. In particular, its relevance to developing country concerns must be considered marginal." may help minimize frictions with major development aid donors or trading partners, while having no major adverse effects domestically). When research capabilities are weak but the country has a substantial productive potential, capable of copying, reverse engineering, adapting, and transforming foreign innovations into marketable products domestically, then there is a strong economic incentive to produce on the basis of foreign innovations. In this case, the country would be better off by disallowing or weakening protection, on the basis that not to do so would lead to a net income (welfare) loss in the form of royalties and rents transferred abroad. Moreover, lower levels of protection would facilitate entry, drive down prices and excess profits, and maximize diffusion rates. It is not by coincidence that countries that were ranked (in 1988) as having the most inadequate IPR regimes were composed solely of those commanding significant industrial capabilities (Brazil, Taiwan, Mexico, Korea, India, China, and others; see International Trade Commission, 1988). There are two problems with this reasoning. First, it is assumed that the country is immune to retaliation and, more generally, that trade and investment flows are not affected by a country's domestic policy choices (a discussion that is taken up below). Second, it is assumed that the absence of protection has no impact on domestic research, to the extent that research capabilities are absent. In fact, rarely is productive capacity totally dissociated from research endeavors. For research itself is an activity characterized by a continuum of subactivities in which producers engage and routinize to the extent of their needs. This set of subactivities can be conceptualized as originating from adaptive efforts: scaling down imported processes and making them consistent with the use of local raw materials; changing product characteristics, including redesign to conform with the local environment; improving productivity through minor innovations and controlling for product quality; replicating and progressively mastering all aspects of the existing technology; and finally, engaging in formal and structured R&D, while engineering the processes to bring new products to the market. Thus, when choosing levels of protection, countries with a productive base face the following closed-economy trade-off: how to establish a structure of incentives (in terms of levels of protection) that will stimulate domestic technological efforts, however frail and tentative, without curtailing productive activity dependent on the use of foreign innovations. It is assumed that these are innovations that can be reverse engineered or copied by other means and that, if patentable, would not be licensed to domestic producers, short of substantial additional costs. When competence in production and research is insignificant, whether or not to protect is immaterial. The country should follow whatever system minimizes.
It appears that for industrializing countries, no protection is not an optimal choice; nor is uniformity the answer. Even with the decision to establish a positive level of protection on economic welfare grounds, the object of protection and the corresponding mechanism are not given ex ante. What then should be the scope and mechanism of protection? How should IPR regimes be tailored to individual sectors or particular technologies, consistent with the stage of development of the country or its particular circumstances? It is important to stress that appropriate instruments of protection already exist. Defining the scope of protection and pairing technologies or product groups with the right instrument would depend on country-specific circumstances. To illustrate, take the case of mechanical technologies, the mastery of which is critical for countries to move beyond the initial stages of industrialization. A particularly appropriate system to stimulate minor inventions or improvements of a mechanical nature (where an "inventive step" is absent) would be utility models or petty patents (as used by Germany, Japan, South Korea, and Brazil, among others). These instruments protect minor innovations, generally obtained as producers attempt to improve the productivity of machinery. Protection levels are lower than they would have been if such innovations were patentable by regular means, but are nonetheless significant (as evidenced by large numbers of utility models patented in countries that allow for such a mechanism). In the biological sphere, an interesting case is that of plant and seed varieties. Most developing countries leave them unprotected, because they are part of the food chain and thus perceived as fulfilling basic needs. Among those countries that consider plants and seed varieties legitimate objects of intellectual protection, most have significant research capabilities in plant genetics and utilize plant breeders' rights (PBRs) as the basis of protection (e.g., South Korea, Argentina, and European nations). The United States makes use of PBRs or regular patents, depending on whether the plants are reproduced sexually or asexually. The advantage of PBRs is that they do not constitute an impediment for farmers or those engaged in research to reproduce protected varieties for their own use. PBRs do require commercial seed producers to pay royalties to breeders. In countries where agricultural production is still concentrated among small and medium farms, such an arrangement might be superior to traditional patents, which suggests again that the particular protective arrangement should reflect country circumstances. For chemicals, pharmaceuticals, and biotechnological inventions, there is the choice of protecting either products, processes, or both. For industrializing countries, it is arguable that the patentability of products might precede that of processes, being more consistent with their small to moderate research capabilities in the engineering of processes, but only marginal ability to introduce new products to the market. The continuing inability of most countries to bring out new products suggests that more or even full protection of foreign product innovations in these areas will not have a detrimental effect, except on those firms that basically copy locally unpatentable products (but have failed so far to develop research capabilities of any significance). To the contrary, greater protection might actually benefit local research institutes specializing in areas that have not attracted the attention of major international producers. Full protection might also be appropriate in the biotechnology area, where research and production are intertwined, and the lag between developing countries and those working at the research frontier is quite large and growing. The presumption so far has been that most research activities take place outside the country and that local producers are capable of appropriating the marketable results of foreign inventions. A different perspective is offered if it is posited that certain innovations cannot be copied because they are kept as unreachable trade secrets, or because they belong to areas that do not command sufficient interest for developed country institutions to allocate significant R&D resources. In this case, protection (however strict) causes no harm, by definition. The country may in fact want to target product groups or technologies in which it is of particular interest to provide additional incentives, by establishing higher levels of protection. An illustrative example would be of that of certain drugs for tropical diseases, for which the market in developed countries is small. Granting full and strict protection could only be to the advantage of the developing country; in fact, once again, the prevailing system in the developed country is quite immaterial from the developing country perspective. The latter may find its interest is to grant higher levels of protection, commensurate with the potential demand for the product (and the inability of local researchers to bring forth the innovation).7 Thus, to the extent that countries have different needs or technological preferences, the free-riding motive in a developing country will compete with the incentive requirements for generating appropriate technologies, and it may be in the country's interest to target protection for such technologies at levels higher than those prevailing in the developed country. Clearly, however, protection should be regarded as only one of the stimuli for innovation. Again, in the development of a new class of drugs for tropical diseases, potential demand may not materialize due to low income levels of those affected. Creating such demand may require a public health program with a strong drug procurement component or, alternatively, well-designed research contracts or a system of prizes. The more general point is that an IPR system should not be regarded as some deus ex machina of innovation; granting full and strict rights may under certain circumstances be decisive and, under others, only an accessory factor.
Once the methods of protection and product coverage are defined, there is still the need to establish the exact configuration of the system of protection. Patent life is a case in point. It has been shown, for example, that length of protection for a given product should be inversely related to the elasticity of demand and the social rate of discount, and positively related to R&D returns.8 It is unlikely that markets in different countries with divergent levels of income and preferences (among other factors) would have similar elasticities or that discount rates or the productivity of the R&D system would be very close in dissimilar societies. In this sense, strict equality in the duration of patents on a cross-country (or a cross-industry basis) would not be justified. Still uniformity might be called for in view of the difficulties of estimating with great precision the parameters that are supposed to determine this point has been elaborated, from a theoretical perspective, in Diwan and Rodrik (1989). A similar point was made, in the discussion of the Brazilian experience, by Frischtak (1989). Optimal patent life (demand elasticities, etc.), and the fact that in the Nordhaus model (among others), "beyond a certain number of years [of patent protection]—usually ten or less—the welfare provided by the patent system cannot be altered significantly" (Nogues, 1990:9). Thus, insofar as "the costs imposed by standardized patent terms are not very significant," patent life should be uniform among industries and technologies at least for the sake of simplifying administrative procedures (see Primo Braga, 1990:32). Yet cross-country uniformity does not necessarily follow from these propositions—much less, a clear justification for adopting a particular patent life standard. From the point of view of the country's welfare, patent life must be made to reflect the degree of industrial and research maturity of the country, and the underlying trade-off between production/diffusion and research incentives. The shorter the time of protection, the sooner will local producers be able to copy and start up (again, on the presumption that they have the necessary imitative capabilities), the faster prices will drop, and ceteris paribus, the quicker the pace of diffusion will be. A similar argument can be made regarding the expiration of patents (through working requirements) or their compulsory licensing to competitors. In all cases, policymakers will be trading off production for research incentives as levels of protection decrease. As always, the probability that two countries will have similar indifference curves and "budget lines" is small; so it would be inadequate for countries to choose similar protection parameters.
Enforcement is a resource-intensive activity. The higher the standards, the larger the resources allocated must be to achieve a given level of compliance, or alternatively, the higher the risks of noncompliance must be. To the extent that enforcement costs are probably larger and budget constraints tighter in developing countries, laws (including those on IPRs), judicial institutions, and practices should naturally diverge from those of developed countries (if countries are attempting to allocate resources efficiently, equalizing the returns at the margin) (Estache, 1990:69 and passim). In sum, it can be argued that cross-country differences in protection levels are justified in view of differences in enforcement costs and available resources.
So far the prevailing assumption has been that trade, investment, and technology flows are invariant with respect to the choice of IPR system. Yet that is hardly a tenable assumption anymore. Except for countries with marginal production systems, all others are being forced to follow fairly high standards of protection as their basic orientation when choosing a particular IPR model. Many industrializing countries (e.g., South Korea, Taiwan, Mexico) have amended their legislation or introduced new laws in the last five years reflecting growing pressures from foreign governments and firms. Two basic threats are driving these changes: that of retaliation, spearheaded by the United States through its own trade legislation (particularly the "super 301") and through the inclusion of IPRs in the General Agreement on Tariffs and Trade framework; and the one posed by foreign firms that refuse to deal (invest or sometimes even trade) in the absence of strong assurances that their proprietary information will not be appropriated without consent. These firms tend to concentrate in research-intensive sectors. The fundamental trade-off that policymakers face in the choice of an optimal IPR regime in the current international environment is simple to state: closed-economy considerations of potential gains from an IPR regime tailored to the country's circumstances may have to be balanced against open-economy income losses from reduced levels of trade and investment, if the choice of IPR regime is inconsistent with the trading/investment partners' interest. Pressure from trade and investment partners is finding less resistance among developing countries with relatively mature production and innovation systems. Additionally, industrializing countries are being induced to change their IPR regimes due to changes in international economic relations (with globalization of production) and an acceleration in the pace of technical progress. Open-economy considerations are becoming more important in policymakers' decisions insofar as the welfare of countries increasingly hinges on their participation in international flows of trade and technology. As the composition of investment and trade moves away from simple commodities toward higher-value-added, more research-intensive goods, the protection of intellectual property grows in importance. The existence of an IPR regime that is similar to those in countries that have a dominant economic position signals to these countries that the industrializing country is a "trustworthy" partner insofar as protection of intellectual property is concerned. This is in fact a critical consideration, for until quite recently, the nature of the IPR regime was either absent or quite hidden in the calculus of trade and investment decisions.9 Even among more traditionally IPR-dependent producers, such as pharmaceuticals firms, it was hard to perceive an noteworthy that throughout the 1970s and early 1980s, foreign investment flows by U.S.-based multinational firms appeared to be quite unaffected in both Brazil and Mexico by their IPR regimes, despite the countries' being rated as having the most "inadequate" regimes according to U.S. foreign investors. Association between the nature of the regime and the intensity of commercial transactions (e.g., the volume of trade, disembodied technology transferred, or investment). In case of investment decisions, in particular, the fundamental considerations were market size and growth prospects, factor supply and costs, and macroeconomic and political stability (country risk). The regulatory regime (investment licensing, foreign remission restrictions, price controls, IPRs) was of secondary importance. Yet that seems to be changing. A combination of greater competition among countries for foreign investment, and considerable shifts in perception regarding the importance of protecting proprietary information due to intensifying technological competition, suggests that IPRs are being brought to the forefront of decisions. Most concerns are still concentrated among the more IPR-sensitive producers (chemicals, pharmaceuticals, biotechnology, scientific instruments, and microelectronics products). An Organization for Economic Cooperation and Development (1987: Table 40) survey, for example, shows that inadequate IPR protection was the greatest disincentive to technology licensing, together with foreign exchange controls and government approval regulations. To the extent that licensing is a form of transaction in which firms have the least control over their technologies, and are therefore most sensitive to IPR issues, it is a "leading indicator" of firms' concerns over IPRs in direct investment and trade. how in an open economy uniform and differentiated IPR systems compare. In the table, each cell specifies the direction and intensity of impact. Thus, for example, in the case of threat of retaliation, the impact of a differentiated system is "adverse"; that is, such a system increases the threat from trade, investment, and technology transfer partner countries, and does so to a "moderate to high" degree, whereas in the case of an undifferentiated regime the impact is "positive" (i.e., the threat is removed) and "significant." The situation is similar for a country's capacity to react, or adapt to changes in technology.
TO SPEAK ABOUT SYSTEMS: A differentiated system, by definition, would allow countries to change their IPR regime as they mature technologically or the technological frontier changes: thus the impact is ''positive and significant." Conversely, for an undifferentiated regime: countries committed to it would be "adversely" affected in a "significant" way by their inability to modify the IPR rules of the game to suit shifts in their technological endowments (e.g., tighten those rules as they progressively accumulate such endowments) or to adapt those rules to exogenous technological changes. How can developing countries respond sensibly to these new forces? An illustration of the complexities involved in deciding about the different trade-offs in an open economy comes from plant breeding in Brazil. In that country, universities and government research institutions have become quite proficient in breeding new and better varieties through classical genetics methods. Most food seed varieties currently used by farmers have been produced by local institutions. Although it has a strong tradition in plant breeding through classical genetics, Brazilian research in genetic engineering is still incipient. With the advent of biotechnologically engineered plant varieties (more than 300 were being tried out in 1990), the country will have to rely to a growing extent on the research results of foreign biotechnology firms. Yet in the absence of protection, foreign investments in the area are confined to firms producing hybrids (such as corn) that are naturally (genetically) protected through the control of lineages. The inability to protect plant varieties by classical or genetically engineered methods appears to have had an overall detrimental effect on local research institutions. Products of their research can be used without compensation. Researchers and local institutions are thus being deprived of an important source of income (in the form of royalties) at a time when the government, due to a chronic fiscal imbalance, has cut public sector wages and budgetary allocations to its agencies, including those engaged in agricultural research. Even more important, domestic researchers cannot interact with their foreign counterparts engaged in biotechnology research: for one, they have nothing "to offer" that is not already freely available and major research institutions in plant genetics have been the Instituto Agronomico de Campinas, Escola Superior de Agricultura Luiz de Queiroz, the Federal University of Vicosa, and Embrapa (Empresa Brasileira de Pesquisa Agropecuaria). despite the potential of combining the two techniques (classical genetics and biotechnology), they will not profit from it. For another, foreign firms, unable to license their biotechnology inventions, will not be willing to make them accessible to local firms and researchers, and will introduce varieties engineered in ways that cannot be copied. In a number of instances, genetically engineered (transgenic) plants may be superior and would substitute for those obtained by classical methods.13 In many others, however, commercial success will be attained by firms combining the two technologies. Access to the innovations of genetic engineering firms (particularly synthetic genes and their expression vectors) would be critical for applying them to genetically improved plants, thus leveraging much of classical genetics work. Granting some form of protection to local breeders—such as joining the International Convention for the Protection of New Plant Varieties (UPOV)14—will allow them to discuss commercial terms and exchange rights with foreign genetic engineering firms, particularly those of small and medium size that would be interested in penetrating a potentially large market for their products. For biotechnology firms, some form of protection (possibly patents) will be required, which at least opens up the possibility of licensing biotechnology products and processes. Cross-licensing and similar arrangements would be natural solutions for research units that lack each other's skills. The absence of protection in this case clearly goes against the national interest. Large international firms in the seed production business, in association with biotechnology firms, will continue to sell in the Brazilian market either through naturally protected hybrids or by combining the two technologies. Thus, they will be introducing new genes and their vectors of expression into varieties that have been developed by local institutions through classical methods and over a considerable period of time, profiting from the results without having to share their profits with those partially responsible for the innovation. The particular form of protection that Brazil should grant to breeders will have to be evaluated carefully; as already suggested, PBRs may have Take the case of delayed ripening of fruits, flowers, and vegetables through genetic manipulation. Researchers have been able to interrupt the natural production of ethylene, the gas responsible for ripening, by altering the gene responsible for releasing the key enzyme in the process—the ACC synthase. Scientists were able to rebuild the RNA of this gene, and used a bacteria to reintroduce the gene into a tomato, the object of the experiment. As a result, 99.5 percent of the production of the gas was blocked for an entire year. Needless to say, the implications of such technique are most significant: perishables can be transported over long distances without any type of refrigeration; more generally, storage losses will be reduced drastically without the need of radiation or agrotoxics for preservation purposes. This may be proposed shortly to the Brazilian Congress, possibly in the context of major reforms of the IPR system. superior features to patents.15 In the case of genetically engineered processes and products, the choice will focus less on the system itself (patents) than on its specifics (term, rules on compulsory licensing, penalties for violations, and disputes settlements). Biotechnology falls within the category of still thin research and production capabilities: moderately stringent rules of protection are unlikely to have adverse effects. In the fundamentally important area of plant variety improvements, strengthening IPRs in biotechnology needs to be accompanied by increased protection in the area of classical plant genetics.
This chapter has argued the proposition that countries should tailor their IPR systems by taking into account their economic needs, productive and research capabilities, and institutional and budgetary constraints. In addition, they should consider how the choice of IPR regime would affect their international economic transactions. Would this approach be at cross-purposes with global welfare maximization? Before attempting to answer this question, it is useful to take a short detour and note the welfare implications of patent protection in a "North-South" context, usually modeled by having research capabilities concentrated in northern firms, while those in the South are able to appropriate research results without cost. In this class of models, free riding by the South generally improves its welfare and correspondingly diminishes that of the North whose firms, of course, always benefit from having their patents recognized in other countries. Note that this result holds as long as the South is a small part of the world market for the good subject to improvement, so that free riding by the South does not constitute a major disincentive to innovators and that cost savings associated with R&D are not substantial; if they were substantial, the gains to the South from additional R&D undertaken by northern firms in response to a strengthening of the South's IPR regime would more than compensate its income losses. Moreover, if the technological preferences of southern consumers are significantly different from those of the North (i.e., if their needs are quite specific, in terms of disease-fighting drugs, for Countries joining the UPOV system recognize the rights of breeders to a special title of protection for a specific plant variety, with the breeder required to authorize the commercial utilization of the variety. A title is awarded if the breeder can both describe genetically the new variety and show that it is characterized by homogeneity and genetic stability example), the welfare gains from extending patent protection to the South may again outweigh income losses to northern firms (in terms of fees, royalties, etc.). When most southern needs must be satisfied by innovations specifically targeting the South's preferences, it may even be in the interest of the South to have a stronger patent protection system than the North, so as to reward R&D efforts targeting smaller and less profitable markets. In this latter case, the trade-off facing the South would be between free riding and stimulating such innovations (Diwan and Rodrik, 1989:9-14). What are the optimal levels of protection from a global welfare perspective? This question can be answered rigorously only at a fairly abstract level and is not often addressed. One of the few papers that does attempt to answer it shows that if one were to value developing countries' welfare gains more highly than those of developed economies (out of an egalitarian concern), then the exact level of protection that maximizes global welfare is indeterminate. Yet numerical simulations suggest that the greater the weight attached to the welfare of developing countries, the lower their level of protection should be (Diwan and Rodrik, 1989:18-19). With a sufficiently large weight, developing countries should be allowed to free ride. Yet just as before, when the welfare of the North and the South were considered separately, these results would be overturned, depending on how specific poorer countries' needs are compared to those countries in which most innovating firms are located. The more distinct their needs or preferences are, the more would global welfare (as well as their own) be enhanced by higher levels of protection in their economies. Finally, in the case of a utilitarian social welfare function, with the welfare of all countries valued equally, one can infer from the model that patent protection should be uniform for global optimality (Diwan and Rodrik, 1989:14-19). The reason is that countries' sizes determine both their relative welfare weights and their importance to innovating firms (i.e., the potential profitability of their markets). To maximize the flow of innovation and aggregate welfare, all countries, developed and developing, should therefore "contribute" to stimulate R&D in direct proportion to their size, which calls for uniform rates of protection. These theoretical or simulation results, derived from recent models examining the welfare economics of patent protection in North-South contexts, do not suggest that global welfare would be increased by a uniform system, except in the case where the welfare of all countries is valued equally and distribution issues are mute. In this latter case, global welfare would improve by having some countries increase their levels of protection to a common minimum. Yet how should this be effected? Clearly, if that were to go against their national interests, they would have to be either bribed (offered appropriate compensation) or pressured. Although proponents of a uniformly tighter IPR system on the basis of global economic well-being have yet to suggest compensating losers, even if they did, one would still have to find the appropriate mechanisms and instruments to effect such transfers. These could involve developed country incentives for corporations to relocate research facilities, or at least some of their activities, to developing countries; or for firms to engage in research projects that are particularly valuable to those countries and to license the results at favorable fees. All this presumes that a move toward greater uniformity on the basis of protection parameters inspired by the more advanced countries would be Pareto superior, that is, making all countries better off and no country worse off after a suitable redistribution of income or endowments. Yet that itself is questionable. The links between IPR protection and innovation have yet to be shown to be on a scale that would justify such social engineering efforts in the name of global welfare maximization. There is no question that firms in developed countries lose potential income by having competitors in industrializing countries produce the fruits of their inventiveness. Although the scale at which this happens may be growing, it is doubtful that it justifies either a global redesign of the system or the political capital being spent on pressuring weaker countries to conform to more uniform and tighter protection standards. The costs of changing the system of protection according to a purely ethical perspective, by taking into account individual countries' needs and relative endowments, would probably be substantial as well. If a Rawlsian criterion of justice were adopted, the objective would be to provide all countries access to innovations to fulfill their basic needs at a cost consistent with their incomes; all other innovations would be licensed on a "fullcost" basis (Rawls, 1971). Needless to say, this would require, in addition to differential rates of protection, a complex mechanism of transfers to stimulate and compensate producers of the basic needs-related innovations. Just as in the case of attempts to mold all countries to a uniform system of protection, it is likely that the cost of implementing such changes would outweigh their benefits.
