Until recently small firms were the invisible man of innovative activity. Most of the measurement and empirical analyses of innovation focused solely on large corporations (Scherer, 1965; and Mansfield, 1983). This reflected a theoretical framework that applied to the innovation process in large corporations but not necessarily small enterprises (Chandler, 1990).
Only within the last fifteen years has the vast degree of innovative activity contributed by small enterprises been uncovered (Scherer, 1991). Systematic, comprehensive empirical studies have provided compelling evidence that small firms generate a significant amount of innovative activity, especially in new and emerging industries (Caves, 1998).
However, there is very little systematic evidence about the role that strategic research partnerships play in small firms. Part of the reason for this paucity of knowledge is theoretical. As was the case for the innovation literature only several years ago, the theoretical frameworks to analyze joint research partnerships are predominately oriented towards large corporate partners. Measurement provides even greater challenges. Small firms have systematically lower rates of survival. In high-technology industries, small-firm survival rates are still lower. At the same time, startup rates are higher. This makes it difficult to even identify firms and track them over time. In addition, small firms are notorious for forsaking formal R&D for informal research, which typically defies measurement (Kleinknecht, 1989a and 1989b; and Roper, 1999). Measuring research partnerships between firms reporting no research is even more challenging.
At the same time, however difficult they are to measure, the importance of research linkages and partners to small firms is undeniable. While it may not make sense for firms that are new and most likely transitory to formalize strategic research partnerships with other firms and institutions, such linkages are clearly at the heart of some small-firm strategies. For example, Saxenian (1994) argues that the rich network of linkages and partnerships in the Silicon Valley region has contributed to a superior innovative performance. According to Saxenian (1990, pp. 96-97) "It is not simply the concentration of skilled labor, suppliers and information that distinguish the region. A variety of regional institutions—including Stanford University, several trade associations and local business organizations, and a myriad of specialized consulting, market research, public relations and venture capital firms—provide technical, financial, and networking services which the region's enterprises often cannot afford individually. These networks defy sectoral barriers: individuals move easily from semiconductor to disk drive firms or from computer to network makers. They move from established firms to startups (or vice versa) and even to market research or consulting firms, and from consulting firms back into startups. And they continue to meet at trade shows, industry conferences, and the scores of seminars, talks and social activities organized by local business organizations and trade associations. In these forums, relationships are easily formed and maintained, technical and market information is exchanged, business contacts are established, and new enterprises are conceived. ... This decentralized and fluid environment also promotes the diffusion of intangible technological capabilities and understandings."
Where systematic evidence does exist, it suggests that strategic research partnerships may, in fact, be more important for small firms than for their larger enterprises. As is documented in this paper, empirical evidence from the biotechnology industry shows that both formal strategic research partnerships as well as less formal linkages among firms, scientists and universities play a central role in the innovative activities of firms.
The purpose of this paper is to draw together disparate strands of literature to draw out what has been learned about the role of strategic research partnerships for small firms. The second section of the paper documents the innovative contribution made by small firms. At the same time, small firms do not undertake high amounts of research activity. The third section reconciles the paradox posed by the high degree of innovative activity combined with the relatively low level of research by suggesting that small firms rely on external knowledge sources, such as strategic research partnerships. The fourth section provides a theoretical framework for analyzing strategic research partnerships for small firms and suggests why such strategic alliances may, in fact, be more important for small enterprises than for large corporations. The fifth section focuses on the role of small-firm strategic research partnerships in high-technology industries. Finally, in the last section a summary and conclusions are provided. In particular, while objective measures of formal agreements are invaluable to fully understand the role that strategic research partnerships play in small firms, future research needs to develop subjective measures using surveys may be invaluable in order to more systematically identify (1) the extent of strategic research partnerships in small firms, (2) their determinants, and (3) their impact on economic performance.
The starting point for most theories of innovation is the firm. In such theories the firms are exogenous and their performance in generating technological change is endogenous (Arrow, 1962). For example, in the most prevalent model found in the literature of technological change, the model of the knowledge production function, formalized by Griliches (1979), firms exist exogenously and then engage in the pursuit of new economic knowledge as an input into the process of generating innovative activity. The most decisive input in the knowledge production function is new economic knowledge. Knowledge as an input in a production function is inherently different than the more traditional inputs of labor, capital and land. While the economic value of the traditional inputs is relatively certain, knowledge is intrinsically uncertain and its potential value is asymmetric across economic agents.[1] The most important, although not the only source of new knowledge is considered to be research and development (R&D). Other key factors generating new economic knowledge include a high degree of human capital, a skilled labor force, and a high presence of scientists and engineers. The most decisive input in the knowledge production function is new economic knowledge. And as Cohen and Klepper conclude, the greatest source generating new economic knowledge is generally considered to be R&D. Certainly a large body of empirical work has found a strong and positive relationship between knowledge inputs, such as R&D, on the one hand, and innovative outputs on the other hand.
The knowledge production function has been found to hold most strongly at broader levels of aggregation. The most innovative countries are those with the greatest investments to R&D. Little innovative output is associated with less developed countries, which are characterized by a paucity of production of new economic knowledge. Similarly, the most innovative industries, also tend to be characterized by considerable investments in R&D and new economic knowledge. Not only are industries such as computers, pharmaceuticals and instruments high in R&D inputs that generate new economic knowledge, but also in terms of innovative outputs (Audretsch, 1995). By contrast, industries with little R&D, such as wood products, textiles and paper, also tend to produce only a negligible amount of innovative output. Thus, the knowledge production model linking knowledge generating inputs to outputs certainly holds at the more aggregated levels of economic activity.
Where the relationship becomes less compelling is at the disaggregated microeconomic level of the enterprise, establishment, or even line of business. For example, While Acs and Audretsch (1990) found that the simple correlation between R&D inputs and innovative output was 0.84 for four-digit standard industrial classification (SIC) manufacturing industries in the United States, it was only about half, 0.40 among the largest U.S. corporations.
The model of the knowledge production function becomes even less compelling in view of the recent wave of studies revealing that small enterprises serve as the engine of innovative activity in certain industries. These results are startling, because as Scherer (1991) observes, the bulk of industrial R&D is undertaken in the largest corporations; small enterprises account only for a minor share of R&D inputs. Thus the knowledge production function seemingly implies that, as the Schumpeterian Hypothesis predicts, innovative activity favors those organizations with access to knowledge-producing inputs—the large incumbent organization.
The model of the knowledge production function becomes particularly weak when small firms are included in the sample. This is not surprising, since formal R&D is concentrated among the largest corporations, but a series of studies (Audretsch, 1995) has clearly documented that small firms account for a disproportional share of new product innovations given their low R&D expenditures.
Knowledge regarding the relationship between firm size and innovation has been largely shaped by measurement. Measures of technological change have typically involved one of the three major aspects of the innovative process: (1) a measure of inputs into the process, such as R&D expenditures, or the share of the labor force accounted for by employees involved in R&D activities; (2) an intermediate output, such as the number of inventions that have been patented; or (3) a direct measure of innovative output.
The earliest sources of data, R&D measured, indicated that virtually all of the innovative activity was undertaken by large corporations. As patent measures became available, the general qualitative conclusions did not change, although it became clear that small firms were more involved with patent activity than with R&D. The development of direct measures of innovative activity, such as data bases measuring new product and process introductions in the market, indicated something quite different. In a series of studies, Acs and Audretsch (1988 and 1990) found that while large firms in manufacturing introduced a slightly greater number of significant new innovations than small firms, small-firm employment was only about half as great as large-firm employment, yielding an average small-firm innovation rate in manufacturing of 0.309, compared to a large-firm innovation rate of 0.202. The relative innovative advantage of small and large firms was found to vary considerably across industries. In some industries, such as computers and process control instruments, small firms provide the engine of innovative activity. In other industries, such as pharmaceutical products and aircraft, large firms generate most of the innovative activity. Knowledge regarding both the determinants and the impact of technological change has been largely shaped by measurement.
Acs and Audretsch (1988, 1990) concluded that some industries are more conducive to small-firm innovation while others foster the innovative activity of large corporations corresponds to the notion of distinct technological regimes—the routinized and entrepreneurial technological regimes.
The breakdown of the knowledge production function at the level of the firm raises the question, Where do innovative firms with little or no R&D get the knowledge inputs? This question becomes particularly relevant for small and new firms that undertake little R&D themselves, yet contribute considerable innovative activity in newly emerging industries such as biotechnology and computer software. One answer that has recently emerged in the economics literature is from other, third-party firms or research institutions, such as universities. Economic knowledge may spill over from the firm conducting the R&D or the research laboratory of a university.
That knowledge spills over is barely dispute. However, the geographic range of such knowledge spillovers is greatly contested. In disputing the importance of knowledge externalities in explaining the geographic concentration of economic activity, Krugman (1991) and others do not question the existence or importance of such knowledge spillovers.[2] In fact, they argue that such knowledge externalities are so important and forceful that there is no compelling reason for a geographic boundary to limit the spatial extent of the spillover. According to this line of thinking, the concern is not that knowledge does not spill over but that it should stop spilling over just because it hits a geographic border, such as a city limit, state line, or national boundary.
Krugman (1991a, p. 53) has argued that economists should abandon any attempts at measuring knowledge spillovers because " ... knowledge flows are invisible, they leave no paper trail by which they may be measured and tracked." But as Jaffe, Trajtenberg and Henderson (1991, p. 578) point out, "knowledge flows do sometimes leave a paper trail"—in particular in the form of patented inventions and new product introductions.
Despite Krugman's warning, a recent body of empirical evidence developing novel measures of knowledge flows clearly suggests that R&D and other sources of knowledge not only generate externalities, but studies by Audretsch and Feldman (1996), Jaffe (1989), Audretsch and Stephan (1996), Feldman (1994a and 1994b), and Jaffe, Trajtenberg and Henderson (1993) suggest that such knowledge spillovers tend to be geographically bounded within the region where the new economic knowledge was created. That is, new economic knowledge may spill over but the geographic extent of such knowledge spillovers is limited.
While the literature on knowledge spillovers has identified that knowledge externalities exist, and are geographically bounded, they shed little light on the mechanisms by which knowledge is transmitted to small firms. One such mechanism is via strategic research partnerships.
According to Kogut (1988), a joint venture occurs when two or more firms pool a portion of their resources within a common legal organization. Conceptually, a joint venture is a selection among alternative modes by which multiple firms can transact. Gomes-Casseres (1997, p. 34) defines alliances more broadly as "an administrative arrangement to govern an incomplete contract between separate firms in which each partner has limited control."
According to Gomes-Casseres (1996), three factors shape the formation of alliances—capabilities, control and context. Capabilities refers to the set of tangible and intangible assets making it feasible for a firm to develop, produce and sell goods and services. Control refers to the authority of the firm to deploy those capabilities. The context refers to the external environment within which the firm operates.
Kogut (1988) emphasizes that if all three of these elements—capabilities, control and context—are present within the firm, there will be no need for it to externally seek a strategic alliance. However, if one of these elements is lacking or weak, the firm has an incentive to seek an external partner or set of partners. If an alliance is formed, the set of capabilities required by the firm shapes the structure of control in the organization. The structure of control similarly shapes the manner in which the capabilities are managed, and the nature of investments made to upgrade the capabilities over time.
Gomes-Casseres (1996) points out that in a context where size bestows a competitive advantage—due to economies of scale or scope—large enterprises will tend to have the competitive advantage. To compensate for this size-inherent cost disadvantage, small firms then have a clear incentive to engage in a strategic alliance in effectively increase their scale and scope.
An implication of the Gomes-Casseres (1996) framework is that not every small firms are at a competitive disadvantage, per se, even if larger and even very large enterprises exist in the same industry. As long as no size-inherent cost disadvantages exist, there will be no compelling reason to participate in a strategic alliance.
In addition, occupying a strategic niche provides small firms with an opportunity for viability in a context where either no scale economies exist, or there are even modest diseconomies of scale. According to Penrose (1959, pp. 222-223), "The productive opportunities of small firms are composed of those interstices left open by the large firms which the small firms see and believe they can take advantage of. The nature of the interstices is determined by the kind of activity in which the larger firms specialize, leaving other opportunities open." Caves and Porter (1978) and Newman (1978) provided compelling empirical evidence for the existence of such strategic niches.
By contrast, when a small firm is at a competitive disadvantage vis-&#agrave;-vis larger competitors developing a strategic partnership or alliance is a mechanism to compensate for size-inherent disadvantages. Gomes-Casseres (1994) provides an example of how a strategic alliance generates compensating competitiveness for small firms. A relatively small computer firm, Mips Computer Systems, operated in the same market as IBM and Hewlett-Packard. Production scale economies and market penetration determined commercial success. Mips produced reduced instruction-set computing (RISC) processors, which required large-scale production. Because of these scale economies, it was clear that only a few of the producers in the market would ultimately survive. This also meant that those designs with the greatest market penetration were likely to be among the survivors. Thus, it was crucial for Mips to obtain a large market share and influence the industry standard. Mips created an alliance including semiconductor partners and a number of systems vendors. These partners contributed production capacity, market presence, technological competencies, and finance. Mips contributed a highly specialized and unique semiconductor design. Along with one of its partners, Sun, Mips was able to attain the scale, scope, and market impact that otherwise would have been unimaginable.
Through the strategic alliance, Mips succeeded in leveraging its small size to a larger unit of competitiveness. Gomes-Casseres (1997, p. 37) observed that "Increasingly, the talk in the industry became one of how the Mips 'camp' was faring versus the camps centered around other firms."
A different factor motivating compensating strategic research partnerships for small firms is the need for finance. As Lerner and Merges (1998, p. 125) note, "Young firms with novel technologies frequently lack the financial resources to effectively introduce a new product and may find it difficult to raise equity or debt due to the informational asymmetries surrounding the project. In many cases, young firms lack complementary assets such as sales forces and manufacturing know-how, which may take many years to develop. As a result, small, research-intensive firms frequently rely on alliances with larger corporations."
In reviewing the role of financial constraints on investment behavior, Chirinko (1993, p. 1902) observed that, "The investment literature has been schizophrenic concerning the role of financial structure and liquidity constraints." As (1988, p. 141) point out, "Empirical models of business investment rely generally on the assumption of a 'representative firm' that responds to prices set in centralized security markets. Indeed, if all firms have equal access to capital markets, firms' responses to changes in the cost of capital or tax-based investment incentives differ only because of differences in investment demand." That is, the financial structure of a firm does not play an important role in investment decisions, since the firm can costlessly substitute external funds for internal capital. Under the assumption of perfect capital markets, then, firm-specific investment decisions are generally independent of the financial condition of that firm.
The assumption of perfect capital markets has, of course, been rigorously challenged. Once it is no longer assumed that capital markets are perfect, it also can no longer be assumed that external capital is a costless substitute for internal capital. An implication of this view is that the availability of internal finance, access to new debt or equity finance, and other financial factors may shape firm investment decisions.
Which view is correct? According to Fazzari, Hubbard and Peterson (1988, p. 142), "Conventional representative firm models in which financial structure is irrelevant to the investment decision may well apply to mature companies with well-known prospects. For other firms, however, financial factors appear to matter in the sense that external capital is not a perfect substitute for internal funds, particularly in the short run."
There are compelling reasons why liquidity constraints become more severe as firm size decreases. Stiglitz and Weiss (1981) pointed out that, unlike most markets, the market for credit is exceptional in that the price of the good—the rate of interest—is not necessarily at a level that equilibrates the market. They attribute this to the fact that interest rates influence not only demand for capital but also the risk inherent in different classes of borrowers. As the rate of interest rises, so does the riskiness of borrowers, leading suppliers of capital to rationally decide to limit the quantity of loans they make at any particular interest rate. The amount of information about an enterprise is generally not neutral with respect to size. Rather, as Petersen and Rajan (1992, p. 3) observe, "Small and young firms are most likely to face this kind of credit rationing. Most potential lenders have little information on the managerial capabilities or investment opportunities of such firms and are unlikely to be able to screen out poor credit risks or to have control over a borrower's investments." If lenders are unable to identify the quality or risk associated with particular borrowers, Jaffe and Russell (1976) show that credit rationing will occur. This phenomenon is analogous to the lemons argument advanced by Akerloff (1970). The existence of asymmetric information prevents the suppliers of capital from engaging in price discrimination between riskier and less risky borrowers. But, as Diamond (1991) argues, the risk associated with any particular loan is also not neutral with respect to the duration of the relationship. This is because information about the underlying risk inherent in any particular customer is transmitted over time. With experience a lender will condition the risk associated with any class of customers by characteristics associated with the individual customer.
The problems of uncertainty, asymmetric information and high transactions cost are exacerbated in innovative small firms highly reliant upon research. Biotechnology is a new industry that is knowledge based and is predominantly produced by new startups and small firms. The industry is characterized by the type of incomplete contracting described by Grossman and Hart (1986), Hart and Moore (1988), Hart (1995) and Aghion and Tirole (1994). The knowledge conditions underlying the biotechnology industry—high uncertainty, asymmetries, and high transactions costs—result in, "Redefining the work when the unexpected happens, as it invariably will. Research is by its very nature an iterative process, requiring constant reassessment depending on its findings. If there is a low risk of unexpected findings requiring program reassessment, then it is probably not much of a research program" (Sherbloom, 1991, pp. 220-221).
The relative small scale of most biotechnology firms may be attributable to the diseconomies of scale inherent in the "bureaucratic process which inhibits both innovative activity and the speed with which new inventions move through the corporate system towards the market" (Link and Rees, 1990, p. 25). Zucker, Darby and Brewer (1998, p. 1) provide considerable evidence suggesting that the timing and location of new biotechnology firms is "primarily explained by the presence at a particular time and place of scientists who are actively contributing to the basic science."
Strategic research partnerships are particularly important in the biotechnology industry (Table 1). These strategic research partnerships and linkages occur between entrepreneurial firms, between the scientists involved with the firms, between the firms and universities, and between corporations and biotech firms.
| Panel A presents the number of publicized alliances by US firms in information technology, and advanced materials between 1980 and 1994. Panel B presents only alliance involving U.S. biotechnology companies between 1981 and 1995 filed with the U.S. Securities and Exchange Commission or with state regulatory bodies that make such information public. Presented are the number of new filed alliances each year, the sum of all promised pre-commercialization payments in the filed alliances that year (the sum of the nominal payments is expressed in millions of 1995 dollars), and the actual payments to a sample of 49 of the largest biotechnology firms in each year (in millions of $1995). |
| Panel A: Inter-firm alliances by US firms in three research-intensive industries 1980-1994 | |||||
|---|---|---|---|---|---|
| Number of new alliances publicized, by national of firms | |||||
| Year | US-US | US-Europe | US-Japan | ||
| 1980 | 42 | 40 | 15 | ||
| 1981 | 48 | 30 | 26 | ||
| 1982 | 57 | 54 | 39 | ||
| 1983 | 51 | 37 | 51 | ||
| 1984 | 88 | 60 | 55 | ||
| 1985 | 86 | 82 | 52 | ||
| 1986 | 118 | 78 | 47 | ||
| 1987 | 133 | 95 | 53 | ||
| 1988 | 141 | 98 | 39 | ||
| 1989 | 122 | 89 | 44 | ||
| 1990 | 121 | 66 | 34 | ||
| 1991 | 106 | 53 | 51 | ||
| 1992 | 155 | 89 | 43 | ||
| 1993 | 192 | 104 | 45 | ||
| 1994 | 235 | 145 | 40 | ||
| Panel B: Intern-firm alliances by US biotechnology firms, 1981-1995 | |||||
| Payments through alliances (millions of 1995 dollars) | |||||
| Year | Number of new filed alliances | Pre-commercial payments promised in new alliances | Actual payments during year to 49 leading firms | ||
| 1981 | 30 | 9 | |||
| 1982 | 35 | 111 | |||
| 1983 | 31 | 152 | |||
| 1984 | 42 | 210 | |||
| 1985 | 57 | 149 | |||
| 1986 | 63 | 184 | |||
| 1987 | 62 | 415 | |||
| 1988 | 64 | 298 | |||
| 1989 | 71 | 205 | |||
| 1990 | 81 | 851 | |||
| 1991 | 115 | 741 | 647 | ||
| 1992 | 75 | 931 | 392 | ||
| 1993 | 113 | 1373 | 806 | ||
| 1994 | 66 | 1772 | |||
| 1995 | 171 | 3421 | |||
Strategic research partnerships between large corporations and biotechnology companies have been particularly important for biotech companies specializing in therapeutics. This is because the cost of developing a new drug, complying with the various layers of regulation, manufacturing the product, and then marketing the product, have required a level of finance that far exceeds the budgets of most small firms. Cullen and Dibner (1993) estimate that the cost of bringing a therapeutic drug from basic research to the market is around $250 million. At the same time, the average budget for research and development of a biotech firm is $12.5 million. To close this gap, biotech firms have engaged in a broad range of marketing and licensing agreements. Under these agreements, the biotech firm provides access to cutting edge technology in exchange from an infusion of capital from their corporate partners.
In documenting the evolution of strategic alliances in biotechnology, Cullen and Dibner (1993, p. 18) conclude that, "The primary strategic goal of small and medium-sized biotechnological companies was to develop products to be marketed by their partners and their primary concern was finding and developing alliances." The obvious advantages to such strategic research partnerships is that they enable a small, new company to concentrate on its core mission—moving from basic research to commercialization through technological innovation. The strategic alliances also enable the biotech company to reduce financial risks as well as operating costs. In addition, the biotech firm is able to better offset the major liabilities associated with biotech startups—acquiring manufacturing capabilities, marketing and sales.
The established firms are generally quite positive and supportive towards biotechnology firms. This is because of the strong complementary nature between biotechnology firms and established firms, particularly in the pharmaceutical industry. There are a number of reasons why such a complementary relationship has evolved between established and biotechnology firms. The first is that the former have recognized that it may be a more efficient structure to engage in an arms length market relationship to obtain new biotechnology products than to produce them internally. The market exchange is apparently more efficient than the internal transaction. The reason for this involves agency problems in undertaking research that is highly uncertain and asymmetric. In addition, the exposure to legal liabilities resulting from biotechnology research is reduced when that research is undertaken at a small firm with limited assets rather than in a large corporation with massive assets.
Sharp (1999) identifies three main phases in the relationship between established firms and biotechnology companies. The first phase involved the formation and incipiency of the biotechnology industries. Sharp (1999, p. 137) reports that "most of the established pharmaceutical companies were uncertain what to make of the new technology and especially of the hype surrounding its development that grew with the small firm sector in the U.S." This uncertainty combined with a considerable degree of skepticism resulted in most established pharmaceutical companies distancing themselves from the fledgling biotechnology industry in this initial phase. At the same time, Sharp points out that most established companies invested in sufficient scientific expertise to enable them to keep abreast of developments in biotechnology and monitor the industry.
The second phase began in the mid-1980s, when the period of watching and waiting ended. The established pharmaceutical recognized that, in fact, biotechnology had a valuable market potential. While strategies pursued by the established enterprises varied, most devised and implemented a strategic biotechnology policy. One common strategy that all companies pursued was to invest heavily to develop an in-house competence in biotechnology. How this was done varied considerably across companies. In some companies, scientific teams were assembled. Other pharmaceutical companies acquired such competence through the acquisition of biotechnology firms or, in some cases through mergers. Another strategy was to engage in external linkages with biotechnology companies. As Cullen and Dibner (1993) document, strategic alliance between biotechnology firms and established enterprises exploded in the mid-1980s.
The third phase, which started around a decade ago involves the commercialization of biotechnology products. The first successful biotechnology products reached the market in the early 1990s. As Juergen Drews, head of R&D at Hoffman LaRoche observed in 1993, "While there are some redundancies among the 150 or so novel proteins in development, about 100 represent truly novel substances that have no precedent in medical therapy. Not all of these proteins will reach the market, but it is fair to assume that their attrition rate will be lower than that for small chemical entities because they should cause few unmanageable toxicological problems. A conservative estimate would expect 30-40 of the recombinant proteins now under development to become successfully marketed products over the next 5-6 years. This means that an average of 5-8 novel proteins should become available each year...If we assume an average sales volume for the forthcoming recombinant proteins equal to the average revenues generated by today's recombinant drugs, the portfolio of recombinant proteins now in clinical trials should amount to $10-$20 billion."
In this third phase, the large established companies take the new biotechnology products developed by biotechnology companies and convert them into large-scale marketed products. For example, Intron A was developed by Biogen but marketed by Schering-Plough, resulting in $572 million of sales in 1993. Humulin was developed by Genetech but marketed by Eli Lilly, for $560 million of sales in 1993. Engerix-B was developed by Genetech but marketed by SmithKline Beecham for $480 million. RecombiNAK HB was developed by Chiron but marketed by Merck for $245 million.
In addition, this third phase has experienced a shift by the established companies away from the broad learning strategies of phase two and increasingly towards a more focused approach, targeting specific technologies. For example, Ciba Geigy reduced its portfolio of interests in biopharmaceuticals in 1989 in order to focus more narrowly on the development of just several targeted products. Ciba Geigy subsequently increased its investment in those targeted areas and engaged in a number of research and licensing agreements with biotechnology companies. Similarly, Bayer reduced its biotechnology research in agro-chemicals while concentrating its focus on pharmaceuticals. Hoffman LaRoche similarly pulled out of agro-biotechnology to concentrate its focus on pharmaceuticals.
| Each column indicates the year and stage at the time the agreement was signed and the primary focus for a different set of agreements. The first column indicates the distribution of all alliances, licensing arrangements, and asset sales involving biotechnology companies between 1980 and 1995 filed with the U.S. Securities and Exchange Commission or state regulatory bodies who make such information public. The second column indicates the distribution of all such agreements summarized by Recombinant Capita. The final column characterizes the final sample of 200 technology alliances initiated between biotechnology and pharmaceutical companies or between biotechnology firms in the 1980-1995 period. |
| All Filed Agreements | All Summarized Agreements | Final Sample | |
|---|---|---|---|
| Time Period: | |||
| 1980-1987 | 20% | 11% | 14% |
| 1988-1990 | 18 | 21 | 21 |
| 1991-1992 | 26 | 26 | 34 |
| 1993-1995 | 36 | 42 | 31 |
| Stage of Product at Signing: | |||
| Discovery/Lead Molecule | 65 | 57 | 64 |
| Pre-Clinical Development | 9 | 11 | 21 |
| Undergoing Regulatory Review | 17 | 23 | 15 |
| Approved for Sale[a] | 9 | 9 | 0 |
| Primary Focus of Agreement: | |||
| Human Therapeutics | 75 | 83 | 92 |
| Human Diagnostics[b] | 18 | 15 | 4 |
| Agricultural or Veterinary Applications | 6 | 2 | 4 |
|
[a] The sample is constructed to include only alliances with a research or a product development component. Thus, many of the agreements in the database involving approved products, which solely entail the marketing or sale of an existing product or process, are excluded from the sample. [b] Many of the agreements involving human diagnostics entail the marketing or sale of an existing product or process developed by a biotechnology company in the course of a program to introduce a new therapeutic. (Because diagnostics tests are frequently of modest economic importance and viewed as tangential to the firm's product development focus, biotechnology firms often sell these outright to major firms specializing in this area.) Because these agreements are not alliances with a research or product development component, they are excluded from the sample. |
|||
Lerner and Merges (1998) use a novel data base identifying biotechnology research alliances complied by Recombinant Capital, a San Francisco-based consulting firms specializing since 1988 in tracking the biotechnology industry. As of December, 1998, Recombinant Capital had identified over 7,000 alliances between private biotechnology firms by examining SEC and state filings, the news media, and press releases (Lerner and Tsai, 2000). Lerner and Merges (1998) drew a random sample of 200 of the alliances to encode. Table 1 shows the distribution by time period, stage of product at signing, and the primary focus of the agreement. Table 3 provides a summary of the characteristics of the research alliances. It should be noted that most of the biotech alliances are arranged at a very early stage. Most of the alliances were signed prior to the beginning of clinical studies.
Another important point from Table 3 is that the biotechnology firms have only modest financial resources. On average, the biotech firm had around $10 million in revenue in the year prior to the alliance. However, given the mean expenditures of over $21 million, mostly on R&D, virtually all of the biotech firms were making losses. The loss corresponds to about one-third of the mean firm's shareholder equity and one-half of its cash and equivalents. The final point from Table 3 is that the strategic partners providing finance are typically much larger than the biotechnology companies, suggesting that the large pharmaceutical companies are providing finance, while the small biotechnology firms provide knowledge.
Lerner and Mergers(1988) use this sample of strategic research alliances between small biotechnology companies and large pharmaceutical companies to examine the determinants of the control rights in the alliances. The control rights consist of:
The empirical evidence suggests that the assignment of control rights between the large pharmaceutical corporation and the small biotechnology company is done in a manner that maximizes innovative output. The exception involves those strategic alliances where the small biotech firm has few resources and little external financing is available.
| The sample consists of 200 technology alliances initiated between biotechnology and pharmaceutical companies or between biotechnology firms in the 1980-1995 period. The table summarizes the financial market conditions around the time of the alliance and the characteristics of the firms in the alliance and the alliance itself. The stage of product, focus or alliance, and characteristics of pair of firms in alliance measures are all dummy variable. The financial condition and alliance payment variables are expressed in millions of 1995 dollars. The date variable is expressed as a decimal (e.g., July 1, 1995 is coded as 1995.5). |
| Variable | Mean | Median | Standard Deviation | Minimum | Maximum |
|---|---|---|---|---|---|
| Stage of Lead Product at Time of Alliance | |||||
| Discovery/Lead Molecule | 0.64 | 0 | 1 | ||
| Pre-Clinical Development | 0.21 | 0 | 1 | ||
| Undergoing Regulatory Review | 0.15 | 0 | 1 | ||
| Focus of Alliance: | |||||
| Human Therapeutics | 0.92 | 0 | 1 | ||
| Human Diagnostics | 0.04 | 0 | 1 | ||
| Agricultural or Veterinary Applications | 0.04 | 0 | 1 | ||
| Condition of Financing Firm: | |||||
| Revenues in Prior Year | 8912 | 5218 | 18649 | 1 | 179601 |
| R&D Expenditures in Prior Year | 588 | 457 | 499 | 2 | 1958 |
| Net Income in Prior Year | 645 | 473 | 623 | -457 | 2232 |
| Cash Flow from Operations in Prior Year | 970 | 668 | 943 | -448 | 5234 |
| Cash and Equivalents at End of Prior Year | 1048 | 644 | 1066 | 1 | 4938 |
| Total Assets at End of Prior Year | 7765 | 5716 | 8210 | 1 | 53632 |
| Shareholders' Equity at End of Prior Year | 3738 | 2851 | 3569 | 0 | 17505 |
| Condition of R&D Firm: | |||||
| Revenues in Prior Year | 11 | 0 | 80 | 0 | 1029 |
| R&D Expenditures in Prior Year | 9 | 5 | 16 | 0 | 171 |
| Net Income in Prior Year | -6 | -5 | 14 | -65 | 134 |
| Cash Flow from Operations in Prior Year | -5 | -5 | 18 | -62 | 171 |
| Cash and Equivalents at End of Prior Year | 16 | 8 | 26 | 0 | 229 |
| Total Assets at End of Prior Year | 36 | 14 | 111 | 0 | 1079 |
| Shareholders' Equity at End of Prior Year | 25 | 11 | 68 | -17 | 665 |
| Age of R&D Firm | 5 | 4 | 3 | 0 | 36 |
| Characteristics of the Alliance: | |||||
| Date of Alliance | Jun 1991 | Dec 1991 | 3.1 Years | Jan. 1980 | Dec. 1995 |
| Minimum Length of R&D Alliance (years) | 3.79 | 3.00 | 2.65 | 0.75 | 31.00 |
| Total Pre-Commercialization Payments | 29.01 | 21.42 | 28.94 | 0.19 | 216.28 |
| Payment at the Time of Signing | 1.76 | 0.51 | 3.02 | 0.00 | 12.00 |
| Previous Alliance Between Firms | 0.06 | 0 | 1 | ||
| Control Rights Given to R&D Firm (out of 25) | 9.22 | 9 | 2.68 | 0 | 16 |
In a subsequent paper, Lerner and Tsai (2000) use the same data set to address two additional questions, (1) Whether success rates differ in agreements that are (i) signed in periods with little external equity financing availability and (ii) cede the bulk of the control to the financing firm; and (2) Whether the less attractive agreements are renegotiated. They find that contracts for strategic research alliances that are signed at times when biotechnology firms are raising little external financing and that assign the most control rights to the large corporation perform significantly worse. These agreements are also more likely to be renegotiated if financial market conditions improve.
Audretsch and Stephan (1995) document the strong research partnerships that exist among universities and biotech firms. These partnerships are crucial because biotechnology companies are strongly defined by their scientists. Many of these scientists, particularly senior scientists with strong reputations, do not work for the biotechnology company full time, but instead are members of university faculties.
For example, Audretsch and Stephan (1999) show that, of 101 founders of new biotechnology firms in the early 1990s, nearly half (50) are from universities. Of these fifty, 35 remain associated with their universities on a part-time basis, while the remaining 15 founders left the university to work full-time for their biotech firm.
These university-based scientists fulfill a variety of roles within biotechnology companies. Some are founders, others serve as members of scientific advisory boards (SAB's), while still others serve as directors. The degree of knowledge provided by university-based scientists varies according to the role played by the scientist. Scientific founders seek out venture capitalists in order to transform technical knowledge into economic knowledge. Scientific advisors provide links between scientific founders and other researchers doing work in the area. They, along with founders, also provide the possibility of outsourcing research into university laboratories staffed by graduate students and post-docs. The concept of scientific advisory boards also provides the firm the option of having, at minimal cost, a full roster of the key players doing research in the firm's area of expertise.
In addition to providing knowledge to newly formed biotechnology companies, university-based scientists also provide a signal of firm quality to the scientific and financial communities. An effective way to recruit young scientists is to have a scientific advisory board composed of the leading scientists in the field. George B. Rathman, president and Chief Executive Officer of Amgen, attributes much of the company's success to an SAB of "great credibility" whose "members were willing to share the task of interviewing the candidates for scientific positions." Rathman goes on to say that the young scientists that Amgen recruited would not have come "without the knowledge that an outstanding scientific advisory board too Amgen seriously" (Burrill, 1987, p. 77).
Certain roles, such as being a founder of a biotechnology firm, are more likely to dictate geographic proximity between the firm and the scientist than are other roles that scientists play. This is because the transmission of the knowledge specific to the scientist and firm dictates geographic proximity. Presumably scientists start new biotechnology companies because their knowledge is not transferable to other firms for the expected economic value of that knowledge. If this were not the case there would be no incentive to start a new and independent company. Because the firm is knowledge-based, the cost of transferring that knowledge will tend to be the lowest when the firm is located close to the university where the new knowledge is being produced. In addition, the cost of monitoring the firm will tend to be minimized if the new biotechnology startup is located close to the founder.
By contrast, the role of scientific advisor to a biotechnology company does not require constant monitoring or even necessarily specialized knowledge. Thus, the inputs of scientific advisors are less likely to be geographically constrained. Furthermore, geographic proximity of all major researchers in a particular scientific field is unlikely given the opportunity cost that universities face in buying into a single research agenda. Thus, if firms are to have access to the technical knowledge embodied in the top scientists in a field, they will be forced to establish links with researchers outside of their geographic area. Scientists whose primary function is to signal quality are also less likely to bet local than are scientists who provide essential knowledge to the firm. Their quality signal is produced by lending prestige to a venture they have presumably reviewed—a task that can be accomplished with credibility from a distance.
To identify the links between knowledge sources, the incentives confronting individual scientists, and where the knowledge is commercialized, Audretsch and Stephan (1996) rely upon a data base collected from the prospectuses of biotechnology companies that prepared an initial public offering (IPO) in the United States between March 1990 and November 1992. This includes a total of 54 firms affiliated with 445 university-based scientists were identified during this time period. By carefully reading the prospectuses, it was possible to identify the names of university-based scientists affiliated with each firm, the role that each scientist plays in the firm, and the name and location of their home institutions. Universities and firms were then grouped into regions, which are generally larger than a single city but considerably smaller than a state. Certain areas, for example, metropolitan New York, cross several state lines.
Only 138 of the 445 links observed between scientists and biotechnology companies are local in that the scientist and firm are located in the same region. This suggests that geographic proximity does not play an important role for links between biotechnology companies and scientists in general. However, the geographic link between the scientist and the founder is influenced by the particular role played by the scientist in working with the firm. Most strikingly, 57.8 percent of the scientist-firm links were local when the scientist was a founder of the firm; 42.1 percent were non-local. By contrast, when the scientist served as a member on the SAB, only 31.8 percent of the links were local, while 68.2 percent were non-local. This disparity suggests that the nature of the knowledge transmitted between the university and the biotechnology firm may be different between scientists serving as founders and those serving on a SAB. Presumably it is the difference in the nature and quality of the knowledge being transferred from the university to the company that dictates a higher propensity for local proximity in the case of the founders, but not for SAB members.
If strategic research partnerships are important to large corporations, they are even more important to small firms. This is because that a small enterprise is more likely than its larger counterpart to be lacking a key component involving control, capabilities and context. As a consequence, small firms may be more dependent upon strategic research partnerships as a mechanism to compensate for size-inherent competitive disadvantages.
Unfortunately, if measurement of strategic research partnerships is challenging for large corporations, it is even more of a problem for small firms. Just as small firms are a more heterogeneous population than large corporations, strategic research partnerships may take on more heterogeneous forms with small firms than with their larger counterparts. Very little comprehensive and systematic empirical evidence exists about the role that strategic research partnerships play for small firms. Just as scholars were slow to measure the innovative activity of small firms, they have been equally slow to measure and analyze the role that strategic research partnerships play for small firms.
While formal agreements clearly play a role in biotechnology, this may be less true in other industries. A virtue of the Recombinant Capital data base is the objectivity in measurement—strategic research alliances are measured externally and reflect contractual agreements. Of course, a cost of that objectivity is the omission of informal research alliances. Just as informal R&D is more important for small firms than for large corporations (Kleinknecht, 1987; Roper, 1999), informal research partnerships may also be of greater significance for small enterprises. These informal research partnerships clearly involve scientists from different firms and institutions working together, scientist mobility, as well as informal linkages among firms. This might suggest that developing subjective measures using surveys may be invaluable in order to more systematically identify (1) the extent of strategic research partnerships in small firms, (2) their determinants, and (3) their impact on economic performance.
[1] Arrow (1962) pointed out this is one of the reasons for inherent market failure.
[2] Macki (1996) points out that these views are not original with Krugman (1991).
|