Alliance Network Presentation-

Slide 1: 

COREY PHELPS UNIVERSITY OF WASHINGTON FOSTER SCHOOL OF BUSINESS

Slide 2: 

An unconventional beginning Research program overview One particular study Research question and motivation Theory Methods Results Contributions AGENDA An Unconventional Beginning Research Overview One Particular Study

AN UNCONVENTIONAL BEGINNING : 

AN UNCONVENTIONAL BEGINNING An Unconventional Beginning Research Overview One Particular Study Is your golf swing this good?

AN UNCONVENTIONAL BEGINNING : 

AN UNCONVENTIONAL BEGINNING An Unconventional Beginning Research Overview One Particular Study If your golf swing was as good as Tiger’s, would you change it? Why / why not? When would you change it?

AN UNCONVENTIONAL BEGINNING : 

AN UNCONVENTIONAL BEGINNING An Unconventional Beginning Research Overview One Particular Study Tiger Woods has radically changed his swing twice in the past 9 years Let’s see what happened…

TIGER WOOD’S PERFORMANCE : 

TIGER WOOD’S PERFORMANCE An Unconventional Beginning Research Overview One Particular Study ANNUAL TOP FINISHES AS A PERCENTAGE OF TOTAL TOURNAMENTS PERCHENTAGE YEAR

TIGER WOOD’S PERFORMANCE : 

TIGER WOOD’S PERFORMANCE An Unconventional Beginning Research Overview One Particular Study Financial performance initially suffers, but ultimately improves YEAR $US AVERAGE MONEY WINNINGS PER EVENT

AN UNCONVENTIONAL BEGINNING : 

AN UNCONVENTIONAL BEGINNING An Unconventional Beginning Research Overview One Particular Study For my exec ed students: What can Tiger Woods’ swing changes teach us about innovation and growth? For you: How does this anecdote relate to my program of research?

Slide 9: 

WHO AM I? An Unconventional Beginning Research Overview One Particular Study My research

RESEARCH FOCUS : 

RESEARCH FOCUS An Unconventional Beginning Research Overview One Particular Study When, why and how… …do organizations create and deploy new resources, particularly technological knowledge, to transform and renew themselves in an effort to adapt, survive and thrive? …do organizations effectively manage the tension between exploitation and exploration? be like Tiger? …do organizations learn from extramural sources of knowledge and translate this learning into innovations? using three particular tools…

RESEARCH FOCUS : 

RESEARCH FOCUS An Unconventional Beginning Research Overview One Particular Study When, why and how do firms use: Corporate venture capital Acquisitions Strategic alliances …to access and learn from extramural knowledge sources?

COMMON THEMES OF RESEARCH : 

COMMON THEMES OF RESEARCH An Unconventional Beginning Research Overview One Particular Study Phenomenon-driven Rigor and relevance Multiple theoretical perspectives Innovation as recombinatory search process Social embeddedness/social capital Heterogeneous diffusion in social networks Economics of innovation; endogenous growth theory RBV/KBV Real options Integrative theoretical development Deductive and inductive research logics Qualitative and quantitative data and analysis Longitudinal designs Panel data estimation techniques Sensitivity to endogeneity Testing and correcting for endogeneity using 2 stage approaches & FIML Multiple levels of analysis Cross-level effects

THE QUESTIONS I ASK ABOUT CVC : 

THE QUESTIONS I ASK ABOUT CVC An Unconventional Beginning Research Overview One Particular Study Antecedents When do firms pursue CVC? (w/ Basu & Kotha; R&R at JBV) Effects on investor innovation When does CVC lead to exploratory innovation? (w/ Wadhwa & Kotha) CVC unit performance How do exploratory subunits achieve ambidexterity? (w/ Basu & Kotha) Effects on new venture performance When does CVC improve new venture performance? (w/ Park) Effects on new venture alliance network evolution Negative affiliation effects: When do initial ties spoil the well for future ties? VC syndication networks When do IVCs partner with CVCs? The evolution of CVC network position How does CVC network position affect CVC investment behavior? Reputation as enabler and constraint Relationship w/ other external venturing modes How do firms choose among CVC, alliance and acquisition?

THE QUESTIONS I ASK ABOUT ACQUISITIONS : 

THE QUESTIONS I ASK ABOUT ACQUISITIONS An Unconventional Beginning Research Overview One Particular Study Prior collaboration and acquisitions When does prior collaboration (alliance or CVC) lead to the acquisition of one partner by the other? (w/ Wadhwa) Prior collaboration and acquisition performance How and when does prior collaboration between an acquirer and target firm influence the acquirer’s post-acquisition innovative performance? Two studies (project funded by NSF grant SES-0521839): Inductive – multiple case study design (in progress) 9 cases; 41 interviews Large sample theory testing (in progress)

THE QUESTIONS I ASK ABOUT ALLIANCES : 

THE QUESTIONS I ASK ABOUT ALLIANCES An Unconventional Beginning Research Overview One Particular Study Effects of alliances on firm innovation When does a firm’s alliance network structure and composition influence its exploratory innovation? (forth., AMJ) When does large scale alliance network structure influence firm innovation? (w/ Schilling; 2007, Mgmt Science) Knowledge transfer within alliances What affects the nature of knowledge transferred between partners in alliances? (w/ Heidl) Multi-party alliances What affects the stability of multi-party alliances? (w/ Heidl)

Slide 16: 

An Unconventional Beginning Research Overview One Particular Study When does a firm’s alliance network structure and composition influence its exploratory innovation? THE QUESTIONS I ASK ABOUT ALLIANCES When does a firm’s alliance network structure and composition influence its exploratory innovation?

RELATED RESEARCH : 

RELATED RESEARCH An Unconventional Beginning Research Overview One Particular Study Firm Innovation When do firms learn from their own knowledge spillovers? (w/ Yang & Steensma; forthcoming, AMJ) When do individuals voluntarily attempt to innovate in their organizations? (w/ Morrison; 1999; AMJ) Exploitation in the pursuit of exploration What affects a late mover’s performance when entering new markets? (w/ Shamsie & Kuperman; 2004; SMJ) Networks and information/knowledge diffusion What do we know about knowledge networks? (w/ Heidl; under review) How do social networks among open source software developers influence the choice of license to govern a new OSS project? (w/ Singh; working paper) How has economics research influenced strategic management research? (Ph.D. seminar project in progress)

Slide 18: 

COREY PHELPS UNIVERSITY OF WASHINGTON An Unconventional Beginning Research Overview One Particular Study Research Questions Motivations Theory Methods Results Contributions A Longitudinal Study of the Influence of Alliance Network Structure and Composition on Firm Exploratory Innovation

PHENOMENON & QUESTION : 

PHENOMENON & QUESTION Research Questions Motivations Theory Methods Results Contributions “An organization that is not adequately enabling and motivating new possibilities is more likely to witness its own decline” (Moran & Ghoshal, 1999) Exploratory innovation the extent to which a firm’s technical innovations embody knowledge that is novel relative to its established domain of technical competence Research Question Under what conditions do a firm’s strategic alliances influence its exploratory innovation?

MOTIVATIONS : 

MOTIVATIONS Research Questions Motivations Theory Methods Results Contributions Prime motive for alliance formation: knowledge creation Interorganizational networks affect firm innovation e.g., Ahuja, 2000; Baum et al., 2000; Schilling & Phelps, 2007; Shan et al., 1994 However, opportunities exist to expand understanding of how and when alliances influence firm innovation…

MOTIVATIONS: EXTANT RESEARCH LIMITATIONS : 

MOTIVATIONS: EXTANT RESEARCH LIMITATIONS Research Questions Motivations Theory Methods Results Contributions Largely ignores network composition In particular: network technical diversity Relevant to on-going debate/conflict… Conflicting theory & results about ego network structure Holes vs. closure Attention to network composition may help resolve… Technologically diverse partners may provide timely access to diverse information Independent of holes/closure Ignores content (e.g., relative novelty) of innovations Research has not examined influence of alliance network structure and network diversity on firm exploratory innovation

THEORETICAL FRAMING : 

THEORETICAL FRAMING Research Questions Motivations Theory Methods Results Contributions Recombinatory search Innovation is costly problem-solving process Solutions (new knowledge) result from novel recombinations of existing knowledge Diversity of knowledge searched will influence exploratory innovation Requires access, detection, transfer and assimilation Structural social capital Instrumentally-valuable resources that inhere in social network e.g., trust, norms, information Provides informal governance, enhancing an actor’s ability to mobilize the actions and resources of other network members Ego network as portfolio Portfolio represents search space (where firms search) Social capital influences efficacy and efficiency of search and transfer

TECHNOLOGICAL DIVERSITY : 

TECHNOLOGICAL DIVERSITY Research Questions Motivations Theory Methods Results Contributions Access to diverse technological knowledge increases potential for exploratory solutions Increases number, variety and novelty of recombinations Challenges existing thinking and beliefs Promotes novel associations and insights Conceptual reframing of problems Analogical cross-pollination Stimulates experimentation with novel combinations Strategic alliances Provide access Better suited than market transactions for repeated exchange of tacit, routine-based knowledge H1:The technological diversity of a firm’s alliance partners will be positively related to its exploratory innovation.

NETWORK DENSITY : 

NETWORK DENSITY Research Questions Motivations Theory Methods Results Contributions Dense networks provide trust and reciprocity (social capital) Density generates “knowledge-based” and “deterrence-based” trust Density facilitates reciprocity exchanges Generalized exchange Trust and reciprocity provide informal alliance governance Reduces threat of opportunism; increases cooperation Greater sharing and transfer of partner knowledge Dense interfirm networks facilitate transfer and integration of complex, tacit knowledge (Ahuja, 2000; Dyer & Nobeoka, 2000; Kogut, 2000) Network density also promotes: Intense social interaction Joint problem solving and experimentation Quick diffusion of alternative perspectives of technical problems Quick identification of appropriate referrals

NETWORK DENSITY AND DIVERSITY : 

NETWORK DENSITY AND DIVERSITY Research Questions Motivations Theory Methods Results Contributions Structural social capital enhances network diversity effect Increasing dissimilarity increases contractual hazards Heightened measurement problems (novelty and tacitness) Coordination problems and incentives for expropriation Informal governance of social capital mitigates Increases partner motivation to “teach” As diversity increases, other benefits of density more valuable Joint problem solving and social interaction Diffusion of alternative interpretations of problems Shared perspective and discourse H2: The relationship between network diversity and exploratory innovation will become stronger (weaker) as density increases (decreases).

SAMPLE AND DATA : 

SAMPLE AND DATA Research Questions Motivations Theory Methods Results Contributions Panel: top 77 public communications equip. mfrs., ‘87-’97 Horizontal tech alliances focused on communications equipment Appropriate context to test hypotheses Frequent use of technology alliances Firms routinely patent their inventions Alliance and M&A data (1980-1996) Systematic archival research SDC, Factiva, Lexis-Nexis, Dialog, Moody’s manuals, reports examined>180,000 articles; ~1200 annual reports/10Ks 7904 alliances, 1967 M&As in database U.S. patent data (Delphion) Constructed “family trees” for each firm-year; used date of application 243,758 unique patents in data set Firm attribute data (financials, industry) CompuStat, ARs, SEC filings, Japan Co. Handbook, WorldScope

MEASURES: DEPENDENT VARIABLE : 

MEASURES: DEPENDENT VARIABLE Research Questions Motivations Theory Methods Results Contributions Exploratory Innovation Extent to which firm’s inventions embody novel knowledge (relative to extant knowledge stock) Citations in firm i’s telecom equipment patents in year t classified as: New citation: not used by firm or to its patents in prior 7 years Used citation: used or to firm patent in past 7 years Exploratory Innovationit = new citesit /total citesit Alternative measure (robustness) Count of new 3-digit tech. classes in which firm i patented in year t

MEASURES: NETWORK DIVERSITY : 

MEASURES: NETWORK DIVERSITY Rodan & Galunic’s (2004) knowledge heterogeneity measure Similar to Bonacich’s (1987) eigenvector centrality Construction: Jaffe’s (1986) measure of technological proximity Similarity of partners’ distribution of patents across tech. classes Range (1-proximity): 0 (max. overlap) - 1 (max. diversity) Dt = annual technological distance matrix (elements: dijt) Computed uniqueness of firm j in year t: Uniqueness of firm j (uj): solution to eigen equation: λU = DU U = eigenvector of D, λ = eigenvalue Network tech. diversityit = Research Questions Motivations Theory Methods Results Contributions

MEASURES: EGO NETWORK DENSITY : 

MEASURES: EGO NETWORK DENSITY Research Questions Motivations Theory Methods Results Contributions Constructed annual adjacency matrices Existence of technology alliances maintained among sample firms at end of year t Used pre-sample alliance formation and termination data Densityit = % of technology alliances that exist among focal firm’s (ego) partners Alternative measures (robustness) Efficiency Constraint

Slide 30: 

CONTROL VARIABLES Research Questions Motivations Theory Methods Results Contributions

SPECIFICATIONS AND ESTIMATION : 

SPECIFICATIONS AND ESTIMATION Research Questions Motivations Theory Methods Results Contributions Period: 1987-1997 RHS variables lagged one year Alternative constructions of DV Linear (untransformed) Log odds transformation (logit) Random effects GLS w/ robust errors Year dummies Significant unobserved heterogeneity Insignificant Hausman tests Insignificant AR1

RESULTS : 

RESULTS Research Questions Motivations Theory Methods Results Contributions Network technological diversity: positive (p<.05) Quadratic term: n.s. H1 supported Interaction Density*Diversity: positive (p<.05) H2 supported Unexpected: positive main effect of density Controls Positive and significant R&D Negative and significant International Current ratio Patent Stock

ROBUSTNESS / ALTERNATIVE EXPLANATIONS : 

ROBUSTNESS / ALTERNATIVE EXPLANATIONS Research Questions Motivations Theory Methods Results Contributions Fixed Effects Dropped time invariant variables Results consistent with random effects Endogeneity of network diversity? Reasons for exogeneity Tech alliances formed for variety of motivations (not solely exploration) Portfolio of alliances (and thus diversity) is difficult to alter in short term Ran 2SLS two-factor f.e. treating diversity as endogenous 1st stage instruments (for identification): firm tech diversity Davidson & MacKinnon (1993) endogeneity test: not significant Endogeneity of structure? Structure is probably exogenous in present analysis Observed networks are formed for a variety of reasons Structures are path dependent (at least in short term) Structure is beyond control of individual firm Ran 2SLS two-factor f.e. treating density as endogenous 1st stage instruments (for identification): firm tech diversity Davidson & MacKinnon (1993) endogeneity test: not significant

ADDITIONAL ROBUSTNESS CHECKS : 

ADDITIONAL ROBUSTNESS CHECKS Research Questions Motivations Theory Methods Results Contributions GEE estimation w/ exchangeable correlation structure (Papke & Wooldridge, 2005) Increases efficiency of estimates Consistent results Alternative measures of density (closure) Both Efficiency and Constraint provided consistent results Alternative measure of exploratory innovation Count of new 3-digit tech. classes in which firm i patented in year t Estimation: panel negative binomial regression w/ random effects Results: qualitatively similar

CONTRIBUTIONS : 

CONTRIBUTIONS Research Questions Motivations Theory Methods Results Contributions Social capital Local structure and access to diverse info/knowledge are distinct Structural holes don’t guarantee access to diverse info Composition matters! Density and diversity can co-exist and combination enhances exploratory innovation Structure (density) enhances influence of composition (diversity) Search Value of diverse recombinatory search space (overcome local search) Network structure enhances search/transfer esp. when accessing diverse alters Alliances and firm innovation Role of partner tech diversity and local network structure Reconciliation of conflicting results? Density and exploration vs. sparsity and exploitation (contingency)

MANAGERIAL IMPLICATIONS : 

MANAGERIAL IMPLICATIONS Research Questions Motivations Theory Methods Results Contributions Network approach to alliance management Diversity AND indirect ties among partners matter Consider alliance network structure in formation and termination decisions Not just technological competences of partners

Slide 37: 

Research Questions Motivations Theory Methods Results Contributions LIMITATIONS & FUTURE RESEARCH Patents are noisy and incomplete measure of (exploratory) innovation Do firms systematically patent explorative innovations for unobserved/unobservable reasons? Firms may patent in anticipation of allying Lag structure reduces reverse causality concerns Alliance survivor bias? Probably not Generalizability? Are causal explanations from interpersonal network research isomorphic to interfirm networks? Wanted: longitudinal qualitative interfirm network research

Slide 38: 

Questions?

Slide 39: 

TEACHING Undergraduate (4.62) Strategic Management Capstone strategy course MBA (4.54) Entrepreneurial Strategy Core course for MBA entrepreneurship certificate and TEC programs Ph.D. (4.60) Economic Foundations of Strategic Management Research Executive Education (4.51) Various custom and certificate programs Topics Business strategy External corporate venturing Internal corporate venturing Technology and innovation management

Slide 40: 

BACKUP SLIDES

Slide 41: 

FIELD STUDY EVIDENCE We use all available tools [including alliances and acquisitions] for making deals. In some instances, the risk profile of the asset is such that an equity stake is the most applicable tool for minimizing risk and maximizing future option value. - Investment professional in corporate venturing group

Slide 42: 

INDUSTRY TECHNOLOGICAL CHANGE Now, the product life cycles are a week. There’s many, many competitors. These dynamics are completely different. Our skill set is chemistry based. Now we’re looking at other areas like computer science and physics [though investments]. It’s just like a complete overhaul. - Investment professional of photographic equipment company (discussing the transition to digital photography)

Slide 43: 

INDUSTRY COMPETITIVE INTENSITY The competitiveness each business unit is feeling or the rate of change of technology in that area speaks to how quickly they want to see what’s happening on the outside or make investments. ... Look at VC expenditures today and [you cannot] possibly compete with them and think you’re going to be developing all the next technologies [internally]. - Investment professional of diversified chemicals company

Slide 44: 

INDUSTRY APPROPRIABILITY Eventually in ’99 the GSM incumbents recognized our IPR in WCDMA. At that point it became a matter of executing on licensing the GSM companies for 3G, WCDMA and making sure that we accelerated the adoption of WCDMA technology as quickly as possible. And so, investments became an important component of executing on that mission. - Investment professional of communications service firm

Slide 45: 

FIRM MARKETING RESOURCES [Our] organization is such a powerful name. It’s the largest pharmaceutical company in the world, so that definitely plays a huge role as far as our deal flow goes. Some of the companies we invest in [we] will actually become a customer of. For other ones, just having us as an investor is sort of a validating event for the company. - Investment professional of pharmaceuticals firm We have great resources that would not only help us, but have proven to help the companies that we invest in as well. Over and beyond [what] a traditional capital group can do, things like clinical regulatory capabilities, technical assessment, operation capabilities, better reading of reimbursement trends and FDA approval patterns, better capabilities in assessing marketing distribution needs on a global level. - Investment professional of medical equipment firm

Slide 46: 

FIRM TECHNOLOGICAL RESOURCES Gee, if you’re a start-up guy and you’ve got a wireless application or a wireless product, wouldn’t you want to work with somebody that has channels, and also has seventy-seven years of technology, that has intellectual property, that has brand, that has access, that has reach? - Investment professional of communications equipment firm The key things that we bring to the CVC program is the immense technical depth in various aspects of wireless that we can tap into and also, the legal talent in the company that is par nobody. We are able to tap into both the patent side where we’re trying to value an IP of a company and the technical side when there’s a reasonably complex and technical solution that the company has developed. Most of the time we are able to handle it within our group because we have technical backgrounds. But when we can’t do that, we can quickly tap into the pool of talented folks within [the firm] to evaluate - Investment professional of communication services firm