power source mismatch and the …...ownership and status, on the effectiveness of...

POWER SOURCE MISMATCH AND THE EFFECTIVENESS OFINTERORGANIZATIONAL RELATIONS: THE CASE OF

VENTURE CAPITAL SYNDICATION

DALI MADrexel University

MOOWEON RHEEUniversity of Hawaii

DAEGYU YANGKyung Hee University

This study explores the impact of the mismatch between two sources of power,ownership and status, on the effectiveness of interorganizational relations. We char-acterize three types of dyadic ownership-status relationships and examine their rela-tive mixes at the group level: (1) power source match (A’s ownership and status areboth higher than those of B); (2) ownership-dominated power source mismatch (A’sownership advantage over B is greater than B’s status advantage over A); and (3)status-dominated power source mismatch (A’s ownership advantage over B is less thanB’s status advantage over A). We found that power source match enhanced the effec-tiveness of venture capital (VC) syndication, and that ownership-dominated powersource mismatch strengthened the positive effect of power source match because thesyndicate could maintain legitimate ownership order and benefit from diverse inputs.By contrast, status-dominated power source mismatch weakened the positive effect ofpower source match, because it may have created disorderly interaction. Moreover,familiarity between participating VC firms diminished the negative moderating effectof status-dominated mismatch, because mutual trust helped organize group interac-tion, and entrepreneurial performance fulfilled a similar function, as the autonomy ofentrepreneurs from the VC syndicate buffered turbulences caused by interaction prob-lems among VC investors.

The fundamental concept in social science is Power,in the same sense in which Energy is the fundamen-tal concept in physics.

The attempt to treat one form of power, say wealth,in isolation, can only be partially successful, just asthe study of one form of energy will be defective atcertain points, unless other forms are taken intoaccount.

–Russell (1938: 4)

Power has been a central topic in classical organ-ization theory (Crozier, 1964; Etzioni, 1961; Pondy,1966). As the open system perspective prevailed inorganizational analysis in the late 1970s (Scott &Davis, 2006), resource dependence theory (Pfeffer &Salancik, 1978/2003) extended the inquiry ofpower into interorganizational relations. However,resource dependence theory has been “marginal-ized as an engine for theoretical advancement and abasis for testable empirical research” (Casciaro &Piskorski, 2005: 167) in recent years. The decline ofthe power account in the analysis of interorganiza-tional relations (Pfeffer, 1997) reflects a shift of

We appreciate the constructive feedback from Associ-ate Editor Tim Pollock and the three anonymous review-ers. We thank Phil Bromiley, John Butler, JacquelineGarner, Steve Kaplan, Alexander Ljungqvist, Mark Miz-ruchi, Sucheta Nadkarni, VK Narayanan, Christian Re-sick, John Schaubroeck, Per Stromberg, Hongwei Xu, LinZhiang, Jonathan Ziegert, and the venture capital profes-sionals that we interviewed for helpful comments andadvice. We acknowledge the financial support of LeBowCollege of Business at Drexel University and the NationalResearch Foundation of Korea grant funded by the Ko-rean Government (NRF-2010-330-B00100; to MooweonRhee). An earlier version of the article was presented in2010 at the annual meetings of the Academy of Manage-ment and the American Sociological Association.

711

� Academy of Management Journal2013, Vol. 56, No. 3, 711–734.http://dx.doi.org/10.5465/amj.2010.0832

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 expresswritten permission. Users may print, download, or email articles for individual use only.

yy

高亮

organization theorists’ attention to institutionaland evolutionary forces (DiMaggio & Powell, 1983;Hannan & Freeman, 1977), but this decline is un-fortunate because institutional actors pursue power(Scott, 1987; Sherer & Lee, 2002), power dynamicsshed light on the evolution of organizational fields(Dacin, 1997), and the inquiry into power may en-hance the impact of organization theory on otherareas, such as financial economics (Hart, 2001).

We join a group of colleagues (Baker, 1990; Cas-ciaro & Piskorski, 2005; Eisenhardt & Schoon-hoven, 1996; Gulati & Sytch, 2007) to reinvigorateresource dependence theory. Although resourcedependence theory hinges on managing the inter-dependence between organizations, Hillman, With-ers, and Collins have noted that it “has not speci-fied which dependencies take precedence overothers if multiple important dependencies exist”(2009: 1417). We suggest that different dependen-cies may play off against each other, such that onesource of power may inhibit the effectiveness ofanother. Whereas most studies of resource depen-dence theory focus on only one source of power orlump different power sources together (Casciaro &Piskorski, 2005; Gulati & Sytch, 2007; Pfeffer &Salancik, 1978/2003), we highlight the imperativeto investigate the mismatch between multiplesources of power.

Specifically, we explore the mismatch betweentwo sources of power: ownership and status. Own-ership is pivotal in both resource dependence the-ory (Pfeffer & Salancik, 1978/2003) and financialeconomics (Hart, 2001), and the inquiry into statusis one of the major streams of contemporary organ-ization theory (Podolny, 2005). We characterizethree types of dyadic relationships between owner-ship and status and examine their relative mixes atthe group level: (1) power source match (actor A’sownership and status are both higher than those ofB); (2) ownership-dominated power source mis-match (A’s ownership advantage over B is greaterthan B’s status advantage over A); and (3) status-dominated power source mismatch (A’s ownershipadvantage over B is less than B’s status advantageover A). We examine the impact of these relation-ships on the effectiveness of venture capital (VC)syndication over a 30-year period (1976–2005).

Hillman et al. (2009) also noticed that, althoughresource dependence theory shares a number offundamental assumptions with contingency the-ory, it has not benefited sufficiently from the latter.We borrow a primary insight of the contingencyperspective (Schoonhoven, 1981) by proposing a

set of interactive relationships concerning powersource mismatch. We underscore the oppositemoderating effects of the two types of power sourcemismatch on the positive association betweenpower source match and the effectiveness of VCsyndication. Whereas ownership-dominated powersource mismatch will strengthen the positive effectof power source match, because the syndicate canmaintain legitimate ownership order while benefit-ing from diverse inputs, status-dominated powersource mismatch will weaken it, because disor-derly interaction may emerge. Further, we proposethat the damage caused by the moderating effect ofstatus-dominated power source mismatch can berelieved by familiarity among VC firms and theperformance of the portfolio company, because mu-tual trust embedded in familiarity helps organizegroup interaction, whereas the autonomy of entre-preneurs that perform well buffers turbulencescaused by interaction problems among VCinvestors.

Our study makes three contributions to organiza-tion theory. First, we use a novel concept of powersource mismatch to revitalize resource dependencetheory. We adopt a contingency approach to eluci-date the moderating effect of different types ofpower source mismatch. The mismatch betweenownership and status links two disparate streamsof inquiry into interorganizational relations: re-source dependence and status (Pfeffer, 2003); themoderating effect of familiarity on the interactionof power source match and status-dominatedpower source mismatch connects the “embedded-ness” argument (Granovetter, 1985) with power re-lations; and the moderating effect of entrepreneur-ial performance depicts the interaction betweendifferent power domains. Second, our study com-plements the recent exploration of social structuralincoherence by organizational theorists and eco-nomic sociologists (Zuckerman, 2004). We anchorour theory on one of the most fundamental subjectsof social sciences, the problem of order (Wrong,1994), stressing that disorderly interaction causedby the interaction of power source match and sta-tus-dominated power source mismatch hinders theeffectiveness of interorganizational relations.Third, our study illustrates the potential impact oforganization theory on financial economics. Al-though venture capitalists intend to be highly ra-tional, they are unlikely to fully discern the conse-quences of complex patterns of interaction amongsyndicate members. Although the highly uncertainenvironment allows venture capitalists to rely on

712 JuneAcademy of Management Journal

both financial contracts and organizational statusto coordinate interaction, these two sources ofpower may create detrimental misalignment. Fi-nancial economists should not only focus on therelationship between investors and entrepreneursbut also analyze in greater depth the interactionamong investors themselves (Chahine, Arthurs,Filatotchev, & Hoskisson, 2012).

THEORY AND HYPOTHESES

Power Source Mismatch

Emerson (1962) suggested that power is a prop-erty of social relations between actors rather thanan actor’s attribute, because it resides implicitly inthe other’s dependency. Furthermore, he notes thatpower “exists nonetheless as a potential, to be ex-plored, tested, and occasionally employed by theparticipants” (Emerson, 1962: 32). Similarly, Lev-inger (1959) defined social power as an actor’s po-tential for influencing one or more other actorstoward acting or changing in a given direction, andWrong stated that “people may react to the pos-sessor or controller of resources by anticipating theeffective use of these resources to control their ownactions” (1979: 126). The notion of “anticipatoryinfluence” illustrates that the sources of power cancoordinate social interaction without activating theexercise of power.

Resource dependence theory also elucidates theimportance of power source in forming, maintain-ing, and changing interorganizational relations. Akey tenet of this theory concerns not how organi-zations exercise power over each other but howthey take actions (e.g., join ventures, interlockingboard of directors, etc.) in anticipation of the use ofresources that they need yet are controlled by oth-ers. Pfeffer and Salancik noted that “a major deter-minant of dependence is extent of discretion overthe allocation and use of a resource possessed byanother social actor,” and “such discretion is amajor source of power” (1978/2003: 47–48).

However, most studies of resource dependencetheory have focused on only one source of power orlump different power sources together (Casciaro &Piskorski, 2005; Gulati & Sytch, 2007; Pfeffer &Salancik, 1978/2003), making it difficult to articu-late the multiplexity of resource dependencies(Hillman et al., 2009). Pfeffer and Salancik (1978/2003) noted that the existence of incompatible de-mands from different coalition participants in-creases the possibility that an organization will be

unable to maintain the necessary coalition of sup-port. The notion of power source mismatch addsnew insights to their observation, because the mis-match arises from multiple dependencies betweenthe same coalition participants and can impact theeffectiveness of their relations.

Our approach of power source mismatch sug-gests that one resource dependency may prevailover another (Hillman et al., 2009), so that onesource of power may inhibit the effectiveness ofanother. Emerson (1962) suggested that multiplepower relations do not necessarily neutralize eachother, but that mismatched dependencies (A de-pends on B along one source of power, yet B de-pends on A along another) may create instability insocial relations. In response to the call by scholarsof power relations (Cook, 1977; Emerson, 1962), weextend the inquiry into interpersonal power rela-tions to interorganizational power relations, artic-ulating the impact of different forms of mismatchbetween two major power sources: ownership andstatus.

Ownership and Status

Ownership has attracted considerable attentionin the study of resource dependence (Hillman &Dalziel, 2003; Katila, Rosenberger, & Eisenhardt,2008; Lincoln, Gerlach, & Takahashi, 1992; Palmer& Barber, 2001), because it is the “means of pos-sessing a resource and therefore controlling it”(Pfeffer & Salancik, 1978: 48). Status refers to anactor’s prestige, which reflects the worth of theactor (Jasso, 2001). Ownership is a source of legit-imate power (French & Raven, 1959), because itinvolves the assertion of a legitimate right to take orretain the possession of an object and the acknowl-edgement of the rightfulness of this claim by others(Kronman, 1983). Status is a source of referentpower (French & Raven, 1959), in that a low-statusactor desires to be associated with a higher-statusone and is willing to assume attitudes or beliefsheld by the latter. Ownership is deal-specific andthus relatively short-term, whereas status is basedon cross-deal relationships and is longer-term innature (Washington & Zajac, 2005).

Pfeffer (2003) noted that there has been littleresearch connecting status inquiry to resource de-pendence theory. We link these approaches by ex-amining the mismatch between ownership and sta-tus in the context of venture capital, a key form ofentrepreneurial financing (Bygrave & Timmons,1992). In addition to providing capital, VC firms

2013 713Ma, Rhee, and Yang

help new ventures (portfolio companies) formulatestrategy, recruit key personnel, build customer re-lations, and communicate with investors (Hallen,2008; Hsu, 2004; Podolny, 2001) to earn capitalgains from the ventures’ successful exit, which is amajor indicator of the effectiveness of VC invest-ment (Hochberg, Ljungqvist, & Lu, 2007; Stuart,Hoang, & Hybels, 1999). VC firms often form co-investment syndicates to accumulate capital,spread risks, and combine complementary exper-tise (Fund, Pollock, Baker, & Wowak, 2008;Podolny, 2001; Sorenson & Stuart, 2001).

In spite of these benefits, coordination betweenVC firms is challenging, because conflict may out-weigh cooperation in interorganizational collabora-tion that lacks a formal authority structure (Ring &Van de Ven, 1992). Team research suggests thatgroup members often disagree over the group’sgoals, the strategies to pursue these goals, and pos-sible solutions to problems, such that groups mustcoordinate members’ behavior to work and makecollective decisions in a peaceful and efficientmanner (Anderson & Brown, 2010). VC syndicatesare not free of such group problems (Chahine et al.,2012); thus, it is important for them to engage inorderly social interaction to process complex infor-mation from both the highly uncertain technologi-cal industries of their portfolio companies and thevolatile financial market, provide coherent adviceto portfolio companies, and reach a timely consen-sus over major actions such as the recruitment ofstart-up management teams, obtaining a new roundof financing, and managing time-to-exit, amongothers (Gompers, Kovner, Lerner, & Scharfstein,2008; Guler, 2007; Hsu, 2004; Lerner, 1994; Wright& Lockett, 2003).

Most financial economics studies assume that VCinvestors act in concert under a financial contract(Chahine et al., 2012), such that signing a deal willensure seamless cooperation among participatingVC firms (Gompers & Lerner, 1999; Lerner, 1994).According to their roles as delineated in the finan-cial contract, VC firms holding higher equity stakeshave more legitimate power (French & Raven, 1959)in making major decisions. Organization theoristsdo not question the importance of ownership butsuggest that cognitive limitation and environmen-tal complexity make it impossible for economicactors to fully anticipate the consequences of mostgoal-oriented behaviors (Guler, 2007; Hallen, 2008;Podolny, 2001; Rider, 2009). When asked whetherVC firms in a syndicate could run into conflict, aventure capitalist said, “It happens and could be

nasty.”1 Wright and Lockett (2003) suggested thatthe syndicated investment agreement is a docu-ment that enshrines the rights of participants ratherthan specifying behavior, implying that socialforces may influence interactions between VC in-vestors in a syndicated deal. Another venture cap-italist insisted that ownership represents the onlyorder in a syndicated deal, but when asked to eval-uate the role of status in a syndicate, she said, “Ifyou are Kleiner Perkins, you speak, we listen.Period.”

A remarkable finding from VC studies is thatstatus decreases the uncertainty that the financialmarket faces when evaluating a portfolio company(Hsu, 2004; Podolny, 2001) and increases the like-lihood of a portfolio company’s successful exit(Stuart et al., 1999), a perspective recently sub-scribed to by financial economists as well (Hoch-berg et al., 2007). Nonetheless, few studies haveexamined the function of status within a VC syn-dicate, and we suggest that status is not only asignal (Podolny, 2005) for the external audiences(e.g., future VC investors, investment banks, insti-tutional investors), but also an important mecha-nism influencing VC firms’ behavior in a syndicate.

Status could be a constraint on low-status VCfirms because they have a high motivational invest-ment in being affiliated with a higher-status VCfirm (“marry-up”; Castellucci & Ertug, 2010; Emer-son, 1962; Fund et al., 2008; Podolny, 2005), whichcan endorse lower-status players in the VC industryand signal the quality of the deal. Therefore, al-though ownership has the legitimacy to be the pre-dominant interaction order in a VC syndicate, itsactual coordinating function may be inhibited bystatus-based dependency. In other words, the effec-tiveness of ownership relies on whether it can suc-cessfully absorb status constraint.

Team research has indicated that different com-positions of the same characteristic could coexistand interact with each other (Gibson & Vermeulen,

1 To better comprehend social and economic dynam-ics in the VC industry, we interviewed eight VC profes-sionals (principals or partners in eight VC firms) in No-vember and December 2008, in the region where the firstauthor resided. The interview questions were open-ended, centering on the following topics: (1) backgroundinformation (for both the individual professional andhis/her firm); (2) experience with VC syndication; (3) thefunction of ownership, prestige, and prior collaborationin VC syndication; and (4) the relationship between VCfirms and entrepreneurs of portfolio companies.


2003; O’Reilly, Caldwell, & Barnett, 1989). Simi-larly, we suggest that matched resource depen-dency may coexist with different forms of mis-matched dependencies. We define power sourcematch as a syndicate member having both owner-ship and status advantage over another member,ownership-dominated power source mismatch as amember’s ownership advantage over another onebeing greater than the latter’s status advantage, andstatus-dominated power source mismatch as amember’s ownership advantage over another onebeing less than the latter’s status advantage. Allthree types of relationships can simultaneously ex-ist within a syndicate, so we can capture theirrelative mixes at the syndicate level, as illustratedby the example in Exhibit 1.

Hypotheses

Power source match. If a VC firm with higherownership also possesses higher status, the statusadvantage will reinforce its ownership advantage.Whereas ownership advantage provides the VCfirm more contractual rights, status advantage al-lows it to receive deference from the lower-statuspartner, thereby strengthening the coordinatingfunction of ownership. At the syndicate level, VCsyndicates with greater power source match willmore effectively coordinate group interaction. Con-sequently, the syndicate can reach timely consen-sus on major decisions (Lerner, 1994), and its port-folio companies can receive clear and coherentguidance from the syndicate. Taken together, theorderly interaction stemming from power sourcematch will benefit the effectiveness of VC syndica-tion. Thus, we hypothesize:

Hypothesis 1. The greater the power sourcematch, the more effective the VC syndication.

Ownership-dominated power source mismatch.In conjunction with power source match, powersource mismatch may also exist in VC syndication.Although power source match represents the coher-ence of multiple dependencies, its impact on the ef-fectiveness of interorganizational relations may de-pend on different forms of power source mismatchthat it interacts with. We investigate the moderatingeffects of power source mismatch in response toHillman et al.’s (2009) call for linking resourcedependence theory with contingency theory, andwe propose that ownership-dominated powersource mismatch will benefit syndicate effective-ness by complementing power source match.

Power source match is a platform that ensuressmooth ownership-based group interaction, andownership-dominated power source mismatch is acontrollable “deviance” that fosters diverse inputs,which may enhance the group’s effectiveness(Hinsz, Tindale, & Vollrath, 1997; Horwitz & Hor-witz, 2007; Milliken, Bartel, & Kurtzberg, 2003).Because status concerns a VC firm’s role in the VCindustry instead of its specific ownership role in asyndicate, high-status VC firms may “cross theline,” provoking different opinions even thoughtheir ownership is lower than a lower-status one.The opinion of these high-status firms may be valu-able because of their extensive social networks,rich experience with technological industries, andtacit knowledge in determining when to make atimely exit (Gompers et al., 2008; Hallen & Eisen-hardt, 2012; Lerner, 1994; Lindsey, 2008).

On the other hand, the value of status also causeslower-status VC firms to depend on high-statusones, leading status to become a constraint on thelower-status firms. If a lower-status VC firm’s own-ership advantage is greater than its status disadvan-tage over another VC firm, it can use its legitimatepower advantage to absorb the latter’s status con-straint by asserting its property rights. As a venturecapitalist from a boutique VC firm commented re-garding working with prestigious VC firms, “Yes,we respect them even though we are the lead, aslong as they don’t go too far.” The comment sug-gests that a low-status VC firm might use owner-ship advantage to claim its leading ownership po-sition. Because the high-status VC firm depends onthe low-status VC firm’s commitment of financialresources in executing the deal, the relatively sub-stantial ownership advantage (vis-à-vis status dis-advantage) will facilitate the enactment of the low-status VC firm’s ownership role, consistent withWright and Lockett’s notion of “residual power”(2003: 2086) in VC syndicates’ decision making.Ownership-dominated power source mismatch notonly lets the VC firm with higher status and lowerownership express different opinions but alsoenables the VC firm with higher ownership andlower status to effectively exercise its contractualownership rights, thereby supplementing the ben-efits of ownership-based interaction from powersource match.

Hypothesis 2. Ownership-dominated powersource mismatch strengthens the positive ef-fect of power source match on the effectivenessof VC syndication.


Status-dominated power source mismatch. Bycontrast, status-dominated power source mismatchmay disrupt the ownership-based interaction bredby power source match, because it becomes veryproblematic when a low-status VC firm’s owner-ship advantage is not substantial enough to absorbthe status constraint imposed by a higher-statuspartner.

The substantial status difference fosters a low-status VC firm’s strong desire to affiliate with high-er-status players (French & Raven, 1959; Podolny,2005), making it difficult for the low-status firm toassert its smaller ownership advantage over a high-er-status firm. Moreover, because the firm with asubstantial status advantage yet relatively smallerownership disadvantage does not depend signifi-cantly on the lower-status firm’s financial contri-bution to the deal, it may neglect the opinion of thelatter (Fiske, 1993) and dampen the latter’s motiva-tion (Keltner, Gruenfeld, & Anderson, 2003), bothof which could diminish the legitimate ownershiporder. The coexistence of power source match andstatus-dominated power source mismatch may alsocreate confusion among syndicate members regard-ing whether the whole group will gauge actionupon ownership or status, resulting in ambiguitieswhen they interact. In addition, due to ineffectivecoordination caused by status-dominated powersource mismatch, entrepreneurs may receive con-flicting directions from the VC syndicate, becausehigher-status VC firms may offer directions withoutconsulting low-status firms, whereas the latter mayalso directly provide guidance to entrepreneursbased on their legitimate ownership advantage.With regard to the importance of order in syndi-cated deals, a venture capitalist reminded the CEOof portfolio companies that “as a CEO, that can be adisaster—come to a crossroads for the companyand suddenly you’re trying to herd a whole bunchof cats into a decision, or fielding anxious phonecalls from all directions. Without a strong lead in-vestor helping to manage her co-investors throughthat process, you as CEO are in for a splitting head-ache.”2

Hypothesis 3. Status-dominated power sourcemismatch diminishes the positive effect ofpower source match on the effectiveness of VCsyndication.

2 This quote can be found in a blog maintained by BradSvrluga, an early-stage venture capitalist (“Beware theHeadless Horseman”). Accessed in December 2011.

EXHIBIT 1

An Example of the Computation of Power Source Matchand MismatchTeam X has four VC firms: A, B, C, and D (sum of status score� 1; sum of ownership score � 1)

VC Firms Status Ownership

A 0.4 0.2B 0.3 0.4C 0.2 0.1D 0.1 0.3

The party with the higher ownership score in a dyad is i;j is the other party of the dyad; O is ownership; S is status;(Oi – Oj) is i’s ownership advantage; (Sj – Si) is j’s statusadvantage.

Power Source Match(A, C), (B, C), and (B, D) are the dyads of power source match.

PMX ��p

k(Oi � Oj) � (Sj � Si)

n

�(0.2 � 0.1) � (0.2 � 0.4)

6�

(0.4 � 0.1) � (0.2 � 0.3)6

�(0.4 � 0.3) � (0.1 � 0.3)

6�

0.3 � 0.4 � 0.36

� 0.17

Ownership-Dominated Power Source Mismatch(A, B) and (C, D) are the dyads of ownership-dominatedpower source mismatch.

OMX ��p


n

�(0.4 � 0.2) � (0.4 � 0.3)

6�

(0.3 � 0.1) � (0.2 � 0.1)6

�0.1 � 0.1

6� 0.03

Status-Dominated Power Source Mismatch(A, D) is the dyad of status-dominated power sourcemismatch.

SMX ��p


n�

(0.3 � 0.2) � (0.4 � 0.1)6

�0.26

� 0.03

All the dyads of team X are as follows:

Dyads Oi Oj Sj Si (Oi – Oj) (Sj – Si)

(A, B) 0.4 0.2 0.4 0.3 0.2 0.1(A, C) 0.2 0.1 0.2 0.4 0.1 �0.2(A, D) 0.3 0.2 0.4 0.1 0.1 0.3(B, C) 0.4 0.1 0.2 0.3 0.3 �0.1(B, D) 0.4 0.3 0.1 0.3 0.1 �0.2(C, D) 0.3 0.1 0.2 0.1 0.2 0.1


Familiarity. We suggest that familiarity stem-ming from prior collaboration among syndicatemembers may moderate the interactive effect ofpower source match and status-dominated powersource mismatch, mitigating damages caused bythe latter. In contrast to the relationship betweennew collaborators, repeated interactions containgreater trust (Gulati, 1995) and more stable roleexpectations (Ring & Van de Van, 1994). Cook andEmerson (1978) conceived commitment betweenexchange partners to be an interpersonal attach-ment leading actors to exchange repeatedly withthe same partners, and Kramer (1999) suggestedthat interactional histories give decision makersuseful information to assess partners’ dispositions,intentions, and motives. In addition, repeated col-laboration also originates from “self-selection,” arational choice view of trust (Kramer, 1999); ven-ture capitalists frequently made the following com-ment in our interviews: “If I worked with someoneonce and found something uncomfortable, I wouldnever work with him again.”

Based on amicable prior interactions, familiarityalso engenders greater psychological safety (Ed-mondson, 1999), a shared belief that a syndicate isa safe environment for VC firms to challenge eachother. As the group’s psychological safety in-creases, they can feel more comfortable aboutclaiming legitimate ownership advantage overthose with higher status but lower ownership with-out being concerned about losing future opportuni-ties to affiliate with the latter. When low-statusfirms make the claim of ownership advantage in afamiliar team, higher-status firms will also not re-duce their motivational investment (Emerson,1962) in the interdependent power relations.Therefore, familiarity enhances the prominence ofthe ownership-based order and facilitates smootherinteraction between VC firms in the syndicate.Thus, we hypothesize:

Hypothesis 4. Greater familiarity among syndi-cate members diminishes the negative interac-tive effect of power source match and status-dominated power source mismatch on theeffectiveness of VC syndication.

Entrepreneurial performance. Because of thepotential agency problems (e.g., entrepreneurs holdasymmetric information against VC investors, en-trepreneurs derive private benefits from VC invest-ment, etc.), VC firms typically acquire controlrights that are disproportionally large compared towhat they would have under the “one share one

vote rule” (Kirilenko, 2001: 565). From a resourcedependence perspective (Pfeffer & Salancik, 1978),the relationship between the VC syndicate and en-trepreneurs highlights the tension between interde-pendence and autonomy (Emerson, 1962). Entre-preneurs need resources and help from VCinvestors but also desire autonomy, potentiallyleading entrepreneurs of better-performing portfo-lio companies to seek greater control (Kaplan &Stromberg, 2003).

We propose that entrepreneurial performance(i.e., the performance of the portfolio company)may also moderate the interactive effect of powersource match and status-dominated power sourcemismatch, helping mitigate problems caused by thelatter. When the portfolio company achieves betterperformance, entrepreneurs will gain more auton-omy in the interdependent relationship with a VCsyndicate because VC firms will be less likely tointerfere with the operations of a venture withstrong performance. VC firms’ concerns about theagency problem are alleviated, because they aremore confident of the entrepreneurs’ capability andcommitment and willing to give more control backto the entrepreneurs. As Beth Seidenberg, a partnerat Kleiner Perkins, noted, “When you get venturecapital money into your company, we will own . . .anywhere from 20% to 60%. . . . For the founders,they want to get more ownership . . . and the onlyway to do that is to increase the valuation of theircompany. So we are all motivated to do the samething.”3 Another venture capitalist commented,“As long as things are going on well, a venturecapitalist will leave you [entrepreneur] alone.”4 Al-lowing entrepreneurs more autonomy will shieldthe portfolio company from the detrimental mod-erating effect of status-dominated power sourcemismatch. Even if entrepreneurs receive conflict-ing directions due to status-dominated mismatch ofthe VC syndicate, they will now have greater au-tonomy in deciding which direction to take in theirmanagerial practice. Therefore, entrepreneurialperformance may buffer against turbulence causedby interaction problems among VC investors. Thus,we hypothesize:

Hypothesis 5. Better entrepreneurial perfor-mance diminishes the negative interactive

3 http://academicearth.org/lectures/who-owns-what.Accessed November 2011.

4 http://www.forbes.com/2004/01/27/0127artofstartmidas04.html. Accessed December 2011.


effect of power source match and status-dominated power source mismatch on the ef-fectiveness of VC syndication.

DATA AND METHODS

Sample and Data

We analyzed all venture-backed biotech compa-nies in the United States from the beginning of1976 to the end of 2005. We collected data from theVentureXpert database, which has been extensivelyused to study the VC industry (Podolny, 2001; So-renson & Stuart, 2001, 2008). Our final sample con-sisted of 1,503 VC firms investing in 1,096 biotechventures. To our knowledge, this is the longestresearch period covered for the study of the USbiotech industry, a major sector for venture capitalinvestment (Stuart et al., 1999). Because our datastarted from the birth year of the biotech industry(the first modern biotech company, Genentech, wasfounded in 1976), we were able to avoid the poten-tial problem of “left censoring” (Collett, 2003) inthe survival analysis, which we report later.

Ownership and Status

Ownership refers to a participating VC firm’s eq-uity share based on its cumulative investment andranges from 0 to 1. Following Hochberg et al.(2007), we used VentureXpert’s estimated amountfor the missing equity data5 and obtained 94 per-cent of the equity information for all the financingrounds in the sample. We included other investors(e.g., banks, insurance companies, endowments) incomputing ownership because they often take aconsiderable portion of ownership, and we added adichotomous variable indicating whether the syn-dicate had investors that were not private VC firms(� 1).

To compute status scores, we constructed net-work matrices in which each participating VC firmis a node, and coparticipation in a prior round iscounted as a tie between two VC firms. We com-puted status scores based on co-investment in allindustries, which better reflect VC firms’ standingin the VC industry. In keeping with previous re-

search (Podolny, 2001; Sorenson & Stuart, 2008),we only included private VC firms in the matrices.Because status is relatively stable (Podolny, 2005;Washington & Zajac, 2005), we used a five-yearwindow (t – 5; McFadyen & Cannella, 2004;Podolny, Stuart, & Hannan, 1996; Sorenson & Stu-art, 2001) to construct the matrix for the currentyear t to smooth potential “sudden” status changescaused by short-term variations.

We used Bonacich’s (1987) centrality to measurea VC firm’s status (Podolny, 2001):

Cit(�, �) ��j

(� � �cij)Rijt�1,

where cit is the centrality of VC firm i in year t andRijt – 1 refers to an element of a co-investment matrixRt – 1, indicating the number of times that VC firmsi and j jointly invest in the same companies over thepast five-year time period. � is an arbitrary scalingcoefficient, and � denotes a weighting parameter,representing the extent to which the centrality ofVC firm i is a function of the centralities of the VCfirms in matrix Rijt – 1.

To obtain the centrality score, we set � equal tozero, which essentially computes degree centrality(i.e., a VC firm’s status is determined by how manypartners with which it is directly affiliated; Bonac-ich, 1987). Degree centrality has been used to mea-sure a VC firm’s status (Hochberg et al., 2007), andthe venture capitalists that we interviewed consid-ered the VC status scores based on the zero value of� to be more appropriate than those based on other� values. In addition, the status scores that weobtained had a .92 correlation with the reputationindex computed by Lee, Pollock, and Jin (2011).6

We divided each VC firm’s status score by the sumof the status scores of all VC firms, such that a VCfirm’s status score ranges from 0–1, comparable tothe range of the ownership score.

5 Approximately 16 percent of the estimated amountwas based on mean substitution (i.e., the total actualamount divided by the number of VC firms in a round),and VentureXpert obtained information for the rest of themissing data via alternative sources.

6 Their reputation index was based on six factors: (1)the total dollar amount of investment funds raised overthe previous five years; (2) the number of investmentfunds founded over the previous five years; (3) the num-ber of start-ups invested in over the previous five years;(4) the total dollar amount of funds invested in start-upsover the previous five years; (5) the number of companiestaken public over the previous five years; and (6) the ageof the VC firm.


Power Source Match/Mismatch

Building from group research on power asymme-try (Van der Vegt, de Jong, Bunderson, & Molleman,2010), our measure of syndicate-level power sourcematch/mismatch is the average score of dyadicpower source match/mismatch. Dyadic powersource match/mismatch refers to power sourcematch/mismatch between any two VC firms in asyndicate, and the syndicate-level power sourcematch/mismatch is the average of all dyadic powersource match/mismatch. Because the distributionof the dyadic power source match/mismatch maybe more skewed for one syndicate than for theother, we used the coefficient of variation (standarddeviation divided by the mean; Harrison & Klein,2007) to control for the distribution of powersource match, ownership-dominated mismatch,and status-dominated mismatch.7

We used the following formula to calculatepower source match or mismatch (see Exhibit 1 foran example):

Powers source match/mismatchr(t)

��p

k (Oi(t) � Oj(t)) � (Sj(t�1) � Si(t�1))n

,

where p refers to a dyadic pair in round r; n is thenumber of all the possible dyadic pairs in round r;k is the number of the dyadic pairs that fall undereach types of power source relations, as we theo-rized; i refers to the party of a dyadic pair that hasthe greater ownership score, whereas j refers to theother party of the pair; Si and Sj refer to the status ofeach party in a dyadic pair; Oi and Oj refer to theownership of each party in a dyadic pair; and trefers to the year of the investment round.

We calculated syndicate-level power sourcematch by using pairs in which Oi is greater than Oj

and Si is greater than Sj. k is the number of pairs inwhich i has the greater scores in both ownershipand status. For ownership-dominated mismatch, kis the number of the dyadic pairs in which a party’sownership advantage (Oi – Oj) is greater than theother’s status advantage (Sj – Si). Regarding status-

dominated mismatch, k is the number of dyadicpairs in which a party’s ownership advantage (Oi –Oj) is less than the other’s status advantage (Sj – Si).

Familiarity. We operationalized familiarity asthe total number of dyadic ties among participatingVC firms in the previous five years (Gulati, 1995;Gulati & Garigiulo, 1999). We used the five-yearwindow because Sorenson and Stuart (2001: 1568)noted that “if two firms have not coinvested withina five-year period, it seems unlikely that their mem-bers remain close confidants.” Consistent with ourmeasure of status, familiarity was based on priorcollaborations in all industries. In a separate anal-ysis, we used the number of ties formed over thepast three years and obtained consistent regressionresults.

Entrepreneurial performance. Patents repre-sent a major form of intellectual property for ven-ture-backed portfolio companies (Ueda, 2004) anda key indicator of their performance (Stuart et al.,1999). Mann and Sager (2007) found a high corre-lation between patent acquisition and several indi-cators of venture capital investment (e.g., numberof rounds and total investment), suggesting thatpatents are a critical resource for entrepreneursseeking to influence the outcome of the negotiationwith VC firms. We thus used the count of patentsthat each portfolio company had prior to an invest-ment round (Somaya, Williamson, & Zhang, 2007)as a proxy measure of entrepreneurial performance.We obtained patent data from two sources: the USpatent database (1976–2006), composed by the Na-tional Bureau of Economic Research (NBER), and adatabase maintained by Recombinant Capital, abiotech research firm, which we used to supple-ment the NBER patent database. We obtained pat-ent data for 236 portfolio companies and replacedthe missing values with zero, because it was likelythat they did not have a patent. Following Hsu,Hannan, and Kocak (2009), we included a binaryvariable indicating the zero replacement, whichavoided a considerable reduction of the samplesize. We also conducted a separate analysis usingthe count of patents without zero replacement; theresults of this analysis were similar to those re-ported.8

Dependent Variable (Time-to-Exit) and7 Although it is not our theoretical focus, we checkedthe prevalence of “power source reverse” (A’s ownershipadvantage equals B’s status advantage) and found that itwas extremely rare; only four rounds (our sample had2,062 rounds in total) had this special case (each roundhad only one dyad experiencing the “reverse”).

8 One may speculate that the quality of patents is alsoimportant, and one possible approach to incorporate thisis to use the number of citations made by subsequent


Model Specification

Although it would be ideal to measure the effec-tiveness of VC syndicates using the returns of VCfunds, no exiting datasets provide historical returninformation (Hochberg et al., 2007). Therefore, weused time-to-IPO (initial public offering) or mergersand acquisitions (M&A) as a proxy for the effective-ness of VC syndication (Gompers & Lerner, 2000;Hochberg et al., 2007). In a separate analysis, weused multinomial logit models to examine whetherthe impact of power source match/mismatch onIPO would differ from that on M&A and did notfind a significant difference, suggesting that it isappropriate to combine IPO and M&A in ourcontext.

We estimated the hazard rate of a portfolio com-pany’s exit using the instantaneous rate, r(t):

r(t)i � lim�t→0

qt(t, t � �t|�t)�t

,

where qt is the probability of a company experienc-ing the exit event in the interval t, t � �t. Toestimate the effects of our independent variables onthe rate of exit, we used the piecewise exponentialmodel. Assuming that the baseline transition rate isconstant within a time period but can change in anunconstrained manner across periods, the piece-wise exponential model does not require a strongassumption regarding the functional form of agedependence (Blossfeld & Rohwer, 1995; Stuart etal., 1999). We also tried the Cox proportional haz-ards model and obtained consistent results.

In our piecewise exponential model estimatingthe exit rates, the age range is split into k points (a1,a2, . . . ak), which, with ak � 1 � �, creates k ageperiods: Il � (t|at � t � at � 1), l � 1, . . . k. Constants(baseline failure rates) are estimated for each ageperiod. Therefore, the piecewise exponential

model takes the following form (Ingram & Baum,1997):

r(t)i � exp(�1Mt � �2Cit � �1�j

PMijt � �2�j

OMijt

� �3�j

SMijt � �4�j

FMijt � �5�j

EMijt)exp(�l),

if t � Il,

where Mt is a matrix of time-varying variables ofenvironmental factors; Cit is a matrix of time-vary-ing variables of portfolio companies’ characteris-tics; PMijt is a time-varying matrix of power sourcematch in round j; OMijt is a time-varying matrix ofownership-dominated power source mismatch inround j; SMijt is a time-varying matrix of status-dominated power source mismatch in round j; FMijt

is a time-varying matrix of familiarity in round j;EMijt is a time-varying matrix of entrepreneurialperformance in round j; �1, �2, �1, �2, �3, �4, and �5

are coefficients to be estimated; and �l is a constantcoefficient associated with the lth age period.9

To accommodate the time-varying covariates, wedivided the time periods into company-quarterspells. All of the time-varying variables were up-dated each calendar quarter. Each quarterly spellwas treated as right-censored, except for the spellin which portfolio companies went to IPO or wereacquired. We computed robust standard errors toaccount for the possible autocorrelation withineach portfolio company (Lin, 1994). Among the1,096 biotech companies in our sample, 256 madeit to IPO, and 74 were acquired by the end of 2005.The remaining companies were treated as right-censored as of December 31, 2005.

Control Variables

We controlled for the average ownership and sta-tus, both of which were in their logarithmic form totreat the skewed distribution, and also controlledfor the coefficient of variation of ownership andstatus and syndicate size, namely the number ofparticipating VC firms (Dimov, Shepherd, & Sut-cliffe, 2007; Guler, 2007; Hallen, 2008). We in-cluded deals with some rounds backed by a single

patents as a proxy for the quality of the current patent(Hall, Jaffe, & Trajtenberg, 2001). We tried this approachand did not find a significant effect of the three-wayinteraction of power source match, status-dominatedpower source mismatch, and the number of citationsmade by subsequent patents, most likely because itwould be very difficult for VC firms to anticipate thenumber of citations a patent may receive. The venturecapitalists that we interviewed suggested that our currentapproach of using the number of patents is appropriate(the video clip by Beth Seidenberg contained similarconfirmation; see the Hypotheses section).

9 We divided the age range at 480 days (approximately1.3 years), 2,100 days (approximately 5.7 years), and4,500 days (approximately 12 years) in our hazard-rateanalysis. We chose these split points because the base-line hazard rate changed substantially at each split point(Kim, Kim & Miner, 2009).


VC firm and used a binary variable to set the effectof multiple-VC rounds apart from that of single-VCrounds. We added another binary variable indicat-ing whether a syndication round was composed oftwo VC firms (� 1) because a two-member syndi-cate is idiosyncratic according to our measure ofpower source match/mismatch (e.g., one member’sownership share equals one minus the share ofthe other).

We controlled for the effects of investment stagesof the latest round by including two types of vari-ables. We first included a binary variable indicatinga later stage (earlier stage is used as the baseline;Sorenson & Stuart, 2001)10 and added dummies ofthe sequential number of each round, which couldimpact VC firms’ decision making (Guler, 2007). Inaddition, we added a variable representing theyears elapsed since 1976 to control for time effectsderived from the evolution of the industry (Rhee &Haunschild, 2006). We also included yearly dum-mies to incorporate time trend effects associatedwith changes in economic and industry environ-ments (Stuart et al., 1999). To control for fast-changing financial market conditions, we addedthe number of biotech companies that had under-gone an IPO or M&A during the prior quarter (Ball,Chiu, & Smith, 2011).

Endogeneity

Like many studies investigating the effects of thechange of network properties over time on eco-nomic outcomes, our study may be susceptible toendogeneity, because certain features of VC firmsor portfolio companies would influence both net-work formation and syndicate effectiveness. Weconsidered two estimations that are known to over-come the problem: (1) a two-stage model incorpo-rating instrumental variables (Heckman, 1979) and(2) a conditional fixed-effect logistic model (Alli-son & Christakis, 2007; Sørensen, 2007). However,a thorough examination suggested that neither wasfeasible for our analysis. The unique construction

of power source match/mismatch makes it difficultto find a good instrument, which prevented us fromusing the first method. The conditional fixed-effectlogistic model must drop the companies thatdid not exit to compose a matching dataset, whichcould severely distort our data.

Following Borgatti and Halgin (2011), we usedthe approach of adding control variables thatwould indicate the quality of VC firms and portfo-lio companies to treat the potential endogeneityeffects (Groysberg, Polzer, & Elfenbein, 2011; Tor-toriello & Krackhardt, 2010). We first controlled forthe cumulative investment amount procured by asyndicate up to each round (in thousands USD),which would likely influence time-to-exit as wellas tie formation. We transformed the variable usinga logarithm to alleviate the skewness of its distri-bution. We included two variables representingtwo means by which VC firms demonstrate theircapability to induce investments: the amount of VCfunds (in millions USD) and the number of the VCfunds that a VC firm had managed up to the point ofeach round (Rider, 2009). Both variables weresummed at the syndicate level, and the amount ofVC funds was logarithm-transformed as well. Be-cause VC firms could enter or leave a syndicateafter recognizing the prospective opportunities orchallenges of the portfolio company, we also gen-erated a mutually exclusive set of dummy variablesto measure four modes of turnover in the latestround: (1) some incumbent members left and nonew members joined; (2) no incumbent membersleft and some new members joined; (3) some in-cumbent members left and some new membersjoined; and (4) no incumbents left and no newmembers joined.

RESULTS

Table 1 presentsthe descriptive statistics and cor-relations for variables in our study. Because of thehigh correlation between logged average ownershipand status, we orthogonalized the two variables(Golub & Van Loan, 1996). We performed a vari-ance inflation factor (VIF) test to investigate multi-collinearity. The VIF scores for each variable rangefrom 1.01 to 7.05, all of which fell below the thresh-old of serious multicollinearity (typically 10). Inaddition, we did a condition index test of multicol-linearity (Belsley, Kuh, & Welsch, 1980), and thecondition index for models 1–6 reported in was15.72, 17.32, 17.49, 18.11, 17.59, and 18.20, respec-tively. Belsley et al. (1980: 105) suggested that con-

10 The VentureXpert database provided informationon six investment stages: (1) seed/startup, (2) early stage,(3) expanding stage, (4) later stage, (5) buy-out/acquisi-tion, and (6) unidentified other stage. After examiningthe definition of each stage, we found that this informa-tion could be rearranged as two stages: (1), (2), and (3) ascomparatively earlier stages; and (4), (5), and (6) as laterstages, in which a syndicate becomes more mature toharvest its investment.


TA

BL

E1

Des

crip

tive

Sta

tist

ics

and

Cor

rela

tion

sa

Var

iabl

esM

ean

s.d

.1

23

45

67

89

1011

1213

1415

1617

1819

2021

2223

2425

1.N

um

ber

ofex

its

inp

rior

quar

ter

3.06

3.25

2.E

lap

sed

year

20.0

86.

98�

.05

3.S

tage

du

mm

y(l

ater

stag

es�

1)0.

240.

43�

.01

.16

4.ln

(Cu

mu

lati

vein

vest

men

t)2.

411.

23�

.03

.52

.29

5.ln

(Am

oun

tof

VC

fun

ds)

13.3

93.

11�

.04

.14

.02

.18

6.N

um

ber

ofV

Cfu

nd

s22

.06

32.1

3�

.07

.36

.06

.35

.48

7.T

eam

size

3.84

2.72

�.0

6.1

2.0

2.3

0.4

5.4

68.

Sin

gle

VC

rou

nd

du

mm

y0.

220.

42.0

5�

.16

.06

�.0

6�

.42

�.2

7�

.55

9.T

wo-

mem

ber

rou

nd

du

mm

y0.

180.

38.0

2�

.02

.03

�.0

7�

.11

�.1

3�

.31

�.2

4

10.D

um

my

ofin

vest

ors

that

are

not

pri

vate

VC

firm

s

0.59

0.50

�.0

4.0

3.0

2.1

4.2

6.1

6.4

7�

.32

�.1

9

11.T

urn

over

du

mm

y(s

ome

firm

sle

ft,

no

firm

sjo

ined

)

0.13

0.34

.03

�.1

2.0

6�

.07

�.1

0�

.16

�.3

2.4

6�

.04

�.2

4

12.T

urn

over

du

mm

y(n

ofi

rms

left

,so

me

firm

sjo

ined

)

0.17

0.38

�.0

1.0

1�

.02

.03

.08

.05

.24

�.2

2�

.04

.13

�.1

8

13.T

urn

over

du

mm

y(s

ome

firm

sle

ft,

som

efi

rms

join

ed)

0.63

0.49

�.0

3.1

1�

.05

.08

.06

.14

.16

�.3

0.0

5.1

0�

.50

�.5

9

14.l

n(A

vera

geow

ner

ship

)�

1.74

0.64

.01

�.0

2�

.22

�.4

8�

.14

�.2

0�

.41

.03

.20

�.2

3�

.03

�.0

1�

.06

15.l

n(A

vera

gest

atu

s)�

1.74

0.56

.01

�.0

3�

.25

�.6

1�

.15

�.2

3�

.46

.02

.22

�.2

5�

.04

�.0

2�

.06

.88

16.C

Vof

stat

us

1.60

0.49

�.0

1�

.01

.12

.24

.02

.01

.15

�.0

2�

.06

.33

.07

�.0

2.0

1�

.31

�.3

417

.CV

ofow

ner

ship

0.77

0.42

.02

�.1

4.2

4.2

7�

.10

�.0

7�

.04

.19

.05

.04

.07

�.0

8.0

2�

.32

�.3

8.2

018

.CV

ofp

ower

sou

rce

mat

ch0.

300.

34�

.03

�.0

2.0

9.3

4.1

8.2

1.4

9�

.04

�.2

9.3

1�

.08

.12

.04

�.4

9�

.56

.31

.15

19.C

Vof

own

ersh

ip-

dom

inat

edm

ism

atch

0.11

0.25

�.0

1�

.06

.07

.21

.10

.16

.39

.01

�.1

5.1

3�

.06

.09

�.0

4�

.37

�.4

2�

.02

.10

.34

20.C

Vof

stat

us-

dom

inat

edm

ism

atch

0.18

0.30

�.0

1.0

6.0

1.3

2.1

5.2

5.4

2.0

2�

.21

.18

�.0

6.1

0�

.03

�.3

9�

.45

�.0

7.0

5.4

1

21.P

ower

sou

rce

mat

ch0.

290.

37.0

2�

.04

.01

�.1

7�

.06

�.1

0�

.18

�.0

1.1

7�

.11

.01

.07

�.0

6.2

7.3

0.0

5.0

5�

.07

�.1

8�

.25

22.O

wn

ersh

ip-d

omin

ated

mis

mat

ch0.

010.

04.0

1.0

2�

.02

.01

.01

�.0

4�

.02

�.0

2.0

1�

.05

�.0

1�

.04

.01

.01

.01

�.0

7.1

6�

.08

.11

�.0

4�

.12

23.S

tatu

s-d

omin

ated

mis

mat

ch0.

090.

17�

.01

.01

�.0

7�

.11

�.0

9�

.04

�.1

8.1

4.0

1�

.07

.11

�.0

3�

.11

.21

.23

.08

�.1

4�

.21

�.1

3�

.04

�.2

9�

.05

24.F

amil

iari

ty1.

122.

29.0

1�

.03

.05

.18

.12

.19

.34

.04

�.1

5.0

4�

.05

.08

�.0

4�

.33

�.3

8�

.22

.09

.30

.51

.46

�.1

5�

.02

�.1

325

.Nu

mbe

rof

pat

ents

0.46

1.86

�.0

2.1

4.1

8.1

7.0

2.0

7�

.01

.02

.01

.07

�.0

1.0

3�

.01

�.1

1�

.12

.15

.10

.06

�.0

6�

.05

�.0

4�

.03

�.0

5�

.05

26.P

aten

td

um

my

0.40

0.49

.01

.01

.08

.05

�.0

3�

.05

.01

.01

�.0

3.0

4�

.04

.03

.03

�.0

6�

.08

.04

.06

.11

�.0

5�

.01

�.0

3�

.02

�.0

6.0

1.3

0

an

�12

,118

;va

lues

grea

ter

than

.02

inco

rrel

atio

nar

esi

gnif

ican

tat

the

p�

.05

leve

l(t

wo-

tail

edte

sts)

.“C

V”

isth

eco

effi

cien

tof

vari

atio

n.

“VC

”is

ven

ture

cap

ital

.

dition indexes of 30–100 indicate moderate tostrong multicollinearity problems; therefore, multi-collinearity does not appear to be problematic inour analysis. We also performed a robustness checkby mean-centering variables in the interactions (Ai-ken & West, 1991), the results of which were con-sistent with the noncentered ones that we report inTable 2.

Table 2 reports the results of our piecewise sur-vival analysis. Model 1 includes control variables.In model 2, we added the main effects of powersource mismatch, ownership-dominated powersource mismatch, and status-dominated powersource mismatch. Power source match significantlyincreased the rate of exit, such that each additionalunit of power source match multiplies the rate ofexit by a factor of 1.78 (e[.58 � 1]), supporting Hy-pothesis 1. Model 3 added the interaction term ofpower source match and ownership-dominatedpower source mismatch as well as that of powersource match and status-dominated power sourcemismatch. Both of the interactive effects were sig-nificant, but in opposite directions. Figure 1 indi-cates that, as Hypothesis 2 predicts, the positiveeffect of power source match on time-to-exit wasstronger when ownership-dominated power sourcemismatch was greater; by contrast, Figure 2 indi-cates that the association between power sourcematch and time-to-exit could even become negativewhen status-dominated power source mismatch in-creased to one standard deviation above its meanvalue, supporting Hypothesis 3. In both figures, thex-axis ranges from 0–2, which is the actual range ofpower source match.

In model 4, we added the three-way interactionof power source match, status-dominated powersource mismatch, and familiarity, which had a pos-itive significant effect and is consistent with Hy-pothesis 4. Figure 3 presents two different scenar-ios of familiarity among VC firms. When none ofthe VC firms in a syndicate had ever previouslycollaborated, status-dominated power source mis-match diminished the positive effect of powersource match on time-to-exit. However, as familiar-ity increased, the positive effect of power sourcematch became stronger, revealing that familiarityrelieved damages caused by status-dominatedpower source mismatch.

Model 5 included the three-way interaction ofpower source match, status-dominated powersource mismatch, and entrepreneurial perfor-mance, which also had a positive significant effect,lending support for Hypothesis 5. Figure 4 presents

two different scenarios of entrepreneurial perfor-mance in terms of the number of patents of a port-folio company, indicating that when the portfoliocompany did not have any patents, status-domi-nated power source mismatch diminished the pos-itive effect of power source match on time-to-exit.Once the portfolio company obtained two patents,the positive effect of power source match becamestronger, implying that entrepreneurial perfor-mance considerably dampened the negative inter-active effect of power source match and status-dominated power source mismatch. Model 6 is thefull model including both three-way interactive ef-fects, which were consistent with those we testedin models 4 and 5.

DISCUSSION

We began this study with the belief that powerremains crucial for organization theory, and that itis important to reinvigorate the inquiry of interor-ganizational power relations (Pfeffer & Salancik,1978/2003). We agree with Russell (1938) that theattempt to treat one form of power in isolation canonly be partially successful, and we extend Frenchand Raven’s (1959) typology of multiple bases ofsocial power to explore how the mismatch betweenmultiple power sources influences the impact ofpower source match on the effectiveness of interor-ganizational relations.

We identified the mismatch between two majorpower sources in interorganizational relations:ownership and status. Whereas ownership-domi-nated power source mismatch may complement theorderly interaction arising from power sourcematch, status-dominated mismatch hinders it dueto potentially conflicting interaction orders (own-ership vis-à-vis status). Furthermore, by incorporat-ing higher-level moderators for the interaction be-tween power source match and status-dominatedpower source mismatch, we demonstrated twomechanisms that relieved the damages caused bystatus-dominated power source mismatch: famil-iarity between VC firms and entrepreneurial perfor-mance. The familiarity effect echoes the analysis ofinterorganizational trust relations (e.g., Gulati,1995), and the impact of entrepreneurial perfor-mance sheds light on the connection between twosets of interdependent power relations: thoseamong VC firms of a syndicate and those betweenthe VC syndicate and entrepreneurs. Taken to-gether, we investigated multiple sources of power,multiple forms of power source mismatch, and


TABLE 2Maximum-Likelihood Estimates of Time-to-Exitsa

Variables Model 1 Model 2 Model 3 Model 4 Model 5 Model 6

Age � 1.3 years �8.41** �8.53** �8.60** �8.64** �8.62** �8.66**(1.33) (1.34) (1.33) (1.32) (1.33) (1.33)

Age 1.3 ~ 5.8 years �6.73** �6.83** �6.95** �6.99** �6.96** �7.02**(1.04) (1.06) (1.06) (1.05) (1.06) (1.06)

Age 5.8 ~ 12.3 years �6.57** �6.66** �6.76** �6.81** �6.78** �6.83**(1.05) (1.07) (1.06) (1.06) (1.07) (1.07)

Age � 12.3 years �7.59** �7.63** �7.75** �7.78** �7.76** �7.79**(1.07) (1.09) (1.08) (1.08) (1.09) (1.08)

Number of exits in prior quarter 0.13** 0.13** 0.13** 0.13** 0.13** 0.13**(0.02) (0.02) (0.02) (0.02) (0.02) (0.02)

Elapsed year �0.35** �0.33** �0.32** �0.33** �0.33** �0.33**(0.11) (0.11) (0.11) (0.11) (0.11) (0.11)

Stage dummy (expansion stage � 1) 0.43** 0.44** 0.43** 0.45** 0.44** 0.46**(0.15) (0.15) (0.15) (0.15) (0.15) (0.15)

ln (Cumulative investment) 0.60** 0.62** 0.62** 0.61** 0.62** 0.61**(0.12) (0.12) (0.12) (0.12) (0.12) (0.12)

ln (Amount of VC funds) �0.02 �0.02 �0.02 �0.01 �0.02 �0.01(0.02) (0.02) (0.02) (0.02) (0.02) (0.02)

Number of VC funds �0.00 �0.00 �0.00 �0.00 �0.00 �0.00(0.00) (0.00) (0.00) (0.00) (0.00) (0.00)

Team size �0.09* �0.05 �0.06 �0.06 �0.06 �0.06(0.04) (0.04) (0.04) (0.04) (0.04) (0.04)

Single VC round dummy �0.43 �0.27 �0.32 �0.26 �0.30 �0.23(0.27) (0.28) (0.28) (0.29) (0.28) (0.29)

Two-member round dummy �0.32 �0.30 �0.30 �0.30 �0.32 �0.31(0.24) (0.24) (0.24) (0.24) (0.24) (0.24)

Dummy of investors that are not private VC firms 0.34* 0.37* 0.37* 0.37* 0.37* 0.37*(0.16) (0.16) (0.16) (0.16) (0.16) (0.16)

Turnover dummy (some firm left, no firms joined) 0.32 0.28 0.30 0.32 0.30 0.32(0.38) (0.38) (0.37) (0.38) (0.38) (0.38)

Turnover dummy (no firms left, some firmsjoined)

0.14 0.15 0.16 0.18 0.16 0.19(0.39) (0.40) (0.39) (0.40) (0.39) (0.40)

Turnover dummy (some firms left, some firmsjoined)

0.20 0.20 0.21 0.24 0.23 0.26(0.37) (0.37) (0.36) (0.37) (0.36) (0.37)

Average ownership �0.06 �0.11 �0.12 �0.11 �0.12 �0.12(0.14) (0.14) (0.15) (0.15) (0.15) (0.15)

Average status �0.63 �1.26* �1.21* �1.21* �1.20* �1.22*(0.52) (0.58) (0.58) (0.58) (0.58) (0.58)

CV of status �0.15 �0.26 �0.24 �0.25 �0.24 �0.24(0.15) (0.18) (0.18) (0.18) (0.18) (0.18)

CV of ownership �0.08 �0.14 �0.18 �0.18 �0.18 �0.18(0.18) (0.19) (0.19) (0.19) (0.19) (0.19)

Familiarity 0.03 0.05* 0.05* 0.08 0.05* 0.08(0.02) (0.02) (0.02) (0.04) (0.02) (0.04)

Number of patents �0.08 �0.07 �0.07 �0.08 �0.10 �0.10(0.05) (0.05) (0.05) (0.05) (0.09) (0.09)

Patent dummy �0.17 �0.20 �0.19 �0.21 �0.20 �0.21(0.14) (0.14) (0.14) (0.14) (0.14) (0.14)

CV of power source match �0.24 �0.10 �0.08 �0.13 �0.11(0.26) (0.28) (0.29) (0.28) (0.29)

CV of ownership-dominated mismatch �0.44 �0.52 �0.55 �0.52 �0.55(0.32) (0.32) (0.32) (0.32) (0.32)

CV of status-dominated mismatch �0.13 �0.08 �0.11 �0.10 �0.14(0.24) (0.24) (0.25) (0.25) (0.25)

Power source match 0.58* 0.58* 0.60* 0.58* 0.59*(0.29) (0.29) (0.29) (0.29) (0.29)

(table continues)


multiple parties (VC firms and entrepreneurs) seek-ing interdependence and control, offering a holisticpicture of interorganizational power relations.

Theoretical Contributions

In response to organization theorists’ call for tap-ping the unrealized potential of resource depen-dence theory (Casciaro & Piskorski, 2005; Pfeffer,2003; Santos & Eisenhardt, 2009), we initiated animportant line of research on power source mis-match to examine interorganizational power rela-tions, underlining that resource dependency alongone dimension may inhibit resource dependencyalong another dimension. Power source mismatchsheds light on a central theme of power relations:the tension between autonomy and interdepen-dence (Emerson, 1962; Pfeffer & Salancik, 1978).While low-status organizations with ownership ad-vantage engage higher-status affiliates to gain atten-tion and status, they may not be cognizant of thefact that the status-based affiliation could constraintheir autonomy. If their ownership advantage issubstantial enough to absorb this status constraint,

the advantages of power source match may be am-plified; if the ownership advantage is too limited toabsorb the status constraint, disorderly interactionmay hamper the effectiveness of the interorganiza-tional relation.

The moderating effects of familiarity and entre-preneurial performance illustrate two mechanismsthat help mitigate the problems caused by status-dominated power source mismatch. Familiarity is aform of “embeddedness” (Granovetter, 1985),which has been rarely linked to power relations inorganization theory (for exceptions, see Baker[1990] and Pollock [2004]). We suggest that “em-beddedness” not only reduces transaction costs(Granovetter, 1985; Uzzi, 1996), but also helps eco-nomic actors form a coherent interaction order byabsorbing the “overflow” of status constraint. Themoderating effect of entrepreneurial performancereveals that power dynamics in one domain (theinteraction between a VC syndicate and entrepre-neurs) may shield the focal organization (the port-folio company) from problems in another domain(interactions among VC firms in the VC syndicate).To gain a more comprehensive understanding of

TABLE 2(Contined)

Variables Model 1 Model 2 Model 3 Model 4 Model 5 Model 6

Onwership-dominated mismatch 1.41 0.54 0.48 0.50 0.44(1.31) (1.29) (1.30) (1.30) (1.30)

Status-dominated mismatch 0.85 1.07 1.18 1.03 1.14(0.59) (0.59) (0.60) (0.61) (0.62)

Power source match � ownership-dominatedmismatch

21.48* 22.70* 22.44* 23.84*(10.89) (11.19) (10.77) (11.05)

Power source match � status-dominated mismatch �9.73* �13.79* �11.42* �15.91**(4.76) (5.74) (5.11) (6.06)

Power source match � familiarity �0.32 �0.32(0.27) (0.27)

Status-dominated mismatch � familiarity �1.49 �1.52(0.99) (0.99)

Power source match � status-dominated mismatch 15.22* 15.80*� familiarity (6.96) (7.05)

Power source match � number of patents 0.01 0.01(0.30) (0.30)

Status-dominated mismatch � number of patents 0.07 0.06(0.29) (0.30)

Power source match � status-dominated mismatch 6.13* 6.75*� number of patents (2.79) (3.05)

N 12,701 12,701 12,701 12,701 12,701 12,701Pseudo-likelihood �268.94 �264.67 �262.11 �260.21 �261.30 �259.30

a Robust standard errors are in parentheses. Year dummies and round sequence number dummies are not reported in the table. “CV”is the coefficient of variation.

* p � .05** p � .01

Two-tailed tests.


resource dependence, organization theoristsshould pay attention to both multiple powersources and their mismatch, and the interactionbetween different power domains (Cook, 1977;Markovsky, Willer, & Patton, 1998).

Because ownership and status reflect two formsof social structure, our inquiry into power sourcemismatch echoes organization theorists’ increasinginterest (Burton & Beckman, 2007; Zuckerman,2004) in exploring the impact of social structuralincoherence on organizational action. The notionof power source mismatch adds a new angle of powerrelations to the framework of incongruence, whichhas fascinated social psychologists (cognitive disso-nance; Festinger, 1957), sociologists (status inconsis-tency; Berger, Norman, Balkwell, & Smith, 1992; Len-ski, 1966), and organization theorists (incoherence ofmarket structures; Zuckerman, 2004). In addition, ourstudy responds to Pfeffer’s (2003) call for linking re-source dependence theory (Pfeffer & Salancik, 1978/2003) with status inquiry (Podolny, 2005), suggestingthat status-based dependency may inhibit owner-ship-based dependency.

Our inquiry into power relations also echoes thereemergence of scholarly interest in power in socialpsychology (Keltner et al., 2003; Magee & Galinsky,

2008; Tjosvold & Wisse, 2009). In spite of this in-terest, contemporary scholars (Guinote & Vescio,2010) share classical authors’ concern (Cartwright,1959) that power is one of the most ambiguousconcepts awaiting theoretical and empirical clarifi-cation. Our analysis suggests that a promising di-rection for such study would be to differentiatebetween multiple sources of power and examinetheir incongruence. As we incorporate insightsfrom social psychologists (Emerson, 1962; French &Raven, 1959) to develop theories about interorgan-izational power relations (Cook, 1977; Pfeffer &Salancik, 1978/2003), we demonstrate that poweris a common mechanism linking micro and macrodynamics, and we hope that future inquiry intointerpersonal power source mismatch will benefitfrom our investigation of interorganizational powerdynamics.

Our analysis of power relations also highlightsthe potential impact of organization theory on fi-nancial economics. We note financial economists’emerging interest in social dynamics (e.g., Hoch-berg et al., 2007; Hong & Kacperczyk, 2009) andemphasize that although venture capitalists intendto be highly rational, it is unlikely for them to fullydiscern the consequences of complex patterns of

FIGURE 1Interaction of Power Source Match and Ownership-Dominated Mismatcha

a Min � 0, max � 2, mean � 0.29, and mean � 1 s.d. � 0.66 on X-axis.


interaction among syndicate members. We suggestthat financial economists not only address the re-lationship between investors and entrepreneurs butalso examine in greater depth the interactionamong investors themselves, a topic that is cur-rently missing from research in financialeconomics.

Limitations and Avenues for Future Research

Due to the limitations of our data, we were un-able to collect some information related to powerrelations. For example, board membership reflectsnot only the power split between VC firms but alsothat between VC investors and entrepreneurs (Ka-plan & Stromberg, 2004); therefore, it may beworthwhile to incorporate this factor into the anal-ysis of power source mismatch. Each VC firm typ-ically assigns one or more professionals to handle asyndicated deal, such that the interaction betweenVC firms in a syndicate may also be influenced bythe individual characteristics of these profession-als. For instance, will similar educational back-grounds among VC professionals facilitate their in-teraction and relieve the damage caused by status-dominated power source mismatch? Will the

prestige of individual VC professionals or entrepre-neurs have a moderating effect on different types ofpowersourcemismatch?Thecombinationofinteror-ganizational and interpersonal analyses will ex-hibit a more comprehensive picture of the dynam-ics of power source mismatch.

Similar to prior research on power imbalance/asymmetry (Casciaro & Piskorski, 2005; Gulati &Sytch, 2007), we used continuous ownership andstatus scores to index power source match/mis-match. Alternatively, a categorical ranking method(e.g., first, second, etc.) may be used if quantitativedifferences in ownership and status do not matter.In our case, if a syndicate has four members—A, B,C, and D—and their status scores are 0.05, 0.1, 0.15,and 0.7, respectively, the categorical ranking differ-ence between A and B would be the same as thatbetween C and D. However, the measure cannotindicate the fact that D’s status advantage over C(0.55) is much greater than B’s advantage over A(0.05); thus, the categorical ranking method is notthe most appropriate one for our context.

It is also possible that VC syndicates may takeactions to solve the interaction problems caused bystatus-dominated power source mismatch. In a post

FIGURE 2Interaction of Power Source Match and Status-Dominated Mismatcha

a Min � 0, max � 2, mean � 0.29, and mean � 1 s.d. � 0.66 on X-axis.


hoc analysis, we used a multinomial logit model toexamine the impact of power source match/mis-match on multiple types of syndication turnover

and found that a syndicate with greater status-dom-inated power source mismatch was more likely todraw new members with no incumbent members

FIGURE 3Interaction of Power Source Match, Status-Dominated Mismatch, and Familiarity


leaving the syndicate. The result implies that VCsyndicates may invite new members to alleviateproblems caused by status-dominated powersource mismatch, and we suggest that future re-

searchers explore the association between powersource match/mismatch and group restructuring.

Our theory may be generalized to other forms offinancial management that concern both ownership

FIGURE 4Interaction of Power Source Match, Status-Dominated Mismatch, and Entrepreneurial Performance


and status, such as securities underwriting (Eccles& Crane, 1988) and joint ventures (Eisenhardt &Schoonhoven, 1996; Gulati, 1995). The investiga-tion of power source mismatch in these contextswill extend the contextual scope of our theory. Inaddition to ownership and status, there may also beother power sources (e.g., brokerage) that may ex-tend the conceptual scope of our analysis. An im-portant benefit of familiarity is trust, which mayalso originate in other sources (e.g., referral); con-sequently, it would be worthwhile to investigatethe moderating effects of multiple forms of trust onthe interactive effect of power source match andstatus-dominated power source mismatch.

Conclusion

We anchor our theory on one of the most funda-mental subjects of social sciences, the problem oforder (Wrong, 1994), suggesting that disorderly in-teraction may hinder the effectiveness of interor-ganizational relations. We highlight a contingencyview of resource dependence in light of socialstructural incoherence, showing that differenttypes of power source mismatch may result in dra-matically different consequences, and that the det-rimental status-dominated power source mismatchmay be moderated by other social forces. We be-lieve that power source mismatch is an importantapproach to reinvigorate the inquiry of interorgan-izational power relations, and we hope that futureresearchers can join us in bringing this line of in-quiry back to the forefront of organization theory.

REFERENCES

Aiken, L. S., & West, S. G. 1991. Multiple regression:Testing and interpreting interactions. New York:Sage.

Allison, P. D., & Christakis, N. 2007. Fixed effects meth-ods for the analysis of non-repeated events. Socio-logical Methodology, 36: 155–172.

Anderson, C., & Brown, C. E. 2010. The function anddysfunction of hierarchy. In B. M. Staw & R. M.Kramer (Eds.), Research in organizational behav-ior, vol. 30: 55–89. Greenwich, CT: JAI.

Baker, W. E. 1990. Market networks and corporate behav-ior. American Journal of Sociology, 96: 589–625.

Ball, E., Chiu, H. H., & Smith, R. 2011. Can VCs time themarket? An analysis of exit choice for venture-backed firms. Review of Financial Studies, 24:3105–3138.

Belsley, D., Kuh, E., & Welsch, R. 1980. Regression di-agnostics. New York: Wiley.

Berger, J., Norman, R. Z., Balkwell, J. W., & Smith, R. F.1992. Status inconsistency in task situations: A testof four status processing principles. American Soci-ological Review, 57: 843–855.

Blossfeld, H. P., & Rohwer, G. 1995. Techniques of eventhistory analysis: New approaches to causal anal-ysis. Mahwah, NJ: Erlbaum.

Bonacich, P. 1987. Power and centrality: A family ofmeasures. American Journal of Sociology, 92:1170–1182.

Borgatti, S. P., & Halgin, D. S. 2011. On network theory.Organization Science, 22: 1168–1181.

Burton, M. D., & Beckman, C. 2007. Leaving a legacy:Position imprints and successor turnover in youngfirms. American Sociological Review, 72: 239–266.

Bygrave, W. D., & Timmons, J. A. 1992. Venture capitalat the crossroads. Boston: Harvard Business SchoolPress.

Cartwright, D. 1959. A field theoretical concepts ofpower. In D. Cartwright (Ed.), Studies in social pow-er: 183–220. Ann Arbor: Institute for Social Re-search.

Casciaro, T., & Piskorski, M. J. 2005. Power imbalance,mutual dependence, and constraint absorption: Acloser look at resource dependence theory. Admin-istrative Science Quarterly, 50: 167–199.

Castellucci, F., & Ertug, G. 2010. What is in it for them?Advantages of higher-status partners in exchange re-lationships. Academy of Management Journal, 53:149–166.

Chahine, S., Arthurs, J. D., Filatotchev, I., & Hoskisson,R. E. 2012. The effect of venture capital syndicatediversity on earnings management and performanceof IPOs in the US and UK: An institutional perspec-tive. Journal of Corporate Finance, 18: 179–192.

Collett, D. 2003. Modelling survival data in medicanresearch. London: Chapman & Hall/CRC.

Cook, K. S. 1977. Exchange and power in interorganiza-tional relations. Sociological Quarterly, 18: 62–82.

Cook, K. S., & Emerson, R. M. 1978. Power, equity andcommitment in exchange networks. American Soci-ological Review, 43: 721–739.

Crozier, M. 1964. The bureaucratic phenomenon. Chi-cago: University of Chicago Press.

Dacin, M. T. 1997. Isomorphism in context: The powerand prescription of institutional norms. Academy ofManagement Journal, 40: 46–81.


DiMaggio, P., & Powell, W. W. 1983. The iron cage revis-ited: Institutional isomorphism and collective ratio-nality in organizational fields. American Sociologi-cal Review, 48: 147–160.

Dimov, D., Shepherd, D. A., & Sutcliffe, K. M. 2007.Requisite expertise, firm reputation, and status inventure capital investment allocation decisions.Journal of Business Venturing, 22: 481–502.

Eccles, R., & Crane, D. 1988. Doing deals: Investmentbank at work. Boston: Harvard Business SchoolPress.

Edmondson, A. 1999. Psychological safety and learningbehavior in work teams. Administrative ScienceQuarterly, 44: 350–383.

Eisenhardt, K. M., & Schoonhoven, C. B. 1996. Resource-based view of strategic alliance formation: Strategicand social effects in entrepreneurial firms. Organi-zation Science, 7: 136–150.

Emerson, R. M. 1962. Power-dependence relations.American Sociological Review, 27: 31–40.

Etzioni, A. 1961. A comparative analysis of complexorganizations: On power, involvement, and theircorrelates. New York: Free Press.

Festinger, L. 1957. A theory of cognitive dissonance.Stanford University Press.

Fiske, S. T. 1993. Controlling other people: The impact ofpower on stereotyping. American Psychologist, 48:621–628.

French, J. R. P., & Raven, B. 1959. The social bases ofsocial power. In D. Cartwright (Ed.), Studies in so-cial power: 150–167. Ann Arbor: Institute for SocialResearch.

Fund, B. R., Pollock, T. G., Baker, T., & Wowak, A. 2008.Who’s the new kid? The process of becoming centralin venture capitalist deal networks. In J. A. C. Baum& T. J. Rowley (Eds.), Advances in strategic man-agement, vol. 25: 563–593. London, UK: Emerald.

Gibson, C., & Vermeulen, F. 2003. A healthy divide:Subgroups as a stimulus for team learning behavior.Administrative Science Quarterly, 48: 202–239.

Golub, G. H., & Van Loan, C. F. 1996. Matrix computa-tions. Baltimore: John Hopkins University Press.

Gompers, P., Kovner, A., Lerner, J., & Scharfstein, D.2008. Venture capital investment cycles: The impactof public markets. Journal of Financial Economics,87: 1–23.

Gompers, P., & Lerner, J. 1999. Conflict of interest in theissuance of public securities: Evidence from venturecapital. Journal of Law and Economics, 42: 1–28.

Gompers, P., & Lerner, J. 2000. Money chasing deals? Theimpact of fund inflows on private equity valuation.Journal of Financial Economics, 55: 281–325.

Granovetter, M. 1985. Economic action and social struc-ture: The problem of embeddedness. AmericanJournal of Sociology, 91: 481–510.

Groysberg, B., Polzer, J. T., & Elfenbein, H. A. 2011. Toomany cooks spoil the broth: How high-status indi-viduals decrease group effectiveness. OrganizationScience, 22: 722–737.

Guinote, A., & Vescio, T. K. 2010. The social psychologyof power. New York: Guilford.

Gulati, R. 1995. Does familiarity breed trust? The impli-cations of repeated ties for contractual choices inalliances. Academy of Management Journal, 38:85–112.

Gulati, R., & Garigiulo, M. 1999. Where do interorgani-zational networks come from? American Journal ofSociology, 104: 1439–1493.

Gulati, R., & Sytch, M. 2007. Dependence asymmetry andjoint dependence in interorganizational relation-ships: Effects of embeddedness on a manufacturer’sperformance in procurement relationships. Admin-istrative Science Quarterly, 52: 32–69.

Guler, I. 2007. Throwing good money after bad? Politicaland institutional influences on sequential decisionmaking in the venture capital industry. Administra-tive Science Quarterly, 52: 248–285.

Hall, B. H., Jaffe, A. B., & Trajtenberg, M. 2001. The NBERpatent-citation data file: Lessons, insights, andmethodological tools. NBER working paper 8498.Cambridge, MA: MIT Press.

Hallen, B. L. 2008. The causes and consequences of theinitial network positions of new organizations: Fromwhom do entrepreneurs receive investments? Ad-ministrative Science Quarterly, 53: 685–718.

Hallen, B., & Eisenhardt, K. 2012. Catalyzing strategiesand efficient tie formation: How entrepreneurialfirms obtain venture capital ties. Academy of Man-agement Journal, 55: 35–70.

Hannan, M. T., & Freeman, J. 1977. The population ecol-ogy of organizations. American Journal of Sociol-ogy, 82: 929–966.

Harrison, D. A., & Klein, K. J. 2007. What’s the differ-ence? Diversity constructs as separation, variety, ordisparity in organizations. Academy of Manage-ment Review, 32: 1199–1228.

Hart, O. 2001. Financial contracting. Journal of Eco-nomic Literature, 39: 1079–1110.


Heckman, J. 1979. Sample selection bias as a specifica-tion error. Econometrica, 47: 153–161.

Hillman, A. J., & Dalziel, T. 2003. Boards of directors andfirm performance: Integrating agency and resourcedependence. Academy of Management Review, 28:383–396.

Hillman, A. J., Withers, M. C., & Collins, B. J. 2009.Resource dependence theory: A review. Journal ofManagement, 35: 1404–1427.

Hinsz, V. B., Tindale, R. S., & Vollrath, D. A. 1997. Theemerging conceptualization of groups as informationprocessors. Psychological Bulletin, 121: 43–64.

Hochberg, Y. V., Ljungqvist, A., & Lu, Y. 2007. Whomyou know matters: Venture capital networks andinvestment performance. Journal of Finance, 62:251–301.

Hong, H., & Kacperczyk, M. 2009. The price of sin: Theeffects of social norms on markets. Journal of Finan-cial Economics, 93: 15–36.

Horwitz, S. K., & Horwitz, I. B. 2007. The effects of teamdiversity on team outcomes: A meta-analytic reviewof team demography. Journal of Management, 33:987–1015.

Hsu, D. 2004. What should entrepreneurs pay for venturecapital affiliation? Journal of Finance, 59: 1805–1844.

Hsu, G., Hannan, M. T., & Kocak, O. 2009. Multiplecategory memberships in markets: AnIntegrative the-ory and two empirical test. American SociologicalReview, 74: 150–169.

Ingram, P., & Baum, J. A. C. 1997. Chain affiliation andthe failure of Manhattan hotels, 1898–1980. Admin-istrative Science Quarterly, 42: 68–102.

Jasso, G. 2001. Studying status: An integrated framework.American Sociological Review, 66: 96–124.

Kaplan, S. K., & Stromberg, P. 2003. Financial contract-ing theory meets the real world: An empirical anal-ysis of venture capital contracts. Review of Finan-cial Studies, 70: 281–315.

Kaplan, S. K., & Stromberg, P. 2004. Characteristics, con-tracts, and actions: Evidence from venture capitalistanalyses. Journal of Finance, 59: 2177–2210.

Katila, R., Rosenberger, J. D., & Eisenhardt, K. M. 2008.Swimming with sharks: Technology ventures, de-fense mechanisms and corporate relationships. Ad-ministrative Science Quarterly, 53: 295–332.

Keltner, D., Gruenfeld, D. H., & Anderson, C. 2003.Power, approach, and inhibition. Psychological Re-view, 110: 265–284.

Kim, J., Kim, J., & Miner, A. S. 2009. Organizationallearning from extreme performance experience: Theimpact of success and recovery experience. Organi-zation Science, 20: 958–978.

Kirilenko, A. A. 2001. Valuation and control in venturefinance. Journal of Finance, 56: 565–587.

Kramer, R. M. 1999. Trust and distrust in organizations:Emerging perspectives, enduring questions. In J. T.Spence, J. M. Darley, & D. J. Foss (Eds.), Annualreview of psychology, vol. 50: 569–598. Palo Alto,CA: Annual Reviews. [DOI: ] [PubMed: 15012464].

Kronman, A. T. 1983. Max Weber. Stanford, CA: Stan-ford University Press.

Lee, P. M., Pollock, T. G., & Jin, K. 2011. The contingentvalue of venture capital reputation. Strategic Orga-nization, 9: 33–69.

Lenski, G. 1966. Power and privilege: A theory of socialstrafication. New York: McGraw-Hill.

Lerner, J. 1994. The syndication of venture capital invest-ments. Financial Management, 23: 16–27.

Levinger, G. 1959. The development of perceptions andbehavior in newly formed social power relation-ships. In D. Cartwright (Ed.), Studies in social pow-er: 183–220. Ann Arbor, MI: Institute for Social Re-search.

Lin, D. Y. 1994. Cox regression analysis of multivariatefailure time data: The marginal approach. Statisticsin Medicine, 13: 2233–2247.

Lincoln, J. R., Gerlach, M. L., & Takahashi, P. 1992.Keiretsu networks in the Japanese economy: A dyadanalysis of intercorporate ties. American Sociologi-cal Review, 57: 561–585.

Lindsey, L. 2008. Blurring firm boundaries: The role ofventure capital in strategic alliances. Journal of Fi-nance, 63: 1137–1168.

Magee, J. C., & Galinsky, A. D. 2008. Social hierarchy:The self-reinforcing nature of power and status. InJ. P. Walsh & A. P. Brief (Eds.), Academy of Man-agement annals, vol. 2: 351–398 Essex, UK: Rout-ledge.

Mann, R. J., & Sager, T. W. 2007. Patents venture capital,and software startups. Research Policy, 36: 193–208.

Markovsky, B., Willer, D., & Patton, T. 1998. Power rela-tions in exchange networks. American SociologicalReview, 53: 220–236.

McFadyen, M. A., & Cannella, A. A. 2004. Social capitaland knowledge creation: Diminishing returns of thenumber and strength of exchange. Academy of Man-agement Journal, 47: 735–746.


Milliken, F. J., Bartel, C. A., & Kurtzberg, T. R. 2003. InP. B. Paulus & B. A. Nijstad (Eds.), Group creativity:Innovation through collaboration: 63–84. NewYork: Oxford University Press.

O’Reilly, C. A., Caldwell, D. F., & Barnett, W. P. 1989.Work group demography, social integration, and turn-over. Administrative Science Quarterly, 34: 21–37.

Palmer, D., & Barber, B. M. 2001. Challengers, elites, andowning families: A social class theory of corporateacquisitions in the 1960s. Administrative ScienceQuarterly, 46: 87–120.

Pfeffer, J. 1997. New directions for organization theory:Problems and prospects. New York: Oxford Univer-sity Press.

Pfeffer, J. 2003. Introduction to the classic edition. In J.Pfeffer & G. R. Salancik (Eds.), The external controlof organizations: A resource dependence perspec-tive: xi–xxix. Stanford, CA: Stanford UniversityPress.

Pfeffer, J., & Salancik, G. R. 1978, /2003. The externalcontrol of organizations: A resource dependenceperspective. Stanford, CA: Stanford UniversityPress.

Podolny, J. M. 2001. Networks as the pipes and prisms ofthe market. American Journal of Sociology, 107:33–60.

Podolny, J. M. 2005. Status signals: A sociological studyof market competition. Princeton, NJ: PrincetonUniversity Press.

Podolny, J. M., Stuart, T. E., & Hannan, M. T. 1996.Networks, knowledge, and niches: Competition inworldwide semicondonductor industry, 1984–1991.American Journal of Sociology, 102: 659–689.

Pollock, T. G. 2004. The benefits and costs of underwrit-ers’ social capital in the U.S. initial public offeringsmarket. Strategic Organization, 2: 357–388.

Pondy, L. R. 1966. A systems theory of organizationalconflict. Academy of Management Journal, 9: 246–256.

Rhee, M., & Haunschild, P. R. 2006. The liability of goodreputation: A study of product recalls in the U.S.automobile industry. Organization Science, 17:101–117.

Rider, C. I. 2009. Constraints on the control benefits ofbrokerage: A study of placement agents in U.S. Ven-ture capital fundraising. Administrative ScienceQuarterly, 54: 575–601.

Ring, P. S., & Van de Ven, A. H. 1992. Structuring coop-erative relationships between organizations. Strate-gic Management Journal, 13: 483–498.

Russell, B. 1938. Power: A new social analysis. London:Allen & Unwin.

Santos, F. M., & Eisenhardt, K. M. 2009. Constructingmarket and shaping boundaries: Entreprenuerialpower in nascent fields. Academy of ManagementJournal, 52: 643–671.

Schoonhoven, C. B. 1981. Reviewed problems with con-tingency theory: Testing assumptions hidden withinthe language of contingency “theory.” Administra-tive Science Quarterly, 26: 349–377.

Scott, W. R. 1987. The adolescence of institutional the-ory. Administrative Science Quarterly, 32: 493–511.

Scott, W. R., & Davis, G. F. 2006. Organizations andorganizing: Rational, natural and open systemsperspectives. Upper Saddle River, NJ: Prentice Hall.

Sherer, P. D., & Lee, K. 2002. Institutional change in largelaw firms: A resource dependence and institutionalperspective. Academy of Management Journal, 45:102–119.

Somaya, D., Williamson, I. O., & Zhang, X. 2007. Com-bining patent law expertise with R&D for patentingperformance. Organization Science, 18: 922–937.

Sørensen, J. B. 2007. Bureaucracy and entrepreneurship:Workplace effects on entrepreneurial entry. Admin-istrative Science Quarterly, 52: 387–412.

Sorenson, O., & Stuart, T. E. 2001. Syndication networksand the spatial distribution of venture capital invest-ments. American Journal of Sociology, 106: 1546–1588.

Sorenson, O., & Stuart, T. E. 2008. Bringing the contextback in: Settings and the search for syndicate part-ners in venture capital investment networks. Admin-istrative Science Quarterly, 53: -266–294..

Stuart, T. E., Hoang, H., & Hybels, R. C. 1999. Interorgan-izational endorsements and the performance of en-trepreneurial ventures. Administrative ScienceQuarterly, 44: 315–349.

Tjosvold, D., & Wisse, B. 2009. Power and interdepen-dence in organizations. New York: Cambridge Uni-versity Press.

Tortoriello, M., & Krackhardt, D. 2010. Activating cross-boundary knowledge: The role of Simmelian ties inthe generation of innovations. Academy of Manage-ment Journal, 53: 167–181.

Ueda, M. 2004. Banks versus venture capital: Projectevaluation, screening, and expropriation. Journal ofFinance, 59: 601–621.

Uzzi, B. 1996. The sources and consequences of embed-dedness for the economic performance of organiza-


tions: The network effect. American SociologicalReview, 61: 674–698.

Van der Vegt, G. S., de Jong, S. B., Bunderson, J. S., &Molleman, E. 2010. Power asymmetry and learningin teams: The moderating role of performance feed-back. Organization Science, 21: 347–361.

Washington, M., & Zajac, E. J. 2005. Status evolution andcompetition: Theory and evidence. Academy ofManagement Journal, 48: 282–296.

Wright, M., & Lockett, A. 2003. The structure and man-agement of alliances: Syndication in the venture cap-ital industry. Journal of Management Studies, 40:2073–2102.

Wrong, D. H. 1979. Power: Its forms, bases, and uses.New York: Harper and Row.

Wrong, D. H. 1994. The problem of order: What unitesand divides society. Cambridge, MA: Harvard Uni-versity Press.

Zuckerman, E. W. 2004. Structural incoherence andstock market activity. American Sociological Re-view, 69: 405–432.

Dali Ma ([email protected]) is assistant professor ofmanagement at Drexel University. He received his Ph.D.in sociology from the University of Chicago, and hisprimary interests are cognitive and economic sociology,status and identity dynamics, social networks, entrepre-neurship, and transitional China.

Mooweon Rhee ([email protected]) is Shidler Dis-tinguished Professor of Management at the University ofHawaii. He received his Ph.D. from Stanford University.His current research interests revolve around organiza-tional learning, corporate reputation, social networks,and Asia-based theories of organizations.

Daegyu Yang ([email protected]) is assistant professor ofmanagement at Kyung Hee University, Seoul. He re-ceived his Ph.D. in management from the University ofHawaii. His current research interests include organiza-tional learning, social networks, and corporate strategies.


Copyright of Academy of Management Journal is the property of Academy of Managementand its content may not be copied or emailed to multiple sites or posted to a listserv withoutthe copyright holder's express written permission. However, users may print, download, oremail articles for individual use.

power source mismatch and the …...ownership and status, on the effectiveness of...

Documents