Leveraging Public Investments with Private Sector Partnerships: A Review of the Economics Literature
Andrew A. Toole and Anwar Naseem
The purpose of this report is to summarize the discussion and findings in the economics literature on collaborative efforts in research and development (R&D). A collaborative R&D arrangement, or equivalently an R&D partnership, unites multiple participants in a structured relationship to conduct research and development activities directed toward one or more objectives. The breadth of this definition presents an immediate problem. How does one organize the tremendous number and variety of collaborative arrangements that fit this definition into a conceptually meaningful structure? Since the answer to this question depends heavily on the analyst’s purpose, it is not surprising that no uniform structure exists in the economics literature. Some of the conceptual structures found in the literature organize R&D partnerships by the number and identity of the participants, by the alternative legal structures governing the relationships among partners, by the type of research and development activities conducted or by the stated objectives of the partnership arrangement.
From the literature, we identified five institutional forms of public-private collaboration: (1) government supported industry consortia; (2) industry-university collaborations; (3) federal laboratory-industry collaborations; (4) government grant programs in support of technology development and commercialization; and (5) global partnerships in health and agriculture. While our focus on public-private collaborations encompasses a diverse set of institutional arrangements and participants, partnerships involving only private participants fall outside the
scope of this report. Hagedoorn (2002) and Hagedoorn et al. (2000) are good entry points into the literature on inter-firm R&D partnerships.
To one extent or another, all the contributions to this literature touch on at least one of four main thematic areas. These are the: (1) motivations of the various parties for participating (objectives, expected benefits); (2) potential risks to the participants (conflicts of interest, comprising public trust, legal liability, loss of proprietary information, compromising the research and educational mission of universities); (3) characteristics of the institutional forms identified above (legal form, intellectual property [IP] rights, governance); and (4) evaluation of the outcomes or perceived success of the institutional forms.
Standing back and looking at the literature as a whole reveals considerable variation in the detail and depth of understanding across these areas. A large descriptive segment of the literature concentrates on the motivations and potential risks of public-private R&D collaboration. A number of studies provide general descriptions of the institutional forms but, overall, they provide very little detail about the specific structure of the relationships within any institutional form. Structural detail would define the role of each partner in various areas such as the decision making hierarchy, the ownership of IP, the funding, the performance of work, and the evaluation of the work. There are some case studies that provide insights into these structural aspects of collaboration. A much smaller group of quantitative studies tries to measure the outcomes or use some indicator of success. Again, these focus on evaluating the alternative institutional forms.
In each of the following two sections, we systematically discuss the findings in the literature related to the thematic areas. For each partnership participant, section 2 summarizes their motivations and risks. Section 3 presents information on the structure and outcomes for the five institutional arrangements identified above. Section 4 concludes the report with some reflections on the key messages that emerge from this diverse literature.
2. MOTIVATIONS AND RISKS
A main thrust in the literature is to understand the incentives that motivate participants to form public-private R&D partnerships. Public and private agents are quite different. National governments, intergovernmental agencies, universities, for-profit firms, foundations, and nongovernmental organizations (NGOs) all answer to different constituencies, operate under different norms of behavior, and frequently pursue different sets of objectives.1 How is it possible that these disparate parties have sufficient incentives to form a collaborative relationship?
The answer, according to the literature, is that each party gets something they value out of the arrangement. This value accrues through a set of participant-specific benefits. Moreover, the benefits can be separate and distinct from the formal objectives of the collaborative agreement. The benefits that flow to firms and universities, for instance, are more strongly linked to their “membership” and the actual performance of the R&D. For governmental entities, foundations, and NGOs, the expected benefits that motivate participation are more closely linked to the achievement of formalized partnership objectives. In rest of this section, we summarize the expected benefits and potential risks to partnership participants identified in the literature.
Why Do Governmental Entities Become Involved in Public-Private R&D Partnerships?
At the broadest level, governments view R&D partnerships with private agents as a mechanism to leverage both their limited financial resources and, when appropriate, their unique research capabilities to achieve social objectives. Governmental entities include intergovernmental agencies, such as the United Nations, World Bank, and World Health Organization (WHO), national governments, and specific agencies within national governments. This is clearly a diverse set of governmental institutions. The literature, however, points to three main social objectives/expected benefits that motivate government involvement in public-private R&D partnerships. These are to: (1) increase industrial competitiveness, (2) foster economic growth by mitigating market failures in research and innovation markets, and (3) more effectively meet agency specific mission-oriented needs through cost and risk-sharing.
The push to increase industrial competitiveness in the United States began in the late 1970s as a response to falling market shares and profits in several key industries, especially automobiles, consumer electronics, and later, semiconductors (Brooks and Randazzese (1998).2 Competitive pressure, particularly from the Japanese, provided the impetus for a series of new pieces of legislation aimed at stimulating collaborative R&D efforts between industry, government, and universities. For instance, the Bayh-Dole Act of 1980 and its amendments allow universities and other performers of federally sponsored research to patent and license their research results with greater ease. By clarifying IP rights, this legislation was intended to increase the flow of knowledge and technology into the private sector. Several other pieces of legislation passed during the 1980s and 1990s established Cooperative Research and Development Agreements
(CRADAs), the Advanced Technology Program, the Small Business Innovation Research Program, and the Small Business Technology Transfer Program. While we do not discuss the historical development of these legislatively mandated programs, Link and Tassey (1989) and Branscomb and Keller (1998) provide good treatments.
A second reason for government involvement in R&D partnerships is to foster economic growth by mitigating market failure in research and innovation markets. Seminal contributions to the economics literature in the late 1950s and early 1960s by Nelson (1959) and Arrow (1962) provide the rationale for government support of R&D. They point out that private firms are likely to under-invest in R&D activities in which the return to society is significantly greater than the firm’s own private return. This wedge between the social and private return to R&D is the result of research “spillovers” that prevent firms from capturing the full stream of benefits from their initial investment. Two forms of spillovers are identified in the literature: knowledge spillovers and consumer surplus spillovers (Branscomb and Florida, 1998). Knowledge spillovers are typically associated with basic scientific and basic technology research. Since the returns to this research are highly uncertain and long-term, firms are likely to under-invest and government can increase social welfare by supporting this type of research. Consumer surplus spillovers are associated with product and process development. In this case, profits to an innovator firm may not be sufficient to justify the required R&D investment; however, from a social perspective, the value of the new product or process exceeds its development cost. Governments may increase social welfare by supporting the development of such a product or process.
Public-private R&D collaboration also serves as mechanism to meet agency specific mission-oriented objectives through cost and risk-sharing. Because this is the most common rationale at the individual agency level, it encompasses the broadest array of mechanisms and objectives, including all five of the collaborative categories identified in the introduction. One important example of this type of cost sharing is the industry/university research center. These centers typically combine state, federal, and university funding of a dedicated research center affiliated with the university. In the late 1970s, the National Science Foundation (NSF) began its Industry/University Cooperative Research Centers (I/UCRC) Program as a means to leverage federal research funding with industry and university funding. The perceived success of this program led the establishment by NSF of the Engineering Research Centers (ERC) Program in 1984 and the Science and Technology Centers (STC) Program in 1987 (Brooks and Randazzese, 1998). Cost-sharing and matching requirements are also being used to supplement traditional cost-reimbursement grant mechanisms at NSF, National Institutes of Health (NIH), and other extramural funding agencies. In section 3, we summarize these and other mechanisms in more detail.
What Are the Risks to Governmental Entities Involved in Public-Private R&D Partnerships?
One of the fundamental risks that government entities face when entering into public-private partnerships is the “R&D contracting problem.” Noll and Rogerson (1998) describe the contracting problem within the context of government-university research grants; however, the same problems plague research-based partnerships. Because research produces new ideas and improved capabilities and competencies, research output is extremely difficult to measure. As a consequence, R&D procurement contracts cannot be written based on measurable outputs. Incomplete contracting introduces risk, because there is no direct incentive for researchers to conduct high-quality research. This point is reiterated in work by Poyago-Theotoky et al. (2002) within the context of university-industry partnerships. They formulate the incentive problem within a principal-agent framework. The funding agency, which is typically a governmental entity, is the principal, while the research performers, either industry or university, are the agents. The problem is that agents pursue their own self-interest and their actions may not be consistent with the best interests of the principal.
There are additional risks to governmental entities that are mentioned, albeit briefly, in the literature. One concern relates to public opinion and trust relationships between governments and their citizenry. Particularly in the context of global partnerships for health and agriculture, some observers believe that private partners will take control of decision making and use public resources for their own gain. Thus, the integrity of the governmental agency comes into question. Moreover, conflicts of interest can emerge from a careless choice of a private partner. For instance, in one of the WHO’s partnerships, it was charged that the appropriate standards for the management of hypertension were jeopardized because of the influence of one private partner that stood to gain from lower standards (Buse and Waxman, 2001). Mowery (1998) also highlights cultural differences as a risk to successful partnering. In his example, different methods of research management created conflicts in a CRADA agreement between a Department of Energy (DOE) laboratory and a private firm. Further, legal liability issues are an additional risk to governmental entities, particularly in R&D partnerships directed toward drug development. Pharmaceutical firms have dedicated legal departments and spend millions of dollars to defend against law suits related to adverse reactions and deaths from drug therapies. As participants in drug development partnerships, government entities expose themselves to similar legal liabilities.
Why Do For-Profit Firms Become Involved in Public-Private R&D Partnerships?
The literature identifies a large number of potential benefits that may accrue to for-profit firms from R&D collaboration. Almost all of the theoretical and
empirical work in this area focuses on inter-firm R&D collaboration rather than collaboration with governments or universities. Nevertheless, it is reasonable to expect that these same benefits carry over to public-private R&D partnerships.
In a recent survey, Hagadoorn et al. (2000) put together a comprehensive list of theoretical benefits from research partnerships. They group contributions into three categories of the literature: transaction costs, strategic management, and industrial organization theory. For transaction cost theorists, R&D partnerships are a hybrid organization form that stands between arm’s length market transactions for knowledge production and in-house knowledge production. They see the emergence of partnerships as an efficient response to problems with market contracts to produce technical knowledge and as a better alternative to building the necessary capabilities within the firm. By forming a partnership, firms are able to establish greater control over knowledge production relative to the market and reduce costs and risk relative to complete in-house knowledge production.
Five alternate perspectives are reviewed from the strategic management literature. First, from a competitive strategy perspective, R&D partnerships allow firms to respond quicker to changing market needs and introduce new technologies faster. Second, partnerships are motivated by strategic network advantages. These networks can increase research efficiency via scale and scope economies, create research synergies by exploiting different organizational competencies, and provide greater power to influence the decisions of rivals. Third, the resource-based view highlights the benefit to firms from increased access to complementary resources external to the firm. Fourth, R&D partnerships allow for greater organizational learning by increasing the effectiveness of knowledge transfer to the firm. Finally, this literature emphasizes a “strategic options” approach in which high levels of uncertainty in knowledge production can be reduced through incremental resource commitments. R&D partnerships allow firms to avoid pre-committing to the full cost of developing a new technology.
Industrial organization theory focuses on the potential market failure in research and innovation markets due to knowledge and consumer surplus spillovers. A standard result in this literature, mentioned above, is that private firms under-invest in R&D from a social welfare standpoint. Hagedoorn et al. (2000) divide this literature, which is heavily game theoretic and mathematical, into non-tournament and tournament models. Non-tournament models, which focus on the extent of innovation, find that cooperative R&D can mitigate problems with under-investment in R&D by reducing spillovers.
Tournament models focus on “races” between firms where the winner captures a monopolistic return. The results in this literature are mixed. R&D investment may or may not increase depending on whether firms undertake substitutive or complementary R&D.
Contributors to the empirical and policy literature identify a number of additional, and sometimes overlapping, benefits to for-profits firms from R&D partnerships. Based on data from inter-firm research cooperation in the video display
terminal industry, Link and Zmud (1984) find that firms want to maintain and increase market share. Mowery (1998) suggests that firms desire access to research results in universities and public laboratories and coordinate with other firms to create a common technological “roadmap” to guide future R&D investment. Hagadoorn et al. (2000) point out that a number of studies emphasize access to complementary research results as well as access to key university personnel, federal laboratory scientists, and a pool of qualified students for recruitment. Drawing on the experience of one firm in a CRADA relationship with a DOE laboratory, Mowery notes that the laboratory offered “unique capabilities, facilities, and equipment that in many cases could not be duplicated elsewhere” (Mowery, 1998:42). Another study by the Government-University-Industry Research Roundtable (GUIRR) adds that the knowledge base of the firm’s employees will be enhanced and expanded through partnerships (GUIRR, 1999). Feller and Roessner (1995) suggest that firms gain methods and tools and not specific products or research findings from their collaborations with ERCs. Finally, firms participating in public-private partnerships can improve their corporate image. This is especially true in the context of global health and agricultural partnerships.
What Are the Risks to For-Profit Firms Involved in Public-Private R&D Partnerships?
In contrast to the volume of work highlighting for-profit firm benefits and motivations for R&D partnerships, there is surprisingly little commentary on the risks to private firms. In a 1998 GUIRR workshop on barriers to collaboration with universities, Francis Via, who is the director of Contract Research for the chemical firm Akzo Nobel, provided the most comprehensive list we could find (GUIRR, 1999). At the top of his list is mistrust among partners. Developing mutual respect and avoiding opportunistic behavior is critical. Other risks he mentions include publication issues, IP, and timing. Firms are very concerned with the potential leakage of proprietary information and with keeping information confidential until sufficient lead-time is developed. IP rights can be a problem when new technologies are jointly developed. There are a number of issues regarding the costs of securing and protecting IP and in establishing agreeable licensing arrangements and royalty rates. Further, Mr. Via stresses differences in time horizons as a risk for firms. Generally, firms work on short-time horizons relative to universities (and probably federal laboratories), and the consequences of missed deadlines can be much greater for a firm than for a public partner.
Why Do Universities Become Involved in Public-Private R&D Partnerships?
There is general agreement in the literature that access to money and technical knowledge are the most important factors driving university involvement in
public-private R&D partnerships (Brooks and Randazzese, 1998; Powell and Owen-Smith, 1998; Jankowski, 1999; Hall et al., 2000; Poyago-Theotoky et al., 2002). Jankowski states, “Not only does such an approach offer opportunities for alternative funding in an increasingly constricted budgeting environment, but such partnership provides an essential means for undertaking work that is becoming evermore complex and multidisciplinary” (Jankowski, 1999:61). Poyago-Theotoky et al. (2002) add that a university with strong ties to industry can leverage these ties to attract “star” scientists to their faculty. Furthermore, they suggest that universities are interested in building long-term relationships that lead to sponsored research, in-kind support, and donations from firms.
A secondary motivation for university R&D collaboration with industry is to enhance student education and job prospects. NSF’s collaborative center programs, particularly the ERC Program, requires an “education program that integrates research results into curricula for precollege and college students and practitioners, and teams undergraduate and graduate students in research and education” (NSF, 2004). Moreover, Stephan (2001) points out that industry-university collaborative research provides a chance for industry and students to get a “pre-employment” look at each other. This serves as a kind of informal “internship” opportunity. Further, she suggests student will have the chance to learn about industry salary and working conditions. On the other hand, Feller et al. (2002) find that the number of firms actually hiring graduate students through ERCs was relatively small. Their interviewees explained that cutbacks in corporate employment and active recruitment by competitors were the primary reasons.
What Are the Risks to Universities Involved in Public-Private R&D Partnerships?
A variety of risks to universities from R&D collaboration with industry have been discussed in the literature. At the broadest level, many observers see increasing reliance on industry funding as a threat to the university’s “open science” norms of behavior. “Open science” refers to the free expression, exchange, and dissemination of new ideas. Threats to the free exchange of ideas from industry collaboration include limitations on the disclosure of research findings in the form of database restrictions or confidentiality agreements, publication delays and decreased communication between faculty or faculty and students. Brooks and Randazzese (1998) point to anecdotal evidence in the New York Times and the Wall Street Journal as well as quantitative research by Blumenthal and colleagues to illustrate these threats. Francis Via, commenting from an industry perspective on publication delays, states, “Many times, any early publication can alert competitors to a new fertile area of research…. Delaying publication for review will provide an 18-month lead for the industry partner” (GUIRR, 1999:19).
In addition to restrictions on the free exchange of ideas, there are concerns about how private research partners are influencing faculty research topics and creating professional conflicts of interest (Cohen et al., 1998). Brooks and Randazzese (1998) and Poyago-Theotoky el at. (2002) cite several studies that find a positive correlation between industry support and the conduct of more applied research and fear that the former is causing the latter. Many observers believe that reallocating the university research portfolio away from basic research toward applied research would be undesirable. Brooks and Randazzese (1998) also cite a study of financial disclosure practices in scientific publications. It finds that more than 33 percent of the authors failed to disclose a direct financial interest in the publication’s results. Further, Harman and Sherwell (2002) provide five interesting case studies illustrating a variety of faculty conflicts of interest. Some of these disagreements eventually involved university administrators in a tangle of legal and political issues.
Stephan considers the possible impacts of faculty ties with industry on students and curriculum. She cautions that such ties have the potential to “divert faculty away from students and curriculum” toward more profit-motivated activities such as securing research funds, patenting, consulting, or commercialization activities (Stephan, 2001:200). Using anecdotal evidence, Stephan points out that the trust relationship between a faculty member and student is jeopardized, sometimes leading to legal action. Moreover, increased secrecy in the laboratory appears to diminish peer learning effects as students are increasingly hesitant to discuss potentially valuable or proprietary information.
Why Do Foundations and NGOs Become Involved in Public-Private R&D Partnerships?
We could not locate any literature describing the motivations for foundations and NGOs to become involved in public-private R&D partnerships. Generally, these organizations are active in very specific areas that are dictated by endowment guidelines or specific charters. At the same time, they are typically very resource-constrained and limited in the organizational and financial contributions they can make to an R&D partnership. One type of potential benefit to these organizations occurs when a foundation uses or “piggybacks” on the peer review process of the federal agency as a project selection mechanism, which saves the foundation the expenses of the application and peer review processes.
What Are the Risks to Foundations and NGOs Involved in Public-Private R&D Partnerships?
We could not locate any literature describing the risks to foundations and NGOs from participation in public-private R&D partnerships. Based on the piggybacking arrangement described above, in which foundations rely on a federal
agency’s peer review process to identify projects to fund, they lose some control over the projects they fund. They run the risk of piggybacking on a flawed or biased project selection process.
3. INSTITUTIONAL FORMS AND EVALUATION
In this section we discuss each of the five institutional forms identified in the introduction. Generally, there is little detail in the literature on the collaborative structures employed within each category. Of course, for many of the federal programs, broad structure is provided by the enabling legislation and the administrative agencies implementing these programs. When available, the best information on structural characteristics is provided by case studies. With respect to outcomes, we summarize the existing empirical work in the literature that attempts to evaluate the success or performance in the five categories. However, there are relatively few such studies. Mowery notes, “… surprising little effort has been devoted to evaluation of any of the legislative or administrative initiatives…” (Mowery, 1998:39).
Government-Supported Industry Consortia
Because most industry consortia are typically industry funded, government-supported industry consortia are a fairly special form of public-private R&D partnership, at least in the United States.3 Under a strict definition, industry consortia are groups of two of more firms in the same industry that are potential competitors (Aldrich and Sasaki, 1995). The best-known U.S. examples of government-supported consortia are the Semiconductor Manufacturing Technology Consortium (SEMATCH) and the Partnership for a New Generation of Vehicles. In Japan, the best known examples are the Very Large Scale Integration Research Project, the Fifth Generation Computer Project, and the Opto-electronics Integrated Circuits Project.4 Regardless of national origin, government-sponsored consortia share three common elements: (1) they are intended to address a high priority national competitiveness issue; (2) they have well defined and specific objectives; and (3) they focus on “pre-competitive” or “generic” research.
In the United States, the legal framework that allows competitors to undertake cooperative R&D was established in the 1980s. The National Cooperative
Research Act of 1984 (NCRA) was intended to reduce the threat of antitrust action against legitimate research joint ventures. The act establishes a rule of reason approach for antitrust proceedings that balances the procompetitive and anticompetitive effects of the research joint venture. The act also protects firms from treble damages in private antitrust suits as long as the research joint venture is registered with the Department of Justice (Scott, 1989; Katz and Ordover, 1990; Mowery, 1998; Hagedoorn et al., 2000). NCRA was amended in 1993 to include cooperative ventures in production.
The academic literature on industry consortia is quite large, even when restricting attention to government-supported industry consortia. While it is not feasible to summarize this literature here, the studies are generally of two varieties: (1) case studies or comparative case studies (Katz and Ordover, 1990; Grindley et al., 1994; Roos et al., 1998; Sperling, 2001; Thornberry, 2002) and (2) quantitative studies based on survey results (Aldrich and Sasaki, 1995; Link et al., 1996; Sakakibara, 1997). Given the complexity and diversity of consortia arrangements, we simply summarize some of the lessons on consortia design and management provided by Grindley et al. (1994) for SEMATECH.
Grindley et al. (1994) provide a detailed discussion of SEMATECH’s evolution and a comparative analysis with other high-technology consortia in Japan and Europe. They highlight three complex design and management challenges that all consortia must face; (1) how to define the research agenda and projects to undertake; (2) how to transfer research results to participants; and (3) how to allow sufficient flexibility to permit change as industry needs and circumstances evolve. In contrast to most European consortia, SEMATECH’s centralized management structure and strong industry control allowed it to address these problems more efficiently. Moreover, they point out that the feasibility and eventual success of consortia-style collaboration in other industries will depend on the structure of the consortium, the political and economic expectations of the sponsors, and the alignment between the research activities of the consortium and the competitive problems in the industry.
As mentioned in Section 2, the NSF introduced the industry-university cooperative research center (I/UCRC) model in the late 1970s. The center model was expanded by the NSF into the Engineering Research Centers (ERC) Program in 1984 and the Science and Technology Centers (STC) Program in 1987. According to the NSF, there are currently more than 50 active I/UCRCs, 11 active STCs, and a total of 41 ERCs have been established since 1984 (NSF, 2004). Moreover, non-NSF centers have grown rapidly through university based initiatives, sometime winning support from state governments through competitions (Adams et al., 2001).
While these centers are quite different in their specific technological focus, objectives, and size, the NSF-supported centers share three broad goals. First, they support generic and precompetitive research relevant to industry needs. Generic R&D has the potential for wide applicability to many different products and processes while precompetitive R&D permits the evaluation of commercial potential but stops short of developing a specific prototype. Second, NSF centers strive to improve education and strengthen the science and engineering workforce. Third, the centers try to promote and accelerate technology transfer from universities to industry.
A recent study by Adams et al. (2001) explores how I/UCRCs influence patenting by and the R&D expenditures of member firm laboratories. Using survey data collected from 202 industry R&D laboratories, the authors find that industrial laboratories that belong to an I/UCRC are over twice as large and more science-oriented than their non-member counterparts. Further, they find I/UCRC member laboratories receive 2 percent more patents, although this effect is not statistically significant. With respect to R&D expenditure, I/UCRC member laboratories spend 2 percent more on average. For both patenting and R&D expenditure, the effects were larger for NSF-supported centers. However, their results are subject to one important qualification. Larger and more productive industrial laboratories may seek membership in I/UCRCs. With their data, the authors are unable to rule out the possibility that their results driven by this alternative direction of causality.
Santoro and Gopalakrishnan (2001) used survey results from 189 firms that are members of I/UCRCs to investigate how various factors like trust, geographic proximity, communication and university IP policies affect technology transfer. Since collaboration involves some loss of control over proprietary resources, a greater degree of trust can facilitate technology transfer. Using an indicator for the extent of technology transfer activities at a center as their explained variable, the authors’ regression results show that greater trust significantly increases technology transfer activities. Geographic proximity and more generous university IP policies are also found to significantly increase technology transfer. Their measure of communication effectiveness, on the other hand, was insignificant.
Feller et al. (2002) studied firms that participate in NSF-funded ERCs. Their primary interest is to investigate the benefits and barriers to technology transfer from ERC participation. In the mid-1990s, the authors collected survey results from 355 firms and conducted telephone interviews with 20 respondents. These firms were participants in one or more of the 18 ERCs active in this period. While many of the benefits they identify were mentioned in Section 2, their survey results provide a ranking of benefits. Firms rated the following benefits as very important or extremely important (the percentage of respondents is given in parentheses): (1) to acquire and access new ideas (80 percent); (2) to be associated with an ERC whose research was close to the company’s research interests (73 percent); (3) to access research expertise at the ERC (65 percent); (4) to keep
up-to-date with university research in the field (58 percent); and to gain access to specific ERC faculty (56 percent). Among these, the extent of alignment of research areas between the firm and ERC that was the most important factor determining the magnitude of benefits reported by the firms.
They identify a number of barriers to deriving benefits from ERC participation. These include company-specific factors and inter-organizational differences. Company strategies and priorities are quite fluid, leading to frequent changes in product lines and personnel. Often times, these changes reduce or eliminate the value from participation. Moreover, they identify several institutional differences between firms and ERCs that act as barriers. These include different value systems, time horizons, and research priorities. Overall, the authors interpret these barriers as a potential threat to the long-term viability of individual ERCs. They note, “…industrial support of cutting-edge academic research appears to be fragile and contingent upon the availability of complementary public sector support” (Feller et al., 2002:473). When NSF support ends, as is required by program design, the leveraging rationale used by firms to justify participation will end as well.
Federal Laboratory-Industry Collaborations
CRADAs are government-industry partnerships designed primarily to commercialize a technology in a federal laboratory. The traditional mechanism of technology transfer has been to simply publicize results of federally sponsored research. Patent licensing, direct research grants, and research consortia are other ways that public sector technology can be disseminated (Day-Rubenstien and Fuglie, 2000). However, in a CRADA, federal laboratories enter into a contractual arrangement with a private firm to develop a technology and are not required to reveal any proprietary information. Moreover, the private firm can be assigned the rights to any IP arising from the partnership, although the federal government maintains a non-exclusive right to license the IP (Ham and Mowery, 1998). CRADAs were instituted under the Federal Technology Transfer Act of 1986, and they have grown in number from about 34 in 1987 to more than 2,500 in the mid-1990s (Guston, 1998). CRADAs are credited for the development of important new technologies, such as the anti-cancer drug Taxol and the AIDS drug AZT.
Although CRADAs have been existence for more than a decade, they have not been subject to much rigorous economic analysis, in part due to a lack of data availability (Cohen and Noll, 1995; Stiglitz and Wallsten, 2000). As a consequence, most prior efforts use the case study approach to analyze CRADAs (Cohen and Noll, 1995; Day and Frisvold, 1993; Ham and Mowery, 1998). In this section we review how CRADAs have been implemented by three federal agencies and the emerging lessons as reported in the studies of Ham and Mowery (1998) for DOE, Guston (1998) for NIH, and Day-Rubenstein and Fuglie (1999, 2000) for the Department of Agriculture (USDA).
CRADAs at DOE
Ham and Mowery (1998) report on CRADAs between DOE’s Lawrence Livermore National Laboratory (LLNL) and industry. They examine in particular five CRADAs at LLNL to identify the management factors that contributed to the success (or failure) of the CRADA and the benefits that were realized by the partners. The five cases were selected to reflect the diversity of CRADAs in terms of project size and duration, size of the participating firm, and the mix of product and process technology.
Several important findings emerge from the Ham and Mowery (1998) study. First, they find that private partners are motivated to participate in the CRADA largely to access the unique capabilities of LLNL such as large specialized facilities and equipment and the ability to put together multidisciplinary teams that focus on specific tasks. This suggests that accessing unique DOE technologies is not the primary motivation for firms to get involved with CRADAs. The firms interviewed stressed that the generic benefits they derived by participating were more important as it improves their long-term scientific and technical capabilities. Second, the authors suggest that CRADAs are most effective if they build on the historic missions and capabilities of the laboratory, rather on the projects that focus on civilian use technologies which are often be distant from the laboratory’s main mission. Ham and Mowery (1998) are critical of the treasure chest view of technology development which assumes that federal laboratories possess unique technologies that need only be further developed and commercialized by private partners. Third, the authors find that CRADAs are not well suited for all projects. In particular, if the project is a co-development project, as was true in four of the five examples they studied, then gaining IP rights for the jointly developed results was not the central motivation of the private partners. Since negotiation over IP issues often delays the implementation of CRADAs, the authors suggest that partners seek other, simpler, mechanisms for collaboration when IP is not a central concern.
CRADAs at NIH
Guston’s (1998) study of CRADAs at NIH discusses some of the mechanisms of project implementation at NIH as well as the emerging lessons of NIH’s experience. As with CRADAs in other federal departments, both public and private partners perceive some mutual benefit from entering into a partnership. For the private firm, a partnership with NIH gives it access to novel gene therapy techniques developed by NIH scientists. Similarly, NIH researchers gain from their private partners access to “proprietary reagents or to commercial-scale facilities for the production of potential new drugs” (Guston, 1998:231).
The criteria for implementing CRADAs at the NIH appear to be more stringent and focused then those in other agencies. First, the CRADA must be related
to the primary mission of NIH in biomedical research. Second, as Guston writes “any CRADA [at NIH] must be a highly focused research plan advancing a scientific purpose that could not be more appropriately achieved through any other mechanism” (Guston, 1998:231). The use of CRADAs to fund normal research activities—such as equipment purchase, support of research fellows, and tests for collaborators—is discouraged.
In 1996, NIH initiated a new type of CRADA—the material transfer agreement CRADA (MTA-CRADA). Under a MTA-CRADA, NIH researchers can acquire proprietary research tools from private partners, but the scope of the agreement is much broader than a simple material transfer agreement. MTACRADA allows collaborations that are primarily over materials without having to negotiate over IP rights that would be required in more interactive collaborations.
Guston identifies several areas of contention regarding CRADAs at NIH. The first issue is access to technologies. CRADAs are generally accessible on a first-come, first-serve basis and do not involve the complexities of procurement and competitive bidding. This informal CRADA selection process opens the door to political difficulties, because firms might question the fairness of process, particularly for high value technologies. As long as the supply of CRADAs is greater than the demand, this is unlikely to occur, but it is potentially problematic if the NIH technology involved is keenly desired by the private sector. A second problem is that some firms may view CRADAs unfavorably because they have the potential of creating competitors especially in mature product markets. Although the evidence for opposition by established firms to new technology created through CRADAs is lacking, Guston still suggests that aggressive marketing of new technology may backfire.
Lastly the issue of fair pricing and IP in the context of CRADA remains unclear. All CRADAs initially contained a fair pricing clause while allowing the licensee to obtain reasonable profits. The clause was eliminated in 1995, partly in response to the uneasiness expressed by private research partners. However, the government maintains nonexclusive rights to license CRADA inventions made by private sector partners “for research or other Government purposes.” More explicitly,
the government retains the right to require third party licensing “on terms that are reasonable under the circumstances,” but only in “exceptional circumstances” where the government determines that health, safety, or regulatory needs require it; such determination is subject to administrative appeal and judicial review (Guston, 1998:237).
Guston further writes that this language on the government’s right to license is too vague and broad and requires clarification.
CRADAs at USDA
Even though agriculture was not the main focus of technology transfer initiatives that led to creation of programs like CRADA, USDA has extensive experience in establishing partnerships as evidenced by the fact that its CRADA program has been operating longer and has more agreements per appropriated dollar than that of any other federal agency (Day-Rubenstien and Fuglie, 2000). CRADAs at USDA have resulted in the commercialization of biopesticides, vaccines for chickens, and a chemical that, when added to water, reduces soil erosion.
As with CRADAs generally, the USDA program requires that the partnership be consistent with the department’s mission, there must be no conflicts of interest, and fairness must be demonstrated in the selection of partners. Scientists at USDA laboratories are generally the ones to initiate a CRADA if they feel that an innovation they have developed has market potential. Private firms can also approach USDA to setup CRADAs if they find a particular technology to be a promising candidate for commercialization (the Agricultural Research Service publicizes its research advances through a variety of channels [conferences, workshops, Federal Register, etc.] and also maintains a database that reports on research).
A criticism of CRADA—and partnerships in general—is that it diverts public research from its central research missions. To address this issue, Day-Rubenstein and Fuglie (2000) study the pattern of research allocation for CRADA partnerships and compare it with the priorities of public and private research activities. Employing USDA’s research classification system, the authors estimate the amount of research resources allocated to five technology areas. Since the five technology areas are broadly representative of agricultural research and show sufficient variation in social and private benefits, the authors assume that a large share of private research will be devoted to those areas with a large private-good component, whereas the public sector will be more focused on areas with high social returns. This leads them to hypothesize that in partnership mechanisms such as CRADAs, the allocation of research resources will “reflect a middle ground between the priorities of each partner.” They find support for their hypothesis because the public share of resources allocated in CRADAs is higher for technologies with relatively higher social returns. However, since the average contribution of private-sector participants is approximately two-thirds of the funding for CRADAs at USDA, the authors suggest that USDA may be underutilizing these partnerships for areas with low private incentives and over-utilizing them for R&D in areas where strong private incentives exist. However, the authors caution in over-interpreting the data, which are based primarily on CRADAs involving small companies.
Government Grant Programs in Support of Technology Development and Commercialization
To supplement traditional grant and contract mechanisms and to promote greater technology transfer and competitiveness, policy initiatives in the 1980s and 1990s created the Small Business Innovation Research Program (SBIR), the Small Business Technology Transfer Program (STTR), and the Advanced Technology Program (ATP).
The SBIR program was established in the Small Business Innovation Development Act of 1982. To be eligible for the program, 51 percent of a firm’s ownership must be held by U.S. citizens and the firm must have fewer than 500 employees. The original legislation mandated all federal agencies with an extramural research budget greater than $100 million to set aside 1.25 percent from their budgets for this program. This budget percentage was phased in over a several-year period. After the reauthorization of the program in 1992, the set-aside was increased to 2.5 percent of each agency’s extramural R&D budget. In the 2000 reauthorization, the set-aside remained at 2.5 percent.
The legislation established three phases to the SBIR program. All applicants must start with a Phase 1 proposal. The Phase 1 project is intended to test the feasibility of a new idea. The feasibility study lasts from 6 to 12 months and the Phase 1 awards can be up to $100,000. Given the preliminary nature of the projects funded in this phase, one would expect a high failure rate. If the results of the feasibility study are favorable, firms may apply for a Phase 2 grant to move their idea into product development. The Phase 2 award is up to $750,000 and lasts for a two-year period. Finally, there is a Phase 3 to the SBIR program. This is an unfunded phase in which the companies are expected to commercialize their product or process. There is no direct government involvement in this phase.
The objectives of the program outlined in the original 1982 legislation have remained intact over the two subsequent reauthorizations, with only minor changes in emphasis. The 1982 Act identified the following four objectives:
To simulate technological innovation
To use small business to meet federal research and development needs
To foster and encourage participation by minority and disadvantaged persons in technological innovation
To increase private sector commercialization of innovations derived from federal research and development.
Although the 1992 reauthorization kept these objectives; it increased the emphasis on commercialization. Archibald and Finifter (2003) explore the extent to which the SBIR program at the National Aeronautics and Space Administration’s (NASA’s) Langley Research Center responded to the new
commercialization emphasis. Based on project survey data, they find that there was a shift to projects with greater commercial potential following the 1992 reauthorization.
The academic literature on the effects of the SBIR program is split. Studies using survey data collected from SBIR participants, either at the project or firm level, consistently find positive program effects across a variety of indicators such as sales, employment and patenting (Audretsch et al., 2002; Archibald and Finifter, 2003; Audretsch, 2003; NIH, 2003). For instance, a national survey sponsored by NIH finds that 39 percent of their SBIR winners have realized sales on their projects (NIH, 2003:3-33). In stark contrast, regression-based evaluations using data on both participant and non-participant firms, such as Lerner (1999) and Wallsten (2000), do not find significant sales or employment effects from participation in the SBIR program. Although the SBIR award indicator is never significant in Lerner’s study, he does find the interaction between awards and regional venture capital investment to be significant. Wallsten’s findings are more pessimistic. In addition to finding no effect on employment in his sample of publicly traded companies, he finds that SBIR awards simply displace a firm’s own R&D spending dollar for dollar.
The STTR program was created by the Small Business Research and Development Enhancement Act in 1992. It is intended to complement the SBIR program and shares the same multiphase structure as SBIR. The most significant difference between the programs is that STTR requires U.S. small businesses to partner with a research institution—a university, federal laboratory or other nonprofit research institution. The research partner receives at least 30 percent of the awarded funds. There are currently five U.S. agencies participating in this program: DOE, Department of Defense (DOD), Department of Health and Human Services, NSF, and NASA. Each agency must set aside 0.30 percent of its extramural research budget for the program. To date, there are no published economic studies evaluating the STTR program.
ATP was established by Congress under the authority of the Omnibus Trade and Competitiveness Act of 1988 and amended by the American Technology Preeminence Act of 1992. It is administered by the National Institute of Standards and Technology (NIST) of the Department of Commerce. The program is designed to increase the competitiveness of U.S. industry by accelerating the commercialization of new scientific and technological discoveries and facilitating the refinement of manufacturing technologies. The program supports collaborative research on generic and precompetitive R&D problems (Hill, 1998).
As the administering agency, NIST is responsible for designing “focused programs,” reviewing proposals, and monitoring awards. NIST laboratories are not allowed to participate in the ATP program. ATP awards are given to single firms or industry-led joint ventures. Single-firm awards are generally granted only to U.S.-owned for-profit companies, although foreign-owned firms may receive awards if there is a clear U.S. interest which is evaluated using a strict set
of guidelines. Industry-led joint ventures must include at least two for-profit firms that meet ATP qualifications under the single-firm guidelines and may include nonprofit organizations, independent research organizations, government laboratories, and universities. Universities and government laboratories may participate in ATP projects as subcontractors or as members of a joint venture; however, they cannot submit proposals on behalf of the joint venture (Department of Commerce, 2004).
The amount, duration, and cost-sharing requirements for ATP awards differ between single-firm awards and joint-venture awards. Awards to single firms may be up to $2 million over a three-year period while joint venture awards may last five years with no funding limitation. Cost sharing may be in the form of cash or in-kind contributions. For single-firm awards, small and medium size companies must absorb the indirect/overhead costs of the R&D while large companies must provide at least 60 percent of total project costs. For joint ventures, cost sharing must exceed 50 percent of the total project costs. All IP that results from the ATP-supported R&D belongs to the for-profit companies and cannot be assigned to government laboratories, universities, or other nonprofits. These institutions may share in the royalties through licensing arrangements (see www.atp.nist.gov).
Noting that only 15 percent of federally registered Research Joint Ventures involve a university partner, Hall et al. (2001) investigate the role of IP rights as a potential barrier to university participation. For 38 ATP projects, the authors supplement data from the ATP program with a survey instrument that asked participant firms if IP rights created and insurmountable barrier. The results show that about 32 percent of the projects examined faced insurmountable IP barriers. Coding this indicator as a dichotomous variable, they use a Probit model to identify those factors systematically related to the existence of IP barriers. The regression results suggest that IP barriers are greater when the there is a higher ATP funding share, when projects have a shorter duration, and when lead participants had previous experience with universities as research partners.
In another paper, Hall et al. (2000) use ATP program data and survey data for 192 ATP projects to investigate the role and consequences of university participation. Before using Order Probit models to analyze their data, the authors review the motivations for industrial firms and the motivations for universities to form R&D partnerships. They posit that for-profit firms seek access to complementary knowledge, eminent researchers, and the reputations of elite universities, while universities, on the other hand, are primarily motivated by financial gain. With this backdrop, the regression results reveal that ATP projects with a university partner have a lower probability of early termination and that the industry partners in these relationships find it more difficult to assimilate basic knowledge required for project completion. Moreover, they find that university partners are not associated with the generation of new applications of the project technologies. On a positive note, university partners, either as subcontractors or joint
venture members, do not seem to introduce any unexpected research problems into the project.
Global Partnerships in Health and Agriculture
In this section we review some of the key studies pertaining to partnerships in other countries, particularly those that are designed to address the significant health and agricultural needs of developing countries. A review of international partnerships can be instructive in the design of domestic partnerships because (1) the institutional and policy environments of international partnerships are often very different from that in the United States and (2) such partnerships are often complex arrangements that involve several participants (governments, multilateral institutions, foundations, and large corporations), with each participant having different mandates and constituents. Examining the genesis of such partnerships and how they are structured, especially in aligning the disparate incentive structures of the participants, may provide some lessons for the U.S. case.
Partnerships in International Health
Public-private partnerships (PPPs) designed to address the research needs of developing country health are a recent phenomenon (Buse and Walt, 2000). Historically, health-related programs such as those that sought to eradicate certain diseases in a developing country were done under the auspice of the national government, often with the help of foreign donors, foundations, and international organizations. Those projects were designed primarily to improve public health by immunizing against a particular disease or educating the most vulnerable population about prevention methods. Few, if any, of these programs were research-oriented with a view of developing new products; rather, they sought to strengthen the health infrastructure and capacity of the country.
The private sector was never a significant player, because raising health awareness and delivering vaccines was regarded as a public good and appropriately the function of government health agencies. Moreover, because the purchasing power of consumers in developing countries is small, the potential market for health products in developing countries was, and remains, commercially unattractive. Private-sector firms, therefore, have generally invested little in research on tropical diseases. For example, of the 1,223 new chemical entities introduced globally in the period 1975-96, only 13 were specific to diseases in the tropics (Webber and Kremer, 2001). However, even though market realities have discouraged private firms from investing in research on tropical diseases, they have discovered, in the process of research on other diseases, several drugs to treat diseases prevalent in the developing world. Private firms also possess many of the important patents and tools needed to undertake pharmaceutical research on tropical disease, although they do not do so because they are constrained by
institutional and market barriers. These include, but not limited to, a weak health infrastructure for drug delivery, inadequate IP protection, and weak markets. To encourage more private sector funding, a variety of pull and push incentives have been suggested (see Webber and Kremer, 2001, for a summary), but they have yet to be implemented.
The scientific and technological capacity of the private firms in developed countries has been strengthened considerably over the years, due in large part to greater technological opportunities and stronger patent rights, but the public research sector in developing countries has remained weak. Funding constraints, low scientific capacity, and problems in gaining access to the most productive research tools has meant that public-sector researchers in developing countries have been unable to provide indigenous solutions to infectious diseases in their countries. This has resulted in a widening disparity in health between rich and poor countries. It has been estimated that in 1990, 80 percent of the disparity in death and disability adjusted life years between rich and poor nations was due to communicable diseases that can be prevented and for which drug treatment exists, although in many cases the drugs would need to be improved to suit local conditions and cultural practices (Widdus, 2001).
It is against this backdrop that PPPs in international health have their origins. The purpose of a PPP is to bring together the public and private sectors in an effort to solve some of the most intractable diseases pressures, which otherwise may not be solved if each sector acted on its own. The private sector, even though it has the necessary resources, is unlikely to undertake R&D, because the private returns to such research are low. The public sector, on the other hand, is handicapped by too few resources, even though such R&D activities have high social returns and serve the public good. By suitably aligning the incentives of the two sectors and leveraging their capabilities, it is felt that resources can be mobilized to conduct the necessary research.
Partnerships are created with different needs and outcomes in mind. As such, there exists a diversity of arrangements, which vary with regard to participants, legal status, governance, management, and operational roles. Widdus (2001) provides six reasons why partnerships in international health have been formed: (1) to develop a new product; (2) distribute a donated or subsidized product to control a specific disease; (3) strengthen health services; (4) educate the public; (5) improve product quality or regulation; and (6) coordinate multifaceted efforts.
Most partnerships are primarily based on meeting the first three objectives. Examples of product development partnerships include the Medicines for Malaria Vaccine (MMV) and the International AIDS Vaccine Initiative (IAVI). Partnerships that are based on the product donation by corporation include the donations of albendzole, eflornithine, and leprosy multi-drug therapy among others. Some of these donation-based partnerships go beyond the donation aspect, and have involved activities to ensure effective distribution and use (Widdus, 2001). Notable among partnerships meant to strengthen the delivery of health services is
the Gates Foundation-Merck-Botswana Comprehensive HIV/AIDS Partnership. Although it is beyond the scope of this paper to discuss the specifics of every partnership—the Initiative on Public-Private Partnerships for Health lists some 70 collaborative relationships (www.ippph.org)—we focus instead on three drug and vaccine development PPPs, specifically MMV, IAVI, and the Global Alliance for TB Drug Development (Global Alliance).
A unique feature of the three partnerships is that they pursue a business model that exploits the venture capital approach to investing (Wheeler and Berkley, 2001). The partners pool their resources and skills around specific projects in an effort to fund research projects that meet a certain socially desirable objective. The term “social venture capital” has been coined to reflect the nature of these partnerships, which primarily focus on high-risk upstream research that seeks to convert scientific basic research into actual drugs. Another important aspect of social venture capital is that it involves multiple nonprofit public and corporate partners to fund competitively the research needed to meet the desired objective. The three partnerships are also different in that they have established themselves as autonomous legal organizations, which gives them greater management and governance control over their research activities. Box D-1 summarizes the key objectives of the three organizations mentioned.
The organizations are rigorous in their evaluation of specific projects and proactively seek to register and license projects that involve for-profit partners. To decide on which projects to fund requires a thorough understanding of the target diseases, available protocols, and constraints (Wheeler and Berkley, 2001). As such, the partnerships have established extensive knowledge databases about each disease that allow for effective identification and prioritization of the projects deemed to be most vital and likely to succeed. As with any venture capital, the partnerships screen potential projects for feasibility and disburse funds to selected projects. The Global Alliance and MMV use competitive calls for project proposals to identify promising research areas, whereas in the IAVI case, staff members seek and develop projects based on information from scientific meetings and the published literature and on the advice of experts.
If drugs are developed, the three projects seek contractual arrangements with private firms to make products available at affordable prices while providing a positive return on the investment. Since partnerships expect that collaborating firms will manufacture and disseminate the final product providing incentives to the firm requires that firms be granted access to IP rights to the product. The contrasting approaches in dealing with IP and affordability in IAVI and MMV are instructive. In the case of IAVI, investments in small biotechnology companies have been made that account for a large share of the recipient firm’s capital, on the assurance that the product will be affordable in low-income countries. IAVI allows the biotechnology firm to retain developmental rights as long as the products are made available to public-sector organizations in developing countries at a reasonable profit which has been set at cost plus no more than 10
International AIDS Vaccine Initiative: To ensure the development of safe, effective, accessible, preventive HIV vaccines for use throughout the world
Partners/Donors include: Foundation Marcel Merieux, Francois-Xavier Bagnould Foundation, National AIDS Trust, AIDS Vaccine Advocacy Coalition, Albert B Sabin Vaccine Institute, World Bank, UNAIDS, Rockefeller Foundation, AP Sloan Foundation, Bill and Melinda Gates Foundation, Department for International Development (DFID), Glaxo Wellcome, Levi Strauss International.
Medicines for Malaria Venture: To discover, develop and commercialize antimalarial drugs at a rate of one new product every five years and at prices that are affordable to the most affected populations
Partners/Donors include: Association of British Pharmaceutical Industries, International Federation of Pharmaceutical Manufacturers Associations, Wellcome Trust, Rockefeller Foundation, WHO, World Bank, Global Forum for Health Research, DFID, and Swiss Development Corporation.
Global Alliance for TB Drug Development: To accelerate discovery and/or development of cost effective new tuberculosis (TB) drugs that will (1) shorten the duration of TB treatment or otherwise simplify it completion, (2) improve the treatment of latent TB infection, and (3) be effective against multi-drug-resistant TB strains.
Partners/Donors include: American Lung Association, American Society for Tuberculosis Education and Research, Association of the British Pharmaceutical Industry, Bill and Melinda Gates Foundation, Centers for Disease Control and Prevention, European Commission, Lupin Laboratories, Novartis India, Ltd, Rockefeller Foundation, DFID, US AID, World Bank, and WHO.
SOURCE: Wheeler and Berkley (2001), with authors’ amendments.
percent. The biotechnology firm, however, retains rights to offer the product to developed county markets without any restrictions on price. If the firm fails to deliver the product at an affordable price to the public sector in developing countries, IAVI retains “march-in” rights, i.e., the right to transfer the technology to another manufacturer. Even if IAVI exercises its march-in rights, the biotechnology firm to which the IP is assigned is allowed to keep its assets and can continue to market the product elsewhere. MMV, on the other hand, has invested in drug-discovery projects done by large firms and where the investment represents only a small fraction of the R&D budget of the firm. Under these circumstances, the expectation of the firm is not greater equity but that it will enter into a product development agreement with MMV in which MMV will have down-
stream rights to the technology which it could license. And although the issue of affordability in MMV has not been specifically addressed, the organization retains the right to develop the product if the commercial partner withdraws or fails to meet its obligations.
If firms are to divert their scarce resource into funding neglected diseases, it is important that such research activities has value to a firm and provides it with access to knowledge, technology, competitive advantage, or markets that they would otherwise not gain. At the same time, the ability to provide the most access to intended beneficiaries in developing countries requires that prices be kept low. A social venture capital organization therefore leverages its investment by negotiating to keep profit margins low. To compensate for lower profits in developing country markets, firms can be given exclusive licenses to market products in developed countries without price restrictions. For example, an HIV vaccine could be sold to high-risk groups in industrialized countries as can a vaccine to tourists and the military. Lastly, firms may be willing to participate in such partnerships if there are nonfinancial benefits as well. For example, partnerships can signal that a firm is a good corporate citizen, and for some small firms, it can be a showcase of its expertise and ability to deliver products. The risk of failure to firms from participating in projects can also be minimized by seeking funds to an array of potential products, allowing large companies to choose the most promising ones.
Partnerships in International Agriculture
The gap between rich and poor countries in the production of knowledge also pervades the agricultural sector. This has resulted in the markedly lower productivity of agriculture in developing countries and has perpetuated poverty in many countries. Furthermore, the diet of many in the developing world does not contain sufficient micronutrients; for example, it is estimated that 250 million children are at risk of vitamin A deficiencies, which leads to learning disabilities and blindness (Rausser et al., 2000). With the advent of agricultural biotechnology, there is much hope that not only can the productivity of staple crops can be increased but also that expression of micronutrients like vitamin A could also be attained. As in the international health area, there is recognition that productivity-enhancing technologies will not be developed without the collaboration of public and private institutions.
This is because many of the key tools of biotechnology necessary for developing novel and productivity enhancing plant varieties are proprietary and in the hands of private firms. While there are several ways that public research institutes or local firms can obtain patented biotechnology genes and tools, partnerships are being increasingly used as a mechanism to transfer proprietary technology from the private to the public domain. In return, the private firm gains access to the public sector’s germplasm, plant variety assessment infrastructure, and the
capacity to undertake upstream research. This suggests that, despite the different underlying incentives facing the private and public research sectors, sufficient common ground exists for agents in the two sectors to partner and develop useful technologies.
For example, the Brazilian Agricultural Research Corporation (EMBRAPA) leveraged its soybean germplasm assets to develop a partnership with Monsanto through which it could obtain Roundup Ready genes and access to plan transformation technology. Together, EMBRAPA and Monsanto have produced a series of herbicide-resistant genetically modified soybeans that Monsanto will sell through its extensive dealer network. Under the terms of the partnership, EMBRAPA receives royalties from the sales, and also a portion of the sales will go back to fund research on sustainable soybean production. A similar type of collaborative arrangement exists in Egypt where the local public research institute and Pioneer-Hi Bred jointly developed a new transgenic Bacillus thuringiensis (Bt) strain. In the collaboration, the Egyptian public system gains access to the expertise needed to develop the local strain of Bt (the innovation) and to educate its staff. The private-sector partner pays the legal costs of patenting the invention and has access to the new Bt strain for use in markets outside in Egypt.
For countries that do not possess a strong scientific capacity, international research centers or IP consortia that partner with private firms may be the only way to access proprietary technology. Such arrangements are thought to significantly reduce transaction costs and risk associated with developing a technology. For example, the Golden Rice Humanitarian Board—a public-private collaboration which includes the International Rice Research Institute, European government laboratories and the Syngenta Corporation—was set up to unravel the overlapping IP claims needed to develop vitamin A-enriched rice (Golden rice) for the poor. By establishing good faith agreements on the use of private-sector IP by academic researchers, the Board significantly reduces transaction costs relative to the case if the public sector had tried to access the technology on its own. Recently, several new multi-country programs have been initiated to obtain access to technology for the poor. The African Agricultural Technology Foundation is a nonprofit corporation funded initially by Rockefeller Foundation. It will license and hold technology from the major biotechnology firms with a humanitarian use license and then provide the technology free to its scientists in poor countries. In addition, the Australian-based institute, CAMBIA, is making information about patented technology more readily available and is developing nonproprietary technologies for biotechnology researchers in poor countries. Another recent initiative is the IP-clearinghouse program in the United States, which has the goal of making IP from universities and government research institutes more readily available. This program seeks to design a toolbox of biotechnologies for public sector researchers in industrialized countries.
It is important to realize that while partnerships may be desirable in many instances, they are not costless. Indeed a survey of partnerships in agriculture
(Spielman and von Grebmer, 2004) finds that transactions costs of establishing a partnership to be “excessively high.” These costs include the direct expenses of meeting legal requirements, such as writing up contracts and enforcing agreements, as well as indirect costs of adapting to different organizational cultures. Such costs can be substantial for public sector partners who often do not possess the legal expertise to negotiate contracts with the private sector. Spielman and von Grebmer (2004) also identify risks to partners who engage in joint collaborative efforts. For the private firm, who is usually the provider of a key technology, the risks include the potential misuse or controversial use of the technology by the partner, end users, or third parties, which could result in legal, financial, or reputational liability for the technology provider. For the public firm, there is risk to its reputation from associating with private firms and developing controversial technologies. Since public sector research organizations are supported by taxpayers, any association with a private firm that is perceived to benefit unduly the private firm may lead to unwelcome scrutiny. In minimizing these types of risks, private and public sector participants incur costs that may diminish the incentives to form a partnership.
Pray and Naseem (2003) identify several characteristics of successful public-private joint ventures in international agriculture. First, both public and private partners have had something to gain from these collaborations. The gains do not have to be financial, although financial gains may provide the strongest incentive. Second, governments had the political will and ability to negotiate with private firms; in many countries this can be very difficult because of ideology and mistrust of the private sector. Lastly, partnerships require a budgetary commitment from the public sector partners, which has been financed by foreign donors.
4. CONCLUDING REFLECTIONS
The economics literature on public-private R&D partnerships is extremely varied and, except for some case studies and legislative guidelines, provides little detail on the particular structures used to organize and carryout partnership arrangements. Nevertheless, this section provides a series of brief reflections on the literature and its relevance for the IOM Committee on Alternative Funding Strategies for DOD’s Peer Reviewed Medical Research Programs.
(1) The focus and character of DOD’s Congressionally Directed Medical Research Programs (CDMRP) are central to defining and organizing potential partnership arrangements. For instance, research projects on topics such as understanding cellular function in cancer propagation, developing a new breast cancer diagnostic method, or conducting a clinical drug trial for a therapeutic candidate must be separated for the purposes of partnership definition and design. First, the fact that each of these examples involves research in cancer immediately limits
the group of potential nonfederal partners. It is unlikely, for instance, that a firm focused on cardiovascular research will be in the potential partnership pool. Thus, the focus area of the research is directly related to the potential partnership pool. Second, the character of the research relates to the degree of “spillovers” and its proximity to the market. Cellular function research is much more basic in character than a clinical study of a potential drug therapy. Consequently, it has a higher degree of uncertainty about expected payoffs and is more distant from the market. As indicated in Section 2, spillovers and proximity to the market relate closely to the incentives for private partners, especially for-profit firms, to become involved in partnership arrangements.
(2) Generally, the role of government in public-private partnerships is usually “research passive,” in the sense that they define mechanisms, review proposals, provide funds, manage the accounting side of research contracts, and sometimes monitor or assess outcomes. The obvious exception to this is the CRADA mechanism in which government laboratories are “research active” partners that engage in the conduct of research. because CDMRP is an extramural research program, the role of DOD for this program will be “research passive” and rules out the CRADA mechanism as a potential institutional form of collaboration.
(3) Consortia, industry-university centers, the SBIR/STTR programs, and the ATP program have a strong industry orientation and extensive industry participation in defining projects and methods. While CDMRP might design a collaborative mechanism based on one of these institutional forms, it seems inevitable that the current structure of vision setting and project selection will need to be changed to incorporate, to some degree, the interests of private partners. Depending on how the collaborative arrangement is structured, any number of risks might be introduced, including conflicts of interest, issues of public trust and program credibility, legal liability, and research reorientation away from high risk basic research toward more developmental and applied objectives.
(4) While Lerner (1999) referred to the SBIR program as “public venture capital,” the social venture capital model that has emerged in the context of international partnerships for health offers another possibility for CDMRP. A key aspect seems to be the creation of a separate legal entity with a different governance structure that can funnel money into projects and programs. This IS similar to the Foundation of the National Institutes of Health (FNIH), which was founded by an act of Congress in 1996 as a nonprofit organization. As a separate institution, FNIH has greater flexibility to accept and direct funds than the NIH itself (Pfizer Journal, 2003).
Adams JD, Chaing EP, Starkey K. 2001. “Industry-University Cooperative Research Centers.” Journal of Technology Transfer, 26(1-2):73-86.
Aldrich HE, Sasaki T. 1995. “R&D Consortia in the United States and Japan.” Research Policy, Vol. 24:301-316.
Archibald RB, Finifter DH. 2003. “Evaluating the NASA Small Business Innovation Research Program: Preliminary Evidence of a Trade-Off between Commercialization and Basic Research.” Research Policy, 32:605-619.
Arrow K. 1962. “Economic Welfare and the Allocation of Resources for Invention.” Pp. 609-625 in The Rate and Direction of Inventive Activity, Nelson RR, editor. Princeton, NJ: Princeton University Press.
Audretsch DB. 2003. “Standing on the Shoulders of Midgets: The U.S. Small Business Innovation Research Program (SBIR).” Small Business Economics, 20:129-135.
Audretsch DB, Link AN, Scott JT. 2002. “Public/Private Technology Partnerships: Evaluating SBIR-Supported Research.” Research Policy, 31:145-158.
Branscomb LM, Florida R. 1998. “Challenges to Technology Policy in a Changing Economy.” Pp. 3-39 in Investing in Innovation, Branscomb LM and Keller JH, editors. Cambridge, MA: MIT Press.
Branscomb LM, Keller JH. 1998. Investing in Innovation. Cambridge, MA: MIT Press.
Brooks H, Randazzese LP. 1998. “University-Industry Relations: The Next Four Years and Beyond.” Pp. 361-399 in Investing in Innovation, Branscomb LM and Keller JH, editors. Cambridge, MA: MIT Press.
Buse K, Walt G. 2000. “Global Public-Private Partnerships: Part I—A New Development in Health?” Bulletin of the World Health Organization, 78:549-561.
Buse K, Waxman A. 2001. “Public-Private Health Partnerships: A Strategy for WHO.” Bulletin of the World Health Organization, 79:748-754.
Caloghirou Y, Tsakanikas A, Vonortas NS. 2001. “University-Industry Cooperation in the Context of the European Framework Programmes.” Journal of Technology Transfer, 26(1-2):153-161.
Cohen LR, Noll RG. 1995. “Feasibility of Effective Public-Private R&D Collaboration: The Case of Cooperative R&D Agreements.” International Journal of the Economics of Business, 2:223-40.
Cohen WM, Florida R, Randazzese L, Walsh J. 1998. “Industry and the Academy: Uneasy Partners in the Cause of Technical Advance.” Pp. 171-199 in Challenges to Research Universities, Noll RG, editor. Washington, DC: Brookings Institution Press.
Day K, Frisvold GB. 1993. “Medical Research and Genetics Resources Management: The Case of Taxol.” Contemporary Policy Issues, 11:1-11.
Day-Rubenstien K, Fuglie KO. 1999. “Resource Allocation in Joint Public-Private Agricultural Research.” Journal of Agribusiness, 17:23-134.
Day-Rubenstien K, Fuglie KO. 2000. “The CRADA Model for Public-Private Research and Technology Transfer in Agriculture.” In Public-Private Collaboration In Agricultural Research, Fuglie KO and Schimmelpfennig DE, editors. Ames, IA: Iowa State University Press.
Department of Commerce. 2004. “ATP Proposal Preparation Kit.” [On-line.] Available: www.atp.nist.gov/atp/kit-04/final-kit.pdf.
Feller I, Ailes CP, Roessner DJ. 2002. “Impacts of Research Universities on Technological Innovation In Industry: Evidence from Engineering Research Centers.” Research Policy, 31:457-474.
Feller I, Roessner DJ. 1995. “What Does Industry Expect from University Partnerships?” Issues in Science and Technology, 11(1):80-84.
GUIRR (Government-University-Industry Research Roundtable). 1999. Overcoming Barriers to Collaborative Research. Washington, DC: National Academy Press.
Grindley P, Mowery DC, Silverman B. 1994. “SEMATECH and Collaborative Research: Lessons in the Design of High-Technology Consortia.” Journal of Policy Analysis and Management, 13(4):723-758.
Guston DH. 1998. “Technology Transfer and the Use of CRADAs at the National Institutes of Health.” Pp. 221-249 in Investing in Innovation, Branscomb LM and Keller JH, editors. Cambridge, MA: MIT Press.
Hagedoorn J. 2002. “Inter-firm R&D Partnerships: An Overview of Major Trends and Patterns since 1960.” Research Policy, 31:477-492.
Hagedoorn J, Link AN, Vonortas NS. 2000. “Research Partnerships.” Research Policy, 29:567-586.
Hall BH, Link AN, Scott JT. 2000. “Universities as Research Partners.” NBER Working Paper 7643, March 2000.
Hall BH, Link AN, Scott JT. 2001. “Barriers Inhibiting Industry from Partnering with Universities: Evidence from the Advanced Technology Program.” Journal of Technology Transfer, 26:87-98.
Ham RM, Mowery, DC. 1998. “Improving the Effectiveness of Public-Private R&D Collaboration: Case Studies at a U.S. Weapons Laboratory.” Research Policy, 26:661-675.
Harman G, Sherwell V. 2002. “Risks in University-Industry Research Links and the Implications for University Management.” Journal of Higher Education, 24(1):37-51.
Hill CT. 1998. “The Advanced Technology Program: Opportunities for Enhancement.” Pp. 143-173 in Investing in Innovation, Branscomb LM and Keller JH, editors. Cambridge, MA: MIT Press.
Jankowski JE. 1999. “Trends in Academic Research Spending, Alliances, and Commercialization.” Journal of Technology Transfer, 24(1):55-68.
Katz ML, Ordover JA. 1990. “R and D Cooperation and Competition.” Pp. 137-203 in Brookings Papers on Economic Activity, Microeconomics. Washington, DC: Brookings Institution Press.
Lerner J. 1999. “The Government as Venture Capitalist: The Long-run Impact of the SBIR Program.” Journal of Business, 72:285-318.
Link AN, Tassey G. 1989. Cooperative Research and Development: The Industry-UniversityGovernment Relationship. Boston, MA: Kluwer Academic Publishers.
Link AN, Zmud RW. 1984. “R&D Patterns in the Video Display Terminal Industry.” Journal of Product Innovation Management, 2:106-115.
Link AN, Teece DJ, Finan WF. 1996. “Estimating the Benefits from Collaboration: The Case of SEMATECH.” Review of Industrial Organization, 11:737-751.
Mowery DC. 1998. “Collaborative R&D: How Effective Is It?” Issues in Science and Technology, 15(1):37-44.
NIH (National Institutes of Health). 2003. “National Survey to Evaluate the NIH SBIR Program,” Department of Health and Human Services, July 2003.
NSF (National Science Foundation). 2004. Various program solicitations, available at www.nsf.gov.
Nelson RR. 1959. “The Simple Economics of Basic Scientific Research.” Journal of Political Economy, 67:297-306.
Noll RG, Rogerson WP. 1998. “The Economics of University Indirect Cost Reimbursement in Federal Research Grants.” Pp. 105-146 in Challenges to Research Universities, Noll RG, editor. Washington, DC: Brookings Institution Press.
Paul JA. 2000. “NGOs and Global Policy-Making,” United Nations, Global Policy Forum, available at www.globalpolicy.org/ngos/analysis/anal00.htm.
Pfizer Journal. 2003. Public and Private Funding in the Quest for New Cures. New York, NY: Impact Communications, Inc.
Powell WW, Owen-Smith J. 1998. “Universities and the Market for Intellectual Property in the Life Sciences.” Journal of Policy analysis and Management, 17(2):253-277.
Poyago-Theotoky J, Beath J, and Siegel DS. 2002. “Universities and Fundamental Research: Reflections on the Growth of University-Industry Partnerships.” Oxford Review of Economic Policy, 18(1):10-21.
Pray C, Naseem A. 2003. “Biotechnology R&D: Policy Options to Ensuring Access andBenefits for the Poor.” Agricultural and Economics Development Analysis (ESA) Working Paper 03-08, Food and Agricultural Organization (FAO), Rome, Italy.
Quintas P, Guy K. 1995. “Collaborative, Pre-competitive R&D and the Firm.” Research Policy. 24:325-348.
Rausser G, Simon L, Ameden H. 2000. “Public-Private Alliances in Biotechnology: Can They Narrow the Knowledge Gaps between Rich and Poor?” Food Policy, 25:499-513.
Roos D, Field F, Neely J. 1998. “Industry Consortia.” Pp. 400-421 in Investing in Innovation, Branscomb LM and Keller JH, editors. Cambridge, MA: MIT Press.
Sakakibara M. 1997. “Evaluating Government-Sponsored R&D Consortia in Japan: Who Benefits and How?” Research Policy, 26:447-473.
Santoro MD, Gopalakrishnan S. 2001. “Relationship Dynamics between University Research Centers and Industrial Firms: Their Impact on Technology Transfer Activities.” Journal of Technology Transfer, 26(1-2):163-171.
Scott JT. 1989. “Historical and Economic Perspectives of the National Cooperative Research Act.” Pp. 65-84 in Cooperative Research and Development: The Industry-University-Government Relationship, Link AN and Tassey G, editors. Boston, MA: Kluwer Academic Publishers.
Sperling D. 2001. “Public-Private Technology R&D Partnerships: Lessons from the U.S. Partnership for a New Generation of Vehicles.” Transport Policy, 8:247-256.
Spielman DJ, von Grebmer K. 2004. “Public-Private Partnerships in Agricultural Research: An Analysis of Challenges Facing Industry and the Consultative Group on International Agricultural Research.” EPTD Discussion Paper No. 113. International Food Policy Research Institute, Washington, DC.
Stephan PE. 2001. “Educational Implications of University-Industry Technology Transfer.” Journal of Technology Transfer, 26(3):199-205.
Stiglitz J, Wallsten, SJ. 2000. “Public-Private Technology Partnerships: Promises and Pitfalls.” In Public-Private Policy Partnerships, Rosenau PV, editor. Cambridge, MA: MIT Press.
Thornberry JB. 2002. “Competition and Cooperation: A Comparative Analysis of SEMATECH and the VLSI Research Project.” Enterprise & Society, 3:657-686.
Wallsten SJ. 2000. “The Effects of Government-Industry R&D Programs on Private R&D: The Case of the Small Business Innovation Research Program.” RAND Journal of Economics, 31(1):82-100.
Webber D, Kremer M. 2001. “Perspectives on Stimulating Industrial Research and Development for Neglected Infectious Diseases.” Bulletin of the World Health Organization, 79:735-741
Wheeler C., Berkley S. 2001. “Initial Lessons from Public-Private Partnerships in Drug and Vaccine Development.” Bulletin of the World Health Organization, 79:728-734.
Widdus R. 2001. “Public-Private Partnerships for Health: Their Main Targets, Their Diversity, and Their Future Directions.” Bulletin of the World Health Organization, 79:713-720.