Innovation in the Pharmaceutical IndustryIn a striking example of concision, the Wall Street Journal has published a 1½-page summary of research on the drivers of innovation that Frank Rothaermel (College of Management at the Georgia Institute of Technology) and Drew Hess (McIntire School of Commerce at the University of Virginia) published two years ago in a 24-page academic paper (pdf).1
Rothaermel and Hess analyzed data covering the pharmaceutical industry in the period 1980-2001 to identify the mechanisms that best enabled firms to develop "dynamic capabilities," i.e., the ability to recognize and adapt to market and technological change, such as the rise of biotechnology.
The researchers began with the general view that "antecedents" (drivers) of innovation can be found simultaneously and interdependently at the individual, firm, and network levels. By "network level," Rothaermel and Hess are referring to external organizations in which a firm might invest resources, e.g, through forming an alliance with another firm or a university, or by making an acquisition of another firm.
To test their theory, Rothaermel and Hess investigated several hypotheses that the theory implies. (You can learn in detail about the rationale for these hypotheses by having a look at their paper.) The hypotheses are:
Hypothesis 1A. A firm's innovative output is a positive function of its intellectual human capital. In the case of pharamaceutical firms, this intellectual human capital was measured in terms of the number of scientists employed.
Hypothesis 1B. A firm's innovative output is a positive function of its star scientists, controlling for nonstar scientists. Star scientists were defined as those with publication and citation records well above average.
Hypothesis 2. A firm's innovative output is a positive function of its R&D capability. R&D was measured in terms of inflation-adjusted dollars (with firm revenues treated as a control variable).
Hypothesis 3A. A firm's innovative output is a positive function of its alliances with new technology providers. These providers include universities, research institutions, and biotechnology firms.
Hypothesis 3B. A firm's innovative output is a positive function of its acquisitions of new technology firms.
Complementarity Hypothesis. Antecedents to innovation located at the intersections between the individual and the firm (Hypothesis 4A), between the individual and the network (Hypothesis 4B), and between the firm and the network (Hypothesis 4C) complement one another, i.e., interactions across levels are positive, and thus increase a firm's innovative output.
Substitution Hypothesis. Antecedents to innovation located at the intersections between the individual and the firm (Hypothesis 5A), between the individual and the network (Hypothesis 5B), and between the firm and the network (Hypothesis 5C) substitute for one another, i.e., interactions across levels are negative, and thus decrease a firm's innovative output.
Rothaermel's and Hess's results included:
- Hypothesis 1A was supported. "... the most effective way to achieve continuous innovation over the long term is to hire and cultivate talented people. Companies that innovate through hiring will have stronger control over their intellectual property and often a steadier pipeline of future inventions because they aren't relying on outside partners for any part of the innovation process."
- Hypothesis 2A was not supported. "The odds of this strategy [investing in talented people] working improve when a company builds a team of both star and nonstar employees. Having people who can prospect for new ideas (star scientists, for example) and people who can turn those ideas into actual products and services (rank-and-file knowledge workers) makes successful innovation more likely."
- Hypothesis 2B was partially supported. "Rather than finding a linearly positive relationship beween R&D expenditures and biotech patenting, ... we find that this relationship is characterized by diminishing marginal returns."
- Hypothesis 3A was not supported, but Hypothesis 3B was. I.e., the data indicate that alliances do not increase innovative output, while acquisitions do. (But see also the findings below for Hypotheses 4B and 5B.)
- Hypothesis 5A was supported (and Hypothesis 4A was not). I.e., "a firm's intellectual human capital (proxied by its total scientists) and a firm's R&D capability are substitutes for one another ... Star scientists and R&D capability also substitute for one another ..."
- Hypothesis 5B was supported (and Hypothesis 4B was not). If a company is using human capital as an innovation strategy, it only makes sense to also invest in alliances if those alliances bring unique assets that the company does not already have access to from its internal talent.
- Hypothesis 4C was supported (and Hypothesis 5C was not). "Alliances and internal R&D spending often complement each other ... because internal knowledge allows companies to more readily identify the most promising research areas, which in turn allows them to select the most promising research partners. Internal knowledge also helps companies better understand the information that is exchanged between partners, increasing the efficiency of such arrangements."
- The most complementary innovation strategies are forming alliances and making acquisitions. "That's because forming a joint venture with a company before trying to buy it gives a company inside information about the target's worth and the value of the research it is seeking to acquire."
firms should, with the help of star scientists, identify an exogenous paradigm shift, and then assemble the requisite human assets in the form of rank-and-file scientists.Their advice to firm managers is to avoid
the "grab bag" approach to innovating, [i.e.,] employing a variety of available mechanisms simultaneously without knowledge of the possible deleterious interaction effects. Our research demonstrates that, due to path dependency and constraints imposed on a firm's financial-, managerial-, and research-related resources, a tandem approach may actually lead to decreases in innovative output. ...Instead, the managers who take a discerning and discriminating approach towards selecting innovation mechanisms will be most successful in building the dynamic capabilities necessary to continuously innovate.__________
1 "Building Dynamic Capabilities: Innovation Driven by Individual-, Firm-, and Network-Level Effects," Frank T. Rothaermel and Andrew M. Hess, Organization Science, Vol. 18, No. 6 (Nov-Dec 2007), pp. 898-921.