1 Recombination Experience: A Study of Organizational Learning And Its Innovation Impact Anindya Ghosh, Univeristy of Pennsylvania Xavier Martin, Tilburg University Johannes M Pennings, University of Pennsylvania & Tilburg University and Filippo C Wezel, University of Lugano SMS 29th Annual Conference Submission February 9, 2009
2 Recombination Experience: A Study of Organizational Learning And Its Innovation Impact ABSTRACT This paper develops an evolutionary theory of the innovation process through the lens of organizational learning. We consider the firm s experience with recombining within and across disparate domains of technological knowledge and attempt to determine the impact of its innovative output as revealed by the acknowledgement of peer firms. We distinguish between recombinations taking place within versus across knowledge domains to describe how complex the search for new knowledge is. We also distinguish among specific and other types of experience, depending on whether past experience is directly related to the focal recombination. Using a longitudinal dataset of patents in the photographic imaging industry, we test the following: a) the stock of specific recombination experience has an inverted U-shaped relationship with the value of an innovation as measured by forward citations of its intellectual property; and b) recombining across knowledge domains has a larger positive effect on that value than recombining within a domain. Keywords: Recombination, Innovation, Knowledge, Imaging, Patents, Organizational Learning
3 INTRODUCTION The notion of innovation as knowledge recombination has been articulated for more than half a century (Henderson & Clark, 1990; Nelson & Winter, 1982; Schumpeter, 1942), and has recently spurred increased interest among scholars of technological change and growth (e.g. Fleming, 2001). In this context, our goal is to shed light on the role of learning on the firm s capability to produce innovations that exhibit a disproportionate impact. In so doing, we aim to develop and test theory about conditions under which organizational learning determines innovative performance. We aim to answer the following research question: Under what conditions does experience with knowledge recombination improve the value of a firm s innovation output? Specifically, we trace theeffects of organizational experience with recombination of technological knowledge in the photographic imaging sector. From the 1970s onwards this sector was marked by the convergence of silver halide, digital, electrical and optical technologies (Munir, 2003). Survival and performance during this time depended on firms capabilities to recombine distinct knowledge domains and produce new technologies to serve evolving productmarkets. This industry is thus highly appropriate to test our theory. THEORY AND HYPOTHESES Literature Review Early studies in the economics of innovation explored the correlation between R&D effort and innovative output (Griliches, 1984), whereas more recent efforts involved questions regarding spatial spillovers and knowledge flows among firms by means of citations (Jaffe, Trajtenberg, & Henderson, 1993). Organizational theorists exploited citations to infer the
4 boundaries of technological niches (Podolny & Stuart, 1995) and to investigate how a structural position in an industry network influences firm survival (Podolny, Stuart, & Hannan, 1996). Direct and indirect ties with incumbent firms as well as structural holes in a firm s network have also been explored as conditions which are conducive to innovative output (Ahuja, 2000; Burt, 1992, 2004); other studies dwelled on explorative search by the bricolage of knowledge dispersed among various actors (Garud & Karnoe, 2003). Knowledge integration more than knowledge itself has been claimed as the critical source of competitive advantage (Grant, 1996) and a critical source of novelty and success of innovations (Hargadon & Sutton, 1997; Hsu & Lim, 2008). Recent innovation research has exploited publicly available patent data as revealing innovative outcroppings of firms and sectors. A wide body of research is now available about diverse mechanisms of recombination, pointing to the intensity and to the rareness of the coupled knowledge (Fleming & Sorenson, 2001), to the usage of scientific evidence (Fleming & Sorenson, 2004; Sorenson & Fleming, 2004), and to scientists mobility across regions or firm alliances (Rosenkopf & Almeida, 2003) as key determinants of the focal innovation s impact and thereby of firms in their competitive domains (Rosenkopf & Nerkar, 2001). Most of the literature examines the link between the success of an innovation (e.g., the number of forward citations received) to either (1) the characteristics of the focal patent (e.g., number of technological classes, inter-personal ties among inventors) or (2) the characteristics of the focal firm (e.g., structural position in the industry, alliances) at the time that the innovation occurs. Few studies have explicitly dealt with the accumulation of recombination experience as a factor in successful recombination, and none to our knowledge addresses our specific research question.
5 Knowledge Recombination - Concepts We define Knowledge Recombination as the bundling of disparate domains of knowledge. We conceptualize knowledge to be comprised of broad domains with smaller sub-domains. For example, the broad domain of chemical knowledge is further subdivided into sub-domains of Coating Chemical knowledge and Resin Chemical knowledge. We further refine knowledge recombination in two categories. First, Specific Knowledge Recombination refers to the bundling of a given set of knowledge domains or sub-domains that are likely to be essential to both derive competitive advantage and survive in a given sector. Second, we define Other Knowledge Recombination as various combinations of knowledge domains or sub-domains other than the focal set. Further, we distinguish between across and within-domain recombination to identify the breadth of recombination of knowledge. Across-Domain Recombination refers to recombination across knowledge domains. An example would be a patent combining optical technologies and chemical technologies. Within-Domain Recombination refers to any combination of which all the components are within a single knowledge domain but that mix knowledge from two or more sub-domains within it. Knowledge Recombination Experience and Impact of Innovation Experience with Specific Recombination As outlined above, the literature argues that recombining knowledge is associated with more substantive innovation, but has largely ignored differences among innovating firms, as well as differences in whether the recombination spans multiple knowledge domains. From an organizational learning perspective, experience (Argote, 1999) and routine development (Nelson et al., 1982) are essential to successfully mastering a given knowledge recombination.
6 Further taking into account the uniqueness of the knowledge recombination, the distinction between within-domain and across-domain recombination is relevant to the extent to which learning makes a difference between firms. Within-domain recombination is comparatively simple and frequent, as it represents the elaboration of knowledge within a unitary domain. As such, it contributes to the deepening of existing knowledge and is akin to incremental exploitation of a single knowledge base (March, 1991), with relatively shallow learning and management processes that may lead to the steady displacement of exploration (Levinthal & March, 1993; Benner & Tushman, 2002). Thus, it is thus prone to diminishing returns: As opportunities within a single domain are repeatedly examined, a firm is likely to find that each subsequent innovation becomes less substantive. We expect a stronger learning effect, at least initially, in the case of across-domain recombinations. Two reasons justify this. First, across-domain recombinations are more complex as synthesize knowledge from multiple technologies. Learning benefits are more substantial as there are more components whose understanding as a set stands to improve with successive projects. Second, the greater complexity of across-domain recombinations prevents simplified routinization of learning and other management processes. Thus, experience accumulation with an across-domain recombination, at least initially, is less likely to displace exploration. The initial learning benefits are especially salient for specific knowledge recombinations, which entail learning to synthesize inputs from multiple but given sciences and technologies. However, as with within-domain recombinations, the firm cannot keep improving indefinitely. Conceptualizing recombination as a combinatorial search for innovation (Fleming, 2001; Fleming et al., 2001; Nelson et al., 1982), learning plateaus as fewer useful innovations remain in
7 the search space; diminishing and even negative returns emerge beyond some point due to opportunity exhaustion. Thus we predict: H1: Experience with a specific across-domain knowledge recombination is inversely U- shaped related to the impact of that specific recombinant innovation. Other Recombination Experience The possibility also exists of substantive spillover effects across innovation projects. Again, the distinction between within-domain and between-domain recombination is relevant. Learning may result from practicing a different combination of knowledge sub-domains within the same domain, albeit with the aforementioned threat that exploitation displaces exploration (Levinthal & March, 1993; Benner & Tushman, 2002) and with the additional risk that the relevance of one sub-domain to another may be overestimated. The fuller power of organizational learning should manifest itself in the case of acrossdomain recombinations. The knowledge being drawn upon may be more diverse, yet applicable to the extent that some domains are explored in common between successive projects. Furthermore, reaching across multiple domains, by itself, stands to develop a discipline of bridging across scientific fields and managing the complex demands of broad search projectsthat encourages exploration and overcomes learning myopia (Levinthal & March, 1993). Though more weakly applicable than the benefits of direct experience with a specific recombination, other across-domain recombination experience may also be less prone to opportunity exhaustion as the search space is potentially much bigger. Thus, though still subject to declining returns, it is an open question whether this type of search exhibits negative returns. Thus we hypothesize that: H2: Experience with other across-domain knowledge recombination is positively related with the impact of an innovation
8 Empirical Setting and Initial Results and Discussion We test our predictions in the photographic imaging sector from 1976-2002. This contains participants from multiple industries including many of its progenitors - the chemically based, silver halide photographic technologies with major players such as Kodak, Agfa, Polaroid and Fuji, for whom digital imaging technology represented a competence destroying discontinuity. By contrast some new entrants originated from the consumer electronics industry (e.g. Panasonic), and leveraged their experience with video. Another group originated from the graphic arts and printing industry and pioneered the use of electronic scanning. Finally, many entrants entered from computer hardware, software and semiconductor industries (e.g. Intel, Hewlett Packard, Adobe) as digital cameras began to be accepted as computer peripherals. Thus, photographic imaging today draws on technological competencies from the semiconductor and electronics industries, computer hardware and software industries, and conventional film based imaging industries. We use the 32,000 most significant imaging patents identified suggested by technology intelligence researchers at Kodak and select firms that have at least one patent per year on average over the window. We focus on recombinations of Chemical, Computer & Communication and Electrical & Electronics knowledge given their importance amidst the technological convergence between chemical and digital technologies. Following extant research, we measure innovation impact and value via forward citations in a multi-year window. Initial results show consistent support for H1: Experience with specific across-domain recombinations is associated with greater innovation impact up to a point, but with lower innovation impact subsequently. H2 receives mixed support, with some but not all across-domain recombinations being beneficial to innovations involving different domain recombinations. The results show several other interesting patterns, including additional effects of recombination
9 breadth, weaker effects (as expected) of within-domain recombinations, and total firm experience. Ongoing analyses promise further insights into conditions for innovative learning. We discuss the importance for innovation researchers of taking into account heterogeneous experience paths across firms. More broadly, given the importance of innovation to the development of competitive advantage (Henderson & Clark, 1990), this opens new avenues for strategy researchers to elucidate both sources of asymmetric competence in dynamic industries, and limitations of these asymmetries over time. We discuss keys to a systematic innovation program that balances search within and across a set of focal and partially overlapping recombinations. We also discuss how this research also initiates avenues to extend the debates on the strategic management of knowledge exploration and exploitation.
10 REFERENCES Ahuja, G. 2000. Collaboration Networks, Structural Holes, and Innovation: A Longitudinal Study. Administrative Science Quarterly, 45(3): 425-455. Argote, L. 1999. Micro underpinnings of organizational learning, Organizational Learning: Creating, Retaining, and Transferring Knowledge. Argyres, N. S., & Silverman, B. S. 2004. R&D, Organization Structure, and the Development of Corporate Technological Knowledge. Strategic Management Journal, 25(8/9): 929. Burt, R. S. 1992. Structural Holes. Cambridge, MA. Burt, R. S. 2004. Structural Holes and Good Ideas. American Journal of Sociology, 110(2): 349-399. Fleming, L. 2001. Recombinant Uncertainty in Technological Search. Management Science, 47(1): 117-132. Fleming, L., & Sorenson, O. 2001. Technology as a complex adaptive system: evidence from patent data. Research Policy, 30(7): 1019-1039. Fleming, L., & Sorenson, O. 2004. Science as a map in technological search. Strategic Management Journal, 25(89): 909-928. Garud, R., & Karnoe, P. 2003. Bricolage versus breakthrough: distributed and embedded agency in technology entrepreneurship. Research Policy, 32(2): 277-300. Grant, R. M. 1996. Toward a Knowledge-Based Theory of the Firm. Strategic Management Journal, 17: 109-122. Griliches, Z. 1984. R & D, Patents, and Productivity: University of Chicago Press. Hargadon, A., & Sutton, R. I. 1997. Technology Brokering and Innovation in a Product Development Firm. Administrative Science Quarterly, 42(4): 716-749. Henderson, R., & Clark, K. 1990. Architectural innovation. Administrative Science Quarterly, 35(1): 9-30. Hsu, D. H., & Lim, K. 2008. The Antecedents and Innovation Consequences of Organizational Knowledge Brokering Capability. Working Paper. Jaffe, A. B., Trajtenberg, M., & Henderson, R. 1993. Geographic Localization of Knowledge Spillovers as Evidenced by Patent Citations. The Quarterly Journal of Economics, 108(3): 577-598. Munir, K. A. 2003. Technological Evolution and the Construction of Dominant Designs in the Imaging Industry: McGill University Doctoral Dissertation. Nelson, R. R., & Winter, S. G. 1982. An Evolutionary Theory of Economic Change: Harvard University Press. Podolny, J. M., & Stuart, T. E. 1995. A Role-Based Ecology of Technological Change. The American Journal of Sociology, 100(5): 1224-1260. Podolny, J. M., Stuart, T. E., & Hannan, M. T. 1996. Networks, Knowledge, and Niches: Competition in the Worldwide Semiconductor Industry, 1984-1991. The American Journal of Sociology, 102(3): 659-689. Rosenkopf, L., & Almeida, P. 2003. Overcoming Local Search Through Alliances and Mobility. Management Science, 49(6): 751-766. Rosenkopf, L., & Nerkar, A. 2001. Beyond Local Search: Boundary-Spanning, Exploration, and Impact in the Optical Disk Industry. Strategic Management Journal, 22(4): 287-306. Schumpeter, J. A. 1942. Capitalism, Socialism and Democracy. New York: Harper. Singh, J. 2007. Asymmetry of Knowledge Spillovers between MNCs and Host Country Firms. Journal of International Business Studies, 38(5): 764-786. Sorenson, O., & Fleming, L. 2004. Science and the diffusion of knowledge. Research Policy, 33(10): 1615-1634.