Information Products: A Challenge to Intellectual Property Theory

It is folly to make predictions about the future, espedally in writing. Nonetheless, I will venture to suggest that, as plastics may have been the "growth industry" of the twentieth century,' information products will be that of the twenty-first. I define information products to mean items like computer technology (including software programs and computerized data bases), "designer genes" 2 and semiconductors: products whose information content vastly exceeds in value the cost of the products on which that information is stored. 3 Just as the industrial revolution dwarfed the importance of agricultural production to the economy by reducing the costs of farming, so too the information revolution will reduce the significance of the industrial sector, by creating information that lowers the costs of manufacturing.

For multinational corporations that wish to preserve their stakes in the economy, the emerging importance of information as a commodity-indeed, a preeminent form of wealth-is problematic. The foremost characteristic of these products is the economics of their development and use. All have extremely high front-end costs, but, once developed, are cheaply reproduced. Accordingly, the need for legal protection against those who might "free-ride" on another's investment is acute. At the same time, however, the protective schemes developed for agricultural and industrial property are inapposite. First, information is not tangible, so standard property laws do not apply. Second, these new information products present difficult conceptual problems that render them unamenable to intellectual property protection under traditional regimes such as copyright, trademark, and patent. 4 Third, because information products are intangible, they flow across national borders more easily than agricultural and manufacturing products. While this simplifies worldwide marketing, fluidity creates a greater need for international protection. 5 The legal system needs time to shape the appropriate response to the challenges posed by these new products. Ostensibly, there are two steps to developing legal rules suitable for protecting novel technologies. In the first stage, new forms of domestic protection are fashioned, either by expanding traditional regimes or by creating new ones. The creation of legal rules furnishes incentives to innovate because it enables innovators to recoup their costs and to earn a profit on their investment and labor.

In the second stage, international agreements are forged. Traditional forms of intellectual property enjoy the benefits of treaties such as the International Convention for the Protection of Industrial Property (the Paris Convention), 6 the Berne Convention 7 and the Universal Copyright Convention (UCC).8 Adherence to these conventions produces many public advantages. The assurance of protection in foreign markets spurs innovative activity by motivating research into areas that would otherwise go unexplored because no single domestic market is lucrative enough to offer investors a sufficient economic return. Moreover, international cooperation facilitates global dissemination by lowering the transaction costs of obtaining and enforcing exclusive rights. 9 Although recognizing the exclusive rights of non-nationals raises the costs of obtaining foreign technology, a special problem for less developed nations, the offer of exclusivity induces outside investment. 10 Unlike some traditional goods, information products often cannot be used fully without significant infrastructure such as computer hardware, communication facilities, skilled operators or complex adaptation to local conditions. These necessary secondary investments are, like the primary investment in the technology itself, spurred by the promise of exclusivity.

Legal protection of information technology is currently in its first stage of development. In the United States, discussion has centered on solving new problems with specialized legislation, a strategy which may ultimately introduce difficult issues on both the domestic and international fronts. This article explores these questions by focusing on the United States' first experiment with sui generis legislation, the Semiconductor Chip Protection Act of 1984 (the Chip Act). 1 After summarizing the legal problems posed by these emerging technologies through the example of the semiconductor industry, it will next explain why the strategy adopted by the Chip Act is likely to appeal to Congress in the future. The article will then focus on a second issue, namely, how LJ. 255 (1984).

the international community should respond to these sui generis schemes.

The article suggests that international accords may take on a wholly new significance in the coming era. Because resort to sui generis legislation means that there will be periods when the right of a particular technology to protection, and the contours of that protection, will be uncertain, a strong presumption of protectability must be put into place to avoid short-term distortions in the pattern of investment in research and development. That presumption could be created effectively through an international agreement providing interim generic protection for new technologies. The exclusivity provided by this agreement would continue until signatory countries worked out domestic protection schemes and reached suitable international arrangements, at which time the new-and presumably better tuned-legislation would supplant the interim agreement. In the end, it may be that the notion of a two-stage process is an anachronism. Because the market for information products is global, international protection mechanisms will largely determine the level of investment in researching and developing such products in the next century.

This section summarizes the problems posed by information products and describes the legislative solution adopted in the Chip Act. Although the discussion is somewhat technical, it is difficult to discuss how the international arena should respond to the Chip Act without an understanding of the tensions that led to its enactment and the means by which these tensions were resolved. If the Chip Act is itself unique, then there is little reason to be concerned about international ramifications. This section demonstrates, however, that other information technologies share the problems that led to this enactment and would benefit equally from the strategies adopted for chips. 12 12. The Chip Act has been discussed extensively elsewhere. See, e.

Every new product is different and thus presents its own unique challenges to the legal system. However, certain commonalities can be observed across the range of products that have developed in the last quarter-century. Be it manmade microorganisms, computer software, robotics, semiconductor chips or computerized data bases, certain distinctive features make legal protection necessary, yet difficult to achieve. Although legal exclusivity is not always necessary to guarantee a return to the developers of new products, nonlegal protections rarely exist in the case of information products.

It has, for example, been said that copyright in certain kinds of books is not necessary because the cost of producing and distributing pirated copies is high enough so that the original publisher's price, even if it includes development expenses, is not much greater than those of the pirate.' 3 Such reasoning does not apply to these new information products. Because development is so costly, the difference in price between what the free rider can charge and what the developer must charge to recoup his costs is substantial. This gap in price will tend to drive consumers to purchase the goods of the free rider. To make matters worse, many of these products change rapidly (data bases become obsolete daily, and even semiconductor chips tend to be commercially significant for only short periods of time). Accordingly, developers cannot bring their selling prices closer to those of their competitors by extending the period during which they recoup their costs.

Developers of traditional products may also enjoy lead time during which they can recapture costs before others enter the market and force down the price.' 4 But the circumstances that produce substantial lead time often do not 263 (1985) (introduction to symposium on the Semiconductor Chip Protection Act of 1984); Wilson & LaBarre, The Semiconductor Chip Protection Act of 1984: A Preliminary Analysis, 67J. PAT. OFF. Soc. 57 (1985).

13. Breyer, The Uneasy Case for Copyright: A Study of Copyright in Books, Photocopies, and Computer Programs, 84 HARV. L. REv. 281, 297-99 (1970).

14. Ide at 299. See also Dreyfuss, Dethroning Lear Licensee Estoppd and the Incentive to Innovate, 72 VA. L. REv. 677, 698 (1986); Kastenmeier & Remington, supra note 12, at 452. exist for information products. With traditional products, lead time occurs because public demand for the product was, at the time of introduction, unforeseen. By the time others enter the field, the developer has recouped his costs. In the case of information products, however, the need for the product is, almost by definition, clear before it is produced. For example, the first step in manufacturing a new semiconductor chip is always the determination of public need. 15 Entry barriers can also assure lead time by preventing competitors from entering the new market quickly. Initially, the semiconductor industry enjoyed an advantage of this type. Although copying is relatively inexpensive, production requires a high level of technical skill. While competitors learned how to build clean facilities and train their staff, many of the early semiconductor companies earned substantial profits.1 6 Although this lead-time advantage may explain why the semiconductor industry did not initially require legal protection to become as successful as it did, such favorable circumstances are rare. The reproduction of computerized information does not require the kinds of technical competence that originally protected semiconductor developers. Furthermore, technical barriers exist only during an industry's infancy. Once the industry matures, the costs of development remain high, while the barriers to piracy are considerably reduced. For the semiconductor industry, for example, there is no longer a technical obstacle to piracy because the skills required for manufacturing chips, technical though they may be, are now widespread and do not vary greatly between designs.

15. Indeed, the need for the entire semiconductor product category was clear before the product was developed. By the end of World War II, electronic products such as televisions and radios were in considerable demand. A great deal of research was devoted to making these products cheaper, smaller and more efficient by reducing the size and cost of the vacuum tubes that were used to isolate and amplify electronic signals. Bell Laboratories, RCA, Texas Instruments, and many others were actively engaged in the research that led up to the semiconductor. Thus, as soon as semiconductors became available, the electronics industry appreciated their value, and the demand was immediate. See Noyce, iMicroelectronits, SI. AM., Sept. 1977, at 262, 263; Wolfe, The Tinkerings of Robert Noyce, Es- Secrecy is another technique that can sometimes be used to ward off competitors. In the course of developing new products, the first mover often learns manufacturing techniques and short cuts that help it produce more efficiendy and keep its prices close to those of free riders. But this kind of protection is rarely available to the makers of information products because the sale of the products themselves usually reveals everything necessary to reproduce them at a cost equal to that of a developer.' 7 The inability of information producers to resort to such nonlegal protections for their innovations would not be so troubling if effective legal protection were available. But traditional legal regimes, such as copyrights and patents, have proved unsuitable for a variety of reasons.

The foremost problem with respect to copyright is conceptual. Copyright law is structured to protect new and original expression; it does not protect the ideas that the creator expresses. 1 8 For many classes of intellectual property, this limitation is not an obstacle. Either the expression represents the whole of the investment, as in most novels, or the expression is easily separated from the ideas, as in household ornaments. 19 Since creators receive exclusivity for at least a substantial part of their investment, the protection of expression, while allowing free copying of ideas, simultaneously protects their economic interests, the public interest in 17. The innovative material stored on the chip, for example, can usually be "reverse engineered" by inverting the process for creating the chip. That is, the chip is photographed, a single layer is stripped off, and that is photographed, and so on through the layers on the chip. These photographs are then enlarged to recreate the masks (or stencils). Similarly, in the course of selling their innovations, developers of computer programs and data bases ordinarily publish everything necessary to copy the product. For bioengineers, the problem may be worse: although the innovator required advance biological techniques to produce the microorganism being sold, any purchaser can simply use natural reproductive techniques to gain more copies.

18. 17 U.S.C. § 102(b) (1984). 19. See, e.g., Mazer v. Stein, 347 U.S. 201 (1954) (finding that the sculptural aspects of a lamp base were separable from the base's function, making copyright protection for the design possible).

motivating their work, and the social interest in allowing others to build freely upon the ideas they have uncovered.

The conceptual limitation is, however, problematic for most information products. Consider, for example, a computerized data base, containing factual material arranged in a particular organizational format. If the data base is well-run, the factual material will be easily retrieved, downloaded and rearranged according to the user's own needs. In other words, the facts in the data base are appropriated while the organizational dimension is discarded. Yet it is only the latter element-the organization, or expression-that is protected by copyright. The facts, which may represent the bulk of the publisher's continuing expenditure, are (like ideas) not copyrightable. 20 For products such as semiconductor chips, the problem is slightly different, but no more solvable. A great deal of effort goes into crafting each mask work. While each layer on a chip can be compared to a work of art, and the entire structure to a sculpture, the aesthetic comparison misses an important element: it is the chip's ultimate function that determines its exact shape. Thus the idea-a chip that meets end-users' needs-merges with the expression-the actual design of the chip.

Copyright has not been extended to protect expression merged with ideas because such protection would overly circumscribe the use of the ideas themselves. 2 ' In the chip context, for example, protection against reproducing the design would prevent others from photographing chips in order to learn the new ideas they contain. But because copyright is unavailable to these functional products, their developers 20. See, e.g., Hoehling v. Universal City Studios, 618 F.2d 972 (2d Cir. 1980) (research into the Hindenberg disaster cannot be protected by copyright), cert. denied, 449 U.S. 841 (1980). Even the expressive features in the organizational technique may not be copyrightable. See, e.g., Matthew Bender & Co. v. Kluwer Law Book Publishers, Inc., 672 F. Supp. 107 (S.D.N.Y. 1987) (method of compiling not copyrightable when there are only a limited number of ways in which the information can be organized).

21. See, e.g., Baker v. Selden, 101 U.S. 99 (1879) (design for bookkeeping ledger cannot be protected because it would prevent free use of the bookkeeping system). Note that it is the Baker principle that prevents data base publishers from using copyright to protect the expressive elements of their systems.

wind up with little, if any, protection for their investment-at least under copyright law.

Patent protection gives the inventor of a new, useful and non-obvious process, machine, manufacture or composition of matter the right to exclude others from making, using or selling the invention for a period of seventeen years. 22 Although patent law, like copyright, tries to leave ideas in free circulation by protecting only particular, original applications of an idea, patent production is available even when the ideas in a discovery merge with their expression, as they do in information products. To be patented, however, an innovation must be more than merely new; it must be so great an advance over all that went before it that one with ordinary skill in the art could not have created it. 23 In mature information industries, this standard cannot always be met. For example, while the major discoveries regarding semiconductors were patented (such as the technique for electrically interconnecting the elements on a chip to an integrated circuit), the embellishments that allow innovators to construct chips to meet specific end-user needs do not attain the degree of novelty required by patent law. The creators of these improvements cannot, therefore, rely upon patent production to gain the exclusivity that they need to compete advantageously. Nor can they rely on the underlying patents, since by the time the industry matures, these have already expired. 24 In some instances, patent law may not cover even the inventions that launch a new technology. In some fieldsbiotechnology is an example-there was initially some controversy over whether the discoveries should be considered patentable subject matter. 25 In other fields, important early advances take place in an academic setting, where results are sometimes not protected, or where protection is lost through publication. 2 6 As with copyright law, when the probable lack of traditional legal protection is coupled with the possibility that nonlegal means of recouping costs will not be available, the disincentives to investment become significant.

The Chip Act is a unique hybrid of patent and copyright law which solves the aforementioned problems for the semiconductor industry through a variety of special devices. The new law protects original mask works by adopting the patent approach of disregarding functionality, 27 but it does not interpose the same novelty requirement contained in the Patent Act. 28 Thus, the Chip Act protects even small advances in semiconductor art, so long as they are original to the applicant.

At the same time, however, the Act is sensitive to the public need to keep ideas in free circulation. It incorporates copyright's prohibition against the protection of ideas, 29 so that even as the design is protected, the ideas embodied in that design are not. To assure that these ideas can actually be utilized without risk of infringement, the Act creates two unique features. First, it defines a category of mask works that are "staple, commonplace, or familiar in the semiconductor industry.... ,,3 0 These basic building blocks of semiity of a human-made microorganism. However, since that decision, Congress has repeatedly considered limitations on patent rights for life forms. See, e.g., H.R. 3119, 100th Cong., 1st Sess. (1987) (placing a two-year moratorium on the issuance of animal patents); H.R. 4970, 100th Cong., 2nd Sess. (1988) (providing a farmer's exception to patent rights).

26. 35 U.S.C. § 102(b) (1984) bars the patenting of an invention that was described in a printed publication more than a year before the application. Because the statute provides a one-year grace period between the time of publication and the time that the patent application must be filed, careful supervision over faculty work can prevent loss of patent rights. The patent laws of many industrialized countries, however, contain no such grace period. See Winner, supra note 9, at 421-22 and n.12.

27. 17 conductor technology are not protectable, even if they otherwise meet the standard of originality.

Next, the Act permits others to discover the ideas embodied in chips by creating a right to "reverse-engineer" them. Under the Act, the process of reverse engineeringwhich would otherwise be considered an infringement, since it reproduces a protected work-is permissible if the purpose is to teach, analyze and evaluate the concepts embodied in the circuitry and to incorporate those concepts into new, original chips. 3 1 Piracy (using photographs of the chip layers to do nothing more than reproduce the chip) is, in short, made an infringement while the industry's ability to exploit the ideas contained in new chips is preserved.

Congress added several other features to the Chip Act to customize the law to the semiconductor industry. First, two provisions were enacted to deal with the rapid pace of change within this industry: copyright's automatic protecdon system was chosen over the cumbersome examination procedure used in patent law, 3 2 and the term of protection was set at ten years, 3 3 rather than the life-of-the-author-plusfifty-years term of copyright law 3 4 or the seventeen-year term of patent law. 35 Second, a section was added to deal with innocent infringement, which is a special problem in the semiconductor industry. To reduce the risk that chip buyers will be injured for innocently purchasing infringing inventory, the Act permits purchasers to use and sell infringing products purchased before the buyer became aware of the infringement. After notice, the purchaser is given a compulsory li-31. 17 U.S.C. § 906(a)(1), (2) (1985). 32. 17 U.S.C. § 904(a) (1985). Copyright subsists as soon as a work is "fixed" in any known tangible medium of expression. 17 U.S.C. § 102(a) (1985). Patents, however, issue only after examination. 35 U.S.C. § § 131-135 (1985).

33. 17 U.S.C. § 904(b) (1985). 34. 17 U.S.C. § 302(a) (1985). For a work created by the employees of a corporation for hire, the term is 75 years from its first publication, or 100 years from its creation, whichever is shorter. 17 U.S.C. § 302(c) (1985).

35. 35 U.S.C. § 154 (1985). The term of patents on certain inventions subject to regulation by the Food and Drug Administration may be extended to compensate the patentee for regulatory delays that prevent timely marketing of the product. 35 U.S. C. § 156 (1986).

cense to continue distribution of good faith purchases. 30 Third, the Act creates a registration system that assures the release of protected ideas to the public, 37 yet permits producers to retain some of their innovations as trade secrets. It requires the deposit of a visually perceptible representation of each layer of the chip. However, the applicant is permitted to withhold two to five layers of masks and to block out sensitive portions of ones that are filed, as long as a major portion of the chip is revealed in the deposit.

Other features of the Chip Act could also be used to demonstrate that the special problems posed by information products can be solved through specialized domestic legislation. By adopting those features of traditional regimes that address the needs of the industry, but abandoning longstanding definitions and conceptualizations that are no longer apt, new laws can strike a workable balance between the private need to recoup high investment costs and the public need to utilize ideas freely. Thus, although the Chip Act was geared to a single industry, it is clear that its strategy is capable of more universal application.

The problem of functionality, for instance, is shared by other technologies, such as industrial design. Although an attempt has been made to accommodate designs in the patent system by relaxing the standard of patentability, 3 8 a copyright-like bar against protecting purely functional elements 3 9 has made this legislation wholly inadequate. 40 A so-36. 17 U.S.C. § 907(a)(1) (1985). 37. 17 U.S.C. § 908 (1985) and 37 C.F.R. § 211.5 (1988) lution to the novelty/function dilemma tailored to the needs of this industry would certainly be welcomed by designers.

By the same token, the computer industry would benefit significantly from greater attention to the problem of keeping ideas in free circulation. When the Copyright Act was extended to computer programs, a provision was added to adjust the reproduction right to the needs of users. 4 ' However, no attempt was made to create a way for programmers to use ideas that others had discovered without risking infringement. This has proved to be a difficult problem for the courts. Most decisions to date have favored the copyright holder, 4 2 with the result that progress in this field has been compromised. 43 Exclusion of staples from protection and a well defined right to "reverse-engineer" may be as beneficial to the computer industry as it is to semiconductor developers.

The Chip Act's reverse-engineering provision serves another important function: it protects the ability of third parties (that is, neither the producers nor the end-users of a protected work) to make products compatible with the protected work. The Copyright Act contains no similar provision, and this too has created difficulties for the computer industry.4 Software producers, for example, cannot write programs that run on a particular computer without access to the operating system of the computer for which they are writing. Similarly, it is not possible to manufacture a computer that runs a series of application programs without knowingly incorporating certain features of the operating system for which those programs were designed. 4 5 It is no 41. 17 U.S.C. § 117 makes clear that those who own copies of a program have the right to reproduce them for archival purposes and in order to run the program on their own computers.

42. 44. See generally Besarovic, The Legal Protection of Computer Programs, 23 COPYRIGHT 144, 148 (1987).

45. It is instructive to note that IBM and Fujitsu have entered into an arbitration agreement to settle Fujitsu's rights to know enough about the wonder, then, that commentators have called for sui generis protection for computer programs along the lines adopted for chips. 4 6 The opportunity to rethink the deposit issue is also an attractive feature of sui generis legislation, for many information products raise difficulties in this area. As with semiconductor chips, publication of a computer program does not always reveal fully the information contained. Although copyright traditionally considers full disclosure a quid pro quo for exclusivity, the Register of Copyrights has made an exception for computer programs and has issued regulations permitting programmers to withhold portions of their work from deposit in order to protect their trade secrets. 4 7 Express legislative consideration of this issue would, however, provide greater certainty to the industry as well as firmer public confidence in the wisdom of the Copyright Office's procedure.

For biotechnology products, the situation is even more problematic. Bioengineered microorganisms are currently protected under general patent law, which has been adapted to require microorganism deposit 48 to compensate for the fact that written specifications do not adequately disclose the ideas contained in these inventions. 49 This adaptation does not, however, work well in practice. If others are given access to deposited samples, there is danger that the material withdrawn will be made to reproduce itself, in a nonmonitorable violation of the patent right; if access is restricted, then the ideas contained in the invention are not, after all, released into the public domain. Many more parallels between the needs of information industries and solutions offered to semiconductor developers can be drawn. They include the Chip Act's rejection of compulsory licenses in most situations 5 ' and its protection of innocent infringers. More generally, the tactic employed by the Chip Act of waiting for a new technology to unfold and then attacking the specific problems that have arisen would be extremely beneficial to other industries. But if that is true and sui generis legislation becomes popular, 52 a new problem will arise. The Chip Act has been successful because the semiconductor industry could afford (due to patent protection and the technological barriers initially facing outsiders) to delay further legal protection until the time when the special legal problems that chips presented were well understood. 53 This delay not only enhanced the law's comprehenan Integrated Policy, 11 RUTGERS COMPUTER AND TECH. LJ. 469, 493 (1985) (noting the cost of complying with deposit requirements internationally).

51. In this regard, the Chip Act should be compared to 17 U.S.C. § § 107-118 (1988), which create a series of compulsory licenses (some of which are compensated, some not) permitting favored uses without authorization.

52. Indeed, the attractiveness of the Chip Act may tell us something about the relative capacity of the various institutions in the government to come to terms with the problems created by information products. In the past, the traditional intellectual property regimes were viewed as covering whatever new inventions came along, and it was the role of the judiciary to manipulate the statutes to meet particular needs. See, e.g., 17 U.S.C.

§ 102(a) (1988) (copyright subsists in works of authorship fixed in any tangible medium of expression "now known or later developed"); Maxtone-Graham v. Burtchaell, 803 F.2d 1253 (2d Cir. 1986) (the fair use doctrine works in a variety of circumstances because "the legislature expressed its intention to give courts the freedom to adapt the doctrine to particular situations on a case-by-case basis, in light of changing technology"). If the Chip Act has any long-term significance, it may be that it stands for the proposition that information products present problems of a kind that cannot be solved by accretion. A new sort of compromise between the public and the parties involved in the development of these products must be achieved, and the greater fact-finding and deliberative capacity of legislators is required to strike the balance that will work best. See generally siveness and effectiveness, it also permitted the development of a class of chips that became unprotected staples under the Act. But not every industry can afford a wait-and-see period during which its needs can crystallize 5 4 Furthermore, if it becomes clear that traditional regimes will not be expanded to encompass new technologies, future industries will not be able to benefit-as the chip industry probably did-from a presumption of protectability. If the likelihood and the details of protection are unclear during the interval between the initial discoveries and the maturation of the industry, there could be a negative impact on investment in research and development. Inventors and investors will tend to devote their energies and money to fields in which a clear right to exclusivity deters free riders.

A trend toward sui generis legislation raises difficulties on the international level as well. However, there are possible solutions to both the international problems raised below and the domestic issues noted above.

The decision to protect chips through specialized legislation has significant international repercussions for both the United States and its trading partners. Had the United States integrated semiconductor chips into either copyright or patent law, certain international obligations would have been triggered. The UCC 55 and the Paris Convention, for example, require signatory countries to provide to the nationals of other signatories the same treatment given their used in connection with the manufacture of, or incorporated in a useful article"). However, the proposal met with little enthusiasm from the industry, which was afraid that ideas would be over-protected by copyright. See H.R. Rep. No. 781, 98th Cong., 2d Sess. 15 (1984). By 1983, nonlegal protections and patent protections had expired, the industry became vulnerable to copyists, and Congress was asked to try again. The Chip Act was the result.

54. See Kastenmeier & Remington, supra note 12, at 454-56 (setting forth reasons why other industries may encounter obstacles more severe than those confronting the chip industry). own citizens. 56 This obligation holds true even when a signatory extends protection to subject matter not covered by other member nations. 5 7 But since the United States rejected copyright and patent regimes, this reciprocal treatment is apparently not required by either convention. 58 The failure to engage international agreements could be considered a benefit. Had the UCC (for example) required the United States to protect the chips of foreign nationals, an imbalance would have occurred: foreign chips would be protected in U.S. markets, but American chip developers would not enjoy protection in foreign markets. Thus, by resorting to special legislation, the United States can use its laws and the value of its markets to influence legal developments abroad.

In the Chip Act, this international leverage is exercised through a series of reciprocity provisions. The Act empowers the President to extend, by proclamation, protection to the nationals of any country that enacts legislation offering substantially the same protection as the Chip Act 59 or that offers to Americans substantially the same protection offered the country's own nationals. 60 In addition, the Act provides interim protection to nationals of countries that have been found by the Secretary of Commerce to be making good faith efforts and reasonable progress toward enactment of adequate protective measures. 61 Unfortunately, the leverage applied by the Chip Act has not had clear international significance. Although the United States credits the Act for developments in those 56. Universal Copyright Convention, supra note 8, art. II, sec. 1; The Paris Convention, supra note 6, art. 2(l).

58. See Budd, supra note 12, at 317 (noting that Congress was told of the chip industry's concern about its ability to compete in world markets).

59. 17 U.S.C. § 902(a)(2)(B) (1988). 60. 17 U.S.C. § 902(a)(2)(A) (1988). 61. 17 U.S.C. § 914 (1988). The Secretary must also find that no one in the foreign nation is engaged in misappropriation of mask works and that issuing the interim order would promote comity. The Act also protects mask works first commercially exploited in the United States, irrespective of the nationality of the developer, 17 U.S.C. § 902(a)(1)(B), and the works of nationals of countries that have entered into chip treaties wisth the United States, 17 U.S.C. § 902(a)(1)(A).

countries that have requested interim protection, 6 2 there is no way to know how their laws would have developed without the Act's influence.

Moreover, the leverage of the Chip Act comes at a cost: the potential disadvantages of legislating outside the confines of international agreements are substantial. For example, the Chip Act puts the United States in the impossible position of essentially supervising the development of the intellectual property law of its trading partners. To date, interim orders have been requested by a fairly large number of countries. 63 Had the Secretary of Commerce or his appointed delegate, the Patent and Trademark Office, 64 assiduously investigated the steps being taken by these countries to enact legislation similar to the Chip Act, these applications would have been burdensome to entertain, since they require difficult substantive evaluations. 6 5 In practice, however, interim orders have been issued rather mechanically. 6 6 The result is that international leverage-supposedly the great advantage of special legislationis lost and the interests of domestic producers are sacrificed to administrative efficiency.

Many difficult issues will arise as applications for presi- 433, 437 (1987).

64. See Interim Protection for Mask Works of Japanese Nationals, Domiciliaries and Sovereign Authorities, 50 Fed. Reg. 12,355, 12,356 (1985).

65. These substantive evaluations include the following: a decision on what features of the Chip Act must be duplicated to make other country's law "substantially similar" to our own; a determination on how much leeway a country should be given in fulfilling these essentials; and, finally, a procedure for evaluating how rapidly a given nation should be expected to move towards actual protection. In evaluating just the first criterion, substantial similarity, it has been suggested that five elements are necessary: protection of the subject matter of chips, the originality standard, the 10year minimum term, limitations on compulsory licenses, and the reverse engineering provision. See REPORT, supra note 62, at 41-43. dential proclamations are processed. The United States will have to react to countries whose chip legislation provides for national treatment, but does not create rights that are in any way similar to those which are available here. 6 7

The procedure used to consider Japan's application for a presidential proclamation 68 raises yet another concern. Rather than process the application through diplomatic channels, the Patent and Trademark Office chose to hold hearings which provided domestic producers with an opportunity to obtain self-serving delays serving their own private interests. 6 9 Such tactics may cause our trading partners to distrust our objectivity and motives. 7 0

The strategy of the Chip Act rests on the assumption that other countries are in a position to respond to its leverage, but this is not necessarily the case: in some countries, not enough chips are manufactured domestically for the Act to provide leverage; in other nations, there is sentiment against creating another opportunity for supracompetiive profits. 7 1 Even in nations that are home to the developers of semiconductors, the local industry may not have the political clout to induce its government to enact legislation that meets the United States' requirements. 7 2 As a result, an extension 67. REPORT, supra note 62, at 16. (July 24, 1986).

69. See REPORT, supra note 62, at 44. 70. Rather than automatically extend protection to the Japanese, the Semiconductor Industry Association (a lobbying group) persuaded the Commissioner of Patents to wait six months in order to see how the legislation would be implemented. See REPORT, supra note 62, at 9-12. Such a probationary period was not contemplated by the Chip Act, and may have diplomatic ramifications resulting from our apparent unwillingness to trust the representations of foreign governments.

71. The developing countries have, for example, been successful in blocking an integrated circuit treaty proposed by the World Intellectual 72. Experience with the protection of computer programs demonstrates that the leverage theory may be wrong. Here, the United States chose copyright protection, thereby foregoing leverage. Other nations nonetheless followed its lead and enacted similar legislation. See Note, The on the three-year window of interim protection has been required.

Finally, the United States' rejection of the applicability of patent and copyright law to chips dilutes the force of international accords. By legislating in a manner that avoids obligations under the UCC and Paris Convention, the United States has in effect declared that these treaties are not as flexible as once thought. 7 4 This, in turn, may influence the manner in which these treaties are construed in the future, to the detriment of new technologies that might otherwise have been seen as clearly falling within their province. 75 Even if the United States' position on the applicability of existing conventions to chips were universally accepted, the victory would be somewhat pyrrhic. If every country fashioned special legislation outside existing conventions to deal with new products, the costs of international protection would become prohibitive. These agreements do more than provide for national treatment: they also lower the cost of 283, 300 n.88 (1985) (noting that the United Kingdom, Australia, Japan, West Germany, France, South Africa, Hong Kong, the Netherlands and Hungary have extended copyright protection to software in the wake of the United States' decision to do so).

73. The PTO has recently promulgated regulations for handling these proclamations. See 37 C.F.R. § 150 (1988); 53 Fed. Reg. 24447 (1988); Semiconductor Chip Protection, Pub. L. No. 100-159 (1987) (extending interim protection untilJuly 1, 1991).

74. Cf id. at 312 (making the same point regardingsui generis legislation for computer software).

75. Moreover, with the enactment of the Chip Act, the United States cannot change its position and decide that international accords cover semiconductors. If the UCC, for example, were construed as flexible enough to encompass chips, then the United States might be viewed as breaching its treaty obligations by failing to extend the Chip Act's benefits to the nationals of its signatories. Id. at 292-93 and 305-07; Wilson & LaBarre, supra note 12, at 69. The Paris Convention allows signatories to decide for themselves what subject matter is protected by patent. Accordingly, the United States would probably not be viewed as breaching this convention by taking chips out of patent protection. See Third World, supra note 72, at 59-60 (WIPO has considered the question whether the Paris Convention applies to chip protection). The UCC, however, mandates a minimum level of protection for subject matter protected. If other countries were to view the UCC as protective of chips, then the level of protection afforded by the Chip Act may not be deemed sufficient. Cf Besarovic, supra note 44, at 159-50 (discussing the duty to protect computer programs).

Imaged with the Permission of N.Y.U. Journal of International Law and Politics [Vol. 20:897 doing business. 76 It is especially ironic that the United States has opted for sui generis protection now, when the intellectual property community is moving to reduce the costs of international protection. The Patent Cooperation Treaty, 77 for example, allows inventors to file and preliminarily to prosecute a single patent application in a central examination office. Centralized prosecution makes it easier for inventors to predict their chances for worldwide protection and saves applicants the cost of translating applications and conforming to certain local prosecution requirements. The European Economic Community offers an even more advantageous opportunity: it has established a central Patent Office with the power to grant patents that are separately valid in each of its member nations. 7 8 Similar (and sometimes even more harmonizing) regional and bilateral treaties exist in other parts of the world.

Presumably, the best that can be hoped for under the Chip Act is that the United States' trading partners individually enact legislation protecting American chip developers or 76. The Paris Convention, for instance, relieves creators of the burden of filing applications in all countries simultaneously. It also facilitates business planning by permitting developers to predict more accurately whether they will succeed in gaining protection in potential markets. By disavowing the applicability of the Paris Convention to chips, the United States has lost this advantage for chip developers. See generally Comment.

Copyright Protection for Computer Firmware in the World Market, 4 Hous.J. Irt' L LAW 203 (1982) (noting the problem of protecting computer hardware in the absence of international agreement on protectability).

77. Patent Cooperation Treaty,June 19, 1970, 28 U.S.T. 7645, T.I.A.S. No. 8733. The United States ratified this treaty in 1975 to permit inventors to file preliminary international patent applications prior to committing resources to the preparation of a formal application, and to receive a preliminary international search report detailing the international prior art bearing on their inventions. Only recently, however, has the United States withdrawn its reservation to Chapter II and enacted implementing legislation, thereby enabling an applicant to obtain an international preliminary examination report that facilitates domestic determinations of patentability. negotiate a separate integrated circuit treaty. 80 Both alternatives would be burdensome, either for developers or for the United States itself (or both). Even if these developments do occur, the timing is likely to be problematic for manufacturers. If protection under these laws is made retroactive, then those who relied on the lack of protection to expand manufacturing facilities will be hurt. If protection is not retroactive, the countries where development occurred first will suffer. 8 ' In the end, the long-range domestic benefits of sui generis legislation may be outweighed by the short-term costs and the other disadvantages that result from legislating outside the confines of an international agreement.

It is likely, however, that countries will legislate for the sole benefit of domestic concerns. If the long-range benefits of specialized legislation outweigh short-term distortions in investment behavior, then it is probable that sui generis legislation will become more prevalent. If this is true, then perhips the international community should respond by forging a generic intellectual property treaty to apply to all new products, or at least a "quasi-generic" agreement to apply to information products, which share special characteristics. 8 2 Such a treaty could minimize the international costs associated with sui generis protective schemes. It also might diminish the dislocations caused by delaying domestic legislation until the needs of new industries become clear.

The World Intellectual Property Organization (WIPO) has been attempting to promulgate such a treaty for some time. It has not, however, met with great success. See Third World, supra note 72, at 59-60 (as with other intellectual property conventions, net importers of the property do not wish to accede to the demands of net exporters).

81. Cf REPORT, supra note 62, at 3 (American chip manufacturers hurt by lack of protection abroad). It is ironic that Fairchild Semiconductor Corp., one of the pioneers in chip development, was nearly bought out by a Japanese manufacturer. See Pollack, Fujitsu Chip Deal Draws More Flack, N.Y. Times, Jan. 12, 1987, at D1, col. 3.

82. The more ambitious undertaking-writing a treaty that would apply to any new technology-would probably eventually suffer from the same defects that currently plague traditional intellectual property regimes. In addition, the generality necessary to produce such a flexible treaty would probably create problems of a different sort. Because information technologies share many characteristics, it would be much easier to create a useful treaty to deal with the problems they uniquely present.

A quasi-generic treaty that accomplishes the goals set out above might contain features of both the Chip Act and the intellectual property treaties currently in force. As identified in Part I, the concerns addressed by the Chip Act are (1) the need to protect developers from pirates so that they can be assured that front-end costs can be recouped, even in the absence of entry barriers and other extra-legal devices that ward off free riders; (2) the interests of the public and the affected ifndustry to unfettered use of newly discovered ideas; and (3) the rights of third parties, such as buyers and those who deal in compatible products, to be free of the risk of infringement.

The international interests identified in Part II include (1) the need to assure foreign developers that they will stand on an equal footing with domestic competitors in all signatory markets; (2) the ability of manufacturers to predict the scope of protection in foreign markets; and (3) the consolidation of application and enforcement procedures in order to lower worldwide transaction costs.

Of course, a generic treaty will not contain all the niceties of existing domestic legislation and treaties. Many provisions in these laws and accords emanate from political compromises between producers, consumers and other players in a particular industry. Thus, they cannot be duplicated easily in a generic measure. The ultimate necessity for finely tailored measures does not, however, make generic legislation unworkable. Rather, it requires that two limitations be accepted. The first is that the protection offered by such a treaty would be temporary, remaining effective for any new technological product only until existing treaties are amended or more specialized legislation is enacted. The second is that a generic treaty would not likely attract the support of nations with radically different needs, such as both developed and developing nations.

Fortunately, these limitations are not critical. Because protection will be temporary, the treaty will be valuable to those countries at the cutting edge of new technological development, whose interests are closely enough aligned that they will agree to a set of mutually binding, beneficial principles. A generic agreement by developed nations may be ad-vantageous to nonsignatories, for the breathing space that the treaty provides will allow the needs of other nations to emerge in time for the drafting of a later treaty addressed specifically to the new technology.

Ideally, the generic treaty will begin with a description of the subject matter protected. This will not be easy, since the subject matter definition in intellectual property legislation tends to be interpreted restrictively rather than inclusively, excluding discoveries that were not foreseeable at the time the measure was enacted. 83 To avoid this result, two solutions are possible. The first would be to adopt a functional definition of subject matter. Since it is the marked divergence between development and production costs that makes protection for information products especially necessary, the treaty could adopt a "sweat-of-the-brow" test, defining products covered by the treaty as works involving high development costs that are not protected by existing intellectual property arrangements. Such a definition has been used in state-law misappropriation cases, which typically protect noncopyrightable materials from piracy when the developer can demonstrate that great labor and expense were employed in creating the work. 8 4 It is a useful definition because it identifies protectable subject matter. Also, it provides a standard of protectability like that in the Chip Act. Although such a test is indefinite, this too may be an asset, for it codifies the notion that the treaty covers products that are not yet in existence. It may, however, create so much uncertainty that the treaty would become useless.

A better approach might therefore be to use the sweatof-the-brow test as a presumption. That is, the treaty could 83. This has been the experience with domestic intellectual property law. See, e.g., White-Smith Music Publishing Co. v. Apollo Co., 209 U.S. 1 (1908) (holding that a perforated piano roll does not constitute a "copy" within the meaning of the Copyright Act of 1897, 29 Stat. 481 ( 1897)). Although this case has been questioned frequently in the lower courts, it is still good law. Collection and Representation of Facts, 76 HARV. L. REV. 1569, 1571 (1963). See also Davidson, Common Law, Uncommon Software, 47 U. Prrr. L. REv. 1037, 1109 (1986) direct member states to recognize the need to protect technologies that have a certain characteristic. Actual protection would, however, depend on more specific actions by the signatory nations themselves. Under this plan, any signatory could declare that a particular new product with a defined degree of novelty falls within the purview of the convention. The interim protection provisions would then be self-executing: all other signatories would automatically protect such works, whether created before or after the subject matter was declared protectable. The convention would continue to cover all similar works until the subject matter received individualized attention, which could take the form of a particularized treaty, sui generis domestic legislation, or integration into existing treaties and legislation. 85 A treaty requiring this degree of deference and trust among signatories may at first appear an unlikely solution. However, it is not unreasonable, since signatories are likely to be highly developed nations with common interests. Nor is the right of a single signatory to act unilaterally cause for concern. If one accepts the view that the reciprocity provisions of the Chip Act have induced chip legislation elsewhere, then the United States has already acted in precisely this manner. This solution has several other interesting and useful features. There are, as we saw, some advantages to tolerating a gap between the time that the technology develops and protection becomes available, for it is during this interval that the basic building blocks of the technology evolve and the needs of the industry become dear. The proposal creates this gap by withholding protection until a member nation decides to act.

At the same time, however, the proposal minimizes the distortion in investment that such an interval creates. Because protection is triggered as soon as a single member recognizes the need for it, the crystallization period is made no longer than absolutely necessary. During the short gap, investment in the field would not be discouraged because the sweat-of-the-brow presumption would act as a promise of protection. Furthermore, the proposal addresses the retro- activity problem that arose in connection with the Chip Act by requiring simultaneous protection by all signatories.

Next, the agreement would create a series of rights that would apply once a particular subject matter was declared protected. These rights must be simple so they can apply effectively to any technology. The Chip Act could serve as a model in this regard. It gives owners the exclusive right to reproduce the work, to import or distribute products embodying the work, and to induce or knowingly cause others to do any of these acts. 8 6 Remedies under the treaty should be limited to monetary damages. Admittedly, injunctions are often necessary to make holders of intellectual property whole. 8 7 However, specific performance would be difficult to administer, would deter competitive conduct that is in the public interest, and would waste resources. a 8 Since a generic statute cannot tailor the right to injunctive relief in a manner that minimizes these dangers, it would be wiser not to make this relief available. 8 9 Instead, the treaty should measure damages by the developer's lost profits and actual damages or by the imputation of royalties based on licenses for comparable works. 9 0 These remedies would offer a degree of deterrence to free riders and assure some return to developers. 9 ' More com- Administration of the treaty should be streamlined. Because protection is temporary, and because many of these technologies are commercially significant for only short periods of time, formalities should be kept to a minimum. A patent-like examination procedure should, accordingly, be rejected. Priority could be awarded under a centralized, firstto-register system. This would be easier and cheaper to administer than a first-to-invent system. Furthermore, registration would encourage early disclosure, provide a mechanism to record transfers, and reduce transaction costs. 92 If kept simple, an existing international organization, such as the World Intellectual Property Organization, could probably be prevailed upon to administer the formal requirements of the treaty. This would keep costs low, both for signatories, who would be saved the cost of quickly adopting special procedures suitable to new art, and for developers, who would be able to get protection in every signatory state for the cost of one application.

Even greater benefits would result from centralizing enforcement proceedings. It is likely that judicial interpretation of the treaty's provisions will be necessary each time a new technology becomes subject to the convention. Centralized adjudication will ensure that disputes are resolved fairly and in a manner binding on all signatories. 9 3

Having provided protection for the developers of new technologies, the generic treaty should also safeguard public access interests. These would be partially protected by a re-92. The United States is one of the few countries that accords patent priority to the first inventor, and it is considering switching to a first-to-file system for all the reasons outlined in the text. See 34 PAT. TRADEMA.RK & COPMIUGHTJ. (BNA) No. 843, at 403 (Aug. 20, 1987) (ABA's Section of Patent, Trademark and Copyright Law approved converting U.S. patent law in this manner in an effort to harmonize patent laws worldwide). Even if not centralized, registration could be used to establish priorities. For example, priority could be based upon the date the creator filed in any country, or in the creator's country.

93. This proposal does not, however, require centralization. Each signatory could designate an office of its government to handle formalities, and domestic courts could enforce rights. Another possibility would be to require the country that first recognized the subject matter to designate an office of its government to administer the treaty for that technology.

gistration system that required deposit, since deposit assures adequate disclosure of the ideas discovered. To ensure that these ideas can be fully utilized, something akin to the reverse-engineering provision of the Chip Act would also be desirable. Unfortunately, this provision may be too technology-specific to be generally useful. It will probably work for semiconductor chips 9 4 because reversing the process for making chips leaves a paper trail that theoretically can be used to document the extent to which the reverse-engineer acted in conformity with the statute. 95 Because other technologies may not share this characteristic, it may be better to rely on a general privilege of fair use, 96 with applications left to case-by-case judicial resolution. Such a privilege could be interpreted broadly enough to protect innocent infringers and third-party interests, such as those of manufacturers of compatible works.

The major safeguard for the public would not, however, lie in fair use. Instead, it would derive from a short duration period, which should be long enough to provide investors a meaningful incentive to innovate, but not so long that the public access interest would be intruded upon severely. Sometime between three and five years, measured from the time of creation or registration, 9 7 would probably be adequate. 98 If a longer period of exclusivity were required for a particular technology, member states would always have the 94. It should be noted that the provisions of the Chip Act have yet to receive sufficient judicial attention to judge their efficacy.

95. That is, to teach, analyze or evaluate its concepts, or to incorporate the results in new, original works. 17 U.S.C. § 906(a) (Supp. IV 1986). See generally Raskind, Reverse Engineering, Unfair Competition, and Fair Use, 70 MINN. L. REV. 385 (1985). Interestingly, the very first (and only) reported case decided under the Chip Act raised questions concerning the interpretation of this provision. See Brooktree Corp. v. Advanced Micro Devices, Inc., No. 88-1750-ECCM Civ. (D.C.S. Cal. Dec. 13, 1988)

Measurement from the time of negotiation would probably be better, since it would provide a clearer line and allow the first developer some time to create a market before the clock started running.

98. For works that change rapidly, three years is likely to be long enough to recoup costs. For works that change more slowly, the free rider problem may not significantly deter investment because investment costs can be spread over more goods.

Imaged with the Permission of N.Y.U. Journal of International Law and Politics [Vol. 20:897 option of extending the term through specific legislation. 99 In fact, the short term provided by the generic treaty would induce signatories to give rapid consideration to protection questions.

There may be situations in which signatories decide that a given class of works should not receive intellectual property protection at all. In that case, further legislation would be required to nullify the effects of the treaty. Under this proposal, these nullification measures would have the effect of withholding protection from works created after the date of nullification. Works created before that date would, however, retain the protection offered by the treaty.

The nullification procedure highlights the major drawback of the proposal: it creates rights in the absence of domestic legislative consideration that cause major competitive consequences. However, the cost is not unreasonable, espedally in light of the proposal's advantages. The costs of the nullification procedure in particular may be more apparent than real. If a technology is important enough to attract a signatory's attention in the first place, it is likely to be important enough to justify some consideration by every nation that signed the treaty. Besides, existing intellectual property legislation is somewhat indefinite. If a nation were in fact determined to release a given technology into the public domain incontrovertibly, some legislative declaration that existing law did not apply would probably be required. Such legislation would be no less burdensome to enact than the nullification of the obligations created by this proposal.

Of course the nonretroactivity of nullification means that some works will receive protection where none "should" have been accorded. Although this is certainly a cost, it is not a severe one. Works that receive protection enjoy it only for a short time, and the number of works that are ultimately accorded exclusivity can be limited through early legislative consideration of the subject matter by member nations.

As to the larger issue, the drawbacks of the proposal are mainly outweighed by the advantages it generates. The 99. The copyright law enacted in 1976 is precedent here. Works protected by prior law found their terms lengthened by the transition provisions of the new law. See, e.g., 17 U.S.C. § 304 (1982). blanket promise of protection for capital-and labor-intensive developments will reduce the risks that early developers face. Thus, it will spur investors and inventors to devote their time and effort to bringing new and better products to the public. The treaty buys nations time in which to find the optimum protection strategy for new technologies. It minimizes the costs that would otherwise accompany resort to sui generis legislation. Most important, a treaty of the sort proposed here lays the ground work for further and better international cooperation.

Intellectual property lawyers have long coped with the problem of change, for their clients uniformly engage in either producing or consuming the fruits of research and development. Recently, however, the possibility of legal innovation has accompanied the certainty of technological innovation. The enactment in the United States of the Semiconductor Chip Protection Act foreshadows the interjection of new layers of complexity into the legal regime. The Chip Act may indicate that in the future, protection tailored to a specific industry or product will be an important goal of domestic legislation.

Such a step would entail considerable costs. For developers who must rely on intellectual property law, a switch to sui generis legislation as the predominant mode for creating exclusive rights means that temporal gaps will be created during which rights will be unclear. These gaps could discourage adequate investment in research and ultimately create inequities between early developers and those who build on their discoveries. The first group will have contributed the most to the new art, but if sui generis legislation is not made retroactive, they will also be the ones who find their inventions unprotected.

As with inventors, so it is with nations. The country that is the site of new discoveries may find itself in a difficult position. The first to enact specialized legislation will face the choice of leaving its nationals at a competitive disadvantage in world markets, or exercising leverage in a manner that generates considerable administrative costs and perhaps political ill will. The proposal made here may seem peculiar since, in a sense, it suggests putting the cart before the horse: creating international obligations in the absence of domestic legislation. But although unusual, this strategy has important advantages. As intellectual products are developed with the international marketplace in mind, international protections assume greater significance. A presumption of protectability in important markets, coupled with a minimal set of rights, would alleviate many of the domestic hardships that sui generis legislation can create. Centralized administration and a mechanism for simultaneous protection in all markets would similarly reduce the international tensions to which these measures give rise.

Admittedly, generic accords carry costs that have not been explored adequately in this article.' 00 They can stultify the development of the law and they can have disparate impacts on different economies. But these disadvantages will be minor as long as the proposal is recognized for what it is: a measure that creates temporary rights in a class of intellectual property (information products) that shares common features, aimed at achieving an accord-in-principle among nations whose interests are closely aligned.

While this article sets out the general contours of a generic convention, many issues remain to be resolved. Nonetheless, the exercise is, hopefully, valuable. It demonstrates that a generic treaty is possible and that issues that loom as major obstacles are not insurmountable (indeed, some carry hidden benefits). If the products of the future turn the world into a single market, then the incentives offered by that market must be of an international dimension. Thus, it may no longer be feasible to rely on domestic legislation followed by international agreement. Placing the cart before the horse may, in the future, be unavoidable.