|The drug industry, powered by the huge profits of the 1980s, had lately embraced the concept of 'strategic alliances.' Fervently in vogue, they were thought to solve a common problem: in a fractionated marketplace spanning countless diseases, an explosion in knowledge, and thousands of laboratories, no big company was big enough and no small one clever enough to go it alone.|
|Berry Werth, The Billion Dollar Molecule|
Biotechnology is a sector where the growth of strategic research alliances has been truly dramatic with over 20,000 alliances formed and an annual average growth rate of 25% percent (Fisher 1996). Hagedoorn (1993) finds that biotech yields the most prolific rate of alliance formation of any sector. Until the early 1980s product development in drugs, chemicals and agriculture followed the classic in-house vertically integrated approach although academic research historically figured prominently in the initial stages of product development (Galambos 1995; Hounshell and Kenly Smith 1988; Swann 1985; Weatherall 1990). The advent of biotechnology, the commercial application of recombinant DNA and molecular genetics technology was a marked departure from the chemically-based expertise of the large firms and created the need for new collaborative research, joint ventures and new forms of cooperative. By its nature, biotech is a very knowledge-intensive industry and progress requires complementary assets that reside in different types of organizations.
There are basically three important actors in biotech research alliances, universities, small entrants and large incumbent firms. Universities and research institutes are the source of scientific knowledge and talent, potentially important breakthroughs and intellectual property and access to the large number of patients required to complete clinical trials. Universities have accepted a new mission of active technology transfer and biotechnology is important to that mission. New Biotech Firms (NBFs) are start-up firms that typically embody the commercial application of university knowledge. There are approximately 1400 NBFS in the US. They are typically small, specialized in the types of products and applications they pursue and in need of financing and expertise. Large established companies have experience in large-scale production, marketing and distribution. Most importantly, they have expertise relevant to navigating the regulatory process required to bring products to the market, and have the substantial resources necessary to complete the process.
Strategic research alliances are formed to bring these actors' complementary competencies together with the goals of advancing the technology and introducing commercial products to the market. Strategic research alliances in biotech cover every possible combination with NBFs partnering with larger firms, NBFs forming alliances between themselves, large established companies joining forces, alternatively both large and small firms partnering with universities, or three-way relationships involving the combination of large companies, NBFs and universities. This review provides an overview of strategic research partnerships in biotech, beginning by outlining the policy issues, then considering data and indicators used to study biotech alliances and concluding with a consideration of what types of data that might help inform the policy debate.
Strategic research alliances have many purposes and Hagedoorn (1993) concludes that biotech alliances encompass most or all of these motivations simultaneously. To the extent that strategic alliances in biotech represents a new system for commercializing science that may become a dominant model for other technically complex emerging sectors, understanding the motivations, incentives and barriers to the formation and operation of biotech strategic alliances informs restructuring towards this formation. We may ask if strategic research alliances are an efficient ways to organize resources for scientific advance. In addition, since public funding is so important in the life sciences we may seek to understand what is the role of the National Institutes of Health and other public entities and how the resulting rents that accrue to successful products are distributed.
NBFs are new entrants that serve as intermediate organizations between universities and large pharmaceutical firms. They typically form around licensees of university intellectual property and involve university researchers as either founders, members of the scientific advisory board, employees or consultants (Audretsch and Stephan 1996; Zucker, Darby and Armstrong 1998). While industry-university collaborations are often strained due to different objectives and time constraints, NBFs are able to license university technology and work with university researchers while being more attuned to commercial pressures (Bower 1992). Thus, NBFs may facilitate the commercialization of academic science and the realization of increased efficiencies. Of course, universities receive licensing fees, royalties and even equity from NBFs in exchange for their intellectual property, yet we do not have a good understanding of how these returns are distributed or if this is an efficient way to organize science and provide incentives for the generation and use of knowledge. We do not know how scientists' financial interest in companies may limit the flows of knowledge that are typically unrestricted from universities. Potentially profitable research findings may be kept confidential, remain unpublished, or be significantly delayed in order to secure proprietary rights.
The literature has documented the types of alliances that accompany biotech research. table 1 provides an overview of academic studies with attention to the focus of the study and the data source. In general, studies focus on firms as the unit of analysis or focused on alliance agreements. While the academic literature has used existing and often proprietary databases, there is typically an emphasis on augmenting the data with data from other sources or conducting complementary case studies to add understanding to the results. Sixty-five percent of these studies used case studies.
The consensus is that firms form thick networks that involve multiple partners in a variety of alliances in order to move research forward. Studies have also either focused on alliance characteristics and the effects on participants or alternatively have looked at firm performance to assess how alliance participation affects firm outcomes such as initial public offerings and market valuation. Each of these deserves mention.
The literature generally accepts that alliances are beneficial for the participants. There appear to be great synergies between the research alliance participants that allow firms to take advantage of their competitive assets, prevent duplication of efforts and promote economic efficiencies. Specifically, NBFs gain much needed revenue and access to specialized resources. Large companies gain new products for their product development pipeline, which also helps them retain their attractiveness in the capital markets. Galambos and Sturchio (1998) find that large pharmaceutical firms are able to establish significant capabilities in new fields because of alliances with NBFs. Interestingly, Bower and Whittaker (1993) find that NBF partnerships may increase knowledge spillover potential by acting as a knowledge conduit indirectly connecting large companies.
Strategic research alliances provide pathways for knowledge spillovers; however, it is difficult to measure the benefits in terms of knowledge generation and refinement. Two important indicators are the rate of innovation and the rate of growth of the participants, measured by market valuation, revenue or employment. Evaluating the economic consequences of biotech strategic research partnerships is more difficult due to the short time frame these alliances have been in existence, the rapid changes in NBF ownership and the general volatility of the market. Some notable results include:
Although the consensus in the literature is that alliances are mutually beneficial, some policy issues warrant discussion. First is the relative absence of research on alliances outside of human therapeutics and diagnostics. Most notably, research alliances in agricultural biotech appear to be different due to a market structure with fewer NBFs and greater market concentration. Second, there are concerns that NBFs may enter into partnerships due to a lack of capitalnot because it is the most appropriate strategy. Third, strategic alliances that limit NBFs to be research boutiques may not be the best strategy for the long-term growth of knowledge in the industry. Fourth, there are growing concerns about the distribution of profits from research largely funded by taxpayers. Finally, we may question the degree to which the strategic research alliances in biotech represents a new model of commercializing university science that may extend to other emerging technology-intensive sectors.
Most research on biotech strategic alliances has focused on medical applications, in large part due to Wall Street investment interests, which in turn, influence the types of data that are readily availability. This focus ignores the importance of biotechnology to the other applications such as agricultural (see Kalaizandonakes and Bjornson 1997 for an exception). Biotechnology is already beginning to improve crop yields and to provide better pest control and new agricultural products, thus reducing farm input costs and benefiting the environment (Service 1998). Advances in agricultural biotechnology have potential to increase agricultural self-sufficiency and economic stability in developing countries. The United States currently leads the world agricultural biotechnology; however, other countries have aggressively moved into this application.
|Author(s)||Unit of Analysis and Alliance Focus||Data Source|
|Arora and Gambardella (1990)||Large U.S. European and Japanese firms with other parties, particularly NBFs and universities||Primary data on the number of agreements for large pharmaceutical and chemical firms (n=81)|
|Barley, Freeman and Hybels (1993)||All organizations involved in biotech alliances||BioScan organizations (n=3056)|
|Audretsch and Stephan (1996)||University-based composition of Scientific Advisory Boards of NBFs||Primary collection of Initial Public Offering (IPO) information on NBFs|
|Baum, Calabrese and Silverman (2000)||NBFs alliances||Original data collection on the universe of 142 biotechnology firms founded in Canada from 1991 to 1996.|
|Bower and Whittaker (1993)||Research partnership of two large firms (Merck and Sandoz) with one NBF (Repligen)||Case study of the Merck-Repligen-Sandoz Network|
|Chang (1998)||Structure of R&D Intensive Firms||Case study of Chiron|
|Cockburn and Henderson (1998)||20 largest pharmaceutical companies with public funded research||Author's compilation of development of 21 drugs; co-authorship of company researchers with university and public researchers using bibliographic citations|
|Deeds and Hill (1996)||NBFs in bio-pharmaceutical product development||BioScan (n=132 NBFs)|
|Deeds, DeCarolis and Coombs (1999)||NBFs in bio-pharmaceutical product development||BioScan (n=94 NBFs) augmented with publication records|
|Estades and Ramani (1998)||Network Structure of 20 NBFs||Case Study of twenty NBFs: ten each in Britain and France|
|Fildes (1990)||NBF collaborations with large firms and other NBFs||History of a biotechnology firm, Cetus (Fildes' company)|
|Freeman and Barley (1990)||Genentech's network of alliances||Detailed Case Study on Genentech|
|Hagedoorn and Schakenraad (1990)||Incidents of inter-firm cooperative agreements||MERIT-CATI|
|Kalaitzandonakes and Bjornson (1997)||All types of agreements in agro-biotech||Collected published data for 1600 collaborative agreements including joint ventures, mergers, acquisitions, licensing agreements and equity investments.|
|Kogut, Shan and Walker (1992)||NBFs agreements with large firms||BioScan (n=114 NBFs with cooperative agreements prior to 1989.|
|Lerner and Merges (1998)||Alliances between NBFs and pharmaceutical firms||Recombinant Capital database (n = 200 alliances)|
|Mang (1998)||NBFs in human therapeutics, diagnostics and vaccines||Original data collection on 81 collaborative projects involving 23 NBFs.|
|McMillan, Narin and Deeds (2000)||Publicly traded NBFs||IPO Prospectuses of 119 NBFs augmented with patent citations.|
|Peters, Groenewegen and Fiebelkorn (1998)||Projects between Public Research Institutes and Private Companies||European Community BRIDGE program joint project participation|
|Pisano (1990)||Development of R&D projects||92 R&D projects of large pharmaceutical firms|
|Powell and Brantley (1992)||DBFs in Human Therapeutics and Diagnostics||1990 Edition of BioScan firms (n=129)|
|Powell, Koput and Smith-Doerr (1996)||DBFs in Human Therapeutics and Diagnostics||Relational database constructed by augmenting BioScan with industry directories, and annual reports, interviews and other sources. (n=325)|
|Prevezer and Toker (1996)||Licensing, marketing, and research alliances for U.S. biotech firms||Institute for Biotechnology information database (U.S. companies to 1980)|
|Segers (1993)||New technology based firms in microelectronics and biotech in Belgium with large established firms||Case studies of two New Biotech Entities and five larger firms|
|Senker and Faulkner (1992)||Public Research Institutes and Private Companies||Seven case studies of collaboration|
|Shan, Walker, and Kogut (1994)||NBFs agreements with large firms||BioScan (n=114 NBFs with cooperative agreements prior to 1989.|
|Stuart, Hoang and Hybels (1999)||Young, venture capital-back firms specializing in human diagnostics and therapeutics.||Relational database constructed with Recombinant Capital, Micropatent Biotechnology Patent Abstracts and other published sources (n=301)|
|Zucker and Darby (1996)||World's top twenty drug-discovery firms||Alliances inferred from discovery of new biological entities, genetic sequence patents and c0-publishing.|
|Zucker and Darby (1997)||Large pharmaceutical company with universities and NBFs.||Case Study of one of the 5 largest U.S. pharmaceutical firms. Quantitative data on co-publishing.|
|Zucker, Darby and Armstrong (1998)||Universities and NBFs||Telephone census of California New Biotech Firms (NBFs).|
Agricultural biotech alliances appear to be different from the human drug and therapeutics market. First, there are relatively fewer NBFs relative to the size of the market. Most of the research alliances appear to involve large firms with universities. Second, the Institute for Biotechnology Information reports that the ag-sector has had a relatively large percentage of legal actions (18%) when compared to the pharmaceutical sector (6 %).
Greis, Dibner and Bean (1995) find that certain barriers to innovation, notably a lack of capital, motivate partnering arrangements among biotech start-ups. One policy concern is the degree to which small firms are forced into alliances due to a lack of capital, external funding opportunities or stock market volatility (Lunzer 1988). This may place the NBF in a disadvantaged bargaining position. Lerner and Merges (1998) find that the allocation of control rights in an alliance increased with the firm's financial resources; thus, financially weak firms may be relatively disadvantaged. As a point of reference, biotech alliances are a major source of revenue, generating $1.35 billion in 1997 for the top 100 NBFs ranked by number of agreements, for a compound annual growth rate of 33%.
One important question raised by Pisano (1997) is that strategic alliances in which the NBF produces the idea and the larger firm undertakes scale-up or large scale production may not be the best long-run strategy due to the specific and specialized nature of the production processes. With an unproven new product, manufacturing process innovation may be critical for developing competence and long-term advantage. In contrast to strategic research partnerships that limit the scope of the NBF, the alternative strategy of becoming a fully integrated operation may be a source of commercially valuable knowledge. Gray and Parker (1997) find that the manufacturing of biotech products has occurred in geographic regions where the pharmaceutical industry has excess capacitynot near the centers where the technology was developed and where the knowledge to increase process productivity, improve product quality and augment the specialized knowledge base (Feldman and Ronzio forthcoming). Thus, strategic alliances may undermine the long-term growth potential of the biotech industry.
There are concerns that strategic alliances with universities allow drug companies to profit from research supported by taxpayers (Gerth and Stolberg 2000). Although the consensus among economists is that, the system of innovation is efficient and that private companies need profit incentives in order to develop commercial products from basic university knowledge (Nelson 1996), the allocation of the return from products developed from research that was publicly funded appears to be developing into a contentious public policy issue.
The most relevant question we may ask is if this model represents a new system for commercializing science that will become dominant for other technically complex emerging sectors. While strategic alliances have proven difficult to manage, their numbers and persistence indicates that the participants must achieve some gains and benefits. Yet, we still have a limited understanding of the most efficient governance structures, contractual terms and monitoring procedures. Increasingly the literature recognizes that the benefits of contracting and outsourcing depend on specific attributes of the technology and the inherent costs of forming and maintaining external partnerships (Hamilton, Vila and Dibner 1990, Pisano 1990, 1991, Greis, Dibner and Bean 1995; Mang 1998). When transaction costs are high, firms pursue in-house research rather than strategic alliances. Biotech start-ups' concerns about the loss of appropriability of intellectual property can limit the firm's willingness to participate in external partnerships (Zeckhauser 1996). Less is known about the contractual relationships that protect the intellectual property interests of the NBFs and universities, reduce moral hazard concerns and minimize transaction costs (Mayer and Nickerson 2000).
The next section considers publicly available data sources that have been used to investigate biotech strategic research alliances.
The consensus from the literature is that firms that are active in biotech have multiple partnerships that cover the range of partnership types. Early studies of biotech alliances were based on case studies of a few companies. Work that is more recent has relied on one of three industry specific data sources: BioScan, Recombinant Capital's Biotech Alliance Database (ReCap), or the Institute for Biotechnology Information (IBI) or has used general alliance data such as MERIT-Cooperative Agreements and Technological Indicators (CATI) maintained by John Hagedoorn. Each of the industry databases will be described in turn. Hagedoorn, Link and Vonortas (2000) provide a description of the MERIT- (CATI) data as well as the U.S. CORE database.
BioScan is perhaps the data source most used to investigate strategic research partnerships in the literature. The database is maintained by American Health Consultants and provides profiles of approximately 1,500 U.S. and foreign companies actively involved in biotechnology research and development. The profiles contain information on strategic alliances, mergers, product acquisitions, new products in development, licensing and R&D agreements, principal investors, financial information, and key personnel. Information available includes address, personnel, history, facilities, financial information, research interests and products in development. This proprietary data is primarily intended for purposes such as generating targeted mailing lists, locating business prospects and researching potential partners, and determining industry agreement details.
Research using BioScan typically augments the data with other sources. For example, Powell, Koput and Smith-Doerr (1996) built an augmented database that filled in missing information from other industry directories, published company information and industry publications such as Genetic Engineering News, which tracks alliance announcements. Despite this diligence, Powell, Koput and Smith-Doerr (1996: 118), in an interview with the CEO of Centocor note that the response that the formal agreements were "the tip of the icebergit excludes dozens of handshake deals and informal collaborations, as well as probably hundreds of collaborations by our company's scientists with colleagues everywhere."
The Institute for Biotechnology Information (IBI) maintains a proprietary database of strategic activities, including alliances, related to the biotechnology industry. Prevezer and Toker (1996) provide an example of a study using this source. For the year 1996, IBI entered 1,368 actions into the database, ranging from marketing and licensing agreements between companies to regulatory approvals and public offerings of individual companies. IBI defines a biotechnology action in most cases as an activity that involves an organization working with genetic engineering or other biotechnologies in their R&D or manufacturing activities. IBI notes the participants involved in the action as well as the type of technology and stage of development involved. Feldman and Ronzio (forthcoming) use this data to examine regional specialization in biotech product applications.
Perhaps the most promising existing publicly available database to investigate biotech research alliances is Recombinant Capital's Biotech Alliance Database (ReCap). The database focuses specifically on alliances and contains summaries of more than 7,900 alliances in biotech that have been formed since 1978. The material is gathered from the U.S. Securities and Exchange Commission (SEC) filings of biotechnology companies, as well as from press releases and other literature and company presentations made at investment conferences and other public meetings. The Alliance Database is principally concerned with alliances for which a biotechnology company partners with a major drug company (drug/biotech), with a university (university/biotech), or with another biotechnology company. In addition, the Database contains many, although by no means all, summaries of alliances of non-biotechnology alliances in the life sciences although there appears to be limited coverage on agricultural biotech. The Alliance Database is full-text indexed and searchable by company name. An example of the type of data that is available from the Alliance Database in provided in the Appendix. Lerner and Merges (1998) have analyzed these data.
A related database maintained by the same company is rDNA.com. This is a very comprehensive proprietary database of alliances that offers an alliance summary (including deal press releases), the full text of the actual contract as filed with the SEC (for deals that have been filed) and contract analyses based on a synthesis of the terms of the alliance. This database would be useful for understanding the terms of the contract, the balance of power between the collaborators and ways in which agreement terms have evolved and adapted to market changes.
In order to understand biotech research alliances, researchers have used co-authorship bibliographic citations to discern the degree of collaboration (See, Cockburn and Henderson 1998). McMillan, Narin and Deeds (2000) use patent citations. The literature is developing rapidly. Our understanding of the policy issues may be facilitated with greater integration of the proprietary alliance databases with other sources of company and university data.
Biotechnology strategic research partnerships represent new methods of conducting science and organizing innovation. In this new era, different types of public policy will become relevant. With new discoveries and development of products and processes moving rapidly, there is little room for errora nation that waits before investing in the requisite infra-technology and infrastructure or that creates barriers to bringing these products to market will be left behind technologically, and in turn will be likely to face slower economic growth (Tassey, 2000). While the practice of strategic research partnerships began in the U.S., European firms have adopted this model and are aggressively pursuing it (Senker and Sharp 1997; Estades and Ramani 1998).
New indicators that would be developed to understand strategic research partnerships should be sensitive to what the questions and policy concerns are in biotech strategic alliances. Current publicly available data, such as the CORE data maintained by Al Link from Federal Register announcements, does not contain information about the companies involved and the biotech sector does not fit well within the existing industrial classification scheme.
 To the author's knowledge the CORE data has not been used to investigate biotech alliances due to the difficulty of identifying biotech within the confines of the Standard Industrial Classification (SIC) code system.
 The website for Bioscan is http://www.ahcpub.com/ahc_root_html/products/newsletters/bsch.html. The cost of the data in either hard copy or digital form is $1395 for one year and includes six bi-monthly updates. Institutional memberships, which allow access by multiple users, are also available.
 The company was started by one of Lerner's former students at the Harvard Business School.