Policy Discussion Paper

Transcription

1 Policy Discussion Paper No. 97/10 University of Adelaide Adelaide SA 5005 Australia IMPLICATIONS OF REGIONAL AND MULTILATERAL AGREEMENTS FOR INTELLECTUAL PROPERTY RIGHTS Keith E. Maskus October 1997

2 CIES POLICY DISCUSSION PAPER 97/10 IMPLICATIONS OF REGIONAL AND MULTILATERAL AGREEMENTS FOR INTELLECTUAL PROPERTY RIGHTS Keith E. Maskus Department of Economics Campus Box 256 University of Colorado Boulder, CO USA ph: fax: October 1997 Forthcoming in The World Economy 20(5), September ii

3 SUMMARY Over the last decade there has been a marked rise in international interest in the protection of intellectual property rights (IPRs) in different nations and the impacts of IPRs on competition, innovation, and international technology transfer. This paper focuses on one understudied aspect of the issue. Specifically, it compares the effectiveness of regional trade agreements, most specifically the North American Free Trade Agreement (NAFTA), and the multilateral approach within the World Trade Organisation (WTO) in fostering stronger and more harmonised levels of intellectual property protection. Preferential trade agreements could generate stronger IPRs within regionally integrated markets than in the global trade regime. Accordingly, the paper also considers some implications of regional harmonisation of IPRs for economic welfare and the trading system. The first task is to compare the WTO and NAFTA agreements on intellectual property in order to illustrate economic issues surrounding the multilateral and regional approaches and the advantages and drawbacks to each. On the multilateral scale, the WTO Agreement on Trade-Related Intellectual Property Rights (TRIPs) is a watershed in global protection of intellectual property. Before its implementation, each country's IPRs regime was largely a matter of choice. However, the inconsistency between the existence of different national IPRs regimes and the need for international exploitation of the gains from innovation through trade and FDI had become well-recognised by the mid-1980s. The TRIPs accord commits all developing countries in the WTO to undertake a significant extension of IPRs protection, albeit with a long phase-in period for the poorest countries, thereby bringing about a substantive convergence in international minimum standards. Despite some shortcomings in the standards and commitments of the TRIPs agreement, at least from the view of technology and product developers in the United States, it represents a substantial strengthening of global IPRs systems. On its own merits, the TRIPs accord could have negative impacts on many nations, at least in the short run, because of higher prices associated with newly protected products. It is interesting to ask, therefore, why such a comprehensive agreement commanded widespread adherence in the Uruguay Round. The most relevant answer is that the Uruguay Round agreements provided enough collateral benefits to developing countries, particularly in terms of market access for agricultural goods, textiles and apparel, along with greater teeth to dispute settlement, that it was worth accepting the short-term (though uncertain) losses from TRIPs. This ability to trade concessions across disparate interests is a key advantage of the multilateral approach. Further, a multilateral approach based on minimum standards and disciplines is preferable to competition among these countries in terms of IPRs to attract FDI and technology. This last observation provides a further powerful argument on behalf of a multilateral approach. The NAFTA agreement on IPRs is stronger than TRIPs in some important ways, though there remain significant exclusions. The successful installation in Mexico of strong IPRs, along with changes in Canadian compulsory-licensing practices, points to one important advantage of the regional approach over the multilateral approach. In the former, it is easier to achieve broad consensus over harmonised standards because of greater commonality of interests associated with regional market integration. In the latter, (at least perceived) economic interests continue to diverge widely and it would have been extremely difficult to procure an intellectual property agreement as strong as TRIPs without the broad potential for trading off market and policy concessions across numerous economic activities. Another salient distinction between NAFTA and TRIPs is that the former agreement provides less leeway to participants in setting national implementation strategies. For example, Mexico would find it more difficult to adopt strong fair-use exemptions and compulsory-licensing procedures under iii

4 NAFTA than under TRIPs. In this important sense the regional approach generally may be counted on to erect stronger minimum standards for protecting IPRs. The paper also reviews recent empirical evidence on the importance of IPRs for international commerce. For example, studies have demonstrated that international variations in the strength of patent laws do have an impact on trade flows, with stronger patents associated with rising imports. Extending one methodology to the case of U.S. exports to Mexico, if the result of NAFTA is to improve Mexican patent protection to the American level, we might expect an increase in trade of perhaps $3 billion per year (in 1984 dollars). There is also tentative empirical evidence that foreign direct investment is attracted by strong IPRs. Again, one methodology suggests that if the IPRs agreement in NAFTA raises Mexican patent protection to U.S. levels, FDI would rise from the United States to Mexico on the order of $893 million, or approximately 22% of 1989 assets. In normative terms, IPRs are inherently second-best policies within each country and it is difficult to specify workable rules that necessarily satisfy conditions for welfare optimality. Rules of thumb, such as providing copyright protection to software but leaving open the possibility of patent protection under particular circumstances, may in some cases be too weak and in other cases too strong. Extended to the international context, these ambiguities are magnified. In terms of economic theory, it is unclear whether TRIPs and NAFTA will be globally pro-competitive or anti-competitive, while questions of international distributional impacts are important. Finally, the article comments on some important differences in the potential international economic effects of multilateral and regional harmonisation of IPRs. The issues relate to the inherently second-best character of both regional trade preferences and IPRs. For example, the essence of regional trade agreements is to provide preferential treatment in commercial policy. Drawing a page from the economic theory of economic integration, define the concepts of intellectual property creation (IPC) and intellectual property diversion (IPD). The former would exist to the extent that more intellectual property is created and marketed within a regional market, such as NAFTA, strictly due to the expansion of markets and the replacement of inefficient technology developers with more efficient ones. The latter would occur to the extent that less intellectual property owned by non-member firms is exploited within the free trade area. These effects are possible even if the intellectual property standards within a free trade area are non-discriminatory. In a related vein, greater harmonisation of both regional and global IPRs standards does not necessarily represent a movement toward less effective variability in world practices. A final, and related, observation is that the trend toward stronger IPRs, within both the WTO and certain regional agreements, has sometimes come with insufficient regard for changes in important complementary regulatory policies. Chief among these is competition policy. This review suggests that, while there are reasons to anticipate net efficiency gains from the new international IPRs systems, the significant differences remaining between TRIPs standards and those emerging under regional trade agreements pose new questions about the role of deep integration in the international regulation of intellectual property use. iv

5 Implications of Regional and Multilateral Agreements for Intellectual Property Rights Keith Maskus 1. Introduction Over the last decade there has been a marked rise in international interest in the protection of intellectual property rights (IPRs) in different nations and the impacts of IPRs on competition, innovation, and international technology transfer. In this paper I focus on one understudied aspect of the issue. Specifically, I compare the effectiveness of regional trade agreements, most specifically the North American Free Trade Agreement (NAFTA), and the multilateral approach within the World Trade Organization (WTO) in fostering stronger and more harmonized levels of intellectual property protection. Preferential trade agreements could generate stronger IPRs within regionally integrated markets than in the global trade regime. Accordingly, I consider some implications of regional harmonization of IPRs for economic welfare and the trading system. In the next section I compare the WTO and NAFTA agreements on intellectual property in order to illustrate economic issues surrounding the multilateral and regional approaches and the advantages and drawbacks to each. In the following section I discuss the general question of global protection for intellectual property and note some recent empirical evidence on the importance of such regulations for international commerce. In the fourth section I comment on some important differences in the potential international economic effects of multilateral and regional harmonization of IPRs. Concluding observations are provided in the final section. 2. Multilateralism and Regionalism in Protecting IPRs The successes that advocates of stronger global IPRs have experienced in recent years are striking. Keep in mind that only twelve years ago the Uruguay Round had not begun, no considerations were under way of folding Mexico into a trade agreement, and pressures on developing countries to change their IPRs practices were only beginning. Yet numerous countries have undertaken significant upgrading of their IPRs systems since For example, the list of nations that adopted stronger patent legislation over that period includes Canada, New 1

6 Zealand, Portugal, Argentina, Brazil, Colombia, Indonesia, Republic of Korea, Malaysia, Mexico, Thailand, Bangladesh, Benin, Burkina Faso, Chile, Ecuador, Paraguay, Hungary, Mali, Egypt, and China (twice). 1 To some unknown extent, these policy shifts reflect both domestic emerging economic interests in these countries and strong external pressures exerted by the United States and the European Union. From the United States such actions range simply from mentioning the countries in the annual USTR National Trade Estimates Report to designating them as a "priority watch" country or the subject of actual negotiations on IPRs. A number of nations, however, did not change their laws in response to this attention (India, Pakistan, and Philippines are prominent examples). Thus, there has been considerable room for broader multilateral and regional negotiating approaches. For example, it is fair to claim that Canada and Mexico adopted stronger laws under the influence of NAFTA, while Egypt's efforts were evidently taken in the context of ongoing negotiations over a Partnership Agreement with the European Union. On the multilateral scale, the WTO Agreement on Trade-Related Intellectual Property Rights (TRIPs) is a watershed in global protection of intellectual property. Before its implementation, each country's IPRs regime was largely a matter of choice, subject to the requirements of any international conventions it might have joined. These conventions required only national treatment and allowed no substantive means for dispute settlement. In brief, the prior global IPRs system was based on mutual recognition in protecting intellectual property and a strong commitment to sovereignty. This system was perceived by American and European innovative interests, which produce a majority of intellectual property worldwide, to be seriously inadequate. The inconsistency between the existence of different national IPRs regimes and the need for international exploitation of the gains from innovation through trade and FDI had become wellrecognized by the mid-1980s. Strong lobbying efforts by firms in pharmaceuticals, software, entertainment, and other sectors placed IPRs at the forefront of both trade policy and trade negotiations. 1 Information taken from World Intellectual Property Organization, Industrial Property, various issues. 2

7 2a. The Multilateral Approach: TRIPs The TRIPs accord commits all developing countries in the WTO to undertake a highly significant extension of IPRs protection, albeit with a long phase-in period for the poorest countries, thereby bringing about a substantive harmonization in international standards. This accomplishment represents a signal victory for intellectual-property interests in the technologyexporting nations and a substantial "concession" (in GATT terms) by the poor countries. It also should have the effect of easing the path toward greater harmonization among the developed economies, within which a number of international controversies over IPRs continue. Table 1 presents in summary form the major substantive requirements embodied in the TRIPs agreement. These issues require significant change in policies in most countries. The most important changes in global norms include copyright protection for computer software and databases, stronger protection for well-known trademarks, the extension of patents to pharmaceutical products and biotechnological inventions, restrictions on the use of compulsory licensing, requirements for trade-secrets protection, and the need for effective border and judiciary enforcement measures. The poorest nations receive a form of "special and differential treatment" through long transition periods. Particularly controversial in the latter context for pharmaceutical firms is the absence of pipeline protection. On the other hand, the administrative costs of implementing and enforcing this agreement are likely to be large for many nations, calling for technical and financial assistance (UNCTAD, 1996). Despite some shortcomings in the standards and commitments of the TRIPs agreement, at least from the view of technology and product developers in the United States, it represents a substantial strengthening of global IPRs systems. On its own merits, the TRIPs accord could have negative impacts on many nations, at least in the short run, because of higher prices associated with newly protected products. It is interesting to ask, therefore, why such a comprehensive agreement commanded widespread adherence in the Uruguay Round. The most relevant answer is that the Uruguay Round agreements provided enough collateral benefits to developing countries, particularly in terms of market access for agricultural goods, textiles and apparel, along with greater teeth to dispute settlement, that it was worth accepting the short-term (though uncertain) losses from TRIPs. This ability to trade concessions across disparate economic categories is the key advantage of multilateral negotiations. 3

8 Table 1. Substantive Requirements of the TRIPs Agreement in the WTO General Obligations Comments 1. National Treatment Applied for persons 2. Most-Favored-Nation Reciprocity exemptions for copyright; prior regionals/bilaterals allowed 3. Transparency Copyright and Related Rights 4. Observes Berne Convention Does not require moral rights 5. Minimum 50-year term Clarifies corporate copyrights 6. Programs protected as literary works A significant change in global norms 7. Data compilations protected similarly 8. Neighboring rights protection for phonogram producers, performers 9. Rental Rights A significant change in global norms Trademarks and Related Marks 10. Confirms and clarifies Paris Convention 11. Strengthens protection of Deters use of confusing marks and speculative well-known marks registration 12. Clarifies non-use Deters use of collateral restrictions to invalidate marks 13. Prohibits compulsory licensing 14. Industrial Designs Minimum term of protection of 10 years Patents 15. Subject matter coverage Patents provided for products and processes in all fields of technology 16. Biotechnology Must be covered but exceptions allowed for plants and animals of traditional methods 17. Plant breeders' rights Effective sui generis system required 18. Exclusive right of importation 4

9 19. Severe restrictions on compulsory Domestic production can no longer be required; non-exclusive licenses with licensing adequate compensation 20. Minimum 20 years patent length from filing date 21. Reversal of burden of proof in process patents Integrated Circuits Designs 22. Protection extended to articles incorporating infringed design 23. Minimum 10 years protection Undisclosed Information 24. Trade secrets protected against unfair methods of disclosure Abuse of IPRs 25. Wide latitude for competition policy to Cannot contradict remainder of WTO control competitive abuses agreement Enforcement Measures 26. Requires civil, criminal measures and Will be costly for developing countries border enforcement Transitional Arrangements 27. Transition periods 5 years for developing and transition economies; 11 for poorest countries 28. Pipeline protection for pharmaceuticals Not required but a provision for maintaining novelty and exclusive marketing rights Institutional Arrangements 29. TRIPs Council Agreement to be monitored and updated 30. Dispute settlement Standard approach with 5-year moratorium in some cases 5

10 While it is accurate to characterize TRIPs as a marked strengthening in international protection for IPRs, analysts often neglect the fact that the agreement itself is somewhat flexible in allowing nations to adopt varying implementation strategies. Countries are given considerable discretion in choosing regulations that maintain a pro-competitive balance in the exploitation of IPRs, so long as they do not unduly frustrate the intentions of TRIPs. For example, little is said in the agreement about fair-use exceptions in copyright for educational and research purposes, meaning that wide exceptions of this kind may be allowed. Further, nothing in the agreement prevents judicious use of compulsory licensing of patents in cases where abusive practices by the patent-holder may be demonstrated under local law, so long as adequate compensation is paid. The agreement also calls for each country to employ competition policies as it sees fit in disciplining abusive exploitation of IPRs, requiring only that such use be based on national treatment. Indeed, given the limited development of competition policies in many developing countries, the implementation period for TRIPs provides a propitious time for them to consider their options in this regard. This flexibility in the final TRIPs document, while frustrating to some economic interests in the developed economies, is also partially responsible for its widespread ratification. 2 Moreover, some developing countries may have feared that a refusal to accede to TRIPs could have subjected them to bilateral disciplines imposed by the United States and other nations. This outcome might have been worse in terms of the development of their IPRs regimes and the evolution of their trade flows. This negative view is misleading, however. Many nations take the view that the adoption and enforcement of TRIPs could have a powerful signaling effect to foreign firms, helping to attract technology and investment. This view is gaining greater currency in the world. Moreover, the higher-income developing countries continue to recognize that a viable IPRs system may well be in their long-run interests in its own right. Accordingly, a multilateral approach based on minimum standards and disciplines is preferable to a competition among these countries in terms of IPRs (or other business regulations) to attract FDI and technology. This last observation provides a further powerful argument on behalf of a multilateral approach, at least as far as the global economy is concerned. 2 See UNCTAD (1996) and Reichman (1993) for detailed expositions of competitive implementation strategies for developing economies. 6

11 Many developing countries have chosen to accelerate their adherence to TRIPs. Most, however, are only now beginning to come to terms with what implementation implies in terms of legislation and administrative costs. It is fair to say that efforts are underway to determine the least-cost means of adopting the various TRIPs disciplines. There is some potential for delays and emerging legislation that will be dissatisfactory for intellectual property developers, while enforcement is sure to remain a problem for some time. 2b. The Regional Approach: NAFTA It was considerably less difficult to reach agreement on IPRs in NAFTA. To a certain extent, this outcome is surprising since NAFTA similarly represented a movement toward significant market integration between a low-wage developing economy and a pair of highincome developed economies. However, NAFTA requires relatively little of the United States in terms of trade-policy changes, while there are substantive potential gains for American firms in the intellectual-property arena. For its part, the government of Mexico views NAFTA as a means of locking in numerous shifts in its commercial policy toward a more outward-looking strategy. A key component of this trend in Mexico lies in cementing the important changes in intellectual property law that took place in 1991, when comprehensive patent and copyright laws were adopted. NAFTA does have the effect of expanding IPRs in Mexico to new areas. Canada, on the other hand, successfully retained rights to protect its cultural industries through variations in its copyright laws. Canada is currently undertaking a major examination of its copyright laws with respect to information industries, including the advisability of maintaining its cultural exemptions to the same degree (Information Highway Advisory Council, 1995). Table 2 presents an overview of the main components of Chapter 17 of NAFTA, the intellectual property rights agreement. As noted, the national treatment principle suffers some injury due to Canada's insistence on exemptions for cultural industries, while Mexico provides limited protection for broadcasts of performances by virtue of invoking reciprocity in treatment. The requirements in copyright go beyond those in TRIPs by clarifying that decoding devices for intercepting satellite transmissions are declared illegal. This last issue remains controversial on a global scale, even among developed nations, and is an issue for future international negotiations. American firms were disappointed that effective opportunities for firms to oppose trademark registration were not required of Mexico. 7

12 Table 2. Substantive Agreements on IPRs in NAFTA General Obligations Comments 1. National Treatment Explicit reciprocity exemption for secondary use (broadcasts) of sound recordings Exemptions for Canadian cultural industries Copyright and Related Rights 2. Observes Berne Convention Does not require moral rights 3. Minimum 50-year term Clarifies corporate copyrights 4. Programs protected as literary works New for Mexico 5. Data compilations protected similarly New for Mexico 6. Sound recordings protected Mexico can withhold copyright for recordings produced before NAFTA in effect 7. Encrypted satellite signals Illegal to produce or use decoding devices without authorization Trademarks and Related Marks 8. Observes Paris Convention 9. Prohibits compulsory licensing and force trademark linkage 10. Trademark registration opposition Not required in Mexico 11. Industrial designs Slightly different standards for protection maintained Patents 12. Subject matter coverage Patents provided for products and processes in all fields of technology 13. Biotechnology Some exclusions from patentability allowed for biotechnological processes and plants and animals of traditional methods 14. Plant breeders' rights Effective sui generis system required; Mexico given two years to adopt UPOV 15. Severe restrictions on compulsory Domestic production can no longer be required; non-exclusive licenses with licensing adequate compensation to serve local market 16. Pipeline protection for pharmaceuticals 17. Minimum 20 years patent length from filing date 18. Reversal of burden of proof in process patents 8

13 Integrated Circuits Designs 19. Based on strongest provisions of 1989 Treaty 20. Protection extended to articles incorporating infringed design 21. Minimum 10 years protection Allows lapsed protection after 15 years 22. Prohibits compulsory licensing 23. Transition period Mexico has four years to comply Undisclosed Information 24. Full trade secrets protection "Gross negligence" standard 25. Confidential test data protected Exclusive use of submitter for marketing purposes for five years Abuse of IPRs 26. Wide latitude for competition policy to Cannot contradict remainder of NAFTA control competitive abuses Enforcement Measures 27. Closely mirrors U.S. system Enforcement problems continue in Mexico Transitional Arrangements 28. Transition period Mexico given three years to comply with enforcement requirements 29. Motion pictures U.S. will protect foreign-made motion pictures it has declared in the public domain 9

14 At Mexico's insistence, NAFTA allows broader exclusions from patentability for biotechnological processes than does TRIPs. On the other hand, during the negotiations for NAFTA, Canada removed its compulsory licensing scheme in pharmaceuticals, a victory for pharmaceutical firms operating there. Further, NAFTA requires pipeline protection for pharmaceuticals, rather than the limited marketing rights provided during the transition periods for TRIPs. Finally, NAFTA provides somewhat stronger protection than TRIPs in standards covering semiconductors, trade secrets, and enforcement requirements. Perhaps most significant, the transition periods are much shorter in NAFTA and, in fact, Mexico may choose not avail itself of the full periods in coming into compliance. Overall, it is fair to claim that NAFTA provides stronger standards for IPRs than does the WTO agreement, though it should be kept in mind that both agreements are comprehensive in setting effective minimum standards. Thus, the NAFTA agreement on IPRs seems stronger than TRIPs in some important ways, though there remain significant exclusions. Presumably, in cases in which NAFTA standards are found to be less rigorous, they would be the prevailing obligation. In any case, the successful installation in Mexico of strong IPRs, along with changes in Canadian compulsorylicensing practices, points to one important advantage of the regional approach over the multilateral approach. In the former, it is easier to achieve broad consensus over harmonized standards because of greater commonality of interests associated with regional market integration. In the latter, (at least perceived) economic interests continue to diverge widely and it would have been extremely difficult to procure an intellectual property agreement as strong as TRIPs without the broad potential for trading off market and policy concessions across numerous economic activities. A final salient distinction between NAFTA and TRIPs is that the former agreement provides less leeway to participants in setting national implementation strategies. For example, Mexico would find it more difficult to adopt strong fair-use exemptions and compulsorylicensing procedures under NAFTA than under TRIPs. In this important sense the regional approach generally may be counted on to erect stronger minimum standards for protecting IPRs. One manifestation of this theme is that the European Union currently seeks to negotiate preferential trading arrangements with many countries in North Africa and the Middle East, with an important component of such agreements being a requirement to accelerate adherence to the conditions of TRIPs. 10

15 It is worth noting briefly that the adoption of strong minimum IPRs standards in NAFTA is analogous to the harmonization efforts within the European Union (Maskus, 1993). Consistent protection of intellectual property throughout the EU lies at the core of ongoing regulatory standardization. Thus, for example, the accession of Spain in 1986 required a rapid convergence of its IPRs regimes with Community standards, including adoption of the European Patent Convention within five years (Baker, 1997). The EU maintains an emphasis on such convergence in negotiating accession with new members. 3. The International Economics of IPRs Both the TRIPs agreement and the IPRs chapter in NAFTA represent a significant alteration in international policy norms. There are important implications for both positive questions concerning the abilities of firms to exploit intellectual property and for normative issues of international welfare. Concerning commercial behavior, global strengthening of IPRs improves the choice set considerably for firms selecting strategies for earning rents from their innovations. That choice set includes: selection of which countries in which to procure a patent or copyright; terms under which to sell licenses to subsidiaries and unrelated foreign firms; location of FDI for purposes of taking advantage of factor costs in producing a new product for local sales, satisfaction of regional markets, and re-exports back to the home country; opportunities for research joint ventures, patent-pooling or cross-licensing agreements with competing foreign firms; and greater latitude in pricing to market as a result of the implicit increase in market segmentation associated with IPRs. International trade in high-technology goods is significant and growing rapidly (Maskus, 1993). The interesting question is whether stronger and more harmonized international IPRs will have an important impact on these trade flows. They could do so through a number of mechanisms. For instance, the strength of IPRs in different markets could affect both price and volume decisions of firms deciding on export markets (Maskus and Penubarti, 1995). In theory, a firm choosing its export strategy to particular markets would charge a higher price in those countries with strengthening IPRs, though the export quantity could be higher (if there is less local infringing activity) or lower (if the law raises market power for the foreign firm). Recent empirical work has demonstrated that international variations in the strength of patent laws do have an impact on trade flows (Maskus and Penubarti, 1995; Primo Braga, 1995, Maskus and Eby Konan, 1994). For example, Maskus and Penubarti (1995) estimated bilateral 11

16 trade equations for 28 3-digit ISIC sectors in 1984, using 22 OECD countries as exporting nations and the same countries plus 49 others, primarily developing economies, as importing nations. In these econometric equations, bilateral imports in each sector were regressed across countries on several variables from a general-equilibrium model of trade, including trade restrictions and economic distortions. They included an index of patent strength in each importing country, corrected for measurement error and endogeneity. This patent index was also interacted with dummy variables for country size in developing markets. The evidence from these regressions is clear that stronger patent laws attract greater amounts of bilateral imports (in comparison with levels expected in equilibrium without patent distortions) into larger (and generally richer) developing countries. The implied elasticities of imports with respect to patent strength are positive, though small. Overall, the results suggest that as global patent strength increases due, say, to the implementation of the TRIPs agreement in the WTO, global trade in patent-sensitive goods could expand by perhaps $15 billion to $25 billion per year. This is a relatively small sum but still a dividend that has not been recognized. Extending the same methodology to U.S. exports to Mexico, if the result of NAFTA is to improve Mexican patent protection to the American level, we might expect an increase in trade of perhaps $3 billion per year (in 1984 dollars), simply on this score. 3 Note that these findings relate to patents; comparable work has yet to be done on the trade impacts of trademarks and copyrights. The activities of multinational enterprises (MNEs) also depend on global IPRs. Firms realize the economic value from intellectual creations in several ways, including FDI in subsidiaries in which production of goods using intellectual property occurs. Parent firms earn license fees and investment income from these affiliates. Further, intellectual property assets may be sold or rented through licensing agreements with unaffiliated firms. It is difficult to establish strong theoretical and empirical linkages between intellectual property rights and FDI and technology trade. Given the economic complexities in these forms of competition, no general claims may be made about the effects of IPRs on such transactions. On one hand, a nation with stronger IPRs could induce more incoming FDI because recognizing the intellectual property of a foreign firm increases the value of that firm's unique asset. In turn, the firm would be more willing to undertake operations abroad. On the other hand, it has been argued that MNEs may undertake greater FDI in countries with limited IPRs in order to 12

17 maintain proprietary information within organizational control. In fact, this factor seems relevant in explaining the importance of foreign investment in the pharmaceuticals industry. Theoretical predictions about impacts on technology transfer are ambiguous as well, in that they depend on cost characteristics of firms, market structure, and other factors. There is some tentative empirical evidence on the relationship between IPRs and FDI. Mansfield (1994) reports survey evidence claiming that IPRs are important for location decisions in FDI, though the importance varies by type of investment. He also asked U.S. firms for their perceptions about the weakness of IPRs in various partner countries, with these perceptions being based on whether IPRs impeded or enhanced decisions to undertake licensing and joint ventures. He then estimated a relationship between U.S. direct investment flows to different countries and their GDP and perceived IPRs protection. He found that the weakness of IPRs was significantly and negatively related to investment decisions across countries, suggesting that countries that strengthen their patent regimes could well attract additional FDI inflows. The situation in Mexico seemed to be of particular interest to firms in this regard. It is possible to improve on Mansfield's model specification by recognizing the joint decisions made by firms to exploit their IPRs and adding further determinants of FDI. In performing this task, I find that a typical result is the following regression equation for the stock of U.S. assets across countries, taken from a four-equation systems approach: Assets = *GDP *DIST *PATD. This equation was estimated across 46 foreign countries in a pooled sample from , using total assets of affiliates as the dependent variable. GDP is the host nation's gross domestic product per capita, DIST is distance of the country's capital city from Washington, DC, and PATD is the previously discussed index of patent strength in developing countries. 4 All variables are significantly different from zero. Thus, it appears that, for two otherwise identical host countries, the country with a one-unit higher level of the patent index attracts additional FDI of around $1.931 billion. To put this in perspective, imagine that the IPRs agreement in NAFTA raises Mexican patent protection to U.S. levels. This equation would predict a rise in FDI from the United States to Mexico on the order of $893 million, or approximately 22% of 1989 assets. Again, this figure does not account for other NAFTA impacts on FDI. If this finding holds up to other specifications it points directly to the importance of IPRs in providing 3 This calculation does not account for other trade-expanding effects of NAFTA, which could increase the figure through certain interaction effects between trade liberalization and IPRs harmonization. 13

18 incentives for inward FDI. Presumably the additional FDI embodies greater levels of technology transfer as well, though the extent of this effect remains an open empirical question. It is important also to consider some dynamic implications of the agreements. There are essentially two dynamic processes that matter here: the pace of innovation and the rate at which new innovation is diffused, or transferred, into the general economy. To assess the system, then, requires investigating the linkages between strength of intellectual property rights, innovation, and imitation. Recent theoretical treatments of the impacts of IPRs on technology diffusion in endogenous growth models are discouraging. In these models, technology is transferred through imitation by firms in developing countries. When IPRs are strengthened in those countries, imitation becomes harder. The rate of imitation slows down and, in turn, the global rate of innovation diminishes. The reason is that if innovative firms expect less rapid loss of their technological advantages they can enjoy higher rents per innovation, thereby reducing the need to engage in R&D (Helpman, 1993; Glass and Saggi, 1995). A more optimistic view comes from studies of international patenting behavior (Eaton and Kortum, 1995). The value of patent rights varies across countries and technology fields, but nonetheless is typically significant in key technologies and products. Additional patent protection in important developing countries, such as Mexico, could enhance this value, inducing further R&D. Moreover, there appear to be considerable spillovers of technological information through patenting and trade in patented products. Eaton and Kortum (1995, p. 1) state that, "...except for the United States, OECD countries derive almost all of their productivity growth from abroad." Indeed, the importance of technology transfer through trade in technologically advanced inputs (machinery, chemicals, software, producer services, and so on), should be emphasized. Evidence is emerging that this channel is responsible for significant amounts of productivity spillovers across borders and a key part of the technology convergence that has happened among the developed economies in recent decades (Coe and Helpman, 1993). In conjunction with the earlier evidence that countries with stronger patent regimes tend to attract more trade in inputs, this finding suggests that developing nations do have a dynamic interest in linking their IPRs regimes with those of the OECD countries. 4 Other variables included tariff rates, R&D expenditures, and incentives and disincentives for FDI, in a fourequation system with assets, exports, sales, and number of patent applications as the dependent variables. 14

19 Turn next to normative issues regarding global IPRs. In economic terms, the contrast between agreements to liberalize barriers to merchandise trade and harmonization of IPRs standards is striking. With regard to trade in goods, strong theoretical arguments and empirical evidence point to the multilateral elimination of trade barriers as a welfare-enhancing policy for all countries concerned. Though distributional problems emerge in each country, the aggregate competitive gains from static and dynamic efficiency tend to outweigh the costs of trade liberalization, save perhaps for some terms-of-trade effects in large countries, which should be largely self-canceling in the multilateral context. In short, global free trade serves as a benchmark of optimality against which to assess the net costs of departures. This point may also be argued on behalf of the elimination of investment barriers, allowing investment to flow to its globally most efficient uses. In contrast, IPRs are inherently second-best policies within each country and it is difficult to specify workable rules that necessarily satisfy conditions for welfare optimality. Rules of thumb, such as providing copyright protection to software but leaving open the possibility of patent protection under particular circumstances, may in some cases be too weak and in other cases too strong. Extended to the international context, these ambiguities are magnified. Each country is liable to have different economic interests in IPRs, which interests could well vary over time. Consider the notion of "global optimality" in IPRs. There is a fundamental tradeoff between static efficiency (each potential user should have the product or technology available at marginal cost) and dynamic efficiency (incentives must be sufficient to induce R&D in socially desirable innovations). Presumably the TRIPs agreement will tilt the balance toward greater innovation and greater exclusion, which is sensible if the world had been under-investing in R&D in the status quo ante. A number of dynamic efficiency gains could ensue from this change, including R&D becoming more closely targeted on projects with the highest expected global demand and financing advantages stemming from global price discrimination. But there might emerge dynamic efficiency losses as well, including excessive and duplicative investments in R&D. To this mix must be added the additional complexity that the location in which IPRs are owned (typically developed nations) may well differ from the location in which they now receive stronger protection (typically developing nations). Thus, a significant issue emerges with respect to international rent transfers. In short, it is unclear whether TRIPs will be globally pro-competitive or globally anti-competitive, while questions of international distributional impacts are important. 15

20 4. Economic Differences between Regional and Multilateral Approaches The experiences of NAFTA and the European Union suggest that IPRs protection is becoming stronger and more harmonized within regional economic groupings than on a global scale. This is consistent with the trend toward deeper integration of regional regulatory regimes. It raises interesting questions about the efficiency implications of strong regional harmonization. Some of these questions, which have been little studied, are stated here in order to stimulate further discussion and analysis, rather than to provide definitive answers. The issues relate to the inherently second-best character of both regional trade preferences and IPRs. For example, the essence of regional trade agreements is to provide preferential treatment in commercial policy. Drawing a page from the economic theory of economic integration, define the concepts of intellectual property creation (IPC) and intellectual property diversion (IPD). The former would exist to the extent that more intellectual property is created and marketed within a regional market, such as NAFTA, strictly due to the expansion of markets and the replacement of inefficient technology developers with more efficient ones. Fully integrating North American markets for technology and product development with harmonized IPRs should have strong rationalization impacts on R&D programs in all three countries. For example, in some industries, such as chemicals, food products, and advanced textiles, plant-level scale economies are high relative to firm-level scale economies. In conjunction with high transport costs to Mexico, this situation would tend to transfer R&D programs southward within North American multinational enterprises. At the same time, in sectors with high firm-level economies associated with knowledge capital, such as transport equipment, machinery, and telecommunications, the effect should be to concentrate R&D efforts in the United States. Such rationalization effects should provide long-term dynamic efficiency gains for all three nations, perhaps at the expense of some transition costs. Perhaps more interesting is the concept of IPD, which occurs to the extent that less intellectual property owned by non-member firms is exploited within the free trade area. Channels through which this could happen include reduced demand for those firms products due to trade preferences, reductions in FDI within the area due to investment diversion, and (market-based) exclusion of foreign firms from regional technology-development programs through joint ventures. Notice that these effects are possible even if the intellectual property standards within a free trade area are non-discriminatory. 16

21 In a related vein, greater harmonization of both regional and global IPRs standards does not necessarily represent a movement toward less effective variability in world practices. It is likely that many developing countries will choose to implement TRIPs at minimal allowable levels, leaving a large economic mass at that stage. Meanwhile, more and larger blocs sustaining regionally distinctive IPRs would impart a discrete character to the international distribution of standards. Thus, while TRIPs is likely to induce more incentives for innovation programs directed to global demands, they could be offset by more-skewed incentives aimed at regional blocs. 5. Concluding Remarks This review suggests that, while there are reasons to anticipate net gains in dynamic global efficiency from the new international IPRs systems, concerns remain about the implications of these systems for the distribution of gains between developed and developing economies. Moreover, significant differences exist between effective standards that will be attained as TRIPs is implemented and standards that are emerging under regional trade agreements. These differences pose new questions about the role of deep integration in the international regulation of intellectual property use. A final, and related, observation is that the trend toward stronger IPRs, within both the WTO and certain regional agreements, has sometimes come with insufficient regard for changes in important complementary regulatory policies. Chief among these is competition policy, which should be intimately related to both intellectual property protection and trade liberalization (UNCTAD, 1996). For example, Spanish accession to the European Union in 1986 resulted in rapid and sharp increases in patent applications and trademark registrations in Spain (Baker, 1997). One reason for this appears to be that firms realized that Spanish membership would expand the region within which parallel imports would be allowed, inducing broader registration of intellectual property. Competition rules for regulating IPRs use across countries remain highly variable, with little effort yet in evidence in most developing countries to develop a coherent and appropriate system to deal with problems that could emerge in the new regime. In this regard, it is noteworthy that NAFTA says nothing about harmonizing competition rules beyond a vague requirement that their use should not counteract the operation of the trade agreement. Neither does the European Union require changes in competition policy in developing countries with which it negotiates Partnership Agreements, though within the Community there is substantial 17

22 harmonization. This means that stronger IPRs could interact with inadequate competition rules to render particular markets less competitive. For example, it is common in the Middle East for governments to allow local distribution of imported branded products and technologies to be monopolized by sole importing agents. As patents and trademarks achieve stronger protection, the implicit monopoly privileges are enhanced for such firms. Thus, a system is emerging in which regional protection for intellectual property could be stronger than global norms while complementary business regulations could be inadequate for either set of policies. It is likely that such a system would result in significant structural inefficiencies in many countries, even as it might raise the global return to innovation. In at least this sense, both regional and global initiatives in IPRs are incomplete in design and execution. 18

23 REFERENCES Baker, Mark E. (1997), Innovation and the European Union: A Case Study of the Economic Effects of Spanish Accession, unpublished MA thesis, University of Colorado. Coe, David and Elhanan Helpman (1993), "International R&D Spillovers," (Cambridge, MA: National Bureau of Economic Research) Working Paper No Eaton, Jonathan and Samuel Kortum (1995), "Trade In Ideas: Patenting and Productivity in the OECD," (Cambridge, MA: National Bureau of Economic Research) Working Paper no Glass, Amy J. and Kamal Saggi (1995), "Intellectual Property Rights, Foreign Direct Investment, and Innovation," (Columbus, OH: Ohio State University), manuscript. Helpman, Elhanan (1993), "Innovation, Imitation, and Intellectual Property Rights," Econometrica 61, Information Highway Advisory Council (1995), The Challenge of the Information Highway (Ottawa: Ministry of Supply and Services). Mansfield, Edwin (1994), "Intellectual Property Protection, Foreign Direct Investment, and Technology Transfer," (The World Bank and International Finance Corporation), Discussion Paper no. 19. Maskus, Keith E. (1993), "Trade-Related Intellectual Property Rights," The European Economy, no. 52, Maskus, Keith E. (1996), "Intellectual Property Rights in the Global Information Economy," in T. Courchene (ed.), Policy Frameworks for a Knowledge Economy, (Kingston, ON: John Deutsch Institute), Maskus, Keith E. and Denise Eby Konan (1994), "Trade-Related Intellectual Property Rights: Issues and Exploratory Results," in A.V. Deardorff and R.M. Stern, eds., Analytical and Negotiating Issues in the Global Trading System, (Ann Arbor: University of Michigan Press), Maskus, Keith E. and Mohan Penubarti (1995), "How Trade-Related Are Intellectual Property Rights?" Journal of International Economics 39, Reichman, J. H. (1993), "The TRIPs Component of the GATT's Uruguay Round: Competitive Prospects for Intellectual Property Owners in an Integrated World Market," Fordham Intellectual Property, Media, and Entertainment Law Journal, 4, United Nations Conference on Trade and Development (1996), The TRIPs Agreement and Developing Countries, (New York: United Nations). 19