Wesley M. Cohen
Stephen A. Merrill
National Research Council
Since 1980, successive changes in patent policy, one of the oldest elements of U.S. technology policy, have expanded intellectual property rights and strengthened the position of patent owners. The establishment of the Court of Appeals for the Federal Circuit (1982), which consolidated all appeals from patent case decisions of federal district courts in a single specialized court, led to a sharp increase in plaintiff success rates in patent infringement law suits. A number of large, widely publicized judgments in infringement cases also suggested a marked rise in the value of patents, at least at the upper end of the distribution.1 Encouraged by a series of court decisions, patenting has been extended to new scientific and technological domains such as life forms, genes, software, and methods of doing business. A federal statute enacted in 1980 encouraged universities and other nonprofit institutions conducting research with public funds to obtain and license patents. Partly as a result of these changes in the policy environment, business strategy in some sectors has placed a greater premium on acquiring and using patents (cf. Kortum and Lerner, 1999). This is especially the case in biomedicine and information technology.
At the same time these policy changes have raised concerns about their impact on innovation and the factors driving innovation. Along with the growth in patenting itself, there has been an increase in patent litigation, which some consider to be an unproductive increment to the cost of innovation. Others see in the
proliferation of software and business method patents a weakening of the standards of novelty and non-obviousness, thereby undermining the purpose of patents to provide an incentive to those who innovate in a genuine way. The proliferation of patents in biotechnology, especially those involving DNA sequences, has raised a different set of concerns—whether intellectual property rights are becoming so fragmented that assembling the rights necessary to commercialize a new therapy or drug is prohibitively costly and whether some promising lines of research are abandoned prematurely.
This volume assembles papers commissioned by the National Research Council’s Board on Science, Technology, and Economic Policy (STEP) to inform judgments about some of these institutional and policy developments made over the past two decades. The chapters fall into three areas. The first four chapters consider the determinants and effects of changes in patent “quality.” Quality refers to whether patents issued by the U.S. Patent and Trademark Office (USPTO) meet the statutory standards of patentability, including novelty, non-obviousness, and utility. The fifth and sixth chapters consider the growth in patent litigation, which may itself be a function of changes in the quality of contested patents. The final three chapters explore controversies associated with the extension of patents into new domains of technology, including biomedicine, software, and business methods.
The style of these contributions varies. Several are based on descriptive and in some cases qualitative data. Others develop and test hypotheses in the manner of empirical economics. And one chapter is a theoretical exploration of the costs and benefits of a proposed institutional change intended to improve patent quality. These contributions are discussed below.
We are interested in questions of patent quality, litigation, and extension into new areas of technology because we are concerned with how the patent system affects the rate and direction of technological change. The trouble with trying to understand the import of the changes in patent policy over the past two decades is that we have a limited understanding of the effects of the patent system to begin with. There has been little systematic empirical analysis of the impact of patents on innovation. Even the narrower question of whether patenting stimulates research and development investment has only recently begun to be studied.
There are theoretical as well as empirical reasons to question whether patent rights advance innovation in a substantial way in most industries. The rationale for patent protection is to augment the incentives to invent by conferring the right to exclude others from making, using, or selling the invention in exchange for the disclosure of its details. Although the prospect of monopoly rents should induce inventive effort, the costs of disclosure can in some circumstances more than offset the prospective gains to patenting (Horstmann et al., 1985). “Strengthening” patent protection enhances the value of not only a given firm’s patents but also those of its rivals who may be able to constrain the original firm’s ability to commercialize its innovations (Jaffe, 2000; Gallini, 2002).
Merges and Nelson (1990) and Scotchmer (1991) argue that where technological advances build upon one another cumulatively, as is increasingly the case, broad patent protection on upstream discoveries may slow the rate of technical change by impeding subsequent innovations. Heller and Eisenberg (1998) suspect that in the domain of genetics patenting has been extended to such finegrained inventions that the intellectual property covering any new drug or therapy may now be so complex and dispersed that heterogeneous patent owners may not agree on the licensing terms necessary to bring a product to market. Cohen and colleagues (2000) point out that in industries such as microelectronics there can be hundreds of patentable elements in one product, with the consequence that typically no single firm ever holds all of the rights necessary for its commercialization. In complex product industries generally and in the semiconductor industry in particular (Hall and Ziedonis, 2001) such mutual dependence commonly spawns extensive cross-licensing. In these cases the kind of breakdown hypothesized by Heller and Eisenberg does not often occur, but the need for patents as bargaining chips to arrive at satisfactory cross-licensing agreements may stimulate expensive patent portfolio races among industry incumbents. Especially if powerful incumbents insist on trading like-for-like in licensing arrangements (Shapiro, 2001), firms with modest or negligible patent holdings may be barred from entry. To the extent that entrants are a vehicle for innovation and their entry is obstructed, technical advance may suffer.
Empirical work by a number of economists over nearly fifty years suggests that patents play a prominent role in stimulating invention in only a few manufacturing industries (Scherer et al., 1959; Taylor and Silberston, 1973; Mansfield, 1986). Surveys of R&D managers by Levin and colleagues (1987) and, more recently, Cohen and colleagues (2000) found that in most industries patents are judged to be less important means of protecting innovations than, for example, being first to market or retaining know-how as trade secrets.2
Although we should therefore not assume that patents invariably induce innovation, neither should we assume the contrary. Firms may rely more heavily on other means of protecting innovations, but patents may still yield a return. Arora and colleagues (2003) recently showed that patents do appear to stimulate R&D across the manufacturing sector, although the magnitude of the stimulus varies greatly from industry to industry. Levin and colleagues (1987), Mansfield (1986), and Cohen and colleagues (2000) all find that pharmaceutical and medical equipment R&D benefits the most from patenting.
The literature on the impact of patents on innovation must be considered emergent. One reason is that the effect of patent policy has many dimensions, some fundamental to understanding the determinants of innovation generally,
and these continue to challenge scholars both theoretically and empirically. For example, although the literature has identified offsetting impacts of intra-industry R&D knowledge flows on R&D incentives (Cohen and Levin, 1989), it is not clear how the protection patents afford and the information they disclose contribute to those flows and associated incentives. Another reason is insufficient data. Although patent and patent citation data are readily available and extensively used in the study of innovation (Jaffe and Trajtenberg, 2002), information on the uses and impacts of patenting is quite limited. For example, without data on the incidence and terms of patent licensing and associated fees and royalties, it is difficult to assess the efficiency and social welfare effects of markets for technology whose growth can depend on the allocation and strength of patent rights (Arora et al., 2001).
Other important data limitations relate to conflicts over patent rights. Although the studies in this volume and others are informative about trends in, parties to, costs of, and determinants of outcomes in formal patent lawsuits, litigation is only one aspect of firms’ maneuvering to exploit patent positions against rivals and to enforce their intellectual property rights. Letters of notification claiming infringement and demanding licensing agreements are far more common than lawsuits and may have a significant effect on firm behavior, including R&D activity and entry. Moreover, the direct cost of prosecuting or defending cases does not represent the full range of litigation costs affecting innovation. Although we know anecdotally that participating in legal strategizing and discovery can consume considerable time of corporate managers and technical staff, we have no data on the associated opportunity costs, let alone on the innovative paths not taken as a result of actual or threatened litigation. In short, scholars gravitate to available data, leaving broad economic impacts only partially examined.
We now turn to reviewing the contents of this volume. Although sharing many of the limitations discussed above, these contributions advance our understanding of the determinants and social welfare implications of patent quality, litigation, and the extension of patenting into new technological domains—some of them in novel ways.
Over the past decade the quality of issued patents has come under attack. The claim that quality has declined in a broad or systematic way has not been empirically tested, although Quillen and Webster’s (2001) claim that patent approval rates are much higher (on the order of 80 to 90 percent or more) than officially reported is consistent with the hypothesis. The conjecture that patent quality is declining has been characterized in two ways. First, Barton (2000) and others have suggested that the standards for patentability—especially the non-obviousness standard—have been relaxed largely as a result of court interpretations. Sec-
ond, other critics have suggested that the USPTO frequently issues patents for inventions that do not conform to the standards for patentability, especially in technology areas that are newly patentable, notably genomics, software, and business methods. Although separable in principle, the notion that standards for patentability are slipping and the notion that the USPTO is failing to apply the legal standards appropriately are difficult to distinguish in practice.
The first two chapters in this volume focus on whether the existing standards have been appropriately applied by the USPTO, not whether the standards themselves have changed. Although Griliches (1990) attempted to estimate a patent office “production function” and concluded that the number of patent examiners is the major determinant of the number of patents issued, these studies are the first to examine the impact of characteristics of the patent examination process and the examiners themselves on different indices of patent quality. This is an avenue of research that needs to be pursued vigorously if we are to have even a modest research basis for making informed judgments about how USPTO organization, management, and resources affect the quality and level of its output.
In both of the first two chapters the measure of quality is the likelihood that a patent’s validity is challenged in litigation and is upheld or overturned by the courts. Merges (1999) suggests that the quality of patent examination, associated with what he calls “front-end costs,” can have an important effect on the likelihood of whether a patent will be litigated, entailing “back-end costs.” In modeling the choice between settlement versus litigation in the case of infringement, Meurer (1989) elaborates the intuition behind this assertion, namely that greater uncertainty over patent validity will lead to a higher incidence of both infringement and litigation. The chapter in this volume by Jonathan Levin and Richard Levin highlights the costs in addition to litigation that are associated with poor quality patents. They argue, “the holders of dubious patents may be unjustly enriched and the entry of competitive products and services that would enhance consumer welfare may be deterred.” Further, they point out that “…uncertainty about what is patentable in an emerging technology may discourage investment in innovation and product development until the courts clarify the law, or, in the alternative, inventors may choose to incur the cost of product development only to abandon the market years later when their technology is deemed to infringe.”
John King in his chapter provides the only evidence to date suggesting a link between the care with which patents are examined and subsequent litigation. In a simple regression controlling for both cost of litigation and USPTO examination (technology-based) groups, he shows a strong negative effect of average examination hours per examination group on the rate at which issued patents were involved in legal complaints. Although not providing evidence of a direct effect of quality on lawsuits, the result suggests that greater effort dedicated to examination does have a shielding effect. With a breakdown by examination group of the annual rate of patents involved in complaints per patent allowed, and esti-
mates of median out-of-pocket legal costs, he even suggests, under plausible assumptions, that the social costs of increasing the resources dedicated to examination may be more than offset by the savings in “back-end” legal costs.
The study by Iain Cockburn, Samuel Kortum, and Scott Stern shifts the focus from the effects of characteristics of the examination process to the effects of characteristics of the examiners themselves. First, they demonstrate considerable heterogeneity across examiners, even within USPTO examination groups, with respect to tenure in the Office, the number of patents they have examined over time, and the degree to which their patents are subsequently cited in other patents. The authors then explore the link between those characteristics and subsequent validity decisions by the Court of Appeals for the Federal Circuit (CAFC) for the small set of patents that are ultimately litigated at that level. Perhaps their most noteworthy results are negative ones. They find that neither the length of experience of examiners, measured either in years or cumulative number of patents issued, or examiner workload, represented by the count of issued patents in the three months prior to issue date of a given patent, appears to have any impact on whether an issued patent is judged valid by the CAFC.
If the result means that fewer examination hours per patent, corresponding to a higher workload, has no impact on measures of patent quality, then we have a finding contrary to King’s, described above. As the authors themselves acknowledge, however, these results warrant skepticism. They are based, for example, on what may be an unrepresentative sample, namely patent suits selected by the CAFC to give the court opportunities to decide significant points of law. To the degree that more representative validity decisions are made at the trial court level, it may be useful to repeat the analysis with a sample of those cases, especially in light of the importance of the relationship being studied for the management of the USPTO.
Cockburn and colleagues arrive at one clear result, that patents issued by examiners who tend to issue patents generating more citations in subsequent patents are more likely to be judged invalid by the CAFC. Although the robustness of their result remains to be seen, the authors interpret it as showing that some examiners systematically approve claims that are broader in scope and that such claims tend to be more vulnerable to invalidity judgments. The authors further interpret their result as suggesting that the courts provide a needed check on the predisposition of some examiners to issue patents with broader claims. This finding also underscores the observation of Gallini (2002), among others, that patent “strength” is an ambiguous and possibly misleading concept. What are commonly considered to be two dimensions of “strength”—claim breadth and enforceability—may well be at odds.
Reflecting growing concerns over patent quality, Merges (1999) has suggested that the United States consider adopting an administrative procedure similar to the post-grant patent opposition process employed in Europe and, more
recently, Japan.3 This volume provides two complementary perspectives on postgrant review of patent validity—first, an empirical comparison by Stuart Graham, Bronwyn Hall, Dietmar Harhoff, and David Mowery of certain features of the European opposition system and the current U.S. patent re-examination procedure and, second, Levin’s and Levin’s theoretical analysis of the social welfare effects of post-grant opposition. In what is almost certainly the first cross-national comparison of patent institutions and procedures, Graham and colleagues show that the European and U.S. systems for reviewing issued patents are very different from one another. The chief differences have to do with the roles of both challengers and patent holders in the proceedings, the grounds for challenge, and whether the outcome limits subsequent litigation. Among other things, the U.S. re-examination process makes it advantageous for the patent holder to request review to accommodate newly discovered prior art and disadvantageous for challengers to initiate re-examination (with the result that the owner-initiated cases constitute 40 percent of the total). This may account for the authors’ finding that only 0.3 percent of U.S. patents were re-examined in the 1981-1998 period whereas 8.3 percent of European patents were subject to opposition. Graham and colleagues also show that in Europe oppositions focus on more commercially important patents and do not appear to be used by large established firms as a competitive weapon against smaller firms.
The most compelling reason to have a post-grant opposition procedure, according to both sets of authors, may be a reduction in patent litigation and its associated costs. An opposition proceeding that lowers uncertainty about a given patent’s validity should make it less likely that the patent will be later litigated, assuming that both infringement and a mutual reluctance to settle patent disputes are associated with uncertainty about patents’ validity. Further assuming that oppositions are on average resolved more quickly and inexpensively than lawsuits, Levin and Levin show that an opposition procedure can offer significant social welfare advantages over the current reliance on litigation to resolve issues of patent validity, especially in newly patented technologies where patent quality is most uncertain.
Graham and colleagues are unable to confirm the Levins’ prediction that the use of opposition should substitute for subsequent litigation over validity if the process is speedier and cheaper. That is partly because European oppositions are relatively unconstrained by deadlines and tend to drag on for extended periods of time. Furthermore, the authors were not able to collect data on the litigation histories of patents that have gone through European opposition proceedings in time to
incorporate in their analysis. Consequently, we do not yet know whether the European experience supports the plausible arguments for an opposition system.
There is another source of uncertainty regarding the social welfare impact of an opposition procedure. If prospective challengers can invalidate patents without incurring the high costs of litigation, there will be more challenges and, conceivably, a greater aggregate cost to society. That may well be the case for pharmaceutical and biotechnology patent disputes. Harhoff and Reitzig (2002) calculate that in Europe over 8 percent of such patents are opposed, while in the United States the litigation rate is just over 1 percent (Lanjouw and Schankerman, 2001). If the price elasticity of demand for post-grant validity checks via either opposition or litigation is roughly unity, then the argument for instituting an opposition procedure must be based on some benefit other than savings on litigation. Levin and Levin acknowledge the point and suggest that there are other substantial gains from a vigorous system of post-grant review of validity. In particular, it would help ensure that society realizes the benefits of conferring monopoly profit only upon those patented inventions representing a genuine technical advance and deserving of encouragement while minimizing the consumer welfare losses that invalid patents may impose.
Because the value distribution of patents is highly skewed (Scherer et al., 1959; Scherer and Harhoff, 2000) and the majority of patents are not even commercialized, an opposition procedure may well be more efficient than devoting additional resources to examining all patent applications more rigorously. Does the European opposition process tend to select for close scrutiny prospectively valuable patents? Consistent with Harhoff and colleagues’ findings (1999, 2002) on the determinants of opposition for European Patent Office (EPO) biotechnology and pharmaceutical patents, Graham and colleagues find that the likelihood of opposition does indeed increase with forward citations, which Trajtenberg (1990) shows to be an indicator of social value.
The almost tenfold growth in patent litigation over the past two decades (Merz and Pace, 1994; Moore, 2001) and the escalating cost of prosecuting and defending an infringement suit (AIPLA, 1997, 2001) have raised concerns about the effects of the cost of patent litigation on R&D incentives and innovation (Barton, 2000). Although some observers have labeled these costs a tax on innovation, it may be that litigation is an essential complement to patenting itself and therefore part of the investment in innovation. Even if that is the case, it is appropriate to ask if the costs of litigation can be contained or reduced, possibly to the benefit of investment in innovation. The two chapters on patent litigation in this volume, by Jean Lanjouw and Mark Schankerman and by Rosemarie Ziedonis, take some initial, complementary steps to addressing these questions.
Building on their earlier work (2001), Lanjouw and Schankerman consider patent litigation in the United States as a whole from 1978 through 1995 and probe the features of patents and their owners that affect decisions to file suits and their outcomes. First they show that litigation rates—defined as decisions to file a suit normalized by numbers of patents—vary substantially across technology fields and that, once disaggregated in that way, litigation rates have not changed over the two decades of rapid growth in patent law suits. In other words, litigation has simply kept pace with patenting. Next they relate a range of characteristics of patents and their owners to decisions to file suit. Having a larger portfolio of patents or a history of repeated interaction with other established firms in the industry seems to reduce the probability of becoming involved in a suit. Consistent with that finding, asymmetry in size affects the probability of filing a suit; it is less likely that a firm that is large relative to likely disputants will file a suit. The authors also find that patents with a larger number of forward citations, indicative of greater value, are more likely to be the objects of suits while patents with more backward citations, reflecting more derivative or incremental innovation, are less likely to be litigated. Patents subject to greater rates of self-citation, interpreted as signaling more cumulative technologies, are more likely to be the subjects of suits, suggesting that the probability of conflict over intellectual property increases when pioneers and followers need to come to terms. This finding is consistent with the arguments of Scotchmer (1991) and Merges and Nelson (1990).
These various characteristics of patents and owners appear to affect only decisions to file suits, not their outcomes, including the likelihood of settlements prior to verdicts. Lanjouw and Schankerman interpret this finding as suggesting that the patent and firm characteristics that reduce suits—asymmetric firm size, large portfolios, and repeat interactions—may lower the social cost imposed by patent litigation. The analysis raises further issues, however. Larger firm size or possession of a large patent portfolio may reduce the direct costs of patent litigation but these characteristics may be associated with other costs. For example, large firms’ threat letters (i.e., letters of notification of infringement) may chill smaller firms’ incentives to undertake innovation in selected markets and may even discourage entry. Although the analysis of firm behavior regarding formal litigation and costs arising from it is extremely useful, we are still far from a complete understanding of the costs and benefits of conflict over intellectual property rights—as far as we are from fully understanding the costs and benefits of licensing.
Building on her earlier work with Hall (2001), Ziedonis’s chapter in this volume presents a detailed picture of litigation trends in a single industry—semi-conductors—between 1973 and 2000. The semiconductor industry has experienced some of the most visible patent settlements over the past two decades as well as a rapid growth in patents per dollar invested in R&D, rising from 0.3 in
1982 to 0.8 in 1997. Some scholars have suggested that this is the result of patent portfolio races in which incumbents aggressively amass patents for use in crosslicensing and to fend off patent infringements suits (Cohen et al., 2000; Hall and Ziedonis, 2001).
A key question is what the pattern of behavior in semiconductors tells us about whether patenting is serving its constitutionally mandated purpose of stimulating innovation. One possibility is that the sharp increase in patent filings reflects little more than a non-cooperative bargaining game wherein rival incumbents are amassing larger and larger portfolios for defensive reasons and the prospect of obtaining patents provides little incentive for R&D beyond that provided by other means of protecting inventions, such as secrecy or the exploitation of lead time advantages. Hall and Ziedonis (2001) call into question whether patenting stimulates R&D in semiconductors to any significant degree. On the other hand, the recent analysis by Arora and colleagues (2003) of the impact of patenting on R&D in major U.S. manufacturing sectors suggests that patenting does stimulate R&D in semiconductors, although not as much as in most other industries.
In describing litigation rates in the industry, Ziedonis emphasizes a different metric than that of Lanjouw and Schankerman. She acknowledges that the fraction of patents involved in legal disputes in the industry has not risen over time. To assess the social costs of patent litigation, however, Ziedonis calculates a litigation rate normalized by R&D expenditures. She shows that, relative to R&D spending, the patent litigation rate in semiconductors has risen 93 percent during the 1986-2000 period. Her interpretation is that: “…semiconductor firms have been directing a larger share of their innovation-related resources towards defending, enforcing, and challenging patents in courts since the mid-1980s….” To say that patent litigation has made innovation more costly is not to suggest that it has actually been a net drag on innovation in the industry, and Ziedonis does not go that far. To address that important question, one would have to estimate the net contribution to innovation that patents and their enforcement have made.
A prominent concern about the contemporary use of patents has to do with barriers to entry in industries such as semiconductors in which large patent portfolios are acquired and used as the basis for cross-licensing and that licensing takes the form of trading “like for like” (Shapiro, 2001). Barriers to becoming an integrated semiconductor manufacturer are no doubt very high, more as a consequence of the huge capital requirements of production than of incumbents’ patent portfolios. But Ziedonis suggests that since the 1980s patent protection has under-pinned the entry and growth of fabless chip design firms, an important component of the industry. Interestingly, fabless design firms tend to be more R&D-intensive and more prone to litigation than integrated manufacturers. The question of the overall impact of patents and their enforcement on the semiconductor industry’s growth, structure, and technological advance remains open.
CHALLENGES POSED BY NEWLY PATENTED TECHNOLOGIES
Apart from institutional changes, an important development in the patent system over the last generation has been the expansion of patentable subject matter. Patenting was extended to life forms with the landmark Supreme Court case of Diamond v. Chakrabarty in 1980 and subsequently to genes and gene fragments. The CAFC also endorsed the limited patentability of software, as an adjunct to a physical process, in its 1981 decision in Diamond v. Diehr. Seventeen years later, in State Street Bank and Trust v. Signature Financial Group, the same court rejected arguments against patents on “methods of doing business” and appeared to dispense with virtually all limitations on software-related subject matter.
The extensions of patents to biotechnology, software, and business methods have aroused controversy on a variety of grounds—the ethical implications of patenting life forms, especially human genetic material, the alleged burden on research and product development of patents on upstream research tools and foundational discoveries, and the proximity of some software and business method developments and DNA discoveries to ideas and information theoretically outside the scope of the patent system.
Three chapters in this volume consider aspects of American experience with the patenting of software, business methods, and biotechnology. They do not go very far in addressing the broad concerns raised by the extension of patenting into these areas, but they do illuminate some challenges they have posed for the patent system and its impact on innovation.
Software and Business Methods
Stuart Graham and David Mowery describe the evolution of software from a “relatively open intellectual property regime to one in which formal protection, especially patents, figures prominently.” They attribute this shift to the diffusion since the 1980s of microcomputing, giving rise to the growth of packaged software and, more recently, the internet. The authors document the rapid growth in patenting activity of packaged software companies but acknowledge that, historically, manufacturers of computers and other electronic systems have been the most aggressive software patenting companies. Their analysis highlights the corresponding decline in the “copyright propensity” of the largest packaged software firms, and they speculate that the shift is a function of the strengthening of patent relative to copyright protection in recent years.
Graham’s and Mowery’s analysis raises several questions for future work. For example, how will the use of software patents continue to evolve? What will be the effect on industry entry of aggressive software patenting on the part of larger incumbents? As software tends to develop cumulatively, what will be the effect of contemporary upstream patents on subsequent innovation in the indus-
try? Finally, a good deal of software development and patenting stands outside of the well defined software technology categories examined by Graham and Mowery and, for the time being, eludes description and analysis.
Both the chapter by Graham and Mowery and the chapter by Allison and Tiller attempt to address the controversial question of patent quality in software generally and business methods in particular. Graham and Mowery use a relative measure of the “importance” of software patents—the forward citations to their sample of the patents obtained by large packaged-software firms relative to citations to all software patents for the technology areas that they examine. They find that citations to the large firms’ patents exceeded those to all software firms’ patents and that this ratio moved modestly upward through 1996. Although not reflective of the importance or quality of software patents in general, their data suggest that the relative importance of patents issued to large producers of personal computer software has not declined during a period in which their patenting rate has accelerated. A similar measure of relative importance suggests the software patents of large electronics firms also have not declined during this period.
Many business methods patents have come under attack for either being obvious or not novel in light of the contemporary practice and teaching of business methods. With regard to novelty, critics have claimed that the USPTO is either inattentive or lacks access to relevant non-patent prior art in business literature. In the first attempt to evaluate this claim empirically, Allison and Tiller compare the number of non-patent prior art references (i.e., backward citations) in a sample of internet business method patents to those found in a sample of all other patents. They find that there were substantially more total references, patent references, and non-patent references in the business methods patents than in the general sample of patents. Nevertheless, Allison and Tillers’ data cannot answer several intriguing questions. For example, is the body of non-patented prior art in the area of business methods so large or diverse that examiners are still missing a good share of it? Does the examination process overlook some business methods that are in common use but not documented in written sources? Notwithstanding these uncertainties, the USPTO appears to be paying more attention to non-patented prior art in the examination of business method patent applications than is widely assumed to be the case.
Economic research has made a convincing case that in at least one area— pharmaceuticals—patents have played a critical role in stimulating technical advance (Scherer et al., 1959; Mansfield, 1986; Levin et al., 1987; Cohen et al., 2000). In recent years, however, a few scholars have speculated that in some circumstances the opposite may be the case—patents may now be impeding drug discovery and development. Two related concerns about the patenting and licensing of biomedical innovations have been articulated. First, Heller and Eisenberg
(1998) posited what they termed a “tragedy of the anticommons,” resulting when heterogeneous players assert numerous property rights claims to separate building blocks for some product or line of research. In these circumstances negotiations to assemble the rights may fail, blocking otherwise promising lines of research or product development. Concern focused initially on access to “research tools” (i.e., inputs into the discovery of new drugs, diagnostics, and therapies), which some firms and many public research institutions were beginning to patent extensively. A related argument, previously developed in general terms by Merges and Nelson (1990) and Scotchmer (1991), is that patents on upstream discoveries if sufficiently broad in scope can impede follow-on discoveries and development if access to the foundational intellectual property is restricted.
In this volume, John Walsh, Ashish Arora, and Wesley Cohen present the first empirical evidence regarding the impact of research tool patenting and licensing on biomedical innovation. The authors draw upon 70 interviews with firms, intellectual property practitioners, and university and government personnel to address two questions: 1) whether an emergent anticommons is in fact impeding the development and commercialization of new drugs, diagnostics, and other therapies; and 2) whether restricted access to patents on upstream, foundational discoveries is blocking important follow-on research and innovation. The preconditions for both results appear to exist. There are now more patents associated with any new therapeutic product. Furthermore, since the passage in 1980 of the Bayh-Dole Amendment, encouraging nonprofit research institutions and small businesses to acquire title to inventions developed with public support, many research universities, the locus of fundamental upstream discoveries, have been patenting and licensing more aggressively.
Walsh and colleagues do not find, however, that these developments are yet impeding the development of drugs or other therapies in a significant way. First, the number of patents required for most projects remains manageable. Most importantly, firms and other institutions have developed a number of “working solutions” that limit the effects of the intellectual property complexities that exist. These range from the normal responses of licensing and occasional litigation to other less visible solutions, including fairly pervasive infringement of patents in the course of laboratory research at a pre-product stage. Such infringement seems to be common in both public research institutions and firms and is informally rationalized as causing no commercial harm and, in any event, shielded from infringement liability by the court-interpreted “research exemption.”4 Finally, the
National Institutes of Health (NIH), other influential research funders, and some scientific publications have encouraged ease of access to important research materials and tools.
An exception to this general finding involves cases where the intellectual property required for follow-on research is the same intellectual property associated with diagnostic tests for genetic predisposition to particular diseases. Walsh and colleagues also suggest that restricted access to upstream discoveries could substantially impede subsequent research and development for particular disease categories and therapies in the future and therefore recommend careful monitoring. Were significant impediments to emerge, a solution would, however, be difficult to devise given the importance of patents to biomedical innovation generally.
Over the past two decades, policy and court decisions have moved patent policy toward a regime of stronger enforcement and extended patents into new domains of technology. At the same time, firms and public research institutions in technology sectors important to the economy now and in the future —biotechnology and pharmaceuticals and computer hardware components and software— have embraced patenting aggressively. These changes have proceeded, however, with a limited understanding of their consequences. The National Academies’ STEP Board’s initiative to support original research on the patent system represents a modest step to illuminate these consequences. The chapters that follow focus on the issue of patent quality, the transactions costs imposed by patent enforcement through litigation, and some of the challenges posed by the extension of patentability to new domains, a process that almost certainly will be repeated indefinitely. Few of these contributions point to particular policy prescriptions and those that are prescriptive are not definitive, but they have informed the findings and recommendations of the STEP Board’s Committee on Intellectual Property Rights in the Knowledge-Based Economy and should be useful in other policy deliberations.
AIPLA. (1997, 2001). “Report of the Economic Survey.” Arlington, VA: American Intellectual Property Law Association.
Arora, A., M. Ceccagnoli, and W. M. Cohen. (2003). “R&D and the Patent Premium.” NBER Working Paper No. 9431.
Arora, A., A. Fosfuri, and A. Gambardella. (2001). Markets for Technology: Economics of Innovation and Corporate Strategy. Cambridge, MA: MIT Press.
Barton, J. (2000). “Intellectual Property Rights. Reforming the Patent System.” Science 287: 1933-34.
Bresnahan, T. (1985). “Post-entry Competition in the Plain Paper Copier Market.” American Economic Review.75(2): 15-19.
Cohen, W. M., and R. C. Levin. (1989). “Empirical Studies of Innovation and Market Structure.” In R. Schmalensee and R. Willig, eds., Handbook of Industrial Organization. Amsterdam: North-Holland Publishers.
Cohen, W. M., R. R. Nelson, and J. P. Walsh. (2000). “Protecting Their Intellectual Assets: Appropriability Conditions and Why U.S. Manufacturing Firms Patent (or not).” NBER Working Paper 7522.
Gallini, N. (2002). “The Economics of Patents: Lessons from Recent U.S. Patent Reform.” Journal of Economic Perspectives 16(2): 131-154.
Griliches, Z. (1990). “Patent Statistics as Economic Indicators: A Survey.” Journal of Economic Literature.28(4): 1661-1707.
Hall, B. H., and R. H. Ziedonis. (2001). “The Patent Paradox Revisited: An Empirical Study of Patenting in the U.S. Semiconductor Industry, 1979-1995.” RAND Journal of Economics.32(1): 101-128.
Harhoff, D., and M. Reitzig. (2002). “Determinants of Opposition Against EPO Patent Grants: The Case of Biotechnology and Pharmaceuticals.” CEPR Discussion Paper DP3645.
Harhoff, D., F.M. Scherer, and K. Vopel. (1999). “Citations, Family Size, Opposition and the Value of Patent Rights.” Discussion paper at http://emlab.berkeley.edu/users/bhhall/harhoffetal99.pdf
Heller, M. A., and R. S. Eisenberg. (1998). “Can Patents Deter Innovation? The Anticommons in Biomedical Research.” Science 280(May 1): 698-701.
Horstmann, I., J. M. MacDonald, and A. Slivinski. (1985). “Patents as Information Transfer Mechanisms: To Patent or (maybe) Not to Patent.” Journal of Political Economy 93: 837-858.
Jaffe, A. (2000). “The U.S. Patent System in Transition: Policy Innovation and the Innovation Process.” Research Policy 29: 531-558.
Jaffe, A., and M. Trajtenberg. (2002). Patents, Citations, and Innovations: A Window on the Knowledge Economy.Cambridge, MA: MIT Press.
Kortum, S., and J. Lerner. (1999). “What Is Behind the Recent Surge in Patenting?” Research Policy 28: 1-22.
Lanjouw, J., and M. Schankerman. (2001). “Characteristics of Patent Litigation: A Window on Competition.” RAND Journal of Economics 32(1): 129-151.
Levin, R. C., A. K. Klevorick, R. R. Nelson, and S. G. Winter. (1987). “Appropriating the Returns from Industrial R&D.” Brookings Papers on Economic Activity 3: 783-820.
Mansfield, E. (1986). “Patents and Innovation: An Empirical Study.” Management Science 32: 173-181.
Merges, R. (1999). “As Many as Six Impossible Patents Before Breakfast: Property Rights for Business Concepts and Patent System Reform.” Berkeley Technology Law Journal 14: 578-615.
Merges, R., and R. R. Nelson. (1990). “On the Complex Economics of Patent Scope.” Columbia Law Review 90: 839-916 .
Merz, J., and N. Pace. (1994). “Trends in Patent Litigation: The Apparent Influence of Strengthened Patents Attributable to the Court of Appeals for the Federal Circuit.” Journal of the Patent and Trademark Office Society 76: 579-590.
Meurer, M. (1989). “The Settlement of Patent Litigation.” RAND Journal of Economics 20: 77-91.
Moore, K. (2001). “Forum Shopping in Patent Cases: Does Geographic Choice Affect Innovation?” North Carolina Law Review 79: 889.
Quillen, C. D., and O. H. Webster. (2001). “Continuing Patent Applications and Performance of the U.S. Patent Office.” Federal Circuit Bar Journal 11(1): 1-21.
Scherer, F. M. (1977). The Economic Effects of Compulsory Patent Licensing.New York: New York University Graduate School of Business Administration.
Scherer, F. M., and D. Harhoff. (2000). “Technology Policy for a World of Skew-Distributed Outcomes.” Research Policy 29: 559-566.
Scherer, F. M., S. Herzstein, Jr., A. Dreyfoos, W. Whitney, O. Bachmann, C. Pesek, C. Scott, T. Kelly, and J. Galvin. (1959). Patents and the Corporation: A Report on Industrial Technology Under Changing Public Policy.2nd Edn. Boston, MA: Harvard University, Graduate School of Business Administration.
Scotchmer, S. (1991). “Standing on the Shoulders of Giants: Cumulative Research and the Patent Law.” Journal of Economic Perspectives 5: 29-41.
Shapiro, C. (2001). “Navigating the Patent Thicket: Cross Licenses, Patent Pools, and Standard Setting.” In A. Jaffe, J. Lerner, and S. Stern, eds., Innovation Policy and the Economy, Vol. 1. Cambridge, MA: National Bureau of Economic Research and MIT Press.
Taylor, C. T., and Z. A. Silberston. (1973). The Economic Impact of the Patent System: A Study of the British Experience.Cambridge: Cambridge University Press.
Trajtenberg, M. (1990). “A Penny for Your Quotes: Patent Citations and the Value of Innovations.” Rand Journal of Economics 21(1): 172-187.