Parent inheritance, founder expertise, and venture strategy: Determinants of new venture knowledge impact

JBV-05710; No of Pages 16 Journal of Business Venturing xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

Journal of Business Venturing

Parent inheritance, founder expertise, and venture strategy: Determinants of new venture knowledge impact☆ Sandip Basu a,1, Arvin Sahaym b,⁎,2, Michael D. Howard c,3, Warren Boeker d,4 a

Department of Management, Zicklin School of Business, Baruch College, City University of New York, One Bernard Baruch Way, New York, NY 10010, USA Department of Management, Information Systems, and Entrepreneurship, College of Business, Washington State University, Pullman, WA 99164, USA c Department of Management, Mays Business School, Texas A&M University, College Station, TX 77843.4221, USA d Department of Management and Organization, Michael G. Foster School of Business, University of Washington, Seattle, WA 98195, USA b

a r t i c l e

i n f o

Article history: Received 30 November 2012 Received in revised form 10 June 2014 Accepted 10 June 2014 Available online xxxx Field Editor: D. Jennings Keywords: New ventures Knowledge creation Knowledge recombination Organizational genealogy Biotechnology industry

a b s t r a c t A genealogical theory of new venture creation posits that “parent” firm routines are transferred to “progeny” ventures founded by the former employees of these parents. This study examines how the knowledge available to a venture from its parent firms and individual founders, as well as its initial technological direction, influences its own creation of impactful knowledge. We argue that new knowledge creation involves the recombination of underlying knowledge elements and hypothesize that the degree to which the venture's knowledge domain overlaps with the parents' knowledge has positive, but diminishing effects on the impact of knowledge created by the venture. We also predict that the breadth of founders' personal knowledge has a positive effect, but that the divergence between individual founders' and parent firm's knowledge domains has a negative effect on the creation of impactful knowledge by the venture. We test our predictions using a sample of 219 biotechnology ventures founded over the eleven year period 1990–2000 and tracked through 2010. Our results contribute to the entrepreneurship, knowledge creation, and genealogical literatures. © 2014 Elsevier Inc. All rights reserved.

1. Executive summary The ability of new ventures to create knowledge that fosters innovation and provides social and economic benefits has been the focus of scholarly attention in the entrepreneurship and strategy literatures for decades. New ventures founded by former employees of established firms (“parent firms”) benefit from founders' own technological expertise and their exposure to the technological knowledge of these parent firms. However, the process of knowledge transfer from parent firms to progeny new ventures has often been assumed by prior research to occur spontaneously and the role of founder choice has not been examined. Moreover, the influence of founder characteristics such as their own expertise and experience has not been examined in conjunction with the influence of inherited parent firm knowledge. To address these limitations, we examine how various elements of parent firm knowledge, venture technological strategy, and individual founder expertise influence the creation of impactful knowledge by the venture. Knowledge impact of the venture is measured as the extent to which the venture's knowledge is subsequently utilized by other firms. ☆ The authors would like to thank Suresh Kotha, Sonali Shah and Kevin Steensma for their insights on the earlier versions of this manuscript. ⁎ Corresponding author. E-mail addresses: [email protected] (S. Basu), [email protected] (A. Sahaym), [email protected] (M.D. Howard), [email protected] (W. Boeker). 1 Tel.: +1 646 312 3636. 2 Tel.: +1 509 335 6365. 3 Tel.: +1 503 914 7130. 4 Tel.: +1 206 543 8731.

http://dx.doi.org/10.1016/j.jbusvent.2014.06.002 0883-9026/© 2014 Elsevier Inc. All rights reserved.

Please cite this article as: Basu, S., et al., Parent inheritance, founder expertise, and venture strategy: Determinants of new venture knowledge impact, J. Bus. Venturing (2014), http://dx.doi.org/10.1016/j.jbusvent.2014.06.002

2

S. Basu et al. / Journal of Business Venturing xxx (2014) xxx–xxx

We argue that new knowledge creation involves the recombination of knowledge elements in new ways. Founders' exposure to the knowledge domains of their parent firms allows them to recognize new recombination opportunities and, in doing so, develop impactful knowledge. However, a venture's founders may choose to diverge from parents' existing technologies to combine novel and unfamiliar knowledge from sources outside the parent in hopes of generating high-impact knowledge. Such divergence from, and therefore a lower degree of overlap with, parent knowledge is beneficial in enabling ventures to search externally for valuable knowledge. However, extremely low levels of overlap with parent firm knowledge may inhibit founders' ability to recombine knowledge elements and dampen the creation of impactful knowledge. We then propose that ventures will be able to create more impactful knowledge if a founder's own technical expertise diverges from the knowledge domains of the parent. Founders' individual knowledge domains are often independent of the knowledge domains that their parent firms specialize in, especially if they played a relatively peripheral role in the respective parent firms' knowledge development. Divergence between the knowledge domains of founders and their parent firms indicates that the founder has a limited ability to recognize such opportunities, therefore negatively influencing the creation of impactful knowledge by the venture. Conversely, founders who have played a more central role in the parent's knowledge development will be in a better position to exploit recombination opportunities utilizing the parent's knowledge. Finally, founders with a greater breadth of personal technical expertise can better recognize recombination opportunities that create more impactful knowledge. We test our predictions using a sample of 219 biotechnology ventures founded over an eleven year period from 1990 to 2000 and followed until 2010. We find support for an inverted U-shaped relationship of venture divergence from parent knowledge on venture knowledge impact. Founder divergence from parent knowledge is also found to have a negative effect on venture knowledge impact as predicted. However, we do not find a significant effect of founder knowledge breadth on venture knowledge impact. Our study contributes to research on entrepreneurship and, in particular, the genealogical view of new venture creation by identifying parent firm and founder-level influences on knowledge transfer from parent firms to new ventures and also examining critical founder choices regarding venture technological strategy. This research not only highlights the value of modest divergence from parents' knowledge for successful knowledge creation in new ventures but also demonstrates the detrimental effect of diverging too far from the knowledge base of parents. We also contribute to the literature on organizational search by applying the concepts of local search and distant search in a unique way to new ventures. 2. Introduction The creation of high-quality, impactful knowledge by new ventures has been a key focus for scholars in the fields of entrepreneurship and strategy (Acs et al., 2009; Ahuja and Lampert, 2001; Henderson and Cockburn, 1996). Ventures that are successful in the development of more significant knowledge and innovation have been shown to be superior in developing new products and markets, overcoming imitators, and successfully adjusting to significant changes in their competitive environment to exploit new opportunities (Leiponen and Helfat, 2009; Shane, 2003).5 There has also been an increasing recognition of the important spillover effects and social benefits arising from the creation of new knowledge by such entrepreneurial ventures (Davidsson et al., 2006; Gilbert, McDougall, and Audretsch, 2008; Shane, 2003). Scholars have called for research on how new ventures' knowledge development can lead to a virtuous cycle of subsequent innovation, encouraging the creation of new technological opportunities and economic growth (Acs et al., 2009; Cliff et al., 2006; Davidsson et al., 2006; Fleming and Sorenson, 2001; Hand, 2007; Yayavaram and Ahuja, 2008). Theoretical and empirical work focusing on the creation of new knowledge has often focused on knowledge impact as a key outcome, with impact judged by the extent to which a firm's knowledge is utilized by others (Kotha et al., 2011; Miller et al., 2007; Trajtenberg, 1990). Prior research has argued that the forward citations of a firm's patents by other firms represent a consistent means of measuring the impact of knowledge created by the focal firm and are one of the strongest indicators of its overall innovative performance (Gilbert et al., 2008; Hoetker and Agarwal, 2007; Miller et al., 2007; Phene et al., 2006). Earlier studies have also considered the possibility that the founding and origins of new ventures are likely to influence their ability to generate high-impact knowledge (Helfat and Lieberman, 2002; Klepper, 2001). For example, ventures in many industries emerge through ‘spawning’ from established firms (Agarwal et al., 2004; Chatterji, 2009; Klepper, 2001),6 that is, are begun by founders coming from established incumbents. Prior experience at an incumbent firm (also called a ‘parent’ firm) often provides both know-how and opportunity for nascent entrepreneurs to begin their (progeny) new ventures (Phillips, 2002). Employees leaving a parent firm to found a new venture take with them routines, capabilities, and technologies developed by the parent that they can then put in place in their venture (Klepper and Thompson, 2006). In this manner, progeny ventures inherit technological experience and expertise from their parents, and such inheritance can play a critical role in their development of new technologies and innovations (Agarwal et al., 2004; Chatterji, 2009; Klepper, 2001). Despite evidence linking the inheritance of knowledge from parent firms to the success of progeny ventures, there are important gaps in our understanding of these inheritance processes. First, prior research has largely assumed that knowledge transfers from parents to progeny occur almost inevitably, with progeny firms naturally following the practices of their parents (Agarwal et al., 2004; Klepper and Sleeper, 2005). There has been little recognition that progeny founders make deliberate decisions about the extent to which to utilize capabilities acquired at their parent firms. Moreover, there is limited examination of whether such critical founding choices influence important venture outcomes, such as those related to subsequent knowledge creation. Studies have not examined 5 In this study “venture” refers to the entire team of employees working in the firm (not just the founding team), any of whom could be involved in knowledge development along the direction established by the founders. 6 In the context of this study, ventures are founded by former employees of established firms (i.e., parent firms) through spawning and benefit from founders’ personal expertise and accumulated knowledge; however, ventures do not have any formal relationship such as alliance, equity, or board membership with the parent firms (e.g., Agarwal et al., 2004; Chatterji, 2009; Klepper, 2001).

Please cite this article as: Basu, S., et al., Parent inheritance, founder expertise, and venture strategy: Determinants of new venture knowledge impact, J. Bus. Venturing (2014), http://dx.doi.org/10.1016/j.jbusvent.2014.06.002

S. Basu et al. / Journal of Business Venturing xxx (2014) xxx–xxx

3

the potential limitations of relying too heavily on the knowledge base of the parents or whether divergence from the practice of parent firms may prove valuable to the venture. Second, prior research in the genealogical tradition does not account for the technological experience of the individual founder in relation to that of the parent firm and how the founder's knowledge and expertise, independent from that of the parent firm, can influence the creation of impactful knowledge in the new venture. Past work has generally assumed that an understanding of firm- and environment-level characteristics are adequate to explain important differences in venture performance such as knowledge creation by new ventures while downplaying the role of individuals (Hmieleski and Baron, 2009; Low and MacMillan, 1988). Although considerable work in the entrepreneurship literature argues that founder capabilities influence the nature and quality of opportunity recognition (Helfat and Lieberman, 2002; Shane, 2000; Shane and Venkataraman, 2000), these insights have not been integrated into the literature on knowledge inheritances from parent to progeny firms. Our research question examines how the technological knowledge available to a venture from its parent firms and its individual founders influences both its initial technological direction and the impact of the knowledge it creates. We evaluate the knowledge impact of a venture as the extent to which other firms draw on the venture's technological knowledge in the future. Our research helps resolve critical gaps in our current understanding of this process, demonstrating limits in the extent to which inherited knowledge can be usefully utilized by a new venture beyond which the value of inheritance diminishes and may even become dysfunctional. We conceptualize knowledge creation as a recombination of underlying knowledge elements (Schumpeter, 1934; Sorenson and Fleming, 2004; Yayavaram and Ahuja, 2008), arguing that the knowledge domains of parent firms and founders represent recombination opportunities that can be utilized by the new venture. We hypothesize that the decision by venture founders to diverge from their parents' knowledge domains creates opportunities to combine more new and unfamiliar knowledge elements, with the potential to generate higher-impact knowledge. However, we propose that such divergence can have a diminishing and even declining effect on the impact of the venture's knowledge. We then focus on the nature of a founder's experience in the parent firm and hypothesize that a more peripheral role played by the founder in the parent's knowledge development will have a negative effect on venture knowledge creation. Finally, we argue that in addition to parent firm knowledge, greater breadth in a founder's technological knowledge is likely to result in a venture founder finding novel ways in which knowledge elements can be combined within these domains. This study conceptualizes founders as being sources of knowledge in their own right, rather than as a conduit for the transfer of parent knowledge, and is the first to simultaneously examine the role of founder's expertise, parent firm knowledge and venture's strategic choices as having important implications for the impact of knowledge developed by new ventures. We investigate these issues through an examination of a sample of biotechnology ventures initially founded from 1990–2000 and tracked for five years after founding. Our research considers whether excessive reliance on parent or founder capabilities, while creating the potential to enhance knowledge transfers from parents firms, might also prove dysfunctional by making ventures myopic and disinclined to enter new technological areas. Our results confirm a negative effect of founder–parent technological divergence on the impact of ventures' knowledge and a diminishing marginal effect of venture–parent technological divergence. Finally, we discuss our contributions to the literatures on entrepreneurship, knowledge creation, and genealogy.

3. Theory development 3.1. The knowledge inheritance process Past studies examining the relationship between parent firms and new ventures have demonstrated that organizational routines and capabilities may be transferred from parents to progeny in a manner similar to the transmission of biological genes (Phillips, 2002; Winter, 1991). Such knowledge transfers occur as founders of new ventures bring with them expertise and experience accumulated through their past employment (Agarwal et al., 2004; Chatterji, 2009; Klepper and Sleeper, 2005). For example, Phillips (2002) illustrates how founders of new law firms in Silicon Valley employed the same routines as the law firms they had come from, and Agarwal et al. (2004) demonstrate that both technological and marketing know-how can be passed from parent to progeny. Klepper and Sleeper (2005) show how founders gained access to technical information through their past experience working at a parent firm and further exploit that experience in their own new ventures in the laser industry. Gompers et al. (2005) similarly show how start-ups use knowledge and practices developed in the parent firm to exploit entrepreneurial opportunities. Chatterji (2009) finds that in addition to technological expertise, progeny ventures also inherit non-technical knowledge related to marketing and regulatory strategy from their parents. Such knowledge inheritances from parent firms typically benefit progeny ventures in their subsequent knowledge creation initiatives (Agarwal et al., 2004) and thereby enhance their overall market value (Chatterji, 2009). While providing evidence that inheritance at birth is beneficial to a progeny venture, extant literature often overlooks the critical role that founders play in such an inheritance process and how their motivation and ability influence the extent to which the venture can exploit valuable parent knowledge. To examine our research question and address this limitation, we draw on a considerable body of literature that conceptualizes knowledge creation as an outcome of the combination and synthesis of underlying knowledge elements (Henderson and Clark, 1990; Schumpeter, 1934; Sorenson and Fleming, 2004). For example, the automobile was conceived by combining knowledge elements from bicycles, horse carriages and internal combustion engines (Sorenson and Fleming, 2004). Similarly, new developments in fields such as microprocessors and pharmaceuticals typically emerge through the combination of elements from diverse knowledge areas (Henderson and Clark, 1990; Henderson and Cockburn, 1996). In this manner, technological knowledge is often produced by either a novel recombination of existing knowledge elements or a synthesis of some existing and some new knowledge elements (Cliff et al., 2006; Sorenson et al., 2006). Please cite this article as: Basu, S., et al., Parent inheritance, founder expertise, and venture strategy: Determinants of new venture knowledge impact, J. Bus. Venturing (2014), http://dx.doi.org/10.1016/j.jbusvent.2014.06.002

4

S. Basu et al. / Journal of Business Venturing xxx (2014) xxx–xxx

Such a conceptualization of knowledge creation extends the genealogical view of knowledge inheritance by further clarifying the role that individuals, given their particular knowledge and background, are able to play in recognizing valuable entrepreneurial opportunities. While greater expertise in a knowledge area often reveals opportunities related to that area, a potentially more influential source of entrepreneurial opportunities lies in the ability of organizational members to bridge multiple knowledge areas that may be technologically distinct and unconnected (Hargadon and Sutton, 1997; Shane, 2000). New entrepreneurs often discover opportunities to recombine existing knowledge elements from both their parent firms and their own experiences in unique ways in addition to combining knowledge derived from parent experiences with that of other founders coming from different parent firms and of other technical employees subsequently hired (Baron, 2006; Vaghely and Julien, 2010). The quality of the entrepreneurial opportunity originally conceived by individual founders based on their experience is likely to determine the impact of technological knowledge created by the new venture (Chatterji, 2009; Shane, 2000). In the venture's early life, founders are focused on executing plans regarding technological development formulated at the pre-founding stage through their own efforts as well as their guidance of other employees (Baron, 2006; Gruber et al., 2008; Vaghely and Julien, 2010). The impact of knowledge created by a progeny venture is thus strongly dependent on the founders' recognition of opportunities to exploit, recombine, and synthesize knowledge elements, while still working in their parent firms. Therefore, we examine the role of individuals that found the venture, namely, their expertise, experience, and early decisions, in influencing the venture's knowledge creation. 3.2. Venture technological strategy and knowledge creation An important strategic consideration for founders of progeny ventures is the extent to which they would want the ventures to work in similar knowledge domains as their parent firms. For a new venture, borrowing from parent firm knowledge elements is the equivalent of “local search,” that is, search conducted in the neighborhood of its existing knowledge domains (Rosenkopf and Almeida, 2003). By exploiting familiar knowledge domains, local search allows firms to avoid large upfront development costs, instead building on routines that have worked well in the past (Winter et al., 2007). However, new ventures do not yet possess a rich knowledge base accumulated through years of operating experience which they can easily exploit. The available alternative for founders is to exploit parent firm knowledge elements they are personally familiar with (Chatterji, 2009; Stuart and Podolny, 1996). Genealogical theory also posits that an important source of founders' existing expertise is their exposure to the knowledge domains of parent firms. While in their previous employment, founders are likely to become aware of the market opportunities and recombination potential of other parent technologies, even when not working with these directly. Founders who choose to maintain a high degree of overlap with the technological domains of their parent firms increase the number of elements that they can subsequently draw upon. Therefore, founders forging a greater overlap with their parent knowledge domains can combine and synthesize this knowledge in more varied and distinct ways (Fleming and Sorenson, 2001) through engaging in “knowledge brokering” — by applying knowledge elements from one domain to another entirely different, and often sparsely connected, domain (Hargadon and Sutton, 1997). As a venture grows over time, additional employees are typically hired (Gilbert et al., 2006) who may then extend the initial technological direction of the venture as conceived by its founders. While past work focusing on inheritance by new ventures has emphasized the benefits of relatedness between the parent firm and the new venture in exploiting such an inheritance (Agarwal et al., 2004), the knowledge recombination perspective suggests benefits to a venture of diverging from its parent technological knowledge (Fleming, 2007). A venture's divergence from the technological direction of its founders' parent firms suggests that the founders' initial intent was to engage in more “distant search” in contrast to local search characterized by borrowing more heavily from parent technologies (Beckman, 2006; Phene et al., 2010). Such technological divergence can help overcome “learning traps” that may arise from a more limited focus on local search and which may inhibit the development of more expansive knowledge creation (Ahuja and Lampert, 2001; Levinthal and March, 1993; Miller et al., 2007; Miner et al., 2001). Past literature and empirical research has highlighted how searching for and accessing heterogeneous resources provides greater opportunities for the creation of path-breaking innovations and new knowledge (Ahuja and Katila, 2004; Rosenkopf and Nerkar, 2001). In our study, by combining familiar parent knowledge with more distant and unfamiliar knowledge, founders can promote the creation of path-breaking and high-impact technological knowledge (Tushman and O’Reilly, 1996). Thus, some degree of divergence from the parents' technological knowledge domains is often desirable. However, as founders rely more heavily on borrowing or developing unfamiliar and distant knowledge relative to exploiting familiar parent knowledge, the opportunities for generating breakthrough knowledge may diminish. An imbalance tilted more heavily to distant search may create fewer opportunities to leverage any inherited capabilities or expertise (He and Wong, 2004; Levinthal and March, 1993). As a result, an increasing focus on technologies in which founders have no previous exposure restricts the potential value they can add to the new venture. Too low a degree of technological overlap with parent firms may preclude a level of understanding that founders need of their targeted opportunities, making it more difficult for innovative knowledge to be produced. Hypothesis 1. The degree of overlap of a new venture with its parents' technological knowledge domains after founding has an inverted-U effect on the impact of knowledge created by the venture.

3.3. Founder role in the parent firm and knowledge creation Going beyond the assumptions of the genealogical literature that founders inevitably leverage parent–progeny inheritances, we examine founders' motivations and abilities to use dominant and valuable parent technologies, which in turn influence the impact Please cite this article as: Basu, S., et al., Parent inheritance, founder expertise, and venture strategy: Determinants of new venture knowledge impact, J. Bus. Venturing (2014), http://dx.doi.org/10.1016/j.jbusvent.2014.06.002

S. Basu et al. / Journal of Business Venturing xxx (2014) xxx–xxx

5

of knowledge created by their new ventures. We recognize that the nature of a founder's experience in the parent firm is likely to influence the extent to which he or she would like to exploit valuable parent firm technologies. In particular, the founder's technological role in the parent will influence the particular venture's creation of impactful knowledge, independent of the individual founder's expertise. An inventor whose technological expertise diverges significantly from that of her parent firm may be less centrally involved in the technological development of the parent and may be relatively more isolated from continued technological developments that may offer a source of new opportunities.7 Such an inventor, with a technical specialization outside the central knowledge domains of the parent, may be less integrated with these technological developments and less able to recognize or fully understand recombination opportunities. It is also plausible that such an inventor's ideas were not well-supported at the parent firm, causing a period of frustration finally culminating in exit (Klepper, 2001). If an inventor from the technological periphery of the parent firm starts a new venture, she is less likely to have the motivation or the ability to transfer the expertise of the parent firm to the venture than if she had been centrally involved in the core technologies of the parent. Thus, in the case of the founder who is not as involved in the central technologies of the parent, there may be fewer opportunities to exploit parent firm knowledge that is of potential value to the new venture. Alternatively, even if such a founder seeks to forge a recombination with certain parent firm knowledge domains, the quality of the recombinatory effort is likely to be poor. In effect, this situation is an exception to our general framework discussed earlier, in which parent firm technological domains represent opportunities to engage in local search. However, parent domains that a venture founder is unfamiliar with are almost as distant to him or her as searching outside of parent firm domains. In contrast, inventors whose own technical expertise closely matches the technologies of the parent firm may be in a better position to exploit these key parent technologies and maximize the advantages of local search. Such inventors are likely to have more knowledgeable insights relating to the potential and future of these technologies and may be in a better position to influence the development of the technologies by the venture (Beckman and Burton, 2008; McGrath et al., 1996). Hypothesis 2. The degree of divergence of founders' knowledge domains from parent firm's knowledge domains at the time of venture founding has a negative effect on the impact of knowledge created by the venture. 3.4. Founder technological expertise and venture knowledge creation From a genealogical perspective, the most important role of founders is to utilize parent knowledge and technological expertise in their new ventures. However, this approach ignores differences in the background and technological expertise of individual founders from the knowledge base of the parent firm. Founders may be more than just conduits for transferring parent firm knowledge and, through their individual expertise, may independently influence knowledge creation by the new venture. Founders are likely to have personal experience and expertise in certain technological and knowledge areas where they have been actively involved in developing the current state of knowledge (Shane, 2000; Vaghely and Julien, 2010). Although the domains of founder technical expertise may overlap with the technological domains of the parent firm, the founder's own expertise represents a separate and potentially unique source of past experience from that of the parent. As noted earlier, the founder may not have worked extensively in many of the technological domains of the parent firm, or may have worked in non-overlapping technological areas prior to taking up employment at the parent firm. Therefore, a founder's personal expertise may be an independent source of new recombination opportunities that can be exploited in the startup venture. In effect, founders engage in local search when they draw on individual knowledge expertise, independent of their transfers of parent firm knowledge. While the effects of individual founder expertise on venture strategy and performance have been recognized by the broader entrepreneurship literature (Hmieleski and Baron, 2009; Shane, 2000), these insights have not been integrated into the parent inheritance literature. For example, past entrepreneurship literature has emphasized how a founder's individual capabilities can influence not only the quality of the initial entrepreneurial opportunity but also subsequent knowledge creation initiatives (Shane and Venkatraman, 2001). Founders with varying pre-founding expertise and capabilities often recognize significantly different commercial applications for the same technology (Shane, 2000). For example, Cliff et al. (2006) find that knowledge breadth of founders, particularly in peripheral knowledge corridors, motivates innovations by enhancing recombinatory capabilities. As a result, founders with a greater breadth of personal knowledge can combine and synthesize their own set of knowledge elements in more varied and distinct ways (Fleming and Sorenson, 2001; Hargadon and Sutton, 1997). They can recognize novel linkages among these knowledge domains which can be exploited to pursue a technological trajectory that is path-breaking and innovative for a new venture (Cliff et al., 2006). In contrast, for founders with very narrow technological expertise or breadth the opportunities available for recombining knowledge may be somewhat limited. Hypothesis 3. The breadth of founders' own technological expertise has a positive effect on the impact of knowledge created by new ventures in addition to their parent firms' technological expertise.

7 Conceptually, the founder’s degree of divergence from parent knowledge is distinct from the venture’s degree of overlap with parent knowledge. An individual founder’s degree of divergence is considered to the point when s/he leaves the parent firm prior to founding of the venture and represents the role s/he played at the parent. The venture’s degree of divergence is considered post-founding and represents the technological strategy of the venture as decided collaboratively by all the founders. Given these conceptual differences, the predicted effects of these two variables on venture knowledge impact are also dissimilar. We do not predict an inverted-U effect for founder–parent knowledge divergence (H2) as we had for venture–parent knowledge overlap (H1).

Please cite this article as: Basu, S., et al., Parent inheritance, founder expertise, and venture strategy: Determinants of new venture knowledge impact, J. Bus. Venturing (2014), http://dx.doi.org/10.1016/j.jbusvent.2014.06.002

6

S. Basu et al. / Journal of Business Venturing xxx (2014) xxx–xxx

4. Methods 4.1. Sample and data sources Our study focuses on new ventures in the U.S. biotechnology industry. The biotech industry is still experiencing rapid growth and has historically been characterized by a large number of start-ups (Rothaermel and Deeds, 2004). To construct our sample, we used the Dow Jones VentureSource proprietary database, which captures founding team membership and other information for all private equity supported entrepreneurial firms. Additionally, we used VentureXpert, NBER and Delphion databases to collect many of our measures, which have also been used extensively by others in past academic research on new ventures (Chatterji, 2009; Dushnitsky and Lenox, 2006; Gompers and Lerner, 1998). We examined all biotech ventures listed in the VentureSource database that were founded in the United States during the eleven-year period 1990–2000. This period represents a period of relatively stable growth for the biotechnology industry, with an increasing volume of start-ups and VC investments (Bergeron and Chan, 2004), therefore somewhat mitigating unique period effects. Moreover, this observation period allowed us adequate time to observe the impact of the technological knowledge that the ventures subsequently created, since such knowledge not only takes time to develop but also to diffuse and influence the technological development of other firms. Because our dependent variable focuses on the impact of knowledge created by the new venture, we examined the impact of patents in our sample for five years after the filing date of the patent, with our data collection ending in 2010. To track parent–progeny linkages for our sample of ventures, we relied on the patenting record of founders. For each founding team member listed in the VentureSource data, we tracked their history of innovation, observing all patents filed with the U.S. Patent and Trademark Office on which the focal founder is listed as an inventor prior to the founding date of the new firm. We utilized the disambiguated list of U.S. inventors developed through the Harvard University patent project (Lai et al., 2009) to ensure that prior patents were appropriately matched to the unique individuals listed in our founder data. Through this method, we were able to establish a systematic record of the technology affiliation of new venture founders to their previous organizations. Consequently, the “parent” organization of a given founder is considered to be the assignee listed on the founder's most recent patent prior to the start date of the new venture. The advantage of this approach to defining a genealogical relationship is that we have clear evidence that the founder was engaged in R&D and technology development activities at the parent firm, thus gaining direct exposure to the parent's knowledge and practices. Following prior research on parent–progeny relationships (Chatterji, 2009; Klepper, 2001) that considers parent firms as incumbent firms where the concerned founders were previously employed, we excluded ventures consisting only of founders coming directly from academia and not-for-profit organizations. We also excluded ventures that had some formal investment or governance relationship with the parent organization since it was not clear whether such ventures were strictly independent spawns of their parent firms. Using these criteria, our final sample consists of 219 ventures with 304 parent–progeny links. 4.2. Dependent variable The dependent variable in our study is venture knowledge impact. Following past research (Trajtenberg, 1990; Yayavaram and Ahuja, 2008), knowledge impact is measured as the total number of external forward citations of a new venture's patents filed for the five year period after its founding.8 The number of external forward citations for a given patent is the total number of subsequent patents that cite this patent that are not assigned to the same venture. Citations by other firms is a common indicator of knowledge impact and perceived utility of the knowledge generated from the perspective of knowledgeable, objective observers (Hoetker and Agarwal, 2007; Miller et al., 2007; Phene et al., 2006; Wry et al., 2010; Yayavaram and Ahuja, 2008) and is often correlated with economic value captured by the inventing firm (Harhoff et al., 1999; Narin et al., 1984; Trajtenberg, 1990). We observed forward citations of all examined patents for a further five years after granting, thereby allowing an adequate window of time for the knowledge embodied in a patent to be disseminated into the future technological development of other firms.9 4.3. Independent variables 4.3.1. Venture technology overlap with parent We compare the knowledge base of the parent organization to that of the entrepreneurial venture to determine the degree of overlap between the two firms. Using the method applied by Ahuja and Katila (2001), we collected all patents and direct backward citations for each firm. We then removed repeat entries from this list. The remaining list of patents is said to comprise the knowledge base of the firm (Ahuja and Katila, 2001). In the case of the parent organization, we included patents with application dates in the five-year period prior to the founding of the new venture. For the focal new venture, we observed patents in the five-year period following founding. To calculate the overlap of the progeny firm with its parent, we counted the total number of common patents that appear in the knowledge base of both firms and divided by the absolute size of the new venture knowledge base. The resulting measure ranges from 0 to 1, with higher values indicating greater technology overlap between the two firms. In the case of firms with multiple parents, we took the average values of the overlap measure.10 Note that we separated founder-invented parent firm patents from the 8

Consistent with our definition of venture, we consider patents assigned to all employees and not just the founders. As an example, for a patent received in 2004 by a venture founded in 1999, we observed forward citations until 2009. 10 Averaging the values of overlap with multiple parents is appropriate since it best represents the theoretical construct of how close a venture is technologically to any of its parents. To ensure the robustness of our results, we performed tests of the subsample of firms that had only one parent; the results of this test were consistent with our baseline findings. 9

Please cite this article as: Basu, S., et al., Parent inheritance, founder expertise, and venture strategy: Determinants of new venture knowledge impact, J. Bus. Venturing (2014), http://dx.doi.org/10.1016/j.jbusvent.2014.06.002

S. Basu et al. / Journal of Business Venturing xxx (2014) xxx–xxx

7

broader set of parent patents. This was done to avoid any possible confounding effects of including founders' patents in the overlap calculation. 4.3.2. Founder technology divergence from parent Following prior research (Colombo, 2003; Rosenkopf and Almeida, 2003), we use patent category information captured through International Patent Classification (IPC) codes to measure this variable. We recorded the IPC code (up to subclass level) for each patent listing a founder as an inventor in the five years prior to founding11 and each patent filed by the corresponding parent in the same period. The degree of divergence from parents was calculated as:   2 1=2 ∑ j ∑i founder patent proportioni – parent j patent proportioni ; where i represents a distinct IPC code and j represents a particular parent firm. The results are standardized to provide a continuous scale from 0 to 1 and averaged across multiple parents, with a value closer to zero reflecting low divergence between the founders and their parents. 4.3.3. Founder technology breadth Past research has used the number of IPC codes as a measure of the breadth of knowledge for a given patent (Lerner, 1994), an inventor, and for an entire organization (Nesta, 2008). We apply this measure to capture the knowledge breadth of patent activity of each founder at the time of her leaving the parent organization. Our measure records the total number of different four-digit primary IPC categories in which the founder is listed as an inventor on filed patents in the five years prior to the founding of the progeny venture. This measure recognizes that founders often have inventing history prior to joining the parent of their venture and focuses on the personal knowledge of the founders, which may be different from the knowledge breadth of the parent firm and that of the venture. To further highlight their conceptual distinctness, a summary of these variables and their timeframe of measurement is shown in Fig. 1. 4.4. Control variables We control for a number of factors that may affect the impact of knowledge developed by a progeny venture. Several characteristics of the parent organizations and their relationships to the new ventures may lead to differences in new venture innovation and the impact of newly developed technologies. If the innovations of the parent firm tend to be more recent, they are likely to represent more recent knowledge that is of greater use to the founders of the new venture. Age of Parent Patents captures the average time that has elapsed since the filing date of patents held by the parent firm. This was measured by observing the application date of each patent in the parent knowledge base, recording the elapsed time between this date and the founding date of the new venture, and calculating the average total age across all parent firm patents. Parent organizations with larger knowledge pools may provide greater opportunities for founders to benefit from exposure to successful patenting in a greater number of technologies. We include two measures to address the extent of the parent firm knowledge pool: Parent Number of Patents is calculated as the average number of successful patent applications by parent organizations in the five-year period prior to the founding of the focal new venture and Size of Parent Knowledge Base is a count of the number of unique backward patent citations referenced in parent firm patent documents (Ahuja and Katila, 2001). With our focus on the impact of knowledge developed by the new venture, it is plausible that the impact of knowledge held by parent firms may have an effect beyond our hypothesized relationships. Thus, we control for Parent Knowledge Impact, measured for each parent firm in the same fashion as described for our dependent variable, in this case, based on patents held by the parent firms. In situations involving new ventures with multiple parent firms, we record the average knowledge impact across parents. We also controlled for Parent Technology Diversity since higher parent diversity might offer more recombination opportunities to venture founders. The technology diversity of the parent firm is calculated as: 1 − ∑(ni/N)2. The cumulative number of patents belonging to IPC class i is represented by ni, while N represents the total number of patents held by the firm (Vasudeva and Anand, 2011). Closer physical proximity to parent organizations may influence knowledge impact by providing founding team members with better access to professional and social networks of contacts within and adjacent to the parent firms (Liebeskind et al., 1996; Saxenian, 1994; Sorenson et al., 2006; Tsai, 2001). To address this issue, we include the control variable, Geographic Proximity to Parents, calculated as the inverse natural logarithm of the average physical distance between postal zip codes of the new venture and each of its parent firms. Various aspects of founder human capital may play a role in determining the impact of innovations developed by the entrepreneurial venture. Founder Number of Patents captures the number of successful patent applications listing the founder as inventor in the five-year period prior to founding. Founders with a greater pool of inventions may have more direct access to technologies, diminishing the role of parent knowledge in shaping the direction of the new firm. Since the extent of the founder's prior experience may influence the impact of knowledge developed in the venture, we control for Founder Length of Experience, measured as the number of years between her first successful patent application and the founding date of the new venture. This measure is a rough but reasonable indicator of a founder's total time period as an active inventor in the industry. If the founder worked for an extended period 11 Our measure of founder divergence reflects its distinctness from venture overlap with parent firm knowledge. Founder divergence considers individual founder inventions prior to founding the venture whereas venture overlap considers all inventions emerging out of the venture after its founding.

Please cite this article as: Basu, S., et al., Parent inheritance, founder expertise, and venture strategy: Determinants of new venture knowledge impact, J. Bus. Venturing (2014), http://dx.doi.org/10.1016/j.jbusvent.2014.06.002

8

S. Basu et al. / Journal of Business Venturing xxx (2014) xxx–xxx

Variable Dependent Variable Knowledge Impact

Measurement

Illustration

The total number of external forward citations (i.e. excluding self-citation) of a new venture’s successfully filed patents for the five year period after its founding.

For a venture founded in the year 2000, we record all successful patent applications through 2005. Next, forward citations of these patents are collected for 5 years hence.Thus, for patents recorded in 2005, we include forward citation data through 2010. Similarly, for the patents received in 2004, we included forward citation data through 2009 and so on.

The proportion of a venture’s total patents for a five year period after founding that exploit the knowledge base of any parent patent.

For a venture founded in the year 2000, we again record all successful patent applications through 2005. Then, the backward citations of these patents were checked to determine if any patent either directly cited a parent firm patent or another patent also cited by a parent firm patent.

Founder technology divergence from parent.

The Euclidean distance between the patent technology class vectors of patents listing founder as inventorand parent firm patents (excluding those assigned to founder) in a fiveyear period prior to venture founding.

For a venture founded in the year 2000, patents successfully filed by parent firm and patents listing founder as inventor between 1995-2000, are included in this calculation. The proportion of each distinct primary IPC code to total patents is computed for parent and founder. The difference in these proportions for each IPC code between founder and parent is squared and summed.

Founder technology breadth.

Total number of distinct fourdigit primary IPC categories in which the venture founder is listed as an inventor on filed patents.

For a venture founded in year 2000, all patents filed between 1995 and 2000 in which the founder is listed as an inventor are included in this calculation. The distinct primary UPC codes are counted for all founders coming from parent firms.

Independent Variables Venture technology overlap with parent

Fig. 1. Description of key variables.

at a different firm in the industry before joining the parent organization, it may also shape her approach toward building on parent technologies in the new venture. Founder Tenure at Firm Prior to Parent captures this effect and is measured as the number of years between her first patent at the prior firm and the first patent at the parent organization. Since a larger founding team may result in greater knowledge inflows and human capital available to a venture (Colombo and Grilli, 2005), we include the Number of Founders as a control variable. Venture Number of Patents controls of the sheer volume of innovation activity in the new firm, a factor that may also influence the impact of knowledge it creates. Greater access to financial resources is likely to influence the impact of knowledge that is produced by a venture. We therefore include Rounds of VC Funding and Venture Resources, the total rounds and amount of venture capital funding received in the first five years after founding, respectively, to capture variance in the financial support available to the firms in the sample. Prior research shows that academic founders may provide unique, distinctive resources to an entrepreneurial venture (Stuart and Ding, 2006). We control for this factor by including the Proportion of Academic Founders in the new venture. Data on prior academic affiliations was captured through the patent record; founders who were listed as inventors on patents assigned to academic institutions were coded academic founders. Once again, we draw from the disambiguated list of U.S. inventors developed through the Harvard University patent project (Lai et al., 2009) in collecting this data. The ability of founders to build on parent firm technologies may be inhibited by legal considerations. To address this issue, we include a binary variable, Non-Compete State, recording whether the genealogical venture is founded in a state that has or has not historically enforced non-compete agreements with prior employers (Marx et al., 2007). Finally, to control for systematic period and industry effects on our DVs, we include year dummies to indicate when the focal venture was founded and industry segment dummies for the four categories of the biotech industry classified by Dow Jones VentureSource: drug delivery, drug development, biotech therapeutics, and pharmaceuticals. Descriptive statistics and correlations of all variables are shown in Table 1. Please cite this article as: Basu, S., et al., Parent inheritance, founder expertise, and venture strategy: Determinants of new venture knowledge impact, J. Bus. Venturing (2014), http://dx.doi.org/10.1016/j.jbusvent.2014.06.002

S. Basu et al. / Journal of Business Venturing xxx (2014) xxx–xxx

9

4.5. Model specification and estimation The dependent variable, Venture Knowledge Impact, takes on non-negative integer values with a distribution that is skewed toward zero, since many of the inventions from newly established technology firms may have a relatively low impact on the industry. A negative binomial regression model is generally appropriate for examining statistical relationships with dependent variables exhibiting overdispersion to zero (Hilbe, 2007). To ensure that further steps were not required to account for overdispersion of the data, we also tested zero-inflated negative binomial models. The Vuong test was not significant in these models, thus demonstrating that the negative binomial model is the appropriate specification for our study. Additionally, we use robust standard errors in all hypothesis tests to address potential issues of heteroskedasticity. 5. Results The results of the tests of study hypotheses are show in Table 2. Model 1 in Table 2 is a baseline model that includes all control variables. Parent technology diversity has a strong, positive effect, indicating that broader technology domains pursued by parents are associated with greater venture knowledge impact. The coefficient for age of parent patents is negative and significant, suggesting that older technologies of the parent firm diminish the experiential benefits obtained by founders. At the same time, geographic proximity to the parent firms has a significant, positive influence. The effects of volume of innovations are also shown to be significant in the controls model — new ventures with more patents also tend to have patents of greater impact. Model 2 of Table 2 includes the direct effects and squared terms of all of the independent variables. Hypothesis 1, which predicted that the degree of divergence of a new venture from its parents' knowledge domains would have an inverted-U effect on the impact of knowledge created by the new venture, was confirmed. The curvilinear term for new venture overlap with parent technology is negative and significant (p b .001), supporting its inverted U-shaped effect on new venture knowledge impact. To interpret these results, we provide a plot of the count of new venture total forward patent citations for various levels of overlap with parent firms in Fig. 2. We calculated the predicted values of venture knowledge impact using the ‘margins’ command in STATA (Williams, 2012) and setting all other covariates equal to their means. For reference, the observed range of venture overlap in our data is from 0 to 0.9. As shown in Fig. 2, the relative maximum venture knowledge impact occurs near the overlap proportion of 0.4, yielding a total knowledge impact of 58. Model 2 demonstrates a significant (p b .05), negative relationship between founder technology divergence from the parent and the impact of the new venture's knowledge, supporting hypothesis 2. This is consistent with our argument that founders who engage in more peripheral technologies during their tenure at the parent receive less benefit that may be applied to innovations at the new firm. Finally, we do not find support for hypothesis 3. The coefficient associated with founder technology breadth is significant (p b .05) but negative, contrary to our hypothesized effects. Furthermore, this effect loses significance in the robustness tests described below. 5.1. Robustness tests and further analyses In the fields of entrepreneurship and management strategy, enodgeneity is a significant concern for econometric analysis (Shaver, 1998). Unobserved factors may influence both independent and dependent variables in a way that confounds the true results, falsely suggesting support for hypotheses. For example, in our context, a venture with certain unobserved attributes might be choosing a strategy of divergence from parent and also creating high-impact knowledge. We pursued a number of steps to reduce the likelihood that endogeneity is a concern in our results (Bednar et al., 2013). First, as described previously, we included a broad set of control variables to address other potential factors that could otherwise explain our findings. Then, we tested directly for the presence of endogeneity in our models. Endogeneity is shown to be present when the residual error term of a regression model is significantly correlated with the independent variables (Bascle, 2008). To check for this, we calculated and retained the residual error term associated with our regression model testing study hypotheses. We then calculated the correlations between this term and our independent variables. The resulting values were 0.005 for Venture technology overlap with parent, −0.001 for Founder technology divergence from parent, and 0.003 for Founder technology breadth. None of these correlations are statistically significant, providing evidence that endogeneity is not a concern. We took a number of other steps in order to ensure the robustness of our results. To verify that there are no multicollinearity issues among the independent variables in our study, we calculated the variance inflation factor for each covariate. The maximum value resulting from this calculation was 1.42, well below the 5.0 threshold established in prior research (Dielman, 1991). As described previously, we conducted a Vuong test to compare the negative binomial distribution to models using a zero-inflated negative binomial approach. The resulting z-test statistic of 1.03 was not significant (Pr N z = 0.15), thus confirming that the use of the negative binomial family specification is warranted. We tested whether our findings are robust to other formulations of knowledge impact. Following in the steps of prior research (Ahuja et al., 2005; Kaplan and Vakili, 2012; Phene et al., 2006; Srivastava and Gnyawali, 2011), we examined whether the firms in our sample have a greater likelihood to develop breakthrough technologies, based on the hypothesized relationships in our study. To do so, we constructed an alternative binary measure of our dependent variable, knowledge impact, by observing patents that were the most highly cited. Those patents that were among the top 5% in terms of external forward citations across all U.S. patents with the same primary IPC class were coded as 1, and all other patents were coded as 0. We built a new data structure using the firm-patent as the unit of analysis, allowing us to test our study hypotheses through a set of logistic regression models. This essentially Please cite this article as: Basu, S., et al., Parent inheritance, founder expertise, and venture strategy: Determinants of new venture knowledge impact, J. Bus. Venturing (2014), http://dx.doi.org/10.1016/j.jbusvent.2014.06.002

10

S. Basu et al. / Journal of Business Venturing xxx (2014) xxx–xxx

Table 1 Descriptive statistics and correlations. Variable 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Mean

S.D.

Venture Knowledge Impact 88.24 247.14 Age of Founder Coinventor Network Patents 752.4 359.5 Founder Co-inventor Network Patents 51.16 87.38 Founder Tenure at Parent 6.27 5.22 Parent Patent Novelty 2.36 7.45 Inventor Migration from Parent 0.68 1.64 Venture Technology Overlap with Parent 0.05 0.13 Founder Technology Divergence from Parent 0.20 0.28 Founder Technology Breadth 5.98 7.07 Parent Technology Diversity 0.66 0.43 Age of Parent Patents 475.0 401.2 Size of Parent Knowledge Base 1131.7 3196.0 Parent Knowledge Impact 5.81 11.16 Geographic Proximity to Parent 0.35 0.27 Number of Founders 1.43 0.72 Founder Number of Patents 9.55 14.51 Venture Number of Patents 25.6 57.6 Parent Number of Patents 528.3 1317.5 Founder Tenure at Firm Prior to Parent 1.88 3.17 Founder Length of Experience 7.90 5.30 Venture Resources 22.40 28.50 Rounds of VC Funding 2.98 2.15 Proportion of Academic Founders 0.37 0.49 Non-Compete State 0.48 0.50

1

2

−.039 −.096 .230⁎⁎ .000 .058 −.116 .002 .137⁎ .068 −.012 .040 −.074 .031 −.033 .118 −.070 .356⁎⁎ −.165⁎⁎ .232⁎⁎ −.075 .148⁎ .152⁎ .052 .050 −.086 −.023 .259⁎⁎ −.078 .211⁎⁎ .082 .082 −.089 .159⁎ −.106 .040 −.139⁎ .249⁎⁎ −.051 .042 .026 .003 −.071 −.119 .074 .155⁎

3

4

.164⁎ .033 .033 .158⁎ .236⁎⁎ .094 .015 −.134⁎ −.145⁎ .336⁎⁎ .726⁎⁎ .288⁎⁎ .100 .079 .012 .242⁎⁎ .041 .018 −.043 −.001 .091 .263⁎⁎ .128 .407⁎⁎ .276⁎⁎ .247⁎⁎ .116 .257⁎⁎ .035 −.036 .023 .158⁎ .214⁎⁎ .186⁎⁎ .020 .017 .059 −.081 .208⁎⁎ .048 −.018

5

6

−.043 .017 .134⁎ .179⁎⁎ −.030 .006 .272⁎⁎ −.032 .121 −.009 −.053 .026 .029 .038 .071 .063 .063 .062 .175⁎⁎ .062 .164⁎ .022 .398⁎⁎ .007 .014 .068 −.021 −.023 .047 −.051 .046 −.016 .103 .023 .136⁎ −.046 .006

7

8

.039 .052 .200⁎⁎ .002 .018 .077 .152⁎ .078 .016 .054 .036 .003 .086 −.012 −.053 −.035 −.058

−.127 .208⁎ .270⁎⁎ −.025 .079 −.082 −.120 −.205⁎⁎ .000 .032 .137⁎ .003 −.092 −.029 −.102 −.083

⁎ Correlation is significant at the 0.05 level (2-tailed). ⁎⁎ Correlation is significant at the 0.01 level (2-tailed).

examines the likelihood that a given patent developed by the focal firm in the period after founding will represent a breakthrough technology. The results of this analysis are fully consistent with the outcome of the regression models reported in Table 3. Again, hypotheses 1 and 2 are strongly supported, while hypothesis 3 does not show significant results. The logistic regression robustness test has the advantages of controlling for variance due to technology class by incorporating dummy variables for each patent category of technologies developed by the new venture.12 To further test the robustness of our genealogical results, we constructed alternate biography-based linkages between ventures and parents for a subset of our sample. Recall that our primary measure of genealogical links between parent firms and their progenies is patent-based: the firm where a given founder was assigned the most recent patent prior to joining the new venture is considered the parent firm (this process is described in detail in the Appendix A). To develop biography-based linkages, we used diverse sources including VentureSource, ZoomInfo, LinkedIn, and similar other professional search platforms to collect biographical data. This process resulted in detailed employment histories of founders with a biography-based matrix linking new ventures to their parents. The resulting subsample consists of 119 ventures with 241 founders founded through 1996–1999. We used this alternative data structure to test hypothesis 1, the influence of venture technological divergence on the impact of knowledge created by new ventures. While this subsample does not focus specifically on founders who engage in patenting and thus does not allow us to test hypotheses 2 and 3, we do find support for hypothesis 1 using this alternate biography-based approach.13 The curvilinear term for a new venture's technological divergence from the parent is negative and significant, consistent with the findings in our analysis of the patent-based parent–progeny sample. We tested whether the results of our logistic regression tests of breakthrough technology might be confounded by a concentration of patents in specific technology areas. After investigating the primary patent codes of all patents in this sample, we observed that US patent category 435-6 accounted for a substantial portion of the observations – a total of 248 out of 3229 patents. We ran a robustness test, omitting all of the patents in this technology category, and our findings remained consistent with the baseline results of Table 3. We also ran tests to ensure that the logistic regression analysis would have sufficient statistical power to provide correct tests of our study hypotheses. Using the ‘powerlog’ function in STATA, we determined that based on our sample size of 3,229 patent observations, our statistical power is greater than 0.90 in testing the effects of venture technology overlap with the parent firm and founder technology breadth, and it is greater than 0.80 in testing the effects of founder technology divergence form the parent firm, with the standard assumption of alpha levels of .05. We also wanted to determine whether any firms in our sample had corporate venture capital investments or another kind of relationship with the parent firm, which might influence our results. We reviewed the venture capital investment history of all the firms in our sample through data available from Thomson One VentureXpert database and did find one observation in which the parent firm (Incyte Pharmaceuticals) made a CVC investment in a progeny firm (Genomic Health). We ran robustness checks omitting this firm in both the negative binomial models and the logistic regression models testing for breakthrough knowledge. In all cases, there were no material changes to the results.

12 13

We thank anonymous reviewers for guiding us in this direction. We do not include the table in the manuscript due to space constraints but it can be obtained from the authors upon request.

Please cite this article as: Basu, S., et al., Parent inheritance, founder expertise, and venture strategy: Determinants of new venture knowledge impact, J. Bus. Venturing (2014), http://dx.doi.org/10.1016/j.jbusvent.2014.06.002

S. Basu et al. / Journal of Business Venturing xxx (2014) xxx–xxx

11

Table 1 Descriptive statistics and correlations. 9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

.122 −.065 .092 −.036 .038 .225⁎⁎ .445⁎⁎ .204⁎⁎ .102 −.029 .155⁎ .104 .088 .134⁎ .062

.460⁎⁎ 0.2697⁎ .177⁎⁎ −.059 .205⁎⁎ .270⁎⁎ .112 .305⁎⁎ −.052 .255⁎⁎ −.080 −.044 −.116 .002

.132⁎ .080 −.053 .060 .052 −.007 .156⁎ .044 .225⁎⁎ −.017 −.044 −.042 .018

.048 −.024 .099 .176⁎⁎ .099 .914⁎⁎ .081 .157⁎ .041 .034 −.101 .024

−.034 .113 .048 .110 .042 .060 .061 −.046 −.012 −.007 .068

−.158⁎ .031 .075 −.046 −.020 .039 −.088 −.074 .096 −.081

.493⁎⁎ .151⁎ .087 .047 −.014 .124 .034 .062 −.020

.209⁎⁎ .175⁎⁎ −.101 .192⁎⁎ .083 .050 .029 −.077

.073 .010 .037 .178⁎⁎ .076 −.001 .060

.086 .150 .062 .035 −.146 .004

.424⁎⁎ −.012 −.005 .006 .091

.048 .078 −.077 .080

.422⁎⁎ −.029 .171⁎⁎

.002 .081

−.123

Finally, implicit in our study is the assumption that creating more impactful knowledge leads to other financial and performance outcomes for new ventures, and we chose to empirically test this assumption among the firms in our sample. Retaining all of the control variables used in our baseline hypothesis tests, we ran logistic regression models with binary outcomes of 1) successful achievement of IPO and 2) firm failure, this time employing venture knowledge impact as the independent variable. The results of this analysis show that higher knowledge impact is significant and positively associated with IPO and significant and negatively associated with firm failure.14 6. Discussion This study sought to examine how parent firm knowledge and the knowledge of founders may influence the impact of knowledge generated by a new venture. By conceptualizing knowledge generation as the outcome of combinations of existing and new elements, we identified the important role that founders play in influencing venture knowledge outcomes beyond merely acting as a conduit to transfer parent knowledge to their new ventures. We argued that a venture's divergence from its parents' knowledge domains represents the founders' choices to combine more new and unfamiliar elements rather than recombine familiar parent knowledge elements and can have an important influence on the impact of the venture's knowledge. Further, in integrating observations from the entrepreneurship literature (Cliff et al., 2006) with research on parent–progeny relationships, we highlight the important role of individual founder expertise and experience in shaping the innovative impact of the new venture. Our results suggest that the extent to which the new venture diverges from the knowledge domains of the parent results initially in higher-impact knowledge for the new venture but that too great a divergence from the parent is ultimately counterproductive. Our findings also demonstrate that greater divergence of founders' knowledge from their parents' knowledge domains has a negative effect on venture knowledge impact, supporting the perspective that founders who work on technologies that are less central to the parent firm provide less benefit for the impact of innovations produced at the new firm. However, we did not find support for our contention that founder's own technological breadth is associated with the creation of impactful knowledge by the venture.

6.1. Theoretical contributions and implications Prior literature has conceptualized knowledge conveyed from parent firms to new ventures by venture founders as a form of genealogical inheritance. Scholars in this tradition have generally assumed that greater parent firm knowledge will have a positive effect on venture performance (Agarwal et al., 2004; Franco and Filson, 2006; Klepper and Sleeper, 2005), but the extent to which the new venture chooses to diverge from this inherited knowledge has not been examined in past research. In addition, most of the past literature on genealogical effects and inheritance assumes a positive impact on the inheritance of parental characteristics and routines. Our 14

Again, the results of this analysis can be obtained from the authors upon request.

Please cite this article as: Basu, S., et al., Parent inheritance, founder expertise, and venture strategy: Determinants of new venture knowledge impact, J. Bus. Venturing (2014), http://dx.doi.org/10.1016/j.jbusvent.2014.06.002

12

S. Basu et al. / Journal of Business Venturing xxx (2014) xxx–xxx

Table 2 Test of study hypotheses. Model Specification

Negative Binomial

Dependent Variable

Venture Knowledge Impact Model 1

Independent Variables Venture Technology Overlap with Parent Venture Technology Overlap with Parent2 (H1) Founder Technology Divergence from Parent (H2) Founder Technology Breadth (H3)

Model 2 8.266⁎⁎ −11.045⁎⁎ −1.177⁎⁎ −0.029⁎⁎

Control Variables Parent Technology Diversity Age of Parent Patents Size of Parent Knowledge Base Parent Knowledge Impact Geographic Proximity to Parent Number of Founders Founder Number of Patents Venture Number of Patents Parent Number of Patents Founder Tenure at Firm Prior to Parent Founder Length of Experience Venture Resources Rounds of VC Funding Proportion of Academic Founders Non-Compete State Year Dummy Variables1 Industry Segment Dummy Variables2 Constant Wald X2 Log Pseudolikelihood

1.368⁎⁎⁎ −0.002⁎⁎⁎ −5.08E-05 0.002 1.232⁎ −0.147 −0.012 0.027⁎⁎ 1.70E-04 0.006 0.006 0.012+ 0.036 0.336 0.451+ included included −1.399⁎ 366.21⁎⁎⁎

Sample Size

219

−821.29

1.442⁎⁎⁎ −0.001⁎⁎ −6.43E-05 −0.005 1.371⁎ −0.331+ −0.006 0 0.025⁎⁎⁎ −0.006 0.0.31 −0.027 0.008 0.064 0.200 0.581⁎ included included −1.632⁎ 441.41⁎⁎⁎ −783.66 219

+

p b .1. ⁎ p b .05. ⁎⁎ p b .01. ⁎⁎⁎ p b .001. 1 1990–99 included, 2000 omitted. 2 Drug Delivery, Drug Development Technologies, and Biotechnology Therapeutics included, Biopharmaceuticals omitted.

research demonstrates that there are critical limits in the extent to which inherited knowledge can be usefully utilized by a new venture beyond which the value of inheritance diminishes and may even become dysfunctional. While previous work has conceptualized knowledge creation as a firm's active search in a particular space for knowledge elements that can be combined in potentially new ways (Katila and Ahuja, 2002; Leiponen and Helfat, 2009), we underscore how such opportunities can be obtained by founders from their previous employment experience. Our use of a genealogical perspective enhances our understanding of how knowledge transfer is influenced by founders' decisions regarding the degree to which the ventures inherit and leverage parent technologies, as captured by the technological overlap of a venture with its parents. We demonstrate the effects of both parent-level and founder-level knowledge on a venture's knowledge impact, conceptualizing founders as being sources of knowledge in their own right rather than just being a conduit for the transfer of parent knowledge. Moreover, the ability of founders to transfer parent knowledge is also constrained by their own prior roles in parent technological development. This study is the first to examine the role of founder's expertise, parent firm knowledge and venture's strategic choices in the creation of impactful knowledge. Our results also have important implications for the broader area of entrepreneurship research. One of our important contributions is to integrate the entrepreneurship literature with research on parent–progeny inheritances. While prior entrepreneurship literature highlights the importance of a founder's background and expertise in influencing the type and quality of opportunities she recognizes, we specifically demonstrate that such valuable knowledge is not only developed within individuals but also carried over from previous employment. Along the lines of Jennings et al. (2009) who examine how novelty/divergence affects new venture performance, we examine the implications of these choices with respect to implementing local versus distant search. Past research has highlighted the benefits of an organization's ability to achieve sufficient levels of both exploitation and exploration and the challenges in doing so (Beckman, 2006; Gibson and Birkinshaw, 2004; He and Wong, 2004; Phene et al., 2010). In our context, a high level of overlap with parent knowledge domains implies that a venture is primarily engaging in local search that might constrain its explorative activity. A low level of overlap implies that the venture is primarily engaging in distant search and is not adequately exploiting its knowledge inheritance. In contrast to current work focusing mostly on challenges for large, established organizations in balancing exploration and exploitation (Tushman and O’Reilly, 1996), our research highlights the tensions that even new ventures can face in trying to pursue both approaches concurrently. Please cite this article as: Basu, S., et al., Parent inheritance, founder expertise, and venture strategy: Determinants of new venture knowledge impact, J. Bus. Venturing (2014), http://dx.doi.org/10.1016/j.jbusvent.2014.06.002

S. Basu et al. / Journal of Business Venturing xxx (2014) xxx–xxx

13

Fig. 2. Variation of knowledge impact with overlap proportion.

The literature on innovation search (Katila and Ahuja, 2002; Leiponen and Helfat, 2009) examines positive and negative effects of various attributes of a firm's search space such as search depth and search scope on its innovation performance. For the purposes of this study, the traditional interpretation of local search as search in the neighborhood of existing knowledge (Ahuja and Lampert, 2001; Winter et al., 2007) is often meaningless for new ventures that do not possess existing stocks of knowledge. Rather, it is the search within parent knowledge that represents local search for such ventures and the search beyond parent knowledge that represents distant search. We therefore complement past search literature that has conceptualized a firm's search space as a product of its own choice by arguing that new ventures also inherit a possible search space for future initiatives from the knowledge structures of their parents. The venture can subsequently choose the extent to which it remains proximal or diverges from its parents' knowledge domains, with such choices having important implications for the impact of knowledge developed by new ventures. We believe that our findings in combination add interesting nuances to the received theory on the value of experiential learning. Our arguments for H1 and H2 highlight the value of building on experience in parent firms but caution against an overdependence on such experience, which might isolate the founders from other potentially valuable sources of knowledge. Our results are consistent with the findings of Greenwood et al. (2002) that deviation from existing templates can promote greater efficiency. Similarly, Reagans and Zuckerman (2001) found that teams having members with diverse tenure are more productive as they recombine different sets of information, and experiences. In the context of progeny ventures, our empirical findings confirm that modest overlap with parent knowledge (representing a significant but not radical deviation from established parent templates) is associated with the creation of impactful knowledge.

6.2. Limitations and future research An important limitation of this study is our reliance on patent-based data to measure our dependent and independent variables. While the limitations of patent-based data as indicators of firm technological capabilities and specializations have been discussed (Alcacer and Gittelman, 2006; Benner and Waldfogel, 2008), there is also consensus that use of such data is appropriate and usually most accurate and comprehensive for measuring knowledge stocks, profiles and transfers (Ahuja, 2000; Miller et al., 2007). Nevertheless, we would like to see future research use survey and interview methodologies to capture venture decisions to mimic or diverge from parent technologies. Moreover, to avoid undue complexity, we chose to restrict our analysis to the venture's immediate parents, that is, the firms where the founders were productive inventors just prior to their founding the ventures in our sample. We recognize that other firms where the founders were employed even earlier may have an important influence on founders' ideas and expertise. Future research could longitudinally track founder's careers to examine questions related to conditions under which founders obtain their dominant influences for starting new ventures. Consistent with our research question, we examine determinants of a venture's knowledge impact, evaluated as the extent to which such knowledge is utilized by other firms. In this study, we do not formally hypothesize about how a venture's knowledge impact influences its eventual economic performance although we do carry out some relevant post-hoc tests. While this relationship may intuitively appear to be positive, there might be negative effects of excessive spillovers to rival firms, thereby weakening the particular venture's intellectual property. We believe that examinations of the contingent effects of venture knowledge impact on its economic performance are not completely understood and therefore promising opportunities for future research. Please cite this article as: Basu, S., et al., Parent inheritance, founder expertise, and venture strategy: Determinants of new venture knowledge impact, J. Bus. Venturing (2014), http://dx.doi.org/10.1016/j.jbusvent.2014.06.002

14

S. Basu et al. / Journal of Business Venturing xxx (2014) xxx–xxx

Table 3 Breakthrough technology analysis. Dependent Variable

Logistic Regression Analysis Breakthrough Technology - Top 5% Impact Model 3

Independent Variables Venture Technology Overlap with Parent Venture Technology Overlap with Parent2 (H1) Founder Technology Divergence from Parent (H2) Founder Technology Breadth (H3) Control Variables Parent Technology Diversity Age of Parent Patents Size of Parent Knowledge Base Average Knowledge Impact of Parent Firms Geographic Proximity to Parent Number of Founders Founder Number of Patents Average Number of Patents of Parent Firms Founder Length of Experience Venture Resources Rounds of VC Funding Proportion of Academic Founders Non-Compete State Primary Patent Class Dummy Variables Wald X2 Log Pseudolikelihood Sample Size

Model 4

Model 5

Model 6

Model 7

−0.008

12.543⁎⁎ −16.976⁎ −1.352⁎ −0.003

0.715⁎ −8.61E-04⁎ −1.86E-04+ 9.93E-04 −1.823⁎⁎ −0.443⁎⁎ 0.019⁎⁎ 3.50E-04⁎ −0.050⁎ 0.004 0.100⁎ 0.515⁎ 0.129 included 188.69⁎⁎⁎ −566.93 3229

0.506 −5.41E-04 −2.38E-04⁎ 0.018⁎ −1.965⁎⁎ −0.505⁎⁎ 0.018⁎⁎ 4.63E-04⁎⁎ −0.055⁎ 0.004 0.084+ 0.619⁎⁎ 0.135 included 198.84⁎⁎⁎ −548.80 3229

13.899⁎⁎ −19.502⁎ 1.443⁎

0.709⁎ −8.06E-04⁎ −1.79E-04+ 0.002 −1.906⁎⁎ −0.446⁎⁎ 0.018⁎⁎ 3.39E-04⁎ −0.053⁎ 0.005 0.097⁎ 0.522⁎ 0.105 included 183.79⁎⁎⁎ 557.08 3229

0.588+ −7.74E-04⁎ −1.64E-04+ 0.003 −1.930⁎⁎ −0.388⁎ 0.017⁎⁎ 3.34E-04⁎ −0.053⁎ 0.004 0.110⁎ 0.575⁎ 0.201 included 196.29⁎⁎⁎ −552.30 ,229

0.622+ −5.23E-04 −2.55E-04⁎ 0.019⁎ −1.983⁎⁎ −0.574⁎⁎⁎ 0.018⁎⁎ 4.73E-04⁎⁎ −0.058⁎ 0.005+ 0.068 0.573⁎⁎ 0.016 included 185.03⁎⁎⁎ −553.01 3,29

+

p b .1. ⁎ p b .05. ⁎⁎ p b .01. ⁎⁎⁎ p b .001.

Founders may come from a number of different functional backgrounds and career experiences. We do not assume that all founders of new ventures are likely to be involved in technology development or patenting in the new venture. Our measure of patentbased genealogical links ensures that the founder was, in fact, engaged in technology development through their prior experience at the parent. We make no claims that founders coming from other backgrounds (e.g., marketing or finance) would be able to transfer technology or knowledge that is useful and impactful to the new venture. In addition, and in keeping with our research question, we observe the effects of parent knowledge structures and venture technological choices for the first five years of the venture's existence. It would be interesting to observe changes in parent-venture technological divergence over a broader window of the venture lifecycle, which would perhaps imply modification of the founders' initial strategies by either the founders themselves or by subsequent managerial hires. Longitudinal data could also be used to address more dynamic questions regarding how such changes result in the evolution of venture knowledge impact over time. As one of the first studies to address the role of founder expertise and choices in the parent–progeny phenomenon, we focused on the most relevant variables consistent with this study's theoretical framing. Other considerations such as inventor networks, collaborations, and publishing activity may also be associated with novel innovations, underscoring the need for additional research that can examine knowledge inheritance from a number of different perspectives, including networks or past collaborations. Finally, it would be interesting and informative to extend this study to other industries and to other countries to examine the extent to which the patterns we observe here also hold in other settings. Earlier investigations of progeny venture performance have devoted primary attention to the transfer of valuable parent knowledge to the venture through its founders. Our examination of knowledge inheritance helps to provide further insights into resource-based view of entrepreneurial activity, since founder knowledge from parent firms is a resource that can aid as well as limit and constrain the venture's development of its own unique technological capabilities. By providing evidence as to how knowledge gained from the parent firm and individual founders influence the impact of knowledge created by the new venture, our research provides a critical theoretical perspective in understanding the emergence and performance of technology-based new ventures. Appendix A. Elaboration of methods To provide further clarification of how we constructed parent–progeny links, here is an example of a particular progeny venture in our sample. Tolerx Inc. was incorporated in 2000 and is based in Cambridge, Massachusetts. It specializes in the development of novel therapies for the treatment of immune system disorders. We track parent–progeny linkages for our sample of new ventures (Tolerx in this case) as follows: First, we identify the patenting record of all the founders of the new venture by mapping founder name with the patents filed with the U.S. Patent and Trademark Please cite this article as: Basu, S., et al., Parent inheritance, founder expertise, and venture strategy: Determinants of new venture knowledge impact, J. Bus. Venturing (2014), http://dx.doi.org/10.1016/j.jbusvent.2014.06.002

S. Basu et al. / Journal of Business Venturing xxx (2014) xxx–xxx

15

Office. We consider only those patents on which the focal founder is listed as an inventor prior to the founding date of the new firm, namely Dr. Douglas Ringler and Dr. Heman Waldmann. In this case, Dr. Ringler's most recent patent activity prior to his participation as a founder of Tolerx occurred in 1995 and 1996. There were three patent applications during that period listing Dr. Douglas J. Ringler as one of the inventors and Millennium Pharmaceuticals Inc. as the sole assignee to the patent. Dr. Waldmann's (the other founder) most recent recorded inventions before founding Tolerx involved patents listing Glaxo Wellcome Co. as the sole assignee. Under our methodology, we thus recognize Millennium Pharmaceuticals and Glaxo Wellcome as genealogical parents to Tolerx. Since the two founders that each worked at different parent firms, we average the values of our explanatory variables. To check the robustness of our analyses, we also ran our models using maximum values for any founder and found consistent results. Our methodology for matching new ventures to parents hinges on the ability to accurately track the prior patenting activity of the founders in our sample. We utilized the disambiguated list of U.S. inventors developed through the Harvard University patent project (Lai et al., 2009) which has established disambiguated data on inventor careers (Aggarwal and Hsu, 2013; Lai et al., 2011). This dataset is able to track patenting records of unique individuals based on their lineage and it is sophisticated enough to identify distinct careers of inventors, even those with the same name. Consistent with the premise that temporal sequence of an individual's career and innovation history can be more accurately tracked with his or her patenting record rather than relying on biographical information, we used this database to match prior patents to the unique individuals (e.g., Douglas Ringler and Herman Waldmann) listed in our founder data. Through this approach, we are able to establish a systematic record of the technology affiliation of new venture founders in terms of intellectual linkages to former employers.

References Acs, Z.J., Audretsch, D.B., Strom, R., 2009. Entrepreneurship, growth, and public policy. Cambridge University Press, New York. Agarwal, R., Echabadi, R., Franco, A., Sarkar, M.B., 2004. Knowledge transfer through inheritance: Spin-out generation, development, and survival. Acad. Manag. J. 47, 501–522. Aggarwal, V.A., Hsu, D.H., 2013. Entrepreneurial exits and innovation. Manag. Sci. 60 (4), 867–887. Ahuja, G., 2000. Collaboration networks, structural holes, and innovation: A longitudinal study. Adm. Sci. Q. 45, 425–455. Ahuja, G., Katila, R., 2001. Technological acquisitions and the innovation performance of acquiring firms: A longitudinal study. Strateg. Manag. J. 22, 197–220. Ahuja, G., Katila, R., 2004. Where do resources come from? The role of idiosyncratic situations. Strateg. Manag. J. 25 (8–9), 887–907. Ahuja, G., Lampert, C.M., 2001. Entrepreneurship in the large corporation: A longitudinal study of how established firms create breakthrough inventions. Strateg. Manag. J. 22, 521–543. Ahuja, G., Coff, R.W., Lee, P.M., 2005. Managerial foresight and attempted rent appropriation: Insider trading on knowledge of imminent breakthroughs. Strateg. Manag. J. 26 (9), 791–808. Alcacer, J., Gittelman, M., 2006. Patent citations as a measure of knowledge flows: The influence of examiner citations. Rev. Econ. Stat. 88, 774–779. Baron, R.A., 2006. Opportunity recognition as pattern recognition: How entrepreneurs ‘connect the dots’ to identify new business opportunities. Acad. Manag. Perspect. 20 (1), 104–119. Bascle, G., 2008. Controlling for endogeneity with instrumental variables in strategic management research. Strateg. Organ. 6 (3), 285–327. Beckman, C., 2006. The influence of founding team company affiliations on firm behaviors. Acad. Manag. J. 49, 741–758. Beckman, C., Burton, D., 2008. Founding the future: Path dependence in the evolution of top management team from founding to IPO. Organ. Sci. 19 (1), 3–24. Bednar, M.K., Boivie, S., Prince, N.R., 2013. Burr under the saddle: How media coverage influences strategic change. Organ. Sci. 24 (3), 910–925. Benner, M., Waldfogel, J., 2008. Close to you? Bias and precision in patent-based measures of technological proximity. Res. Policy 37, 1556–1567. Bergeron, B.P., Chan, P., 2004. Biotech industry: A global, economic, and financing overview. John Wiley & Sons, Inc., New York. Chatterji, A.K., 2009. Spawned with a silver spoon? Entrepreneurial performance and innovation in the medical device industry. Strateg. Manag. J. 30, 185–206. Cliff, J.E., Jennings, P.D., Greenwood, R., 2006. New to the game and questioning the rules: The experiences and beliefs of founders who start imitative versus innovative firms. J. Bus. Ventur. 21, 633–663. Colombo, M.G., 2003. Alliance form: A test of the contractual and competence perspectives. Strateg. Manag. J. 24, 1209–1229. Colombo, M.G., Grilli, L., 2005. Founders' human capital and the growth of new technology-based firms: A competence-based view. Res. Policy 34, 795–816. Davidsson, P., Delmar, F., Wiklund, J., 2006. Entrepreneurship and the growth of firms. Edward Elgar, Chelterham. Dielman, T.E., 1991. Applied regression analysis for business and economics. PWS-Kent Publishing Company, Boston, MA. Dushnitsky, G., Lenox, M.J., 2006. When does corporate venture capital investment create firm value? J. Bus. Ventur. 21, 753–772. Fleming, L., 2007. Breakthroughs and the “long tail” of innovation. Sloan Manag. Rev. 49 (1), 69–74. Fleming, L., Sorenson, O., 2001. Technology as a complex adaptive system: Evidence from patent data. Res. Policy 30, 1019–1039. Franco, A.M., Filson, D., 2006. Spin-outs: Knowledge diffusion through employee mobility. RAND J. Econ. 37, 841–860. Gibson, C.B., Birkinshaw, J., 2004. The antecedents, consequences, and mediating role of organizational ambidexterity. Acad. Manag. J. 47, 209–226. Gilbert, B.A., McDougall, P., Audretsch, D., 2006. New venture growth: A review and extension. J. Manag. 32 (6), 926–950. Gilbert, B.A., McDougall, P.P., Audretsch, D.B., 2008. Clusters, knowledge spillovers and new venture performance: An empirical examination. Journal of Business Venturing 23 (4), 405–422. Gompers, P.A., Lerner, J., 1998. The determinants of corporate venture capital success: Organizational structure, incentives, and complementarities. Working paper 6725. NBER. Gompers, P., Lerner, J., Scharfstein, D., 2005. Entrepreneurial spawning: Public corporations and the genesis of new ventures, 1986 to 1999. J. Financ. 60, 577–614. Greenwood, R., Suddaby, R., Hinings, C.R., 2002. Theorizing change: The role of professional associations in the transformation of institutionalized fields. Acad. Manag. J. 45, 45–57. Gruber, M., MacMillan, I.C., Thompson, J.D., 2008. Look before you leap: Market opportunity identification in emerging technology firms. Manag. Sci. 54, 1652–1665. Hand, J.R.M., 2007. Determinants of the round-to-round returns to pre-IPO venture capital investments in U.S. biotechnology companies. J. Bus. Ventur. 22, 1–28. Hargadon, A., Sutton, R.I., 1997. Technology brokering and innovation in a product development firm. Adm. Sci. Q. 42, 716–749. Harhoff, D., Narin, F., Scherer, F.M., Vopel, K., 1999. Citation frequency and the value of patented inventions. Rev. Econ. Stat. 81, 511–515. He, Z.L., Wong, P.K., 2004. Exploration versus exploitation: An empirical test of the ambidexterity hypothesis. Organ. Sci. 15, 481–494. Helfat, C.E., Lieberman, M.B., 2002. The birth of capabilities: Market entry and the importance of pre-history. Ind. Corp. Chang. 11 (4), 725–760. Henderson, R.M., Clark, K.B., 1990. Architectural innovation: The reconfiguration of existing product technologies and the failure of established firms. Adm. Sci. Q. 35, 9–31. Henderson, R., Cockburn, I., 1996. Scale, scope, and spillovers: The determinants of research productivity in drug discovery. RAND J. Econ. 27, 32–59. Hilbe, J., 2007. Negative binomial regression. Cambridge University Press, New York. Hmieleski, K.M., Baron, R.A., 2009. Entrepreneurs' optimism and new venture performance: A social cognitive perspective. Acad. Manag. J. 52, 473–488.

Please cite this article as: Basu, S., et al., Parent inheritance, founder expertise, and venture strategy: Determinants of new venture knowledge impact, J. Bus. Venturing (2014), http://dx.doi.org/10.1016/j.jbusvent.2014.06.002

16

S. Basu et al. / Journal of Business Venturing xxx (2014) xxx–xxx

Hoetker, G., Agarwal, R., 2007. Death hurts, but it isn't fatal: The postexit diffusion of knowledge created by innovative companies. Acad. Manag. J. 50, 446–467. Jennings, J.E., Jennings, P.D., Greenwood, R., 2009. Novelty and new firm performance: The case of employment systems in knowledge-intensive service organizations. J. Bus. Ventur. 24 (4), 338–359. Kaplan, S., Vakili, K., 2012. Identifying breakthroughs: Cognitive vs. economic. Druid Conference, June, Copenhagen. Katila, R., Ahuja, G., 2002. Something old, something new: A longitudinal study of search behavior and new product introduction. Acad. Manag. J. 45, 1183–1194. Klepper, S., 2001. Employee startups in high-tech industries. Ind. Corp. Chang. 10, 639–674. Klepper, S., Sleeper, S., 2005. Entry by spinoffs. Manag. Sci. 51, 1291–1306. Klepper, S., Thompson, P., 2006. Submarkets and the evolution of market structure. RAND J. Econ. 37, 861–886. Kotha, R., Zheng, Y., George, G., 2011. Entry into new niches: The effects of firm age and the expansion of technological capabilities on innovative output and impact. Strateg. Manag. J. 32, 1011–1024. Lai, R., D’Amour, A., Fleming, L., 2009. The careers and co-authorship networks of U.S. patent holders since 1975. Harvard Business School, Harvard Institute for Quantitative, Social Science (http://en.scientificcommons.org/48544046). Lai, R., D Amour, A., Yu, A., Sun, Y., Torvik, V., Fleming, L., 2011. Disambiguation and coauthorship networks of the US Patent Inventor Database. Harvard Institute for Quantitative Social Science, Cambridge, MA, 2138. Leiponen, A., Helfat, C.E., 2009. Innovation objectives, knowledge sources, and the benefits of breadth. Strateg. Manag. J. 31, 224–236. Lerner, J., 1994. The importance of patent scope: An empirical analysis. RAND J. Econ. 25, 319–333. Levinthal, D.A., March, J.G., 1993. The myopia of learning. Strateg. Manag. J. 14, 95–112. Liebeskind, J.P., Oliver, A., Zucker, L., Brewer, M., 1996. Social networks, learning, and flexibility: Sourcing scientific knowledge in new biotechnology firms. Organ. Sci. 7, 428–443. Low, M., MacMillan, I., 1988. Entrepreneurship: Past research and future challenges. J. Manag. 14, 139–161. Marx, M., Strumsky, D., Fleming, L., 2007. Noncompetes and inventor mobility: Specialists, stars and the Michigan experiment. Harvard Business School working paper 07-042, Cambridge, MA. McGrath, R.G., Tsai, M.H., Venkataraman, S., MacMillan, I.C., 1996. Innovation, competitive advantage and rent: A model and test. Manag. Sci. 42 (3), 389–403. Miller, D.J., Fern, M.J., Cardinal, L.B., 2007. The use of knowledge for technological innovation within diversified firms. Acad. Manag. J. 50, 308–326. Miner, A.S., Bassoff, P., Moorman, C., 2001. Organizational improvisation and learning: A field study. Adm. Sci. Q. 46, 304–337. Narin, F., Carpenter, M.P., Woolf, P., 1984. Technological performance assessments based on patents and patent citations. IEEE Trans. Eng. Manag. 31, 172–183. Nesta, L., 2008. Knowledge and productivity in the world's largest manufacturing corporations. J. Econ. Behav. Organ. 7, 886–902. Phene, A., Fladmoe-Lindquist, K., Marsh, L., 2006. Breakthrough innovations in the US biotechnology industry: The effects of technological space and geographic origin. Strateg. Manag. J. 27, 369–388. Phene, A., Tallman, S., Almeida, P., 2010. When do acquisitions facilitate technological exploration and exploitation? Journal of Management first published on May 7, 2010 as http://dx.doi.org/10.1177/0149206310369939. Phillips, D.J., 2002. A genealogical approach to organizational life changes: The parent–progeny transfer among Silicon Valley law firms, 1946–1996. Adm. Sci. Q. 47, 474–506. Reagans, R., Zuckerman, E., 2001. Networks, diversity, and productivity: The social capital of corporate R&D teams. Organ. Sci. 12, 502–517. Rosenkopf, L., Almeida, P., 2003. Overcoming local search through alliances and mobility. Manag. Sci. 49, 751–766. Rosenkopf, L., Nerkar, A., 2001. Beyond local search: Boundary spanning, exploration, and impact in the optical disk industry. Strateg. Manag. J. 22, 287–306. Rothaermel, F.T., Deeds, D.L., 2004. Exploration and exploitation alliances in biotechnology: A system of new product development. Strateg. Manag. J. 25, 201–221. Saxenian, A., 1994. Regional advantage: Culture and competition in Silicon Valley and Route 128. Harvard University Press, Cambridge. Schumpeter, J.A., 1934. Theory of economic development. Harvard University Press, Cambridge, MA. Shane, S., 2000. Prior knowledge and the discovery of entrepreneurial opportunities. Organ. Sci. 11, 448–469. Shane, S., 2003. A general theory of entrepreneurship. The individual–opportunity nexusEdward Elgar, Cheltenham. Shane, S., Venkataraman, S., 2000. The promise of entrepreneurship as a field of research. Acad. Manag. Rev. 25, 217–226. Shaver, J.M., 1998. Accounting for endogeneity when assessing strategy performance: Does entry mode choice affect FDI survival? Manag. Sci. 44, 571–585. Shane, S., Venkataraman, S., 2001. Entrepreneurship as a field of research: A response to Zahra and Dess, Singh, and Erikson. Academy of management review 26 (1), 13–16. Sorenson, O., Fleming, L., 2004. Science and the diffusion of knowledge. Res. Policy 33, 1615–1634. Sorenson, O., Rivkin, J., Fleming, L., 2006. Complexity, networks and knowledge flow. Res. Policy 35, 994–1017. Srivastava, M.K., Gnyawali, D.R., 2011. When do relational resources matter? Leveraging portfolio technological resources for breakthrough innovation. Acad. Manag. J. 54, 797–810. Stuart, T.E., Ding, W.W., 2006. When do scientists become entrepreneurs? The social structural antecedents of commercial activity in the academic life sciences. Am. J. Sociol. 112, 97–144. Stuart, T.E., Podolny, J.M., 1996. Local search and the evolution of technological capabilities. Strateg. Manag. J. 17, 21–38. Trajtenberg, M., 1990. A penny for your quotes: Patent citations and the value of innovations. RAND J. Econ. 21, 172–187. Tsai, W., 2001. Knowledge transfer in intraorganizational networks: Effects of network position and absorptive capacity on business unit innovation and performance. Acad. Manag. J. 44, 996–1004. Tushman, M.L., O’Reilly, C.M., 1996. Ambidextrous organizations: Managing evolutionary and revolutionary change. Calif. Manag. Rev. 38 (4), 8–30. Vaghely, I.P., Julien, P.A., 2010. Are opportunities recognized or constructed? An information perspective on entrepreneurial opportunity identification. J. Bus. Ventur. 25, 73–86. Vasudeva, G., Anand, J., 2011. Unpacking absorptive capacity: A study of knowledge utilization from alliance portfolios. Acad. Manag. J. 54 (3), 611–623. Williams, R., 2012. Using the margins command to estimate and interpret adjusted predictions and marginal effects. Stata J. 12, 308–331. Winter, S., 1991. The nature of the firm. Oxford University Press, Oxford, UK. Winter, S.G., Cattani, G., Dorsch, A., 2007. The value of moderate obsession: Insights from a new model of organizational search. Organ. Sci. 18, 403–419. Wry, T., Greenwood, R., Jennings, P.D., Lounsbury, M., 2010. Institutional sources of technological knowledge: A community perspective on nanotechnology emergence. In: Phillips, N., Sewell, G., Griffiths, D. (Eds.), Technology and organization: essays in honor of Joan Woodward. Research in the Sociology of Organizations, vol. 29. Emerald Group Publishing Limited, pp. 149–176. Yayavaram, S., Ahuja, G., 2008. Decomposability in knowledge structures and its impact on the usefulness of inventions and knowledge-base malleability. Adm. Sci. Q. 53, 333–362.

Please cite this article as: Basu, S., et al., Parent inheritance, founder expertise, and venture strategy: Determinants of new venture knowledge impact, J. Bus. Venturing (2014), http://dx.doi.org/10.1016/j.jbusvent.2014.06.002