In this paper, we provide an ex-ante explanation for why some technologies such as James Watt’s steam engine move successfully across broad technological fields, while other technologies do not. Using a sample of VC-backed biotechnology firms, we examine firm knowledge exploration choices along three dimensions—the decision to build from technologies across broad fields, the decision to explore application domains that are new to the firm, and the decision to mix these two options at the same time. We argue that firm-level invention decisions find differing responses when received by the selection environment. We find evidence of a “breadth-of-impact frontier” for technologies, wherein the choice of whether a firm should enter into a new application domain than those of the past should be informed by the degree to which the technology is citing prior work narrowly or broadly. The findings suggest that the belief that broad sourcing diversity will always result in greater citation diversity requires some caveats. The results contribute to the understanding of not only how entrepreneurial firms evolve but also how individual firms contribute to collective progress.
For some reason, some technologies such as James Watt’s steam engine move successfully across broad technological fields, while other technologies do not. In the case of Watt’s engine, the technology moved into fields as widely divergent as water pumping in the coal mining industry to propelling locomotives in the railroad industry (Diamond, 1999). As Levinthal (1998) describes, the porting of existing technology to a new domain of application triggers the transformation of technology across domains. In the case of these boundary-spanning technologies, insights and discoveries from different organizations (or organizational units) are merged into new products or new technical solutions (Rosenkopf and Nerkar, 2001).
It should be no surprise then that scholars traditionally have attributed the emergence of such boundary-spanning technologies to particular abilities of the individual firm. A firm’s so-called “second-order competence” (Rosenkopf and Nerkar, 2001) is the ability of the firm to create new knowledge through recombination of knowledge across boundaries. Alternatively, a firm’s “architectural competence” (Henderson and Cockburn, 1994) represents the multidimensional ability to access new knowledge outside the boundaries of the organization and to integrate knowledge flexibly across disciplinary class boundaries within the organization. Both firm-level abilities may play an important role in creating technologies with powerful breadth-of-impact (i.e., the generality in utilization of an invention) in the economy (Hall et al., 2001). In such cases of second-order or architectural competence, entrepreneurial firms can impact technological progress in an individual way, where discoveries are made by the entrepreneurial firm and these discoveries and their benefits are internalized (Fairtlough, 2000). What these traditional views tend to set aside, however, is the role of the ability of others—not just the focal individual inventing firm—to act in the technological community, i.e., the selection environment (Levinthal, 1998). Discoveries are picked up, utilized, and improved upon by other firms in the technological community, so that in addition to impacting technological progress in an individual way, entrepreneurial firms also impact technological progress in a collective way. Thus, boundary spanning technologies evolve over time through the actions within the selection environment.
In this paper, we attempt to shed light on the interplay between individual and collective aspects of invention by investigating the choices made by the individual inventing firm and the choices available to other firms in the selection environment as a result. We contribute to the literature in three ways. First, we contribute to the understanding of not only how entrepreneurial firms evolve but also how individual firms contribute to collective progress. Firms have choices that can impact not just the individualized invention process (e.g., Freel, 2005), but the cumulative invention process that occurs outside the firm as well (Holmen et al., 2007). We build to the logic that firm-level decisions can either limit or enhance whether others are able to build on a given invention, which feeds into how much breadth-of-impact a given technology can muster. Using a sample of VC-backed biotechnology firms, we examine firm knowledge exploration choices along three dimensions—the decision to build from technologies across broad fields, the decision to explore application domains that are new to the firm, and the decision to mix these two options at the same time. Second, we contribute that knowledge exploration choices at the inventing firm level mix with aspects of the selection environment to yield differing degrees of breadth-of-impact for a given technology. The belief that broad sourcing diversity will always result in greater citation diversity requires some caveats.
The paper is structured as follows. First, we describe the prior work on selection environments for technologies and then describe what the process of entering a new application domain means for a technology in the selection environment. We follow with a discussion of how sourcing diverse pools of knowledge may draw broad attention to a technology, and then describe how these two main effects may interact. Thereafter, we present our methodology, our measures and our model specification. We review the findings of the empirical model, and then conclude with a discussion of the theoretical and practical implications of our findings. Specialization provides depth for the firm, but it may not lead to larger overall breadth-of-impact. At the same time, a willingness to step out of specialization may harm the ability of a firm to leverage past successes, but it may create an avenue by which others may build upon that knowledge. We find evidence of a breadth-of-impact frontier for technologies, wherein the choice of whether a firm should enter into a new application domain than those of the past should be informed by the degree to which the technology is citing prior work narrowly or broadly.
