when do relational resources matter? leveraging portfolio ... - CiteSeerX

4 downloads 0 Views 97KB Size Report
Collectively, these elements of a capa- bility-based perspective suggest that collaboration with resource-rich partners provides firms with ac- cess to knowledge ...
娀 Academy of Management Journal 2011, Vol. 54, No. 4, 797–810.

WHEN DO RELATIONAL RESOURCES MATTER? LEVERAGING PORTFOLIO TECHNOLOGICAL RESOURCES FOR BREAKTHROUGH INNOVATION MANISH K. SRIVASTAVA Michigan Technological University DEVI R. GNYAWALI Virginia Polytechnic Institute and State University We examine the paradox of capabilities: although portfolio resources contribute to innovation success, and technologically capable firms have the ability to gain more such resources, firms’ “competency traps” and the tension between value creation and value protection reduce benefits from portfolio resources for such firms. Results show that the quality and diversity of portfolio technological resources contribute to breakthrough innovation. The benefits are greater for firms with low internal strength and low internal diversity, thus suggesting positive synergy between portfolio and internal resources for such firms. Technologically strong firms, however, benefit from the quality of their portfolio resources when they overcome some of their competency traps.

A central question in the literature on alliances and networks is this: When do relational resources matter (Sampson, 2007; Stuart, 2000)? The issue of when becomes even more critical if a firm is in pursuit of generating high-impact or breakthrough innovations using knowledge-based resources in a dynamic environment. The resource-based view (Barney, 2001), the relational view (Dyer & Singh, 1998), and related literature on the role of alliances and networks on innovation (Ahuja, 2000a; Sampson, 2007) all suggest the importance of internal and relational resources for innovation. Moreover, the widely held notion of absorptive capacity (Cohen & Levinthal, 1990) stresses that firms with strong internal capabilities will benefit more from relational resources than do firms that lack these capabilities. Collectively, these elements of a capability-based perspective suggest that collaboration with resource-rich partners provides firms with access to knowledge (Dyer & Singh, 1998) that can be combined with internal knowledge in pursuit of breakthrough innovations requiring novel combinations (Ahuja & Lampert, 2001).

Two critical factors, however, can inhibit a firm’s ability to gain from external resources. First, firmspecific factors such as “competency traps” (Levitt & March, 1988), “core rigidities” (Leonard-Barton, 1992), and a “not-invented here” attitude (Katz & Allen, 1982) inhibit a firm’s willingness to reach out and integrate external knowledge. Second, the stronger the firm’s knowledge base, the greater the tendency to use safeguards against knowledge expropriation (Heiman & Nickerson, 2004) that can limit knowledge flow (Norman, 2002). Further, although internal capabilities are helpful in attracting external resources (Ahuja, 2000b), they may also make firms inwardly focused (Cohen & Levinthal, 1990). These potential benefits of and barriers to external resources would magnify in the context of breakthrough innovation because payoffs from success, risks from knowledge leakage, and organizational barriers (Ahuja & Lampert, 2001) are high in breakthrough innovations. We focus on this capability paradox.

THEORY AND HYPOTHESES The risk and uncertainty inherent in breakthrough innovation (Teece, 1996) pose several challenges to firms. These challenges are particularly high for firms in dynamic technological environments because their internal resources may be insufficient and even inappropriate for achieving breakthrough innovation, requiring them to acquire external resources and to combine a wide variety

We gratefully acknowledge the highly valuable comments and suggestions from Associate Editor Wenpin Tsai and three anonymous reviewers on the earlier versions of this article. Editor’s note: The manuscript for this article was accepted during the term of AMJ’s previous editor, Duane Ireland. 797

Copyright of the Academy of Management, all rights reserved. Contents may not be copied, emailed, posted to a listserv, or otherwise transmitted without the copyright holder’s express written permission. Users may print, download or email articles for individual use only.

798

Academy of Management Journal

of resources and capabilities in an ongoing manner (Ahuja & Lampert, 2001; Phene, FladmoeLindquist, & Marsh, 2006). Alliance portfolios provide firms access to a wide range of valuable resources from different partners, help in managing risk and uncertainty, and enable firms to uniquely benefit from the specific resource contributions of partners (Wassmer, 2010). Our alliance portfolio theory of breakthrough innovation is built on the basic tenets of financial portfolio theory (Markowitz, 1952): collection of assets, relationships among the assets, and diversification of such assets. Portfolios are important to understanding how economic actors facing environmental uncertainty manage various investments by considering the risk and return associated with each investment and its relationships with other investments. Financial portfolio theory provides an explicit means for understanding the relationships among all of the individual components of a system and the overall structure and “behavior” of that system (Chandra & Shadel, 2007). Thus, portfolio theory provides a useful lens through which to evaluate trade-offs based on the elements of a portfolio and their mix and to examine synergy among them. A superior portfolio with a diverse and unique mix of valuable resources helps a firm to cope with uncertainties and risks in innovation. Our notable departure from financial portfolio theory, however, is asserting the criticality of the focal firm—the portfolio holder—and how it interacts with the elements of the portfolio. A firm generates portfolio effects through the acquisition of portfolio resources, codevelopment with partners, and creation of interpartner synergy among the portfolio elements. The Paradox of Firm Capabilities A firm’s internal capabilities are important in generating breakthrough innovations (Ahuja & Katila, 2004; Ahuja & Lampert, 2001) and also play a critical role in leveraging portfolio resources. Internal technological capabilities are reflected in the firm’s repertoire of strong and diverse technological resources. Collaborations with partners with rich resources provide firms with access to valuable knowledge (Dyer & Singh, 1998) that they can combine with internal knowledge in pursuit of breakthrough innovations. Moreover, firms with stronger internal capabilities benefit more from external resources (Cohen & Levinthal, 1990). However, firm capabilities may become core rigidities (Leonard-Barton, 1992), contribute to the not-invented-here attitude (Katz & Allen, 1982), and lead to persistent use of inferior resources

August

and procedures (Levitt & March, 1988). These factors together create competency traps (Levitt & March, 1988) whereby the firm is unwilling and unable to reach out and integrate external knowledge. Competency traps have three types of underlying elements: cognitive, organizational, and behavioral. The cognitive element involves overvaluing internal capabilities, undervaluing external capabilities, and consequently having a dismissive attitude toward external resources and capabilities. Further, since a firm’s members understand internal capabilities better than external ones, the firm resorts to more familiar capabilities (Ahuja & Lampert, 2001), especially in an uncertain and complex environment. The organizational element involves internal routines that are well established, practiced rigorously, and therefore reinforced over time until they become difficult to change (Nelson & Winter, 1982). Such routines are impediments when firms need to explore and use external resources. The behavioral element is about risk taking. To a firm with strong internal capabilities, these familiar and proven capabilities appear less risky than external capabilities, which are less well understood, less controllable, and perhaps not directly and immediately applicable. In this context, an analogy based on prospect theory (Tversky & Kahneman, 1974) emerges. According to prospect theory, firms with stronger financial resources tend to show lower appetite for risk (Fiegenbaum & Thomas, 1988); this suggests that firms with stronger internal capabilities will also have lower risk-taking propensity. Thus, the cognitive element suggests a firm’s unwillingness, the organizational suggests its inability, and the behavioral suggests its avoidance of external resources and capabilities. The paradox of capabilities becomes evident when one considers portfolio and internal resources jointly. Two conflicting forces shape a firm’s willingness and ability to attract, acquire, and leverage portfolio resources: (1) need for both value creation and value protection and (2) what the firm can do versus what it actually does. Simultaneous concerns for value creation and value protection lead to a “value tension.” Although a firm’s strong internal capabilities make it a more attractive partner (Ahuja, 2000b) and enable it to bring in more relational resources for value creation, strong capabilities also raise value protection concerns for the firm. With these concerns, the firm makes extra efforts to protect its core resources by instituting safeguards and barriers (Heiman & Nickerson, 2004). Since reciprocity is expected in alliances, partners respond by instituting their own safeguards and barriers to protect their core resources.

2011

Srivastava and Gnyawali

With barriers erected on both sides and value tension intensified, resource flows are adversely affected (Heiman & Nickerson, 2004; Norman, 2002). The second tension, which we call “can-but-does not,” is between a firm’s ability and its actual behavior. Internal capabilities provide the ingredients necessary to help identify, comprehend, and apply external knowledge, but they are essentially just the potential. To leverage external resources, a firm needs both the willingness and the ability to notice, identify, comprehend, and use valuable knowledge (Arora & Gambardella, 1994; Cohen & Levinthal, 1990). When willingness and ability are conflicting, willingness becomes more critical than ability; in other words, what a firm does matters more than what it can do. If the firm does not value external knowledge, it is unlikely to make efforts to access and comprehend it. Before applying external knowledge, the firm needs organizational mechanisms for bringing in the knowledge, and internal acceptance of the knowledge. Only then can recombinatory ability effectively kick in and generate breakthrough innovations. Internal capabilities push the firm’s willingness and ability in opposite directions. High internal capability enhances the ability to identify, comprehend, and deploy external resources; at the same time, it makes the firm more inward looking and attenuates willingness to notice, identify, and deploy external resources. The capability paradox, which intensifies with increasing value tension and canbut-does-not tension, impedes the creation of portfolio effects. Portfolio technological resources that are rich, credible, and easily deployable are critical for generating innovations. Moreover, breakthrough innovation requires “bundling” heterogeneous resources (Patel & Pavitt, 1997). These critical considerations guided our choice of the characteristics of portfolio technological resources to study, and accordingly we selected quality and diversity as the key dimensions. Given the role of internal technological capability in generating innovations and in leveraging external resources, we used it as a moderator and selected two of its dimensions: technological strengths and technological diversity. Effects of Diversity of Technological Resources Diversity of portfolio resources refers to the mix or variety of technological resources held by portfolio partners. Breadth of portfolio resources helps a focal firm engage in a greater degree of exploration. Exposure to partners’ diverse technologies broadens the firm’s perspective and increases its ability to see fruitful opportunities that may arise

799

at the confluence of several technologies. Our focus is on portfolio-level diversity rather than dyad-level diversity (Sampson, 2007) and how the effects of such diversity on breakthrough innovation are contingent on a firm’s internal technological diversity. Diversity of internal technological resources refers to the breadth or range of technologies possessed by a focal firm (Patel & Pavitt, 1997). As internal diversity increases, the benefits of external diversity diminish, because the firm becomes less willing to leverage external diversity and faces significant challenges in pursuing both internal and external explorations. First, with its strong internal search routines, established technological trajectories (Cimoli & Dosi, 1995), and preference for using a limited set of resources and competences (Ahuja & Lampert, 2001), an internally diverse firm sees less value in external exploration and may even consider it more risky. High external and internal diversities together increase cognitive complexity, and simultaneous pursuit of diverse external and internal resources fragments innovation efforts. A preference for known internal resources, the perceived riskiness of external exploration, and the complexity of pursuing both types of exploration reduce the willingness of internally diverse firms to reach out and acquire external resources. Second, even if a firm exhibits willingness to engage in both explorations, it is extremely challenging for it to do so because of the different routines and organizational mechanisms needed for these explorations. As each firm has its own idiosyncratic search routines (Nelson & Winter, 1982), a technologically diverse portfolio implies diversity of search routines. An internally diverse firm that already has diverse internal search routines will struggle in creating effective interorganizational routines. Pursuing diverse and complex external technologies also requires open exchanges with partners, but the fear of unwanted knowledge exploitation by the partners limits openness and exploration. All these forces together simultaneously increase internal exploration and decrease external exploration and therefore limit the gains from external resources. On the contrary, internally less diverse firms face weaker internal exploration pressure because search options are few, and they face stronger pressure for external exploration because it has greater attractiveness for them. As a result, they exert much stronger efforts to engage in external exploration and institute mechanisms for identifying and assimilating external knowledge with greater urgency. Their extra drive for external exploration generates greater benefits from the di-

800

Academy of Management Journal

verse external resources and consequently contributes to more breakthrough innovations. Hypothesis 1. The greater the technological diversity of a focal firm, the lower the positive impact of portfolio resource diversity on its rate of breakthrough innovation. Effects of Quality of Technological Resources Quality refers to the reliability of portfolio partners’ technological resources. Portfolio resource quality is high when partners have high volume of resources and when other firms have lent credibility to those resources by building on them (Cattani, 2005). High-quality portfolio resources contribute to breakthrough innovation for various reasons. First, access to reliable and directly applicable resources alleviate a firm’s resource constraints (Ahuja & Lampert, 2001), allowing it to pursue breakthrough innovations that are resource intensive and risky. Access to quality resources changes risk perception (Sitkin & Weingart, 1995), and the ready applicability of partners’ resources reduces perceived risk and improves the probability of project success. Together these factors increase the focal firm’s motivation to engage in breakthrough projects. Second, time is a critical factor in determining whether or not an important innovation becomes a breakthrough one. When relational resources are reliable and useful, a firm saves time that may otherwise go into either refining and improving these resources or internally developing them. Finally, access to high-quality resources and the possibility of working together with well-endowed partners increase the use of joint problemsolving arrangements (McEvily & Marcus, 2005) for breakthrough projects and diminish the necessity of changing internal routines and practices (Dougherty, 1996). We focus on how the effects of portfolio resource quality on breakthrough innovation are contingent upon a firm’s technological strength. Technological strength refers to the internal resource competence of a focal firm relative to that of other firms in its industry. A firm’s technological strength is high if it has traditionally generated more technological innovations than others. The capability paradox becomes critical when a technologically strong firm engages in alliances with other strong firms. First, increase in internal capability intensifies value tension as the firm tries simultaneously to leverage portfolio resources and to institute strong mechanisms to protect its proprietary resources. Out of fear of knowledge expropriation, the firm devises governance structures that increase control (Heiman & Nickerson, 2004) or re-

August

duce alliance scope (Oxley & Sampson, 2004). As the firm approaches alliances relationships with a protective mind-set, its partners are likely to behave the same way, thus limiting effective flow and integration of partners’ knowledge. These risk-mitigating strategies and suspicions inhibit breakthrough innovation, which requires openness and an exploratory mind-set (Fleming, 2001). Second, competency traps are a critical force for technologically strong firms. Pursuit of external resources involves great risk. Drawing on the analogy from prospect theory (Tversky & Kahneman, 1974), we argue that technologically weaker firms tend to take more risk and put forth more intense effort to acquire and use external resources than technologically stronger firms (Fiegenbaum & Thomas, 1988). The technologically weaker firms also have fewer concerns about the potential loss of proprietary knowledge. On the other hand, stronger firms will avoid going out of their “comfort zones” and eschew risky opportunities to leverage portfolio resources. Also, stronger firms will have a dismissive attitude toward external resources, as they have become inwardly focused (Cohen & Levinthal, 1990) and prefer to stay on their own established technological trajectories (Cimoli & Dosi, 1995). Additionally, stronger firms have limited motivation to engage in external search for knowledge and other resources (Ahuja, 2000b), because they often innovate by combining their own internal ideas and technologies in new ways (Katila, 2002). Thus, cognitive and behavioral factors coupled with limited motivation reinforce competency traps. Although these firms have the capability to benefit from portfolio resources, competency traps and their strong focus on knowledge protection impede realizing that capability. The behavioral force prevails over the opposing capability force. When a firm is unwilling to identify and leverage external knowledge, the potential of internal capability remains underutilized. Hypothesis 2. The greater the technological strength of a focal firm, the lower the positive impact of portfolio resource quality on its rate of breakthrough innovation. Effects of Portfolio Resource Leveraging The competency traps of technologically strong firms and their muted motivation keep them from benefiting from alliance portfolio resources. However, when stronger firms can overcome these impediments, they will benefit more from the quality of portfolio resources than technologically weaker

2011

Srivastava and Gnyawali

firms will. When stronger firms are able to overcome the motivational and attitudinal barriers with some selected partners, they are likely to work more effectively with them. Use of some partners’ technological resources by a focal firm implies that it has been able to partly overcome its motivational and attitudinal barriers and is already capable of acquiring and leveraging external resources. We name this capability portfolio leverage. The intensity of portfolio leverage increases as the focal firm builds upon the resources of a greater proportion of its portfolio partners. Actual building on portfolio resources means that firm members have cognitively realized the importance or benefits of partners’ resources and overcome prohibitive routines and processes to reach out, and thus the firm has become less inwardly focused (Cohen & Levinthal, 1990). When the firm values some partners’ resources, it will intensely channel its internal strengths so that it can successfully acquire and effectively leverage them. The firm can create portfolio effects both by enriching its internal resources with the high-quality external resources and by combining these resources. It can also work together with the selected partners to jointly pursue risky and resource-intensive projects and therefore improve the outcomes of such projects (Rosenkopf & Nerkar, 2001). Because strong firms are attractive partners (Ahuja, 2000b) and have the necessary internal capability to acquire and assimilate external resources (Cohen & Levinthal, 1990), they act like “hawks”: they concentrate, acquire, and leverage the identified portfolio resources. Increased motivation to build on external knowledge and reduction of competency traps better align behavioral and capability forces and release the can-but-does-not tension. The firm becomes able to mobilize its already strong capabilities to actually benefit from its partners’ resources. Thus, compared with technologically weaker firms, stronger firms, with their already greater potential to create portfolio effects, combined with increasing leverage intensity, will generate more breakthrough innovations. Hypothesis 3. The greater the technological strength of a focal firm, the higher the positive impact of portfolio leverage intensity on its rate of breakthrough innovation. METHODS We tested our hypotheses using longitudinal (1986 –2000) data from publicly traded firms in the U.S. semiconductor industry (SIC code 3674), an appropriate context given the strong patenting ten-

801

dencies of firms and prevalence of strategic alliances in the industry (Hall & Ziedonis, 2001). We started off with all 258 publicly traded dedicated semiconductor firms listed in the Compustat database. To ensure that our sample consisted of predominantly semiconductor firms, we excluded diversified firms such as Toshiba, Motorola, IBM, NEC, and Samsung. We also dropped firms that did not exist as public firms or appeared for one year only during the observation period. The final estimation sample consisted of 180 firms, including 24 foreign firms whose shares were listed on NYSE or NASDAQ. Alliance Portfolio Data We collected alliance portfolio data of semiconductor firms using the SDC database, which is the most comprehensive alliance database available. Using this database, we identified 900 unique partner firms for our 180 sample firms. However, this database includes information based on alliance announcement rather than on actual formation, and data on items such as duration of alliances are missing (Phelps, 2003). We therefore used Factiva to trace and verify the existence of the announced alliances. Following Schilling and Phelps (2007), we took a conservative approach and assumed alliance duration of three years but also did sensitivity analysis using four and five years. Following Ahuja (2000a), we converted multilateral alliances into a set of bilateral alliances. However, we distinguished these converted bilateral alliances from the true ones, and assigned different alliance weights, as described later. Patent Data We used the updated NBER patent database (Hall, Jaffe, & Trajtenberg, 2001) to construct the patent data of the focal firms and their portfolios of alliances. This database contains patent number, assignee name, assignee number, filing year, and grant year but does not contain the CUSIP numbers (used in the Compustat database) of the assignee firms. We therefore also used the name-matching database, a bridge between the NBER patent and Compustat databases provided by Hall and colleagues (2001), in which the names of assignee firms are “standardized” and matched with the names in the Compustat database. We refer to this as the “coname database.” However, we could find patent information for only 210 partner firms out of 900 firms. We could not determine whether the rest of the firms had no patents or had patents but did not appear in the coname database. We pursued

802

Academy of Management Journal

two steps to overcome this limitation. First, we collected patent data on additional 151 firms using the Google patent database and cross-checked patent information for other firms using both the Google and NBER patent databases. Next, we wrote a name-matching program and used it to get patent information for additional 98 firms. With these procedures, we gathered patent data for a total of 459 partner firms and ensured that the rest of the firms had indeed no patents. We used the same procedure to obtain patent information for the 180 sample firms, of which 30 had zero patents. Measures Dependent variable: Rate of breakthrough innovation. Patents are often used to measure the outcome of a firm’s innovation activities (Ahuja, 2000a). A patent for an invention is the grant of property rights to the inventor issued by the United States Patents and Trademarks Office (USPTO). Patents are granted for both products and processes. As patents go through a rigorous examination process before they are awarded and confer certain property rights on the assignee, they have external validity and economic significance. Patent citations are important as they acknowledge the “prior art” that has contributed to a new patent and define the boundaries of property rights. Hall and colleagues (Hall, Jaffe, & Trajtenberg, 2005) showed that a firm’s market value increased by 3 percent for every patent citation it received. Therefore, citation is a credible measure of a patent’s quality and impact (Cattani, 2005). Following prior research (Ahuja & Lampert, 2001; Hall et al., 2001; Phene et al., 2006), we measured a firm’s rate of breakthrough innovation as the number of highly cited or impactful patents it was granted per year. Because the number of forward citations (citations by subsequent patents of a given patent) considerably varies by technological subcategories (Teece, 1996) and by the duration of time lapsed after a patent is granted (Hall et al., 2001), we followed the fixed-effects approach suggested by Hall and colleagues (2001). Accordingly, we first scaled the forward citations’ counts of patents by dividing the counts by the mean value of citations based on all patents in their technological subcategory granted in the same year in the NBER database. By scaling citation counts across technological subcategories and grant years, we made these patents, to a certain extent, insensitive to their technological subcategory and the number of years they had been out. This approach allowed us to identify the leading patents within and across technological categories after removing the year ef-

August

fects (including truncation effects and systematic changes in citation tendencies over time) and yeartechnology field interaction effects (Hall et al., 2001). After the scaling, we selected the top 3 percent (1,497 patents) out of the 49,900 patents of all the sample firms during the observation period as breakthrough patents (Phene et al., 2006). The number of patents of a focal firm appearing in the list of 1,497 breakthrough patents in a given year was used as an indicator of the rate of breakthrough innovation for that year for the firm. Independent variables. We used all patents of all partners of a focal firm to measure the underlying technological resources in the firm’s portfolio. We measured portfolio resource quality as the sum of the citation-weighted count of all alliance partners’ patents successfully applied for in a given year. Many alliances involved multiple value chain activities (R&D, manufacturing, marketing, etc.) and contained bilateral and multilateral ties. We therefore used two considerations in computing portfolio resource quality. First, following Contractor and Lorange (1988) and Nohria and Garcia-Pont (1991), we used a weighting scheme to assign higher weights to alliances that contained a “high intensity” of relationships. We slightly modified the weighting scheme to account for alliance participation in multiple value chain activities and assigned higher weights to areas that are more important in the semiconductor industry (such as R&D and manufacturing). Second, we weighted real dyadic alliances higher than multilateral alliances that were converted to dyadic alliances (as noted earlier), using the following computation: alliance weight (ALw) ⫽ (4 ⫻ R&D ⫹ 3 ⫻ manufacturing ⫹ 2 ⫻ cross-licensing ⫹ licensing ⫹ marketing ⫹ technology transfer) ⫻ partner weight, where partner weight (ⱖ.25, ⱕ1.0) factored in the number of partners. If a multilateral alliance had three partners (apart from the focal firm), each partner received a weight of 0.70 (instead of the weight of 1.0 it would have received if these were treated simply as three bilateral alliances with the focal firm). We calculated the portfolio resource quality of each focal firm i in year t using the following formula: Portfolio resource quality it

冘 ALw 冘 Cw ⫻ p , n

⫽␥ ⫻

m

ijt

j⫽1

k

kt

k⫽1

where ALw is the weight of the alliance with partner j, n is the total number of partners of the focal firm in year t, Cwk is the citation weight of the pkth patent of partner j, m is the total number of patents

2011

Srivastava and Gnyawali

successfully filed by j in t, pk is one of those m patents, and ␥ is a scaling constant. The focal firm’s citation of its partners’ patents is indicative of the firm’s building upon its partners’ technologies. We defined portfolio leverage intensity as the extent to which a firm leveraged its partners’ resources, measuring this variable as the proportion of portfolio quality resources that a firm leveraged. Thus, portfolio leverage intensity varied from 0 to 1. If a focal firm did not cite any of its alliance partners’ patents, its portfolio leverage intensity was 0; if it cited patents of all of its portfolio partners, portfolio leverage intensity was 1. For computing the variable portfolio resource leverage intensity, we introduced lever weight as a multiplier. We assigned the lever weight a value of 0 if the focal firm did not cite and a value of 1 if it cited its alliance partner’s patents during the window t – 3 to t ⫹ 3 around alliance year t and computed the variable using the following formula: Portfolio leverage intensityit ⫽



冘 Lw n

j⫽1

冘 Cw ⫻ p m

ijt

⫻ ALw ijt

k

kt

k⫽1

册冒

冘 ALw 冘 Cw ⫻ p . n

m

ijt

j⫽1

k

kt

k⫽1

Lwij is the lever weight for the partner j. Lwij is 0 when focal firm i does not cite partner j’s patents during the t – 3 to t ⫹ 3 window, and is 1 when the firm does cite partner j’s patents. We measured portfolio resource diversity (H) using the Herfindahl-Hirschman index as applied by Lin, Chen, and Wu (2006): H ⫽ 公1 ⫺ 兺i37⫽ 1(Xi/兺Xi)2, where Xi is the number of patents of the alliance portfolio in the technological subcategory i. There were 37 technological subcategories across which alliance portfolio patents of the sample firms were distributed. The value of H resulted in a missing value for observations when the sum of alliance portfolio patents (兺Xi) was zero. Following the procedure suggested by Cameron and Trivedi (1998), we imputed a low value for H (⫽ .05) for such observations. Moderator variables. Given the importance of patenting in the industry, a firm’s technological assets and capabilities (e.g., scientists, engineers, technological infrastructure, and intellectual capabilities) will be manifested in its patenting profile. We chose to focus on diversity of internal technological resources and internal technological strength in order to match them with the portfolio technological resources of interest to us. Internal

803

technological diversity was measured using the range of focal firm’s patents across technological subcategories (Miller, Fern, & Cardinal, 2007) and was computed using the Herfindahl-Hirschman index as described earlier. The technological strength of the focal firm was measured as the firm’s share of patents among all the patents of the sample firms in the industry in a given year. Control variables. The presence of unobservable permanent heterogeneity in innovation models may produce biased estimates (Blundell, Griffith, & Vanreenen, 1995). We followed Schilling and Phelps (2007) and controlled for the knowledge stock of focal firms that may account for permanent unobserved heterogeneity among the sample firms. We included the variable presample patents (Ahuja & Katila, 2001) as the sum of patents obtained by the focal firm during the five years prior to entering the sample. Since R&D expenditure importantly impacts a firm’s innovative output (Ahuja & Lampert, 2001), we controlled for the logged value of R&D. We also controlled for firm age, measuring it as the difference between the year of observation and the year of firm incorporation; firm size, using the log of sales; and number of assignees (Damanpour, 1991; Miller et al., 2007). We also controlled for type of firm (such as integrated manufacturing, “fabless,” and foundries) as these types may differ in their patenting strategies. Statistical Analysis Since our dependent variable, the number of breakthrough innovations, is a count variable, has skewed distribution owing to the large number of zeros, and suffers from overdispersion, we used negative binomial regression (Long & Freese, 2006) for our analysis. The Hausman test we conducted suggested that it was safe to use the random-effects model, but we also computed the fixed-effects model and confirmed that the results were very similar. Robustness Checks We took several steps to ensure that our findings were robust. We performed a sensitivity analysis using top the 1 percent (499), 2 percent (998), and 4 percent (1996) of patents as breakthrough patents. Because the number of claims shows the ex ante importance of patents in view of the assignee and is an important indicator of patent quality (Lanjouw & Schankeman, 2004), we used it as an alternative measure of breakthrough innovations. We used alternative estimation models in terms of Poisson regression and fixed-effect negative binomial re-

804

Academy of Management Journal

gression. We also ran alternative models without the presample patent control variable to ensure that the correlation between presample patent and technological strength was not influencing our results. Because our models were nonlinear, we followed Corredoira and Rosenkopf (2010) and converted the estimation model into a log-linear form to confirm the significance of the interaction effects. All these analyses produced similar and consistent results, thus lending credence to our findings. RESULTS Table 1 provides descriptive statistics and correlations. To ensure that collinearity among variables was not hurting our estimations, we followed the multicollinearity test procedure described in Belsley, Kuh, and Welsch (1980) and used the “coldiag” procedure in Stata (Wang & Zajac, 2007). The condition number for all the control and independent variables was 12.77. Although this value was above 10 (a desirable condition number to conclude absence of multicollinearity), it was well below 30 (an indication of a severe multicollinearity problem). Further investigation suggested that the higher condition number was largely due to high correlation between sales and R&D. High correlation between control variables does not influence estimates of the independent variables (Wooldridge, 2009: 98). The condition number without R&D was 6.12 and without the sales was 8.11, both well below the desirable number 10. Table 2 contains results of the negative binomial regression random-effects model. Model 1 displays the effects of control variables only. Models 2 and 3 respectively include diversity and quality of portfolio resources. Model 4 shows the interaction effects of portfolio diversity and internal diversity, and model 5 shows the interaction effects of portfolio quality and internal strength. Model 6 includes portfolio leverage intensity, and model 7 shows the interaction effect of portfolio leverage intensity and internal strength and is also the full model. We subjected all these models to stronger likelihood-ratio tests showing the significance of various models (Gould, Pitblado, & Sribney, 2006) as compared to different base models. The beta coefficients of portfolio diversity and quality are both positive and significant, as shown in models 1 and 4, suggesting that these portfolio resources do contribute to breakthrough innovation. The coefficient of the interaction term between portfolio diversity and internal diversity in model 4 is negative and statistically significant (p ⬍ .05), thus providing support for Hypothesis 1.

August

Hypothesis 2 predicts that the rate of benefits from portfolio quality will be lower for internally strong firms. Model 5 shows that the coefficient of the interaction term between portfolio quality and internal strength is negative and significant (p ⬍ .05), thus lending support for the hypothesis. In Hypothesis 3, we posit that the impact of portfolio leverage intensity is higher for firms with high internal strength. The coefficient of the interaction term between portfolio leverage intensity and internal strength in model 7 is positive and significant (p ⬍ .05). Results thus support Hypothesis 3 as well. Because portfolio leverage intensity is based on a subset of portfolio quality resources, we also tested the effect of portfolio leverage intensity without the portfolio quality variable in the model, and the results were stronger. Post Hoc Analysis To explore the relative importance of portfolio and internal resources and the possibility of complementarity and substitution effects, we conducted an additional analysis (not reported here). Following the approach of Whittington, OwenSmith, and Powell (2009), we removed both internal strength and internal diversity from the control model and examined changes in the coefficients of the variables by entering the internal and portfolio variables sequentially and in the reverse order. This analysis suggested that both internal and portfolio diversities are important, but internal diversity is more important than portfolio diversity. Effects of both of are positive, but their individual effects become weaker in the presence of the other, suggesting a partial substitution effect. Engagement in simultaneous internal and external explorations, though attractive, is taxing for firms, results in a modestly negative synergy. Further, portfolio quality is more important than internal technological strength. More critically, the effectiveness of internal technological strength increases in the presence of portfolio quality, thereby indicating their complementarity effects as well. DISCUSSION AND IMPLICATIONS Our study demonstrates that technologically weaker firms benefit at a higher rate from increasingly knowledge-rich portfolio resources, and technologically less diverse firms benefit at a higher rate from increasingly knowledge-diverse portfolio resources. We found a positive synergy between portfolio resources and internal resources for technologically weaker firms but a negative synergy for technologically stronger firms. Further, stronger

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16.

Mean

s.d.

Min.

Max.

Rate of breakthrough patents 1.1 5.61 0 10.2 Log of sales(t ⫺ 1) 4.17 1.95 ⫺3 10.1 Pre sample patents 24.71 116.3 0 1,018 Firm age(t ⫺ 1) 15.41 11.6 0 69 Log of R&D(t ⫺ 1) 2.03 1.9 ⫺5.1 7.98 Number of assignees 1.22 0.76 0 7 Foreign 0.08 0.26 0 1 Fabless 0.32 0.46 0 1 Tester 0.02 0.14 0 1 Foundry 0.13 0.34 0 1 Discrete 0.01 0.08 0 1 Technological strength(t ⫺ 1) 0.01 0.04 0 0.43 Internal tech diversity(t ⫺ 1) 0.4 0.38 0 0.95 Portfolio tech resource diversity(t ⫺ 1) 0.34 0.4 0 0.97 Portfolio tech resource quality(t ⫺ 1) 1.7 7.22 0 90.3 Portfolio leverage intensity(t ⫺ 1) 0.14 0.33 0 1

Variable

.34 .32 .11 .32 ⫺.02 ⫺.02 ⫺.12 ⫺.03 .13 ⫺.02 .55 .23 .25 .46 .22

1

.40 .31 .83 .21 .24 ⫺.31 .02 ⫺.02 ⫺.08 .48 .58 .42 .37 .33

2

.50 .40 ⫺.01 ⫺.03 ⫺.14 ⫺.03 ⫺.07 ⫺.02 .82 .26 .23 .57 .14

3

.13 .14 ⫺.14 ⫺.38 .17 .13 ⫺.03 .37 .10 .05 .30 .05

4

.21 .18 ⫺.09 ⫺.07 ⫺.16 ⫺.18 .48 .65 .46 .37 .35

5

TABLE 1 Descriptive Statistics and Correlations

.20 ⫺.13 ⫺.06 ⫺.07 ⫺.13 .00 .14 .04 .00 .04

6

⫺.05 .05 ⫺.03 ⫺.02 ⫺.01 .04 .02 ⫺.05 .10

7

⫺.09 ⫺.27 ⫺.06 ⫺.16 ⫺.12 ⫺.10 ⫺.12 ⫺.05

8

⫺.05 ⫺.01 ⫺.04 ⫺.03 ⫺.06 ⫺.03 ⫺.02

9

⫺.03 .02 ⫺.07 .03 ⫺.04 .05

10

⫺.02 ⫺.07 ⫺.06 ⫺.02 ⫺.04

11

.34 .30 .60 .21

12

.43 .26 .33

13

.35 .60

14

.28

15

Standard errors are in parentheses. † p ⬍ .10 * p ⬍ .05 ** p ⬍ .01 *** p ⬍ .001

a

10.19***

⫺3.82*** (0.60) 1,203 180 ⫺697.02 13 130.28***

0.22 (0.15) 0.00** (0.00) ⫺0.05*** (0.01) ⫺0.01 (0.15) 0.14 (0.15) ⫺0.65 (0.50) ⫺0.77* (0.32) ⫺0.40 (1.16) ⫺0.14 (0.35) ⫺14.34 (3,354) 2.16† (1.13) 2.56*** (0.49) 0.67** (0.22)

0.26† (0.15) 0.00** (0.00) ⫺0.05*** (0.01) 0.01 (0.15) 0.15 (0.16) ⫺0.64 (0.51) ⫺0.77* (0.32) ⫺0.45 (1.17) 0.02 (0.35) ⫺17.74 (18,237) 1.62 (1.13) 2.70*** (0.49)

Log of sales(t ⫺ 1) Pre sample patents Firm age(t ⫺ 1) Log of R&D(t ⫺ 1) Number of assignees Foreign Fabless Tester Foundry Discrete Technological strength(t ⫺ 1) Internal tech. diversity(t ⫺ 1) Portfolio tech. resource diversity(t ⫺ 1) Portfolio tech. resource quality(t ⫺ 1) Portfolio tech. resource diversity(t ⫺ 1) ⫻ internal tech. diversity(t ⫺ 1) Portfolio tech. resource quality(t ⫺ 1) ⫻ technological strength(t ⫺ 1) Portfolio leverage intensity(t ⫺ 1) Portfolio leverage intensity(t ⫺ 1) ⫻ technological strength(t ⫺ 1) Constant Number of firm years Number of firms Log-likelihood Degrees of freedom Wald chi-square Likelihood-ratio test Base (model 1) Base (model 2) Base (model 3) Base (model 4) Base (model 5) Base (model 6) ⫺3.77*** 1,203 180 ⫺702.12 12 116.07***

Model 2

Model 1

Variables

39.79*** 29.61***

⫺3.47*** (0.57) 1,203 180 ⫺682.22 14 192.78***

0.13 (0.15) 0.00* (0.00) ⫺0.06*** (0.01) 0.13 (0.15) 0.16 (0.16) ⫺0.67 (0.50) ⫺0.85** (0.31) ⫺0.08 (1.15) ⫺0.11 (0.36) ⫺19.45 (49,620) 2.94* (1.28) 2.44*** (0.48) 0.49* (0.22) 0.02*** (0.00)

Model 3

44.65*** 34.46*** 4.85*

⫺4.07*** (0.68) 1,203 180 ⫺679.79 15 180.97***

0.14 (0.15) 0.00* (0.00) ⫺0.05*** (0.01) 0.11 (0.15) 0.15 (0.16) ⫺0.70 (0.50) ⫺0.86** (0.32) ⫺0.11 (1.16) ⫺0.13 (0.36) ⫺16.83 (17,282) 2.92* (1.30) 2.96*** (0.58) 1.58** (0.54) 0.02*** (0.00) ⫺2.00* (0.92)

Model 4

(0.03)

48.94*** 38.76*** 9.15* 4.3*

⫺4.11*** (0.68) 1,203 180 ⫺677.65 16 187.76***

⫺0.07*

0.15 (0.15) 0.00** (0.00) ⫺0.05*** (0.01) 0.10 (0.15) 0.18 (0.16) ⫺0.67 (0.50) ⫺0.82** (0.32) ⫺0.06 (1.16) ⫺0.05 (0.36) ⫺13.56 (3,368) 2.98** (1.15) 2.95*** (0.57) 1.54** (0.54) 0.03*** (0.01) ⫺2.01* (0.91)

Model 5

TABLE 2 Negative Binomial Regression Analysis Using Random Effect for Breakthrough Innovationa

(0.20)

(0.03)

51.79*** 41.6*** 11.99** 7.14* 2.84†

⫺3.90*** (0.70) 1,203 180 ⫺676.22 17 189.68***

0.35



⫺0.07*

0.14 (0.15) 0.00** (0.00) ⫺0.06*** (0.01) 0.08 (0.15) 0.19 (0.16) ⫺0.79 (0.50) ⫺0.91** (0.32) ⫺0.07 (1.17) ⫺0.06 (0.36) ⫺15.35 (7,739) 2.78* (1.10) 2.97*** (0.57) 1.42** (0.55) 0.03*** (0.01) ⫺1.98* (0.91)

Model 6

(0.24) (2.05)

55.55*** 45.37*** 15.76** 10.91* 6.61* 3.77†

⫺3.87*** (0.70) 1,203 180 ⫺674.34 18 195.91***

0.12 4.10*

⫺0.10** (0.04)

0.13 (0.15) 0.00*** (0.00) ⫺0.07*** (0.01) 0.06 (0.15) 0.20 (0.16) ⫺0.72 (0.50) ⫺0.91** (0.32) ⫺0.02 (1.16) ⫺0.20 (0.36) ⫺14.65 (5,371) 2.37* (1.07) 3.12*** (0.59) 1.55** (0.56) 0.04*** (0.01) ⫺2.15* (0.93)

Model 7

2011

Srivastava and Gnyawali

firms suffer more from competency traps. But once they overcome their behavioral inhibitors, they do indeed create a positive synergy, despite the continued presence of the value tension. This also implies that can-but-does-not tension is more critical than value tension. Furthermore, firms weak on the diversity dimension of technological capability can overcome that by configuring knowledgediverse portfolios. We also found that portfolio quality is more important than internal technological strength, whereas internal diversity is more important than portfolio diversity. More broadly, our results suggest that internal capabilities of firms influence their behavior in ways that diminish stronger firms’ likelihood of benefiting from increasing portfolio resources. Our primary theoretical contribution lies in the development of an alliance portfolio theory of breakthrough innovation. Resource-rich firms have the potential to create greater portfolio effects, yet competency traps and concerns about unwanted leakage of knowledge internally inhibit such effects. Simultaneous consideration of the characteristics of a firm’s portfolio and internal resources, along with underlying behavioral and capability forces, help to reveal the capability paradox. In the conflict between the capability and behavioral forces, the behavioral force prevails and becomes the bottleneck. When the firm’s behavior is misaligned with its capability, the potential of internal capability to leverage external knowledge remains largely unutilized. However, firms that can overcome the behavioral inhibitors and channel their strong internal capabilities to leverage portfolio resources would benefit more from their alliance portfolio. We contribute to the alliance literature by providing empirical evidence that knowledge-rich and knowledge-diverse alliance portfolios enhance breakthrough innovation by enabling firms to overcome some of the inherent barriers. Our theory and findings together suggest that internal technological strength on its own fails to overcome breakthrough innovation barriers. Firms, can, however overcome these barriers by leveraging portfolio resources. More importantly, the effectiveness of internal technological strength increases in the presence of portfolio quality resources. We thus empirically demonstrate that portfolio effects are created directly through portfolio resources and indirectly through enrichment of internal resources. Our study illustrates how competency traps (Leonard-Barton, 1992; Levitt & March, 1988) hinder leveraging of portfolio resources and extends their importance in the interorganizational context. We bridge the gap between absorptive capacity and

807

competency trap literatures by demonstrating how high internal capability may make a firm more inward looking. We thus underscore the need to examine firm behavior together with the potential provided by their resources. Taken together, these findings illustrate the negative aspects of resources not explicitly addressed by the resource-based view and bridge research on competency traps and the resource-based view. Owning and having access to valuable resources provide the potential, but behavioral forces need to be aligned with capability forces for firms to realize that potential, thus suggesting that firm behavior is an important link between resources and outcomes. Further, incorporation of firm behavior and managerial role would be an important condition for realizing competitive advantages from the resources. We also contribute to the advancement of an emerging relational resource-based theory (Ahuja, 2000a, 2000b; Dyer & Singh, 1998; Gulati, 2007; Lavie, 2006; Tsai & Ghoshal, 1998) by underscoring the importance of focusing on characteristics of relational resources along with a focal firm’s resources, by explicating how portfolio effects may be created, and by explaining how internal capability plays a critical role in facilitating or inhibiting the creation of the portfolio effects. Interesting insights regarding the paradox of capabilities emerge when we examine the underlying tensions that build up when a firm with strong capabilities is embedded in a resource-rich relational neighborhood and how these tensions influence the interplay of internal and relational resources and their effects. Some important limitations of this study also illustrate directions for future research. First, owing to data limitations, we could not directly measure the cognitive and behavioral forces that underlie competency traps. Also, a possible alternative explanation of the results could be that firms strong in internal resources do not need external resources. However, our results demonstrate that internal strength is not sufficient, portfolio resources have direct effects, and internal resources become more effective in the presence of external ones. Future research could tease out competency trap and complementarity effects to further advance understanding of the role of internal capabilities. Second, we were unable to account for the extent of conflict in portfolios, which might impact the mobilization of portfolio resources. An important way to assess tension in a portfolio would be as the degree of competitive tension (Chen, Su, & Tsai, 2007) among the portfolio partners and between the focal firm and its partners. Future research could examine the moderating impact of portfolio con-

808

Academy of Management Journal

flict. Third, in our measure of breakthrough innovation, we could not take into account the technological trajectories involved. It is likely that the paradox of capabilities would be stronger for firms pursuing breakthrough innovations along different technological trajectories. Future research could build on our initial work on the paradox of capabilities and examine how the implications would differ for firms facing different technological trajectories and shocks. In conclusion, we contribute to the literature by demonstrating the effects of relational resources on breakthrough innovation under varying internal resource conditions, by illustrating potential complementarity between internal and relational resources, and by articulating how and why simultaneous considerations of internal and relational resources lead to interesting insights about firm behavior and outcomes. Our examination of the paradox of capabilities provides a strong foundation for subsequent research in this intriguing area.

REFERENCES Ahuja, G. 2000a. Collaboration networks, structural holes, and innovation: A longitudinal study. Administrative Science Quarterly, 45: 425– 455. Ahuja, G. 2000b. The duality of collaboration: Inducements and opportunities in the formation of interfirm linkages. Strategic Management Journal, 21: 317–343. Ahuja, G., & Katila, R. 2001. Technological acquisitions and the innovation performance of acquiring firms: A longitudinal study. Strategic Management Journal, 22: 197–220. Ahuja, G., & Katila, R. 2004. Where do resources come from? The role of idiosyncratic situations. Strategic Management Journal, 25: 887–907. Ahuja, G., & Lampert, C. M. 2001. Entrepreneurship in the large corporation: A longitudinal study of how established firms create breakthrough inventions. Strategic Management Journal, 22: 521–543. Arora, A., & Gambardella, A. 1994. Evaluating technological information and utilizing it—Scientific knowledge, technological capability, and external linkages in biotechnology. Journal of Economic Behavior & Organization, 24: 91–114.

August

count data models of technological innovation. Economic Journal, 105: 333–344. Cameron, A. C., & Trivedi, P. K. 1998. Regression analysis of count data. Cambridge, U.K.: Cambridge University Press. Cattani, G. 2005. Preadaptation, firm heterogeneity, and technological performance: A study on the evolution of fiber optics, 1970 –1995. Organization Science, 16: 563. Chandra, S., & Shadel, W. G. 2007. Crossing disciplinary boundaries: Applying financial portfolio theory to model the organization of the self-concept. Journal of Research in Personality, 41: 346 – 373. Chen, M., Su, K., & Tsai, W. 2007. Competitive tension: The awareness-motivation-capability perspective. Academy of Management Journal, 50: 101–118. Cimoli, M., & Dosi, G. 1995. Technological paradigms, patterns of learning and development: An introductory roadmap. Journal of Evolutionary Economics, 5(3): 243–268. Cohen, W. M., & Levinthal, D. A. 1990. Absorptive-capacity—A new perspective on learning and innovation. Administrative Science Quarterly, 35: 128 – 152. Contractor, F. J., & Lorange, P. 1988. Why should firms cooperate? The strategy and economics basis for cooperative ventures. In F. J. Contractor & P. Lorange (Eds.), Cooperative strategies in international business: 1–30. Lexington, MA: Lexington Books. Corredoira, R. A., & Rosenkopf, L. 2010. Should auld acquaintance be forgot? The reverse transfer of knowledge through mobility ties. Strategic Management Journal, 31: 159 –181. Damanpour, F. 1991. Organizational innovation—A meta-analysis of effects of determinants and moderators. Academy of Management Journal, 34: 555– 590. Dougherty, D. 1996. Organizing for innovation. London: Sage. Dyer, J. H., & Singh, H. 1998. The relational view: Cooperative strategy and sources of interorganizational competitive advantage. Academy of Management Review, 23: 660 – 679. Fiegenbaum, A., & Thomas, H. 1988. Attitudes toward risk and the risk-return paradox: Prospect theory explanations. Academy of Management Journal, 31: 85–106.

Barney, J. B. 2001. Is the resource-based “view” a useful perspective for strategic management research? Yes. Academy of Management Review, 26: 41–56.

Fleming, L. 2001. Recombinant uncertainty in technological search. Management Science, 47: 117–132.

Belsley, D. A., Kuh, E., & Welsch, R. 1980. Regression diagnostics: Identifying influential data and sources of collinearity. New York: Wiley.

Gould, W., Pitblado, J., & Sribney, W. 2006. Maximum likelihood estimation with Stata: College Station, TX: Stata Press.

Blundell, R., Griffith, R., & Vanreenen, J. 1995. Dynamic

Gulati, R. 2007. Managing network resources: Alli-

2011

Srivastava and Gnyawali

809

ances, affiliations, and other relational assets: New York: Oxford University Press.

diversified firms. Academy of Management Journal 50: 308 –326.

Hall, B. H., Jaffe, A. B., & Trajtenberg, M. 2001. The NBER patent citation data file: Lessons, insights and methodological tools (NBER working paper series). Cambridge, MA: National Bureau of Economic Research.

Nelson, R. R., & Winter, S. G. 1982. An evolutionary theory of economic change. Cambridge, MA: Harvard University Press.

Hall, B. H., Jaffe, A. B., & Trajtenberg, M. 2005. Market value and patent citations. RAND Journal of Economics, 36: 16 –38. Hall, B. H., & Ziedonis, R. H. 2001. The patent paradox revisited: An empirical study of patenting in the US semiconductor industry, 1979 –1995. RAND Journal of Economics, 32: 101–128. Heiman, B., & Nickerson, J. 2004. Empirical evidence regarding the tension between knowledge sharing and knowledge expropriation in collaborations. Managerial and Decision Economics, 25: 401– 420. Katila, R. 2002. New product search over time: Past ideas in their prime. Academy of Management Journal, 45: 995–1010. Katz, R., & Allen, T. 1982. Investigating the not invented here (NIH) syndrome: A look at the performance, tenure, and communication patterns of 50 R&D project groups. R&D Management, 12: 7–19. Lanjouw, J. O., & Schankeman, M. 2004. Patent quality and research productivity: Measuring innovation with multiple indicators. Economic Journal, 114: 441– 465. Lavie, D. 2006. The competitive advantage of interconnected firms: An extension of the resource-based view. Academy of Management Review, 31: 638 – 658. Leonard-Barton, D. 1992. Core capabilities and core rigidities—A paradox in managing new product development. Strategic Management Journal, 13: 111– 125. Levitt, B., & March, J. 1988. Organizational learning. In W. R. Scott, J. Blake, & G. H. Elder Jr. (Eds.), Annual review of sociology, vol. 14: 319 –338. Palo Alto, CA: Annual Reviews. Lin, B. W., Chen, C. J., & Wu, H. L. 2006. Patent portfolio diversity, technology strategy, and firm value. IEEE Transactions on Engineering Management, 53: 17–26. Long, J. S., & Freese, J. 2006. Regression models for categorical dependent variables using Stata (2nd ed.). College Station, TX: Stata Press. Markowitz, H. M. 1952. Portfolio selection. Journal of Finance, 7: 77–91. McEvily, B., & Marcus, A. 2005. Embedded ties and the acquisition of competitive capabilities. Strategic Management Journal, 26: 1033–1055. Miller, D. J., Fern, M. J., & Cardinal, L. B. 2007. The use of knowledge for technological innovation within

Nohria, N., & Garcia-Pont, C. 1991. Global strategic linkages and industry structure. Strategic Management Journal, 12: 105–124. Norman, P. M. 2002. Protecting knowledge in strategic alliances: Resource and relational characteristics. Journal of High Technology Management Research, 13: 177–202. Oxley, J. E., & Sampson, R. C. 2004. The scope and governance of international R&D alliances. Strategic Management Journal, 25: 723–749. Patel, P., & Pavitt, K. 1997. The technological competencies of the world’s largest firms: Complex and pathdependent, but not much variety. Research Policy, 26: 141–156. Phelps, C. C. 2003. Technological exploration: A longitudinal study of the role of recombinatory search and social capital in alliance networks. New York: New York University Press. Phene, A., Fladmoe-Lindquist, K., & Marsh, L. 2006. Breakthrough innovations in the US biotechnology industry: The effects of technological space and geographic origin. Strategic Management Journal, 27: 369 –388. Rosenkopf, L., & Nerkar, A. 2001. Beyond local search: Boundary-spanning, exploration, and impact in the optical disk industry. Strategic Management Journal, 22: 287–306. Sampson, R. C. 2007. R&D alliances and firm performance: The impact of technological diversity and alliance organization on innovation. Academy of Management Journal 50: 364 –386. Schilling, M. A., & Phelps, C. C. 2007. Interfirm collaboration networks: The impact of large-scale network structure on firm innovation. Management Science, 53: 1113–1126. Sitkin, S. B., & Weingart, L. R. 1995. Determinants of risky decision-making behavior: A test of the mediating role of risk perceptions and propensity. Academy of Management Journal, 38: 1573– 1592. Stuart, T. E. 2000. Interorganizational alliances and the performance of firms: A study of growth and innovation rates in a high-technology industry. Strategic Management Journal, 21: 791– 811. Teece, D. J. 1996. Firm organization, industrial structure, and technological innovation. Journal of Economic Behavior & Organization, 31: 193–224. Tsai, W., & Ghoshal, S. 1998. Social capital and value creation: The role of intrafirm networks. Academy of Management Journal, 41: 464 – 476.

810

Academy of Management Journal

Tversky, A., & Kahneman, D. 1974. Judgment under uncertainty: Heuristics and biases. Science, 185: 1124 – 1131. Wang, L., & Zajac, E. J. 2007. Alliance or acquisition? A dyadic perspective on interfirm resource combinations. Strategic Management Journal, 28: 1291– 1317. Wassmer, U. 2010. Alliance portfolios: A review and research agenda. Journal of Management, 36: 141–171. Whittington, K., Owen-Smith, J., & Powell, W. 2009. Networks, propinquity, and innovation in knowledgeintensive industries. Administrative Science Quarterly, 54: 90 –122. Wooldridge, J. M. 2009. Introductory econometrics: A modern approach (4th ed.). Mason, OH: SouthWestern Cengage Learning.

August

Manish K. Srivastava ([email protected]) is an assistant professor at the School of Business and Economics, Michigan Technological University (Michigan Tech). He received his Ph.D. in strategic management from Virginia Tech. His current research interests include knowledge structures of firms and strategic alliances, evolution of alliance portfolios, and technological innovations. Devi R. Gnyawali ([email protected]) is a professor at the Pamplin College of Business, Virginia Polytechnic Institute and State University (Virginia Tech). He received his Ph.D. in strategic management from the University of Pittsburgh. His current research examines the role of intangible relational and internal resources on firm innovation, co-opetition, and competitive dynamics.