spillovers across organizational architectures: the role...

Strategic Management JournalStrat. Mgmt. J., 33: 710–733 (2012)

Published online EarlyView in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/smj.1965

Received 1 February 2010; Final revision received 26 July 2012

SPILLOVERS ACROSS ORGANIZATIONALARCHITECTURES: THE ROLE OF PRIOR RESOURCEALLOCATION AND COMMUNICATION INPOST-ACQUISITION COORDINATION OUTCOMES

RAJSHREE AGARWAL,1* JAIDEEP ANAND,2 JANET BERCOVITZ,3

and RACHEL CROSON4

1 R. H. Smith School of Business, University of Maryland, College Park, Maryland,U.S.A.2 Fisher College of Business, The Ohio State University, Columbus, Ohio, U.S.A.3 College of Business, University of Illinois, Champaign, Illinois, U.S.A.4 School of Management and School of Economics, Political and Policy Sciences,University of Texas, Dallas, Richardson, Texas, U.S.A.

We integrate insights from organization design, economic game theory, and social psychologyto examine the role of prior resource allocation and communication in alleviating behavioraluncertainty arising in interunit coordination settings. We use the context of post-acquisitioncoordination, focusing on the extent to which routines created under one organizational archi-tecture (i.e., interorganizational alliances) may transfer to another organizational architecture(i.e., internal divisional structures via acquisition of alliance partners). Using a randomizedexperimental design, we find that prior resource allocation decisions in the absence of priorcommunication lowers post-acquisition performance due to the development and transference ofpre-acquisition stage routines that may be inappropriate post-acquisition. Post-acquisition per-formance is aided, however, by the formation of noncompetitive routines in the pre-acquisitionstage in the presence of communication. Copyright 2012 John Wiley & Sons, Ltd.

INTRODUCTION

A central issue in organizational design is thedevelopment of optimal architectures that supportthe efforts of multiple actors in resource invest-ment, knowledge sharing, and economic valuecreation. Organizations are faced with the gen-eral problem of coordinating tasks, in both intra-and interorganizational settings (Argyres, 1995;

Keywords: strategic alliances; post-acquisition perfor-mance; experiments; organizational architecture anddesign; behavioral uncertainty and coordination*Correspondence to: Rajshree Agarwal, R.H. Smith School ofBusiness, University of Maryland, 4512 Van Munching Hall,College Park, MD 20 742, U.S.A. E-mail: [email protected]

Ghoshal and Bartlett, 1990; Puranam, Singh, andChaudhuri, 2009). Tushman and Nadler (1978)note that interunit task interdependence increasesthe uncertainty managers face and, thus, increasesinformation processing requirements. Informationneeded to support coordination among divisionalmanagers can flow from resource allocation activ-ities and direct communication (Galbraith, 1973;Noda and Bower, 1996). The information providedby these sources to a manager has two compo-nents—one that is objective and relates to the taskrequirements and incentives themselves and theother that is ‘socialized’ relating to trust and expec-tations regarding potential cooperative behavior byother managers (Gulati, 1995).

Copyright 2012 John Wiley & Sons, Ltd.

Spillovers Across Organizational Architectures 711

Against this background, the literature on clas-sic organization design and congruence theory hasexamined how appropriate selection of internalstructural elements may achieve higher coordi-nation in the presence of task (and role) uncer-tainty (Burns and Stalker, 1961; Galbraith, 1973;Lawrence and Lorsch, 1967; Miller, 1986). Thisliterature has extensively examined the role oftask uncertainty (Burton and Obel, 2004; Donald-son, 1987, 1990; Lawrence and Lorsch, 1967) onorganizational design choice and subsequent per-formance. Yet, the role of behavioral uncertainty,despite a long history (e.g., Barnard 1938; Marchand Simon, 1958; Simon, 1947) is not well under-stood. This is problematic as behavioral uncer-tainty stemming from lack of confidence that theother managers will make decisions that are con-gruent with their own cause the focal manager tohave both coordination and appropriation concerns(Gulati and Singh, 1998), even if all managers havecomplete information on the task requirements.Specifically, we know little about the economic orsocial factors that enhance or mitigate behavioraluncertainty, particularly as organizational architec-tures evolve from one form to another.

Behavioral uncertainty is particularly salient incontexts where divisional managers have to inter-act with other managers in redefined structures.Routines created in prior organizational structuresmay be ‘sticky’ and transfer over to the newstructural contexts (Gulati and Puranam, 2009).For example, when one firm acquires another,there are numerous challenges in achieving coor-dination of activities between the acquired andacquiring entities (Anand and Singh, 1997; Birkin-shaw, Bresman, and Hakanson, 2000; Graebner,2004; Karim and Mitchell, 2000; Puranam, Singh,and Zollo, 2006). Scholars have noted that a keysource of heterogeneity in acquisitions relates towhether the acquired entity was a prior alliancepartner (Benson and Ziedonis, 2010; Dyer, Kale,and Singh, 2004; Higgins and Rodriguez, 2006;Vassolo, Anand, and Folta, 2004; Zaheer, Hernan-dez, and Banerjee, 2010). Thus, post-acquisitionsettings provide a rich context within which wecan examine the effects of behavioral uncertaintyon coordination. If prior experience in an alliancesetting permits divisional managers to have alreadydeveloped trust, common understanding, andshared routines (Agarwal, Croson, and Mahoney,2010; Gulati, 1995; Gulati and Nickerson, 2008;Sarkar, Aulakh, and Madhok, 2009; Zhao and

Anand, 2009), the reduction in behavioral orpartner-specific uncertainty may result in greatervalue creation relative to outright acquisitions(Higgins and Rodriguez, 2006; Zaheer et al.,2010). Anecdotally, this is exemplified by acquisi-tions undertaken by Cisco, where managers explic-itly acknowledge the use of alliances to ‘evaluatefirms to determine if acquisitions will work’ (Dyeret al., 2004: 114), and Siemens, whose venturearm has stated that ‘every investment is a potentialacquisition’ (Wieland, 2005: 1).

However, superior post-acquisition performanceis not guaranteed when the acquisition is amongprior alliance partners. For example, AOL’s acqui-sition of prior alliance partner BEBO resulted ina subsequent divestiture at a loss of more than$1 billion (Helft, 2010). Consistent with this exam-ple, Benson and Ziedonis (2010) found that out-right acquisitions may outperform acquisitions ofalliance partners, and Zaheer et al. (2010) failedto find a positive main effect of acquisition ofprior alliance partners relative to outright acqui-sitions. Therefore, it is important to understandhow the information flow mechanisms highlightedin the organizational design literature, namelyresource allocation decisions and direct communi-cation (Galbraith, 1973; Noda and Bower, 1996),may address behavioral uncertainty and influencedivisional managers’ inferences regarding coordi-nation and appropriation costs (Gulati and Singh,1998).

We draw on a parallel literature stream based onboth social psychology and economic game theo-retic logic that has examined the role of behavioraluncertainty on successful coordination in interor-ganizational settings such as alliances (Khanna,Gulati, and Nohria, 1998, Gulati, Khanna, andNohria, 2000; Agarwal et al., 2010). Notably,these scholars have examined the role played byresource allocation and communication in pro-viding managers with valuable information thatshapes inferences regarding partner intent andimpacts the success or failure of their cooperativeendeavors.

Specifically, we examine two related researchquestions. First, how does the presence or absenceof pre-acquisition resource allocation and commu-nication in alliances impact post-acquisition per-formance? Second, do routines that are formedin the alliance stage impact post-acquisition per-formance? To do so, we identify the main andadditive effects of information flows through prior

Copyright 2012 John Wiley & Sons, Ltd. Strat. Mgmt. J., 33: 710–733 (2012)DOI: 10.1002/smj

712 R. Agarwal et al.

resource allocation and prior communication. Ouruse of experimental methods (Croson, Anand,and Agarwal, 2007; Agarwal et al., 2010; Knezand Camerer, 1994; Song, Calantone, and DiBenedetto, 2002; Weber and Camerer, 2003) allowsus to set aside the effects corresponding to non-random selection of alliances that are followed byacquisitions (as well as task uncertainty) and fur-ther isolate the effects of such mechanisms thatoften coexist in real-world settings through the cre-ation of treatments that relate to each mechanismalone.

Our study finds that divisional managers makeinferences about the likelihood that the other man-agers will cooperate based on whether positive ornegative routines are formed during the interac-tions in alliance settings. Prior resource allocationalone makes it more likely that negative routineswill be formed, but prior communication allowsmanagers to create positive routines through moreequitable sharing of the value created. Thus, ourresults underscore the importance of using ‘rich’communication media among divisional managers(Lengel and Daft, 1984), not only for the resolutionof task uncertainty, but also due to the beneficialeffects on behavioral uncertainty.

Our study makes important contributions to theacademic literature on organizational architecturesand design and offers key managerial implica-tions. We refine core concepts in organizationaldesign and illuminate how post-acquisition orga-nizational reconfiguration may interact with sticki-ness of routines established in prior organizationalrelationships to impact performance. By abstract-ing away from issues related to selection and endo-geneity, we focus on the causal mechanisms thatoccur during the alliance to unpack some reasonsfor the contrary findings regarding value creationin acquisition of prior alliance partners relativeto outright acquisitions. By highlighting issues ofbehavioral uncertainty and integrating perspectivesfrom economic game theory with social psychol-ogy approaches, our research reveals the benefi-cial effects of informal mechanisms on manag-ing the tensions between cooperation and com-petition that can arise in economic exchange. Ofparticular significance are intertemporal spilloversacross organizational architectures; we show howantecedent interorganizational relationships havea dynamic and path dependent effect on theinteractions within organizations and the ability

to manage ongoing tensions in an intraorgani-zational framework. From a managerial perspec-tive, our results temper the conventional wisdomthat pre-acquisition alliances always improve post-acquisition performance. Further, they highlightthat alliances are not just opportunities for selec-tion, but also for creation of potentially success-ful targets. Doing so requires managers to investin ‘rich’ mediums of information exchange, suchas direct communication and feedback (Daft andLengel, 1986). Even though this may be expensivein terms of managerial time and attention (McCannand Galbraith, 1981), the returns may be worth itin the longer term when one takes into considera-tion the positive spillover effects across organiza-tional architectures. Additionally, our study pro-vides a cautionary note to those managers whoseek to acquire partners in failing alliances toimprove performance through the dissolution offirm boundaries—competitive and negative rou-tines that are created in unsuccessful alliance expe-riences are also likely to hamper value creationpost-acquisition.

THEORY

Theoretical backdrop

At the core, organization design grapples with twocomplementary problems: (1) how to best partitiontasks across organizational players; and (2) howto reconnect these organizational elements to bestrealize the organization’s strategic goals (Burton,DeSanctis, and Obel, 2006; Lawrence and Lorsch,1967). In the context of the latter problem, orga-nizational architectures have to enable effectivegathering and processing of information to reduceuncertainty—due to an incomplete description ofthe world (Arrow, 1974) and ambiguity—arisingfrom ‘the existence of multiple and conflictinginterpretations about an organizational situation’(Daft and Lengel, 1986: 556) Interunit task inter-dependence increases uncertainty facing divisionalmanagers and, thus, increases information process-ing requirements (Tushman and Nadler, 1978).

Only by processing information can organiza-tional decision makers observe the situation, ana-lyze existing problems, make efficient choicesabout alternative actions, and convey these choicesto others (Burton et al., 2006). The informationmechanisms are varied and range from routines



and codified information through direct communi-cation (Burns and Stalker, 1961; Daft and Lengel,1986; Galbraith 1973, 1977; Miller and Friesen,1982). Further, there are two components to therequisite information—one that is objective andrelates to the task requirements and the other thatis ‘socialized,’ relating to trust and expectationsregarding potential cooperative behavior by othermanagers (Gulati, 1995). The former involves theinvestment of effort and/or exchange of goods andservices according to the agreed upon and specifiedrules and rewards (Macneil, 1978). The latter, con-versely, is relationally based to address potentiallyconflicting interpretations or incentives across mul-tiple actors and is associated with social processesthat can generate norms of flexibility, solidarity,and information exchange (Macneil, 1980; Heideand John, 1992; Zaheer and Venkatraman, 1995)to help establish ‘common ground’ (Clark, 1996).Such common ground allows interacting parties tosynchronize and adjust their actions to each other(Becker and Murphy, 1994; Schelling, 1960). Ofcourse, both components are intertwined in real-ity and successful coordination occurs when theinformation mechanisms provide managers boththe knowledge of the task requirements and theassurance that other managers will undertake theirassigned actions.

While the former component of task uncertaintyhas been widely studied in the organization designliterature (Burns and Stalker, 1961; Burton andObel, 2004; Donaldson, 1987, 1990; Galbraith,1973; Lawrence and Lorsch, 1967; Miller, 1986),relatively less attention has been paid to howmanagers approach the resolution of behavioraluncertainty (Grote, 2009; Leifer and Mills, 1996;Tushman and Nadler, 1978).1 This is problematicbecause even if all managers have full informa-tion on what is required for successful completionof the task, uncertainty regarding whether othermanagers will behave to achieve the necessary

1 The potential importance of behavioral uncertainty is alluded toin the organization design literature, even though it has not beena subject of significant scholarly attention. For example, Tush-man and Nadler (1978) highlight interunit task interdependenceas a key source of work-related uncertainty with structural impli-cations. They acknowledge the potential for diverse interests andmiscommunications between units, but focus mainly on task-based joint problem solving issues. Similarly, Leifer and Mills(1996: 116) draw attention to the interpretation challenges giventhe numerous meanings that can be attributed to the informationprovided by various organizational players who approach prob-lems with ‘different experiences, cognitive perspectives, goals,values, and priorities.’

coordination along with value appropriation con-cerns (Gulati and Singh, 1998) may cause them toact in a manner contrary to achieving success.

In this context, it may be worthwhile to integrateinsights from parallel literature in economic gametheory and social psychology related to intra- andinterorganizational coordination (Agarwal et al.,2010; Khanna et al., 1998; Zeng and Chen, 2003).2

As Nadler and Tushman (1997: 14) note, ‘the goalof organization design is to fashion a set of for-mal structures and processes that, together with anappropriate informal operating environment, willgive people the skills, direction, and motivationto do the work necessary to achieve the strate-gic objectives.’ Using economic game theoreticand social psychology logic, the critical organiza-tional design issue can be thought of as configuringboth structural and motivational solutions to thecoordination problem within a ‘social dilemma’ orassurance game setting. As such, the coordinationchallenge is to assure the other partner of one’sown cooperation so that they are more likely tocooperate themselves, thus enhancing the likeli-hood the partners will jointly invest in activitiesthat create more value than each can create alone(Agarwal et al., 2010; Gulati et al. 1994).

Consider a hypothetical assurance game settingwhere two division managers in a firm have tocoordinate for joint value creation. We illustrate theassurance game in Table 1, which provides infor-mation on the ‘structural’ or economic incentivesfor potential coordination in the form of ‘payoffs’to each manager based on alternative outcomes.To create value, each manager needs to allocateresources to the joint activity. The opportunity cost

Table 1. Assurance game

Manager 2

Cooperate Do notcooperate

Manager 1 Cooperate (10000, 10000) (3200, 6800)

Do notcooperate

(6800, 3200) (8000, 8000)

2 For a discussion on why assurance games (with multiple Nashequilibria and no dominant strategy) are more representativeof inter- and intraorganizational cooperation than prisoner’sdilemma games (typically characterized by a single Nash equi-librium given symmetric dominant strategies), please refer toGulati et al. (1994) and Agarwal et al. (2010).



of contributing to the joint activity is the fore-gone ability to invest in division-specific activitieswhose payoffs are not dependent on the other divi-sion’s cooperation. The rewards of the investmentsare contingent on whether the other division man-ager also contributes the requisite resources to thejoint activity. The assurance game, thus, representswhat Thompson (1967) would classify as recipro-cal task interdependence. Given the payoff matrixin Table 1, the optimal decision for each man-ager is dependent on the other manager’s decision(Agarwal et al., 2010; Harsanyi and Selten, 1988).

There are two potential Nash equilibria in thisgame. The first equilibrium outcome—payoffdominant/cooperative—occurs when both man-agers invest resources in the joint activity toalign actual and potential value creation and earn$10,000 each as payoffs. This equilibrium is calledpayoff dominant because it maximizes individ-ual and collective payoff. Since each managerexpects the other to cooperate, cooperation is thebest strategy. This outcome is a Nash equilib-rium since no manager is better off deviating fromtheir strategy, given the other manager’s strat-egy. The second equilibrium outcome—risk dom-inant/noncooperative—occurs when neither man-ager invests in the joint activity, resulting in eachearning a payoff of $8,000. This equilibrium iscalled risk dominant because it minimizes theexposure of the managers to the risk that the othermay not contribute; however, each manager earnsless than they would in the payoff-dominant equi-libria. The risk-dominant outcome is also a Nashequilibrium: not cooperating is the best strategyunder the expectation that the other manager willnot cooperate, and there is no incentive for eithermanager to deviate from their strategy given theother’s strategy.

We note that this task interdependency prob-lem has no task-related uncertainty, in the sensethat the outcomes are completely determined con-tingent on each manager’s actions. Yet, there isbehavioral uncertainty stemming from the fact thateach manager is unsure of whether the other man-ager will act cooperatively. While economic incen-tive alignment predisposes toward greater valuecreation through joint activity, the lack of a dom-inant strategy and unique equilibrium implies thathigher potential payoffs from the joint activity maynot be sufficient for ensuring actual value creation(Agarwal et al., 2010; Zeng and Chen, 2003). The

information conveyed via structural interaction ele-ments (i.e., resource allocation) alone may be inad-equate to catalyze cooperation, as it arrives absentsupporting explanatory data (Daft and Lengel,1986). In such circumstances, motivational solu-tions that facilitate coordination (such as commu-nication, articulation of common goals, and devel-opment of shared trust) may enhance actual valuecreation. These ‘richer’ modes of interaction canaid interpretation and subsequently increase under-standing of, and congruence in, partner actions(Russ, Daft, and Lengel, 1990). Consistent withthis argument, Agarwal et al. (2010) found thatenabling communication among alliance partnersdoubled the probability of achieving successfuloutcomes relative to economic incentive alignmentalone.

Though the assurance game settings have beenused largely to examine interorganizational coor-dination (Agarwal et al., 2010; Gulati et al., 1994;Zeng and Chen, 2003), they are equally appli-cable to intraorganizational coordination settings.Thus, against this theoretical backdrop, we beginour formal theorizing regarding coordination out-comes that occur after a merger or an acquisition,which is the specific phenomenon through whichwe highlight the role of information mechanismsin addressing behavioral uncertainty.

Post-acquisition performance

Post-acquisition performance is determined bymany factors: the underlying reasons for the acqui-sition, firms’ capabilities and their synergies, theprice paid for the target firm, and the post-acquisition coordination of the firms (Haspeslaghand Jemison, 1991). In this study, we focus onpost-acquisition coordination for two importantreasons.3 First, surveys of Fortune 500 CEOs

3 Given variation in acquirers’ value creation models, there canbe significant variance in the extent and nature of the post-acquisition coordination desired. As Haspeslagh and Jemison(1991), Puranam et al. (2006), and Puranam et al. (2009) haveargued, under certain conditions, too much coordination can beunnecessary and even harmful. However, there needs to be atleast some degree of interdependence established among theacquired and existing organizational subunits. At a minimum,such interdependence can come from participating in an internalcapital market, though often there is also an internal labor marketor other forms of resource exchanges even under ‘holding’or ‘preservation’ models (Haspeslagh and Jemison, 1991). Thesimple act of bringing a transaction internal to an existing firmhas ramifications for both incentives and administrative controls.



find that missteps in post-acquisition coordina-tion consistently rank among the top reasons formerger failure (Schmidt, 1999). This has been cor-roborated by extant academic research (Haspes-lagh and Jemison, 1991; Marks and Mirvis, 1998;Nahavandi and Malekzadeh, 1988; Lubatkin et al.,1998; Puranam et al., 2006; Schweiger and Goulet,2000; Ranft and Lord, 2002; Weber and Camerer,2003). Second, the degree of organizational designchanges required when two firms merge depends,in part, on the preexisting organizational relation-ships between the firms.4

We highlight the nature of pre-acquisition inter-actions between the target and the acquiring firmand explore the effects of these interactions onpost-acquisition performance. We investigate howboth pre-existing resource allocation decisions andthe ability to communicate provide informationthat shapes the routines developed and the conse-quences of such routines on post-acquisition per-formance.

Pre-acquisition organizational architectures

The prior relationships between targets and acquir-ing firms vary greatly. At one extreme, an acqui-sition may occur in the absence of any priorrelationships. For example, in seeking strategicrenewal, an organization may enter a new productor service area by acquiring a previously unasso-ciated firm (Agarwal and Helfat, 2009; Anand andSingh, 1997; Karim and Mitchell, 2000; Capronand Mitchell, 2009; Anand, Oriani, and Vassolo,2010). Facing competitive pressures to rapidly gaina position in a new emerging opportunity, man-agement may decide to acquire a firm with whichit has had no economic resource allocation-basedinteractions and also little opportunity to communi-cate via an extended evaluation process. In the caseof outright acquisition, the two firms start from

4 Post-acquisition coordination between acquired and existingdivisions is an important issue, with theoretical and practicalsignificance. Conceptually, this phenomenon represents a naturalexperiment in dissolution of firm boundaries. When coordinationis ‘complete,’ it implies that two previously autonomous firmsnow behave as a single economic and social entity, thus, havingremoved their interfirm boundary. It is a good context to pro-vide insights into the theory of the firm. In terms of practicalsignificance, poor coordination has been identified as a frequentculprit behind poor acquisition performance (e.g., Schweiger andDeNisi, 1991). While firms with successful coordination can reapbenefits from acquisitions, other firms suffer from managerialturnover, poor decision making, and internal feuds, all leadingto financial losses.

a ‘clean slate.’5 For examples, Cisco’s numerousoutright acquisitions include those of Pure DigitalTechnologies and Starent Networks (Vance, 2009).

Alternatively, the acquisition may follow a com-plete alliance, where acquiring and target firmshad fully developed joint resource allocation andcommunication-based interactions that provideinformation and guide partner behavior in thepre-acquisition period. The task-based incentivesdrive resource allocation choices by the partners.However, these choices may be further influ-enced by (or interpreted in light of) the informalorganizational norms that incorporate contextualinformation gained from communications between,and learning about, partners. For example, IBMacquired Cognos, Inc., in 2007 after several yearsof a successful alliance partnership where the twofirms collaborated in a service-oriented architec-ture (Austen, 2007). During the alliance stage,the two firms met corporate customers’ needs byallocating their complementary resources so thatCognos provided business intelligence and con-sulting services, while IBM provided integratedhardware and software solutions. These resourceallocation decisions were also accompanied by sig-nificant communications where both parties devel-oped a common understanding and trust regardingpartner abilities and intentions. The presence ofthese structures in a complete alliance setting, thus,implies that the post-acquisition coordination chal-lenge would be vastly different than in the outrightacquisition setting, as depicted in Table 2.

Table 2. Pre-acquisition organizational interaction

Prior communication

Yes No

Prior resourceallocation

Yes Completealliance

Allocation only

No Communicationonly

Outrightacquisition

5 We note that such outright acquisitions may occur due to dif-ferences in the extent to which managers at different hierarchieshave been involved in prior negotiations. For our purposes, evenif there may have been extensive communication between the topmanagement team of the two firms, relative absence of interac-tion between the mid- and lower-level managers of the relevantdivisions may still result in the ‘clean slate’ outcome.



Table 2 also highlights the possibility of the non-diagonal cells of allocation only and communica-tion only. While not plausible in the real world,6

these cells are theoretically important since theyallow us to isolate the informational effects ofresource allocation and communication-based link-ages, where one occurs in the absence of theother. It should be noted that these linkages varyin the ‘richness’ of information they carry. Ascompared to communication, resource allocationactivities are more circumscribed and provide lim-ited emotional cues and fewer opportunities forrapid feedback. (Daft and Lengel, 1986; Lengeland Daft, 1984). Direct communication is a richermedium that enables the transfer and assimila-tion of nuanced messages and offers the meansto confirm understandings and correct misinter-pretations (Russ et al., 1990). Yet, as highlightedby McCann and Galbraith (1981), direct feed-back mechanisms are the most expensive in termsof managerial time and attention. Identifying themain and additive effects of these linkages permitsa better understanding of the relative importanceof these two critical dimensions of organizationaldesign to achieving post-acquisition coordination.Importantly, it also enables a better understand-ing of what types of prior interactions are morelikely to transfer from one organizational struc-ture (alliance) to the other (acquisition). Thus, thefour types of pre-acquisition relationships depictedin Table 2 flow from the potential combinationsof organizational architectural elements: an orga-nizational structure associated with the ability tocommunicate and an organizational structure asso-ciated with resource allocation. We now turn to thehypothesized effects of each of the organizationalcomponents on post-acquisition performance.

6 In the context of allocation only, there is a possibility thatfirms may engage in an ‘arms-length’ alliance; where the tasksand economic incentives are set and the partners then engage inresource allocation, with only limited communication or one-on-one interaction. Thus, partners rely on their independentand individual cost-benefit calculus while making their deci-sions, forming impressions of their partners based solely on therevealed resource allocation decisions in the alliance. In theseallocation only interactions, the resource allocation experiencemay create routines that influence later interactions. The real-world plausibility of communication only is far lower, thoughone may envision a case where communication enables thetwo firms to learn about each other and the tasks required forsynergistic outcomes, prior to any ‘economic’ transactions orcombined operational tasks. Such dialogue and observation mayenable managers to gain an awareness of the dominant opera-tional culture as well as catalyze social ties formation (Smith-Doerr and Powell, 2005).

PRE-ACQUISITION ORGANIZATIONALARCHITECTURE ANDPOST-ACQUISITION PERFORMANCE

Prior resource allocation experience

We begin by a comparison of Rows 1 and 2 inTable 2: both complete alliance and allocation onlyscenarios provide prior resource allocation experi-ence to the managers of the two firms that willultimately be integrated in a single organization.In contrast, the managers of the two firms in Row2 do not gain comparable experience through jointallocation of resources. For these managers, effortsto coordinate value-creation activities occur onlywithin the integrated entity. Compared to Row 1,the decision makers in Row 2 face a delay in gain-ing hands-on knowledge of the task and developingpartner-specific absorptive capacity (Zaheer et al.,2010), which translates into a corresponding lag inefforts to move down the learning curve. Thoughthe timing and, thus, the organizational contextwill differ, both sets of players will face similaroperational challenges—inducing sufficient contri-butions across partners to achieve greater mutualbenefit.

The challenge of coordinating to support value-creation is not trivial. Even with no task uncer-tainty, managers face the problem of multipleequilibria, as depicted in the previously discussedassurance game setting (Agarwal et al., 2010;Harsanyi and Selten, 1988). If the acquired andacquiring firms have no prior resource alloca-tion experience, they have not had the oppor-tunity to either learn about the task or developproclivity toward either cooperation (the payoff-dominant equilibrium) or noncooperation (the risk-dominant equilibrium). However, if the two firmshave engaged in prior resource allocation in thepre-acquisition setting, the knowledge gained fromthis experience will likely spill over with implica-tions for post-acquisition coordination.

Importantly, an interorganizational context isdifferent from coordination within organizations,since the former are more likely to experiencegreater tensions between cooperation and com-petition (Hamel, 1991). While competition mayexist in intraorganizational settings as well, thetensions are more salient in an interorganizationalsetting (Khanna et al., 1998; Williamson, 1991).Thus, prior resource allocation experience amongalliance partners may set them on one of two



divergent paths. If the prior resource allocationexperience results in cooperative action being cho-sen, there may be positive spillovers between thepre-acquisition activities and post-acquisition per-formance. Alternatively, if the prior resource allo-cation experience results in noncooperative action,there may be negative spillovers for the new orga-nizational design choice. Since each path is viable,in lieu of offering competing hypotheses, we leavethe performance ramifications of prior resourceallocation as an empirical question.

Prior communication experience

In contrast, the comparison between Columns 1and 2 of Table 2 illuminate the effect of commu-nication between the acquiring and acquired firmand the potential for development of some sharedculture and social norms. Prior communicationcan provide information that alleviates both opera-tional and behavioral challenges to value creation.First, through communication, the firms can pro-vide information about their respective positions,gain insight into their potential struggles in manag-ing the interdependent task and improve economicreturns post-acquisition (Bastien, 1987; Shanley,1988; Napier, 1989; Bohl, 1989; Schweiger andDeNisi, 1991).

Communication facilitates a flow of informationthat can clarify expectations and causal connec-tions between individual actions and group out-comes (Kogut, 2000). Communication can alsoreduce behavioral uncertainty in the presence ofbounded rationality (Simon, 1947; Zeng and Chen,2003). Direct communication supports the trans-fer of ‘rich information’ enhancing the capacityof managers to process complex subjective mes-sages and resolve ambiguity in a timely manner(Lengel and Daft, 1984). Insights from the socialpsychology ‘contact hypothesis’ illuminate howcontact and communication may reduce tensionbetween groups (Allport, 1954; Pettigrew, 1971;Brewer and Brown, 1998). Thus, prior communi-cation can help the acquiring and acquired firmsalleviate problems that may arise post-acquisition.By reducing the possibility of surprises, com-munication can provide convergent expectationsthat enhance the coordination and cohesion of thegroup (Malmgren, 1961; Williamson, 1975). Evenin the absence of resource allocation interactionsthat may occur when the target is a former alliance

partner, communication plays a vital role; the eval-uation process is not only important in ensuringthat the right partner is selected, but also criticalto ensuring common ground between the two firmsis established (Jemison and Sitkin, 1986; Bastien,1987; Haspeslagh and Jemison, 1991). Further,when the prospective target is an alliance part-ner, research on strategic alliances has underscoredthe role of communication among decision makersfor increasing performance (Agarwal et al., 2010;Rodan and Galunic, 2004; Zaheer and Venkatra-man, 1995). Communication may engender trustbetween the transacting parties, increase familiar-ity with one another, strengthen personal ties, andconfirm the goodwill of both organizations (Gulati,1998, 1999; Dyer and Chu, 2000). With commu-nication, the potential gains to cooperative actionare perceived as more certain as such interactionsenable the parties to come to a shared understand-ing on ambiguous issues (Russ et al., 1990). Relat-edly, prior communication provides opportunitiesfor moral suasion, aiding in the development ofgroup identity and raising the salience of groupinterests, which may increase the likelihood ofcooperation (Dawes, van de Kragt, and Orbell,1988; Komorita and Parks, 1994; Zeng and Chen,2003).

In sum, when acquired and acquiring firms haveengaged in prior communication, the increasedflow of information and creation of social normsenable a smoother transition to activities per-formed within the integrated organizational design.In the language of social dilemma and assurancegames, communication enhances the probability ofthe payoff-dominant relative to the risk-dominantequilibrium (Agarwal et al., 2010; Gulati et al.,1994; Harsanyi and Selten, 1988). So:

Hypothesis 1 (H1): Prior communication expe-rience will result in higher post acquisition per-formance than no prior communication experi-ence.

Joint effects of prior resource allocation andprior communication

The previous two sections relate to the ‘maineffects’ of prior resource allocation and communi-cation; we now examine the impact of one in thepresence/absence of the other. Understanding theseresults, we argue, will help us understand how thetwo experiences combine to affect performance inthe post-acquisition organizational framework.



While prior resource allocation creates the struc-tural framework for economic exchange, priorcommunication provides additional motivationalsolutions to the social dilemma problem (Agar-wal et al., 2010; Zeng and Chen, 2003). Purelystructural linkages may be incapable of provid-ing the necessary breadth of information and, thus,may result in poor coordination (Nadler and Tush-man, 1997). Some of the gaps in the former typeof structure can be filled as the parties communi-cate and develop a shared set of norms that fur-ther direct partner behavior.7 With future decisionsguided by a more comprehensive base of sharedknowledge (which includes not only the outcomesof one’s partner’s decisions but information aboutthe reasons for these decisions), successful out-comes are more likely. Both Gulati et al. (1994)and Orbell, Dawes, and van de Kragt (1990) high-light the need for multilateral promises to aug-ment formal decision rules that may flow fromthe underlying incentive structure. These commu-nication efforts may exert influence on decisionmakers that reinforces identification with the coop-erative system (Barnard, 1938; Simon, 1947) andincreases the likelihood of a payoff-dominant equi-librium.8

Similarly, prior communication can be aug-mented by prior resource allocation experience.First, communication is more effective when it isconcomitant with knowledge of the critical issuesthat need to be resolved. Challenges encounteredduring the prior resource allocation process mayprovide insights on what issues are most salientfor discussion. Second, communication withoutresource allocation may result in ‘cheap talk,’ since

7 Classical organization theory argues that effective coordinationis a product of both monetary incentives and nonmonetaryawards and highlights the role of both formal and informalmanagerial communication can play in increasing performance(Barnard, 1938).8 In a landmark study using a field experiment, Schweiger andDeNisi (1991) showed that communication can ameliorate thenegative effects of an acquisition. They showed that there isa heightened sense of uncertainty and greater perceived riskfollowing an acquisition, which tend to increase with time, lead-ing to reduced trust and commitment to firm goals. However,realistic communications during and after the acquisition signif-icantly reduced such outcomes. Further, research in proceduraljustice (e.g., Greenberg, 1987) shows that even when people areunhappy with an outcome, they experience fewer dysfunctionswhen they understand the process through open communica-tion. Besides the useful informational content in such commu-nications, there is also a useful symbolic component to suchcommunication.

it may not carry significant payoff-relevant infor-mation, or worse, may be strategically misleadingas partners assert intentions that are not backed byactions (Crawford, 1998; Ledyard, 1995; Farrelland Rabin, 1996).

In sum, by capturing economic incentives, thestructure of an alliance can establish a foundationfor value creation. However, value may not be real-ized due to partner miscalculation or misbehavior,which is where the shared culture and social normsenabled by communication may facilitate cooper-ative outcomes. Thus, the concatenation of boththese mechanisms can jointly enhance task-relatedlearning and, equally importantly, alleviate oppor-tunism concerns that may arise post-acquisition, aswell as increase performance. Simply put, havingboth prior resource allocation and prior commu-nication opportunities increases the likelihood ofgetting the payoff-dominant Nash equilibria, ratherthan defaulting to the risk-dominant Nash equilib-ria.9

Hypothesis 2 (H2): The acquisition of a com-plete alliance partner (presence of priorresource allocation and communication) willresult in higher performance than the acquisi-tion of an allocation only partner (presence ofprior resource allocation alone).

Hypothesis 3 (H3): The acquisition of a com-plete alliance partner (presence of priorresource allocation and communication) willresult in higher performance than the acquisi-tion of a communication only partner (presenceof prior communication alone).

Formation and transfer of routines fromalliance to post-acquisition

Formation of routines for value appropriation

Implicit in our hypotheses relating interorganiza-tional designs during the pre-acquisition stage tothe post-acquisition performance is the assump-tion that routines created during the alliance-setting

9 We note that H2 is closely related (but not identical to) H1.Both H2 and H1 suggest that communication will lead to higherperformance. However, these hypotheses are not logically equiv-alent. For example, we might see a main effect of communicationwhen comparing all four treatments (supporting H1), but noeffect of communication in the presence of prior resource allo-cation (rejecting H2). Alternately, we might see an effect ofcommunication in the presence of prior resource allocation (sup-porting H2), but not overall (rejecting H1).



transfer to the post-acquisition stage. We now turnto the subset of acquisitions that involve prioralliance partners, treatments where managers fromboth Firm A and Firm B make resource allocationdecisions in the pre-acquisition periods, to examinethis issue in-depth.

As noted earlier, while strategic alliances areformed with the intent that both partners will reapbenefits accruing to combinational activity, thereis nonetheless a competitive element as partnersmake decisions with an eye to their own privateeconomic benefits. These two elements correspondto value creation and value appropriation, respec-tively. Simply put, exchange partners can pursuetheir own interest at the expense of their partner’sby engaging in economic holdup and/or learningraces in a manner that affects both value creationand value appropriation (Doz, 1996; Khanna et al.,1998; Das and Teng, 2000). Even in the absenceof opportunistic behavior, miscalculations and mis-understanding can result in alliances falling shortof their potential in value creation (Kale, Dyer,and Singh, 2002). Thus, the alliance stage can befraught with situations where there is coordinationfailure and lack of contribution.

Agarwal et al. (2010) provide evidence thatcommunication enhances the likelihood of payoff-dominant equilibria being chosen over risk-dominant equilibria in alliance settings, thusincreasing the likelihood of value creation. Weargue that in addition to enhancing value cre-ation (creation of the pie), communication mayalso facilitate more cooperative value appropri-ation (sharing of the pie) among the decisionmakers.

In Table 1, we provided a simple example ofan assurance game setting, where the decision isa simple binary variable regarding the choice ofcooperation. In more realistic assurance game set-tings, partners have to additionally choose theirlevel of effort or amount of resource allocation.Accordingly, there are many payoff-dominant Nash equilibria that represent differentlevels of costs and, thus, share of the value, appro-priated by each alliance partner. Economic/gametheoretic logic would argue that rational part-ners (should) consider both issues simultaneously,determining their level of contribution to insurethat sufficient resources are committed to createvalue in the joint alliance activity while optimiz-ing private payoffs. Thus, each partner prefers apayoff-dominant equilibrium in which he or she

contributes relatively little while his/her partnerscontribute relatively more. For acquisitions thatachieve payoff-dominant outcomes, an examina-tion of the relative contributions of the variousparties helps determine whether the value is beingappropriated equally or unequally.

We argue that interorganizational architecturesthat rely solely on economic incentive alignmentand resource allocation as the primary form ofinteraction among alliance partners are more likelyto result in unequal value appropriation. In con-trast, interorganizational architectures that com-bine communication with economic incentives andresource allocation engender social norms andsocial identity that support more ‘equitable’ contri-butions of resources and sharing of the value cre-ation (Deutsch and Gerard, 1955; Zeng and Chen,2003). Accordingly, we posit:

Hypothesis 4 (H4): In alliances, communicationwill result in more equal value appropriationthan will no communication.

Transfer of routines post-acquisition

Previous research suggests that post-acquisitionperformance may be enhanced by the existenceof pre-acquisition alliance activity (Reich andMankin, 1984; Doz, Hamel, and Prahalad, 1986;Haspeslagh and Jemison, 1991; Bleeke and Ernst,1995; Hagedoorn and Sadowski, 1999). Pre-acquisition alliances may serve as an importantscreening mechanism for potential targets, as isargued in the real options and related literatures(Kogut, 1991; Arend, 2004). If only a ‘selected’group of alliances are followed by acquisitions,one may observe a positive relationship betweensequential strategies and performance without theexistence of a causal link between them.

Aside from selection, distinct processes formedduring the alliance stage may enhance or exac-erbate post-acquisition performance. Moving froman alliance to an acquisition entails a governancestructure transition. Though, post-acquisition, thepartners continue to contribute and combineresources, they do so under a lower-poweredincentives regime (Williamson, 1991). The poten-tial spillovers from the alliance stage to the post-acquisition stage stem from learning from/withpartners about the task or learning about part-ners and their level of opportunism (Ring and van



de Ven, 1994; Anand and Khanna, 2000). Thus,the creation of partner-specific absorptive capacity(Zaheer et al., 2010) can result in higher post-acquisition performance. Organizational architec-ture design choices of the alliance, as it contributesto this learning, may have significant consequencesin setting routines that partners carry into thepost-acquisition stage. This is particularly salientgiven inertness of organizational contractual com-mitments, which may result in governance insepa-rability and constrain future organizational designchoices (Argyres and Liebeskind, 1999).

In some instances, the routines establishedthrough alliance experience are expected to havepositive performance effects post-acquisition.Through successful alliance activity, partners gainresource exchange know-how and resource inte-gration experience (Porrini, 2004). Consistent withwhat Pisano (1994) refers to as ‘learning beforedoing,’ post-acquisition performance can be facil-itated by the opportunity that the two firms havehad to learn about how best to use each other’sresources and obtain potential value-creating com-binations in the alliance phase. Broadly, success invalue creation creates common ground—an under-stood set of shared knowledge that supports coor-dination (Clark, 1996; Puranam et al., 2009). Iftask requirements remain similar, partners with ahistory of success are likely to rely on establishedroutines, continuing to cooperate in the new post-acquisition organizational context.

Players that adopted cooperative norms in thealliance phase will find themselves more suitablypositioned to operate within the lower-poweredincentive environment. Argyres (1995), for exam-ple, provides detailed case studies of how success-ful alliance experience translated to superior post-acquisition performance, as with IBM’s acquisitionof Lotus Notes. However, the opposite outcome isalso possible—when prior experience is fraughtwith competitive rather than cooperative routines,there can be a negative effect due to the transferof competitive routines from the alliance stage tothe acquisition stage (Haleblian and Finkelstein,1999; Barkema and Schijven, 2008). Due to com-petitive pressures and opportunism concerns thatexist in alliances (Khanna et al., 1998), some part-ners will struggle to create value in the alliancestage. Particularly because alliances and acquisi-tions represent different governance modes, ele-ments of competition and opportunism in the rou-tinized interactions of the alliance stage should be

muted post-acquisition, since acquisitions result inintrafirm settings that represent joint maximizationproblems (Williamson, 1975). But once these rou-tines are set up between alliance partners, theymay persist (owing to their stickiness) even afterthe acquisition has dissolved the interfirm bound-aries and the alliance partner has been internalized(Argyres, 1995; Argyres and Liebeskind, 1999). Ifthe competitive routines developed in the alliancestage persist post-acquisition, they may result inlower performance.

Thus, the transference of inappropriate rou-tines implies that alliance partners may experiencenegative effects. Routines created under incen-tive structures where the resource allocation isbased on competition among independent organi-zations may remain sticky, thus causing frictionspost-acquisition when divisional members need toreconfigure their relationships in the context of amore hierarchical governance structure.

Hypothesis 5 (H5): Post-acquisition per-formance will increase with the level of pre-acquisition alliance performance.

Hypothesis 6 (H6): Resource allocation rou-tines (value allocation) established in a pre-acquisition alliance will be retained post-acquisition.

EMPIRICAL METHODOLOGY

Experiments as a research method

We test our hypotheses using experimental method-ology (Kahneman, Knetsch, and Thaler, 1990;Smith, 2000). The earlier theoretical section iden-tified prior resource allocation and prior communi-cation as separate drivers of post-acquisition per-formance. In addition to cleanly isolating theseindividual effects, our empirical methodologyavoids selection issues, given random assignmentof subjects to treatments, and permits the devel-opment of clear dependent measures of post-acquisition performance. Experiments have beenused for many generations in psychology, sociol-ogy, and economics to distinguish between com-peting theories, testbed possible policies, andexplore anomalies in the field (see Falk and Heck-man, 2009, for a review and methodological dis-cussion). In management, experiments have been



used to investigate issues as diverse as agency con-tracts (e.g., Tosi, Katz, and Gomez-Mejia, 1997;Parks and Conlon, 1995; Conlon and Parks, 1990),compensation and allocation decisions (e.g., Freed-man 1978; London and Oldham, 1977; Fossum,1979) and negotiating behavior (e.g., Tenbrunsel,1998; Neale and Bazerman, 1985; Bhappu, Grif-fith, and Northcraft, 1997; Thomas-Hunt, Ogden,and Neale, 2003). More specific to corporate strate-gies related to acquisitions and alliances, exper-imental design has been implemented by Weberand Camerer (2003) to examine the effect of cul-tural differences in merger failures and by Agar-wal et al. (2010) to study the relative effects ofincentive alignment and communication on successof alliances. Also, Croson et al. (2007) provide areview of the methodology for specific applicationsto corporate strategy.

The use of the experimental methodology isparticularly valuable in situations where compet-ing theories offer alternative explanations for anobserved effect. For example, in our setting,improved post-acquisition performance might becaused by selection issues (firms acquire only theirhighest-value alliance partners) or by the forma-tion of positive routines. Our experiment sidestepsselection issues and allows us to focus on thedevelopmental processes as potential causes ofthe observed relationship between pre-acquisitionactivity and post-acquisition performance. The useof an experiment controls for endogeneity in whichalliances become acquisitions—endogeneity thatis not easy to control with the use of field data(Hamilton and Nickerson, 2003). In our experi-mental setting, partners are randomly assigned toeach other and, importantly, there is no choiceabout the acquisition. This enables us to focuson the decisions in the pre-acquisition stage as acausal driver of post-acquisition performance. Ina laboratory experiment like this one, the theorycan be tested directly by controlling for extraneousfactors, much as a physics experiment might con-trol air pressure in measuring an atomic reaction.A second advantage of the experimental methodol-ogy is that it allows us to separate alternate theoriesand predictions that might not otherwise be sepa-rable (or even observed) with naturally occurringdata. In our setting, this enables us to untanglethe confound between prior resource allocationand ability to communicate that is present in thefield and tease apart the competing causal expla-nations for post-acquisition performance. Thus, we

can obtain clean measures of our independent vari-ables in a way that would not be possible usingobservational data.

Similarly, the use of experiments permits a clearmeasure of our dependent variable of interest,post-acquisition performance, without relying onnoisy or possibly biased data like stock marketreturns. Finally, since experiments are replicable,other researchers can reproduce the experiment andverify the findings independently.

While experiments provide some advantages,they have limitations as well.10 These primarilyinclude an abstraction from reality and the relianceon subjects whose incentives may not reflect thoseof real-world actors. We accommodate these lim-itations by ensuring that the relevant factors fromthe field are captured in the experimental design. Inparticular, we ensure that the experimental partic-ipants’ incentives match as closely as possible theincentives of the real-world actors. Similarly, thetask assigned to the participants captures impor-tant aspects of post-acquisition coordination, whileretaining the simplicity needed to provide a cleanoutcome measure for analysis. Finally, we rely onsubjects who have managerial experience to proxyfor real-world decision makers.

Experimental design and procedure

The experiment was designed so participants couldreplicate the role of managers in firms engagingin pre-acquisition/post-acquisition activities. In our

10 Critics of experiments argue that since the laboratory situationis necessarily abstract and unrealistic (in that it contains fewerconsiderations, dimensions, and confounds than the real-worldsituation), no results from the lab can be used to predict behaviorin the field. We disagree. Zelditch (1969) discusses this issuein-depth in his amusingly titled article Can you really studyan army in the laboratory? His argument acknowledges thatthe laboratory setting is different from any naturally occurring,real-world setting one is likely to find. However, he argues,the bridge between the lab and the field is the theory beingdeveloped to explain observed behavior and being tested in thelaboratory design. Theories are developed to predict and explainreal-world observations. These theories should also predict andexplain behavior in laboratory settings. If they do not, it isnot the fault of the experiment, but a flaw in the theory. Thisargument is also made by Plott (1991) in the context of marketexperiments. He argues that the theory of market equilibrium,if true, should predict behavior in a laboratory market just as itshould predict behavior on the London Stock Exchange. In fact,it should perform better in the lab, since confounding factorsnot incorporated in the theory are absent in that setting. If thetheory doesn’t predict in the clean, uncomplicated environmentof the lab, how likely is it to predict in the cluttered, confoundedenvironment of the field?



experimental setting, we simulated two firms, FirmA and Firm B. Firm A is represented by two divi-sional managers who interact with each other. Thetwo divisions share profits. Firm B is representedby a single manager. When the experiment begins,Firm B is involved (or not involved) in an alliancewith the two divisions of Firm A. After a fewinitial periods, Firm A acquires Firm B and themanager of Firm B becomes the third divisionalmanager in the now ‘merged’ firm. As a result, allthree divisions now share profits.

The managers choose how many resources toallocate toward their own division/firm’s produc-tion and toward joint production (combined activ-ities) with the other divisions/firm (which couldbe conceptualized as joint production, R&D, tech-nology transfer, or marketing). Resources allo-cated toward production in their own division/firmgenerate a profit stream commensurate with theinvestment. Resources allocated toward combinedactivities, described as firm (or alliance) produc-tion, generate returns only if a minimum thresh-old is met, to capture synergistic gains from thesecombined activities. If this threshold of resourcesallocated toward combined activities is met, theproject is successful, value is created, and eachfirm earns additional profit. Each manager has dif-ferent monetary benefits from the joint produc-tion’s success, which affects their decisions con-cerning how much of their resources to contribute.If the resources allocated toward combined activ-ities fail to meet the threshold level, the projectfails, no value is created, and contributed resourcesare lost. The experimental design ensures that thethreshold requires more resources than are con-trolled by any one manager, thus ensuring taskinterdependency and adhering to the principle thatsuccessful combined activities requires participa-tion of multiple divisions and/or firms.

While the implemented experimental design mir-rors the simple threshold social dilemma/assurancegame example in our theoretical backdrop, severalfeatures that increase complexity and correspon-dence to the real world are worth noting. First, theparticular parameters are chosen to induce a setof efficient and stable outcomes (payoff-dominantequilibria) in which combined activities are suc-cessful. Second, the resource allocation is efficient;the sum of the profits that result from success-ful combined activities is larger than the cost offoregone profits from private production. Third,each manager’s bonus is sufficiently large so as

to induce allocation of resources toward the com-bined activities, given that they believe others willdo so as well. Finally, there is an asymmetry ofcosts of resource allocation across the multiplepay-off dominant equilibria, which results in dif-ferences in the extent of value appropriated byeach manager. Each payoff-dominant equilibriumrequires that managers not only have to agree onwhether the threshold will be met, but also onthe amount of resources that each will allocate toachieve the threshold level. This is important sinceeach manager would prefer the payoff-dominantequilibrium in which he or she contributes rela-tively little while his/her partners contribute rela-tively more.

Appendix 1 describes the parameters used forthe treatments, both pre- and post-acquisition.Managers engage in multiple rounds of interaction(described as financial quarters). To avoid endgameeffects and simulate the unknown endpoint featureof the real-world setting, we implement a finitegame with an unknown end. The participants donot know when the game will end; instead theyare informed that after each quarter there is an 80percent chance that the game will progress to thenext quarter (and a 20 percent chance that it willend). The continuation probabilities are indeed asdescribed; no deception is used in any feature ofthis experiment.

Treatments

Our experimental design manipulates pre-acquisition interaction to test if post-acquisitionperformance is enhanced by the existence of pre-acquisition interactions of different types.Accordingly, there are two stages in the experi-ment. Stage 1 (the pre-acquisition stage) consistsof three quarters. At the end of the third quarter,an announcement of the acquisition is made andStage 2 (the post-acquisition stage) commences.Mirroring Table 2 in the earlier theoretical section,we distinguish between the competing causes ofhypothesized differences: prior resource allocationand prior communication. We implement a 2x2design that manipulates the pre-acquisition rela-tionship between the two divisions of Firm A andFirm B. The two dimensions of the design are priorresource allocation (yes, no) and prior communi-cation (yes, no).

The first manipulation focuses on prior resourceallocation (Row 1 versus Row 2 in Table 2) and



is implemented based on whether the manager ofFirm B is able to make resource allocations inthe pre-acquisition quarters. In the communicationonly and outright acquisition treatments, duringthe first three quarters only the two divisional man-agers of Firm A make resource allocations. In thecomplete alliance and allocation only treatments,all three managers make resource allocation deci-sions. The second manipulation focuses on priorcommunication (Column 1 versus Column 2 inTable 2) and is implemented via a chat box. Inthe treatments without prior communication withFirm B (allocation only and outright acquisition),only the two divisions within Firm A can dis-cuss the situation and their decisions. In the treat-ments with prior communication (communicationonly and complete alliance), the manager of FirmB can observe and participate in the discussions.

In the post-acquisition stage, the experience isthe same for all participants in all four treat-ments. New parameters values are introduced (asdescribed in Appendix 1) to represent payoffs con-sistent with the post-acquisition setting; in particu-lar, Firm B now represents a division that engagesin profit sharing with the other two divisions.However, the task of resource allocation remainsthe same. The three divisions have the opportu-nity to allocate resources toward (and reap ben-efits from) combined activities. We gauge perfor-mance by measuring coordination success and rela-tive contributions achieved in this post-acquisitionphase. Note that all groups experience an acquisi-tion, even those who have had unsuccessful pre-acquisition alliances. This eliminates the selectionendogeneity present in field data.

Participants

Our experiment involved 213 participants, all ofwhom were MBA students at a major businessschool and had prior work experience. Addition-ally, some of the experiments were conductedusing participants from the Executive MBAclasses—these participants represent people withcurrent managerial experience in a variety of cor-porate settings. There was no significant differencein the results across the regular MBA and execu-tive MBA pool.

To induce participants to take tasks seriously,we used the induced valuation methodology ofexperimental economics (Smith, 1982, 2000). Thisinvolves paying participants in cash based on the

profits earned in the experiments. The importantdimension of this payment is that it is responsiveto the decisions that participants (and their counter-parts) make. We designed the incentives to reflectrealistic pre- and post-acquisition earnings of man-agers in the field, and participants were (privately)paid those incentives in cash at the end of theexperiment. When the experiment ended, partic-ipants were debriefed and dismissed. The exper-iment involved no deception, thus contaminationeffects are not a major concern. Nonetheless, par-ticipants were asked to not discuss the experimentwith others. The data was collected in three ses-sions scheduled across three days.

Procedure

Participants were randomly assigned to treatmentand role within the treatment. Consistent with ourtheoretical focus and the likelihood that some treat-ments are more likely to occur in the real worldthan others, we oversampled the diagonal cellsof complete alliance and outright acquisitions.Specifically, we ran 25 games of complete alliance,20 games of outright acquisition, 14 games of allo-cation only, and 12 games of communication only.When participants arrived at the lab, they wereshown to a computer terminal, read and signeda consent form, and were given a copy of role-specific instructions to read. Once everyone hadarrived, a composite version of the instructionswas read aloud (see Appendix 2 for the compos-ite version). After the instructions, but before theexperiment began, participants completed a quiz toensure understanding of the decisions required andthe resulting payoffs. The entire experiment wascomputerized, run via the Web using a Java appli-cation; participants input their allocation decisionswhen appropriate, engaged in electronic chat, andwere given feedback entirely electronically. Afterthe experiment ended, participants completed anexit survey describing their experiences.

Variable definitions

Dependent variables

Post-acquisition performance is measured usingpost-acquisition success as the dependent variablefor H1-H3 and H5. Success in coded as ‘1’ if thetotal resource allocations by all divisions meetsor exceeds the threshold amount required for syn-ergistic profits from combined activities and ‘0’



otherwise. Whereas H1-H3 and H5 explore value-creation outcomes, the remaining two hypotheses(H4 and H6) explore value appropriation whenthere is successful value creation. As noted earlier,there are many payoff-dominant Nash equilibriafor which the value creation threshold is met (i.e.,success = 1), but which differ in the relative con-tribution levels (and, hence, value appropriationlevels) adopted by the partners. H4 focuses on thelink between communication and pre-acquisitionvalue appropriation. The dependent variable usedto test H4 is the pre-acquisition value appropria-tion ratio, calculated based on the benefits receivedby Firm B relative to Firm A in the successfulpre-acquisition rounds. The higher this ratio is, theless equitable the division of value. H6 investigatesthe influence of pre-acquisition contribution rou-tines on contribution choices post-acquisition. Thedependent variable for H6 is the post-acquisitioncontribution ratio of the manager of erstwhile FirmB relative to managers in Firm A, calculated basedon the contributions of each manager during suc-cessful post-acquisition rounds.

Explanatory variables

For H1 and H4, the main explanatory variable ofinterest is prior communication, which takes thevalue of ‘1’ if the treatment allows for Firm Bto engage in communication in the pre-acquisitionstage. In the model investigating H2 and H3, indi-cator variables for the organizational architecturetreatments are used (e.g., complete alliance = 1 forobservations drawn from that treatment and 0 oth-erwise). The outright acquisition treatment is theomitted category. To explore the effect of routinesestablished via pre-acquisition alliance activity onpost-acquisition performance, we use two indepen-dent variables. The first is prior success, whichtakes the value of ‘1’ if the threshold value wasmet during the pre-acquisition stage. The secondis the pre-acquisition contribution ratio, which isconstructed similarly to the post-acquisition con-tribution ratio.

In all models, quarter— denoting the period inwhich the decision is being made—is used as acontrol for task-related learning with repetition.To control for unobserved heterogeneity betweenthe different groups of managers, we use randomeffects logistic regression analysis in models wherethe dependent variable is success and randomeffects regression analysis for models where the

value appropriation ratio or the contribution ratiois the dependent variable.

RESULTS

The tests for Hypothesis 1, and for Hypotheses2 and 3, are reported in Tables 3 and 4, respec-tively. We note that repeated interactions in thepost-acquisition stage have a beneficial effect oncoordination, given that the effect of quarter ispositive and significant in both tables. Control-ling for this learning over time, we find thatprior resource allocation has a significant adverseeffect on post-acquisition success. This result sug-gests that alliance partners may have a proclivityto choose risk-dominant equilibria over payoff-dominant equilibria when making resource allo-cation decisions in an interorganizational context.On the other hand, prior communication has asignificant beneficial impact on post-acquisitionsuccess. The probability of success is higher incontexts where managers have the ability to com-municate prior to the acquisition to mitigate oper-ational and/or behavioral challenges to value cre-ation. Thus, H1 is supported.

The analysis reported in Table 4 allows compari-son across the alternative pre-acquisition organiza-tional architectures. To facilitate comparison withTable 3, we retain outright acquisition as a base-line. While not hypothesized, we note that the coef-ficient of complete alliance is not different fromthe baseline in either statistical or economic signifi-cance, which is consistent with the assurance gamelogic mentioned earlier and the findings of Zaheeret al. (2010). In support of H2, the negative and

Table 3. Effects of prior resource allocation and priorcommunication on post-acquisition coordination

Success (logistic)

Coeff s.e.

Intercept −0.52 0.48Prior resource allocation −0.73 0.37∗∗

Prior communication 0.76 0.38∗∗

Quarter 0.33 0.06∗∗∗

Number of observations 568chi-sq 36.07∗∗∗

∗∗ p < 0.05; ∗∗∗ p < 0.01, two-tailed test.Random effects panel regression by group.Observations consist of the eight post-acquisition quarters acrossall 71 sets.



Table 4. Organizational architecture and post-acquisition success

Success (logistic)

Coeff s.e.

Intercept −0.41 0.48Allocation only −0.97 0.48∗∗

Communication only 0.43 0.56Complete alliance −0.01 0.46Quarter 0.33 0.06∗∗∗

Number of observations 568chi-sq 36.46∗∗∗

∗∗ p < 0.05; ∗∗∗ p < 0.01, two-tailed test.Random effects panel regression by group.Observations consist of the eight post-acquisition quarters acrossall 71 sets.

significant coefficient of allocation only impliesthat post-acquisition success is less likely follow-ing the acquisition of an allocation only partnerrelative to a complete alliance partner (test fordifferences in coefficients χ 2 = 6.42, p < 0.05).This suggests that there are performance benefitsto adding communication to resource allocation inthe pre-acquisition stage. It appears that the neg-ative routines that develop in mixed-motive, priorresource allocation situations can be mitigatedthrough the use of communication to help increasethe likelihood of a payoff-dominant equilibrium.Interestingly, we find no support for H3—the coef-ficients of complete alliance and communicationonly are not significantly different from each other(test for differences in coefficients χ 2 = 0.71, p =0.70). The lack of support for H3, in conjunctionwith the support for H2, is interesting—it impliesthat while prior resource allocation in the absenceof prior communication is detrimental to success,the converse is not true. Allowing managers tohave prior resource allocation experience, whenthey already have communication channels estab-lished, does not increase the likelihood of successover and above communication only. A potentialexplanation for this result might be that absenttask uncertainty (since all managers have full infor-mation regarding the rules, incentives, and out-comes), there is no additional benefit from havingprior resource allocation experience; and that com-munication alone is sufficient to resolve behav-ioral uncertainty and to assure divisional managersthat the everyone will cooperate. Taken together,the results in Table 4 also highlight the fact thatcompetitive routines are most likely created when

managers have to make resource allocations in theabsence of communication. Communication per-mits cooperative routines to be formed, whetherin the absence or presence of resource alloca-tion. Importantly, the findings indicate that outrightacquisitions, where managers are starting with aclean slate, outperform settings where managersmake resource allocation decisions without com-munication, thus showing that no prior routinesare better than competitive routines.

We delve further into this issue in the remain-ing analyses, presented in Table 5 and Table 6,which explore the development of routines in pre-acquisition alliances (when managers from bothFirm A and Firm B make resource allocation deci-sions) and the subsequent impact of these rou-tines on post-acquisition performance. Though notexplicitly hypothesized, the data shows that theprobability of success in the pre-acquisition periodis greater when the partners communicate. Thisis consistent with the findings of Agarwal et al.(2010) and supports the contention that communi-cation reduces behavioral uncertainty and movespartners toward achieving payoff-dominant equi-libria over risk-dominant equilibria. Consistentwith H4, we find that communication also facil-itates the adoption of more equitable value appro-priation routines by the decision makers. As canbe seen in Table 5 (Column 2), communicationin the alliance stage results in a more equitabledistribution of benefits (i.e., lower ratio of Firm

Table 5. Organizational architecture and developmentof pre-acquisition routines

Success(logistic)

Successfulquarter firmB/firmA benefits

Coeff s.e. ratio (OLS)

Intercept −3.43 0.93∗∗∗ 0.39 0.04∗∗∗

Communication 0.98 0.62∗∗ −0.07 0.04∗∗

Quarter 1.02 0.32∗∗∗ 0.02 0.01∗∗

Number ofobservations

117 43

chi-sq/f-stat 11.62∗∗∗ 8.38∗∗

r-sq 0.16

∗∗ p < 0.05; ∗∗∗ p < 0.01, one-tailed test.Random effects panel regression by group.Observations consist of: (a) the pre-acquisition quarters for thecomplete alliance and allocation treatments (39 sets) for thelogistic regressions on success; and (b) the subset of thesetreatments that achieved success for the OLS regressions onbenefits ratio.



Table 6. Stickiness of routines: effect of alliance activity on post-acquisition performance

Success (logistic) FirmB/firmA contribution ratio (OLS)

Coeff s.e. Coeff s.e.

Intercept −2.08 0.68∗∗ 0.39 0.03∗∗∗

Success in alliance stage 1.530 0.49∗∗

Contribution ratio in alliance stage 0.27 0.07∗∗∗

Communication 0.66 0.48∗ −0.10 0.04∗∗∗

Quarter 0.36 0.070∗∗∗ −0.002 0.002

Number of observations 312 287chi-sq/f-stat 47.1∗∗∗ 17.92∗∗∗

r-sq 0.21

∗ p < 0.10; ∗∗ p < 0.05; ∗∗∗ p < 0.01; one-tailed test.Random effects panel regression by group.Observations consist of the eight post-acquisition quarters for the complete alliance and allocation only treatments (39 sets).

B’s benefits relative to Firm A’s benefits), whichis significant at the 5 percent level. Finally, inTable 6, we find strong support for the contentionthat routines established in the alliance period arecarried over into the post-acquisition stage.11 Ashypothesized in H5, pre-acquisition value-creationsuccess begets post-acquisition value-creation suc-cess. Similarly, and providing support for H6, theresource allocation routines (contribution ratio ofthe managers) set in the alliance context is pos-itively and significantly related to the resourceallocation routines adopted in the new, integratedorganizational context.

DISCUSSION AND CONCLUSIONS

Successful organization design requires selectionamong multiple media for effective informationprocessing during task execution (Galbraith, 1973;Daft and Lengel, 1986; Russ et al., 1990). Ini-tial design choices, however, influence the interac-tion among decision makers and thus may subse-quently impact the ability to achieve coordinationas organizational structures are redesigned. Thisresearch relates these critical issues to behavioraluncertainty by examining a specific context andasking a deceptively simple question: what arethe consequences of pre-acquisition architectural

11 This is consistent with anecdotal field observations. For exam-ple, managers report that it takes about 12–18 months to buildtrust with partners and assess their collaborative potential (Dyeret al., 2004). For example, Cisco worked with NETSYS Tech-nologies for 20 months before acquiring the firm for its networkinfrastructure and software technologies in 1996.

choices on post-acquisition performance? Morespecifically, how does the presence or absence ofpre-acquisition information flows via resource allo-cation and/or communication-based interactionsshape both value-creation and value-appropriationroutines? And further, are these routines sustainedwith a move from one organizational form (interor-ganizational alliance) to another (integrated firm)?Thus, our fundamental research question relates toidentifying and teasing apart the creation and trans-fer of routines from one organizational architectureto another, particularly as it relates to behavioraluncertainty and the need for assurance that theother managers will act consistent to expectationsfor successful coordination.

Existing research provides important clues, yetdoes not answer these questions completely. Theo-retically, many alternative mechanisms have beenproposed (Reich and Mankin, 1984; Doz et al.,1986; Haspeslagh and Jemison, 1991; Kogut, 1991;Balakrishnan and Koza, 1993; Bleeke and Ernst,1995; Hagedoorn and Sadowski, 1999; Chang andRosenzweig, 2001) linking pre-acquisition allianceactivity with subsequent performance. Similarly,the empirical evidence is unable to provide a uni-form answer to this question or to distinguishamong competing causes of performance differ-ences that are proposed or observed. For example,Higgins and Rodriguez (2006) study 160 phar-maceutical acquisitions, of which 28 percent hadprior alliances, and report a positive effect ofprior alliances on market reaction to acquisitionannouncements. However, Benson and Ziedonis(2010) study 530 acquisitions by 61 top investorsand report a negative effect of prior corporate



venture capital investments on market returns atthe time of announcement. Zaheer et al. (2010)find no main effects of prior alliances on post-acquisition performance, but do find that interna-tional alliances or alliances with strong ties resultin superior post-acquisition performance.

In this study, we integrate insights from organi-zation design, economic game theory, and socialpsychology to illuminate the role of prior resourceallocation and communication in alleviating behav-ioral uncertainty that arises in interunit coordina-tion settings within and across organizations. Bothmechanisms are related to learning (task specificand partner specific) and development of socialprocesses and norms. However, as highlighted inorganization design literature, both mechanismsvary in their ‘richness’ and concomitant costs interms of managerial time and effort (Daft andLengel, 1986; Russ et al., 1990). By examining themain and additive effects, we are able to shed lighton which one is more important in the creation ofpositive routines. Using a randomized experimen-tal design, we find that prior resource allocationreduces, rather than enhances, post-acquisition per-formance. This finding builds on and complementsprevious work in organizational design that high-lights the differential ability of resource allocationactivities and communication interactions to con-vey subjective/socialized information. Simply, wefind that richer information flows are needed topromote cooperation in the presence of behav-ioral uncertainty (H1 and H2). Sans such flows,it is likely that competitive, rather than cooper-ative, routines will develop between managers inthe mixed-motive interorganizational environment.In this context, we note that the lack of support forH3 is interesting in its own right. It indicates thatin situations where there is no task uncertainty (asin our framework), resource allocation does notadd any additional information beyond commu-nication—thus, communication is relatively moreimportant in the resolution of behavioral uncer-tainty than is resource allocation and prior experi-ence with resource allocation.

Once an acquisition has occurred, these compet-itive routines, though no longer appropriate in thecombined entity, appear to be retained (H4-H6).Our results indicate, however, that such compet-itive routines are less likely to be established inthe presence of communication. When managerscommunicate before the acquisition, they have theopportunity to share individual positions and can

develop a mechanism for understanding each oth-ers’ positions. Such interactions support both suc-cess in value creation and equity in value appropri-ation. The effect of prior communication is unam-biguously positive, as predicted by existing theory(Gulati, 1995). Further, and as expected, perfor-mance also improves with experience. In sum, ourresearch shows that competitive and confronta-tional routines that develop in an alliance settingmay have adverse transfer effects when the alliancepartners merge and dissolve the interorganizationalboundaries.

We note several limitations—which also repre-sent future research opportunities—of our study.First, our use of experiments permits us to dis-entangle the effects of underlying mechanisms,but at some cost of realism. In particular, weabstracted from the effect of task uncertainty(either technological or market driven) on the post-acquisition performance outcomes, while focusingon the behavioral uncertainty resulting from man-agers interacting with each other. Future researchexamining the main and interaction effects of taskand behavioral uncertainty on the transition oforganizations from interorganizational to intraor-ganizational settings would be very fruitful. Also,while we identify and isolate the effects of resourceallocation and communication as aggregate-levelinformation processing mechanisms, research thatfurther disaggregates the effect of task-orientedand partner-specific learning and creations ofnorms and culture is warranted. Further, researchthat examines how firm capabilities may affect therelationship between pre-acquisition resource allo-cation or communication on post-acquisition per-formance would enhance our understanding of howorganizational capabilities and organizational eco-nomics influence organizational design outcomes.For example, it would be interesting to examinenot only whether the specific alliance designs helpor hurt future acquisition performance, but alsothe processes that are used by the allying firmsto initiate and negotiate their alliance arrangement.Given the path dependence that is revealed in suchsetup procedures, it would be nice to identify earlymarkers of future success or failure in corporatedevelopment activities.

Another limitation of our study relates to the useof students rather than professionals. We undertookseveral steps to minimize the effect of this lim-itation. First, all participants in our experimentswere pursuing professional graduate-level degrees



and had at least four years of work experience.Further, there was no significant difference in theresults obtained for groups in which the man-agers were enrolled in the executive MBA pro-gram who were older, had more experience, andhad participated in alliances and post-acquisitioncoordination activity as part of their job descrip-tion. Finally, we rely on the rich tradition of theuse of experiments in related areas of experimentaleconomics, social psychology, and organizationalbehavior, where researchers have consistently doc-umented the equivalence of results of experimentsusing students versus ‘real-world’ decision makers(Dyer, Kagel, and Levin, 1989; Croson and Dono-hue, 2006). However, additional research wherenonstudent samples are used would clearly provideadditional value to the field.

Since field studies confound multiple causalmechanisms, our study provides an important com-plementary methodology that isolates the mecha-nisms that are at play in the alliance stage andabstracts away from exogenous uncertainty, orselection effects, about partner quality. As weargued earlier in the section on empirical method-ology, endogeneity and sample selection havebecome recognized as important issues in strategyresearch (Hamilton and Nickerson, 2003). Experi-ments provide a powerful alternative to two-stagetype models using field data. Such a methodologycan generate data pertaining to cases that do notnaturally exist in the field.

Importantly, this study contributes to the exist-ing organizational design literature by movingbeyond the traditional focus on environmental andtask uncertainty to explicitly consider the roleof behavioral uncertainty in the selection andsubsequent performance of alternative organiza-tional architectures. As divisional managers pro-cess information regarding interdependent tasksand create codified routines that guide futureactions, a critical determinant of these routines iswhether they embody expectations of cooperativeor competitive behavior from the other managers.Classic organization design theory and congruencetheory pay little attention to how managers coor-dinate when task interdependence is complicatedby internal competition and significant behavioraluncertainty. In such situations, design of appropri-ate structural and informal mechanisms must takeinto consideration the ‘history’ of the interdepen-dent units, as it may be necessary to adapt designs

to overcome previously established counterproduc-tive routines. This issue is of particular saliencefor integration events involving the dissolution offirm boundaries where managers who were for-merly ‘outsiders’ are now are part of the firm. Ourstudy employs an under-researched route of exam-ining inter- and intraorganizational design usingthe social dilemma paradigm (Khanna et al., 1998;Agarwal et al., 2010) and advances understand-ing of how resource allocation and communicationmay result in persistence of positive or negativeorganizational outcomes.

Since post-acquisition coordination may be seenas a natural experiment in the dissolution of firmboundaries, the context we examine also relatesto issues related to the theory of the firm. Forexample, what kinds of decisions are made differ-ently by divisions within a firm versus autonomousdivisions? What kinds of factors cause the inter-firm boundaries to dissolve effectively? Our resultsreveal the complementary role of ‘soft’ issues likecommunication (Schweiger and Goulet, 2000) andthe ‘hard’ issues like resource allocation (Ghoshand Ruland, 1998). By examining the effect ofvariance in these factors, we can identify thecontribution of each set and, more importantly,the additive effect among them. As per previousresearch (Kogut and Zander, 1996; Conner andPrahalad, 1996), our research shows the comple-mentary roles of these factors.

In terms of implications for practice, thisresearch enhances our understanding of post-acquisition coordination and the actual mecha-nisms that may drive successful post-acquisitionperformance. Our study illustrates that competitiveroutines develop even in simple settings—ratherthan the complex settings of the field, the onlychange in our experimental task between pre- andpost-acquisition was in the incentive structure orthe profit-sharing rule (representing a shift froman alliance to acquisition setting). Also, our studyunderscores the fact that alliances are not justopportunities for selection, but also for creationof potentially successful targets. Creation of coop-erative resource allocation routines helps in thelong run creation of value when alliance partnersare subsequently acquired. Our research suggeststhat the early cooperative engagement with thepotential partner can have long-term effects, evenlasting beyond an eventual acquisition, so firmsshould take into account such effects even at earlystage. Firms may consider starting this process



keeping in mind how it might end. Finally, thestudy is a cautionary note to managers who seekto overcome failure in alliances by acquiring part-ners with the intent to use the dissolution of firmboundaries to achieve cooperation. Negative rou-tines, once formed, are hard to overcome. Thus,it behooves managers to weigh the positive andnegative effects of sequential strategies when theydesign their corporate development efforts. Specifi-cally, it points out that in pre-acquisition alliances,managers should mitigate the formation of com-petitive routines and that managers should engagein wide and open communication for the develop-ment of appropriate norms and culture, both beforeand after the acquisition. Finally, it provides impli-cations for a better understanding of the conceptsof acquisition and alliance capabilities.

ACKNOWLEDGEMENTS

We gratefully acknowledge research funding fromthe Office of Research, University of Illinois andthe University of Michigan Business School. Shra-van Gaonkar provided valuable research assis-tance. We also thank the special issue editors andanonymous reviewers, Jay Barney, Seth Carnahan,David Croson, David Greenberger, Samina Karim,Joanne Oxley, Roberto Weber, Rosemarie Ziedo-nis, and participants in the preconference devel-opment workshop on experimental methods at the2003 Academy of Management meetings, the 2004Strategic Management Society meetings, the con-ference on alliances at IESE in 2005, and the Har-vard Business School conference on organizationalarchitectures in 2010 for their input and sugges-tions. All remaining errors are ours.

REFERENCES

Agarwal R, Croson R, Mahoney J. 2010. The role ofincentives and communication in strategic alliances:an experimental investigation. Strategic ManagementJournal 31(4): 413–437.

Agarwal R, Helfat C. 2009. Strategic renewal oforganizations. Organization Science 20(2): 281–293.

Allport GW. 1954. The Nature of Prejudice. Addison-Wesley: Reading, MA.

Anand BN, Khanna T. 2000. Do firms learn to createvalue? The case of alliances. Strategic ManagementJournal 21(3): 295–315.

Anand J, Oriani R, Vassolo RS. 2010. Alliance activityas a dynamic capability in the face of a discontinuous

technological change. Organization Science 21(6):1213–1232.

Anand J, Singh H. 1997. Asset redeployment, acquisi-tions and corporate strategy in declining industries.Strategic Management Journal, Summer Special Issue18: 99–118.

Arend RJ. 2004. Conditions for assymetric informationsolutions when alliances provide acquisition optionsand due diligence. Journal of Economics 82(3):281–312.

Argyres NS. 1995. Technology strategy, governancestructure, and interdivisional coordination. Journal ofEconomic Behavior & Organization 28(3): 337–358.

Argyres NS, Liebeskind JP. 1999. Contractual commit-ments, bargaining power, and governance inseparabil-ity: incorporating history into transaction cost theory.The Academy of Management Review 24(1): 49–63.

Arrow KJ. 1974. The Limits of Organization: FelsLectures on Public Policy Analysis . W.W. Norton &Co.: New York.

Austen I. 2007. I.B.M. acquires Cognos, maker ofbusiness software, for $4.9 billion. The New YorkTimes , 13 November.

Balakrishnan S, Koza M. 1993. Information asymmetry,adverse selection, and joint ventures: theory and evi-dence. Journal of Economic Behavior & Organization20: 99–117.

Barkema HG, Schijven M. 2008. How do firms learn tomake acquisitions? A review of past research and anagenda for the future. Journal of Management 34(3):594–634.

Barnard CI. 1938. The Functions of the Executive.Harvard University Press: Cambridge, MA.

Bastien DT. 1987. Common patterns of behavior andcommunication in corporate mergers and acquisitions.Human Resource Management 26: 17–34.

Becker GS, Murphy KM. 1994. The division of labor,coordination costs, and knowledge. The QuarterlyJournal of Economics 107(4): 1137–1160.

Benson D, Ziedonis RH. 2010. Corporate venture capitaland the returns to acquiring portfolio companies.Journal of Financial Economics 98(3): 478–499.

Bhappu A, Griffith T, Northcraft G. 1997. Media effectsand communication bias in diverse groups. Orga-nizational Behavior and Human Performance 70:199–205.

Birkinshaw J, Bresman H, Hakanson L. 2000. Manag-ing the post-acquisition integration process: how thehuman integration and task integration processes inter-act to foster value creation. Journal of Management37(3): 395–425.

Bleeke JA, Ernst D. 1995. Is your strategic alliance reallya sale? Harvard Business Review 73: 97–105.

Bohl DL. 1989. Tying the Corporate Knot . AmericanManagement Association: New York.

Brewer MB, Brown RJ. 1998. Intergroup relations. InHandbook of Social Psychology (Vol. 2), Gilbert DT,Fiske ST, Lindzey G (eds). McGraw-Hill: New York;554–594.

Burns T, Stalker GM. 1961. The Management ofInnovation . Tavistock: London, U.K.



Burton RM, DeSanctis G, Obel B. 2006. OrganizationalDesign: A Step-by-Step Approach . Cambridge Univer-sity Press: Cambridge, U.K.

Burton RM, Obel B. 2004. Strategic OrganizationalDiagnosis and Design: The Dynamics of Fit . SpringerScience+Business Media: New York.

Capron L, Mitchell W. 2009. Selection capability: howcapability gaps and internal social frictions affectinternal and external strategic renewal. OrganizationScience 20(2): 294–312.

Chang SJ, Rosenzweig PM. 2001. The choice of entrymode in sequential foreign direct investment. StrategicManagement Journal 22(8): 747–776.

Clark H. 1996. Using Language. Cambridge UniversityPress: Cambridge, U.K.

Conlon EJ, Parks JM. 1990. Effects of monitoring andtradition on compensation agreements: an experimenton principal-agent dyads. Academy of ManagementJournal 33: 603–622.

Conner KR, Prahalad CK. 1996. A resource-basedtheory of the firm: knowledge versus opportunism.Organization Science 7(5): 477–501.

Crawford V. 1998. A survey of experiments oncommunication via cheap talk. Journal of EconomicTheory 78(2): 286–298.

Croson R. 2006. The use of students as participants inexperimental research. Working paper, University ofTexas, Dallas.

Croson R, Anand J, Agarwal R. 2007. Using experimentsin corporate strategy research. European ManagementReview 4(3): 173–181.

Daft RL, Lengel RH. 1986. Organizational informationrequirements, media richness, and structural design.Management Science 32(5): 554–571.

Das TK, Teng B-S. 2000. Instabilities of strategicalliances: an internal tensions perspective. Organiza-tion Science 11: 77–101.

Dawes RM, van de Kragt AJC, Orbell JM. 1988. Notme or thee but we: the importance of groupidentity in eliciting cooperation in dilemma situations:experimental manipulations. Acta Psychologica 68:83–97.

Deutsch M, Gerard HB. 1955. A study of normativeand informational social influences upon individualjudgment. The Journal of Abnormal and SocialPsychology 51(3): 629–636.

Donaldson L. 1987. Strategy and structural adjustment toregain fit and performance: in defence of contingencytheory. Journal of Management Studies 24(1): 1–24.

Donaldson L. 1990. The ethereal hand: organizationaleconomics and management theory. Academy ofManagement Review 15(3): 369–381.

Doz YL. 1996. The evolution of cooperation in strategicalliances: initial conditions or learning processes.Strategic Management Journal , Summer Special Issue17: 55–83.

Doz YL, Hamel G, Prahalad CK. 1986. Strategic partner-ships: success or surrender? The challenge of com-petitive collaboration. Paper presented at AIB annualmeeting, London, U.K.

Dyer D, Kagel JH, Levin D. 1989. A comparison ofnaive and experienced bidders in common value offer

auctions: a laboratory analysis. The Economic Journal99(394): 108–115.

Dyer JH, Chu W. 2000. The determinants of trust insupplier-automaker relationships in the U.S., Japan,and Korea. Journal of International Business Studies31(2): 259–285.

Dyer JH, Kale P, Singh H. 2004. When to ally and whento acquire. Harvard Business Review 82: 108–115.

Falk A, Heckman JJ. 2009. Lab experiments are a majorsource of knowledge in the social sciences. Science326: 535–538.

Farrell J, Rabin M. 1996. Cheap talk. The Journal ofEconomic Perspectives 19(3): 103–118.

Fossum JA. 1979. The effects of positively and negativelycontingent rewards and individual differences onperformance, satisfaction, and expectations. Academyof Management Journal 22: 577–589.

Freedman SM. 1978. Some determinants of compensationdecisions. Academy of Management Journal 21(3):397–409.

Galbraith Jr. 1973. Designing Complex Organizations .Addison-Wesley: Reading, MA.

Galbraith Jr. 1977. Organization Design . Addison-Wesley: Reading, MA.

Ghosh A, Ruland W. 1998. Managerial ownership, themethod of payment for acquisitions, and executive jobretention. The Journal of Finance 53(2): 785–798.

Ghoshal S, Bartlett C. 1990. The multinational corpora-tion as an interorganizational network. The Academyof Management Review 15(4): 603–625.

Graebner M. 2004. Momentum and serendipity: howacquired leaders create value in the integrationof technology firms. Strategic Management Journal25(8/9): 751–777.

Greenberg J. 1987. A taxonomy of organizational justicetheories. The Academy of Management Review 12(1):9–22.

Grote G. 2009. Management of Uncertainty: Theoryand Application in the Design of Systems andOrganizations . Springer-Verlag: London, U.K.

Gulati R. 1995. Familiarity breeds trust? The implicationsof repeated ties. Academy of Management Journal 38:85–112.

Gulati R. 1998. Alliances and networks. StrategicManagement Journal 19(4): 293–317.

Gulati R. 1999. Network location and learning: theinfluence of network resources and firm capabilitieson alliance formation. Strategic Management Journal20(5): 397–420.

Gulati R, Khanna T, Nohria N. 1994. Unilateral com-mitments and the importance of process in alliances.Sloan Management Review 35(3): 61–69.

Gulati R, Khanna T, Nohria N. 2000. The economicmodeling of strategy process: ‘Clean models’ and‘dirty hands.’ Strategic Management Journal 21(7):781–790.

Gulati R, Nickerson J. 2008. Interorganizational trust,governance choice, and exchange performance.Organization Science 19(5): 688–708.

Gulati R, Puranam P. 2009. Renewal through reorganiza-tion: the value of inconsistencies between formal and



informal organization. Organization Science 20(2):422–440.

Gulati R, Singh H. 1998. The architecture of cooperation:managing coordination costs and appropriationconcerns in strategic alliances. Administrative ScienceQuarterly 43(4): 781–814.

Hagedoorn J, Sadowski B. 1999. Transition from strate-gic technology alliances. Journal of Management Stud-ies 36(1): 87–107.

Haleblian J, Finkelstein S. 1999. The influence oforganizational acquisition experience on acquisitionperformance: a behavioral perspective. AdministrativeScience Quarterly 44(1): 29–57.

Hamel G. 1991. Competition for competence andinterpartner learning within international strategicalliances. Strategic Management Journal, SummerSpecial Issue 12: 83–103.

Hamilton BA, Nickerson JA. 2003. Correcting forendogeneity in strategic management research.Strategic Organization 1(1): 53–80.

Harsanyi JC, Selten R. 1988. A General Theory ofEquilibrium Selection in Games . The MIT Press:Cambridge, MA.

Haspeslagh PC, Jemison DB. 1991. Managing Acquisi-tions: Creating Value Through Corporate Renewal .Free Press: New York.

Heide JB, John G. 1992. Do norms matter in marketingrelationships? Journal of Marketing 56: 32–44.

Helft M. 2010. AOL sells social site to investment firm.The New York Times, 17 June.

Higgins MJ, Rodriguez D. 2006. The outsourcing ofR&D through acquisitions in the pharmaceuticalindustry. Journal of Financial Economics 80(2):351–383.

Jemison DB, Sitkin SB. 1986. Corporate acquisitions: aprocess perspective. Academy of Management Review11: 145–163.

Kahneman D, Knetsch J, Thaler R. 1990. Experimentaltests of the endowment effect and the Coase theorem.Journal of Political Economy 98(6): 1325–1348.

Kale P, Dyer JH, Singh H. 2002. Alliance capability,stock market response, and long-term alliance success:the role of the alliance function. Strategic ManagementJournal 23(8): 747–767.

Karim S, Mitchell W. 2000. Path-dependent and path-breaking change: reconfiguring business resourcesfollowing acquisitions in the U.S. medical sec-tor, 1978–1995. Strategic Management Journal21(10/11): 1061–1081.

Khanna T, Gulati R, Nohria N. 1998. The dynamicsof learning alliances: competition, cooperation, andrelative scope. Strategic Management Journal 19(3):193–210.

Knez M, Camerer C. 1994. Creating expectational assetsin the laboratory: coordination in the ‘weakest-link’games. Strategic Management Journal, Winter SpecialIssue 15: 101–119.

Kogut B. 2000. The network as knowledge: generativerules and the emergence of structure. StrategicManagement Journal 21(3): 405–425.

Kogut B, Zander U. 1996. What do firms do? Coordina-tion, identity, and learning. Organization Science 7(5):502–518.

Kogut BA. 1991. Joint ventures and the option to expandand acquire. Management Science 37(1): 9–24.

Komorita SS, Parks CD. 1994. Social Dilemmas . Brown& Benchmark: Dubuque, IA.

Lawrence PR, Lorsch JW. 1967. Organization andEnvironment . Harvard Business School Press: Boston,MA.

Ledyard J. 1995. Public goods: a survey of experimen-tal research. In The Handbook of Experimental Eco-nomics , Kagel JH, Roth AE (eds). Princeton Univer-sity Press: Princeton, NJ; 111–181.

Leifer R, Mills PK. 1996. An information processingapproach for deciding upon control strategies andreducing control loss in emerging organizations.Journal of Management 22(1): 113–137.

Lengel RH, Daft RL. 1984. An exploratory analysisof the relationship between media richness andmanagerial information processing. Technical report,Texas A&M University.

London M, Oldham G. 1977. A comparison of groupand individual incentive plans. Administrative ScienceQuarterly 20: 34–41.

Lubatkin M, Calori R, Very P, Veiga J. 1998. Managingmergers across borders: a two nation exploration of anationally bound administrative heritage. OrganizationScience 9(6): 670–684.

Macneil IR. 1978. Contracts: adjustment of long-termeconomic relations under classical, neoclassical, andrelational contract law. Northwestern University LawReview 72(6): 854–901.

Macneil IR. 1980. The New Social Contract: An Inquiryinto Modern Contractual Relations . Yale UniversityPress: New Haven, CT.

Malmgren HR. 1961. Information, expectations, and thetheory of the firm. Quarterly Journal of Economics75(3): 399–421.

March JG, Simon HA. 1958. Organizations . Wiley: NewYork.

Marks ML, Mirvis PH. 1998. Joining Forces . Jossey-Bass: San Francisco, CA.

McCann JE, Galbraith Jr. 1981. Interpersonal relations.In Handbook of Organizational Design , Nystrom PC,Starbuck WH (eds). Oxford University Press: NewYork; 60–84.

Miller D. 1986. Configurations of strategy and structure:towards a synthesis. Strategic Management Journal7(3): 233–249.

Miller D, Friesen PH. 1982. Structural change andperformance: quantum versus piece-meal incrementalapproaches. Academy of Management Journal 25(4):867–892.

Nadler D, Tushman M. 1997. Competing by Design:The Power of Organizational Architecture. OxfordUniversity Press: New York.

Nahavandi A, Malekzadeh AR. 1988. Acculturation inmergers and acquisitions. Academy of ManagementReview 13: 79–90.

Napier NK. 1989. Mergers and acquisitions, humanresource issues and outcomes: a review and



suggested typology. Journal of Management Studies26: 271–289.

Neale MA, Bazerman MH. 1985. The effects of framingand negotiator confidence on bargaining behaviors andoutcomes. Academy of Management Journal 28(1):34–39.

Noda T, Bower JL. 1996. Strategy making as iteratedprocesses of resource allocation. Strategic Manage-ment Journal, Summer Special Issue 17: 159–192.

Orbell JM, Dawes RM, van de Kragt AJC. 1990. Thelimits of multilateral promising. Ethics 100: 616–627.

Parks JM, Conlon EJ. 1995. Compensation contracts:do agency theory assumptions predict negotiatedagreements? Academy of Management Journal 38(3):821–838.

Pettigrew TF. 1971. Racially Separate or Together?McGraw-Hill: New York.

Pisano GP. 1994. Knowledge, integration and thelocus of learning: an empirical analysis of processdevelopment. Strategic Management Journal, WinterSpecial Issue 15: 85–100.

Plott CR. 1991. Will economics become an experimentalscience? Southern Economic Journal 57(4): 901–919.

Porrini P. 2004. Can a previous alliance between anacquirer and a target affect acquisition performance?Journal of Management 30(4): 545–562.

Puranam P, Singh H, Chaudhuri S. 2009. Integratingacquired capabilities: when structural integration is(un)necessary. Organization Science 20: 313–328.

Puranam P, Singh H, Zollo M. 2006. Organizing forinnovation: managing the coordination-autonomydilemma in technology acquisitions. Academy ofManagement Journal 49(2): 263–280.

Ranft AR, Lord MD. 2000. Acquiring new technolo-gies and capabilities: a grounded model of acqui-sition implementation. Organization Science 13(4):420–441.

Reich RB, Mankin ED. 1986. Joint ventures with Japangive away our future. Harvard Business Review 64:78–86.

Ring PS, van de Ven AH. 1994. Developmental pro-cess of cooperative interorganizational relationships.Academy of Management Review 19(1): 90–118.

Rodan S, Galunic C. 2004. More than network structure:how knowledge heterogeneity influences managerialperformance and innovativeness. Strategic Manage-ment Journal 25(6): 541–562.

Russ GS, Daft RL, Lengel RH. 1990. Media selectionand managerial characteristics in organizational com-munications. Management Communication Quarterly4(2): 151–175.

Sarkar M, Aulakh PS, Madhok A. 2009. Process capa-bilities and value generation in alliance portfolios.Organization Science 20(3): 583–600.

Schelling TC. 1960. The Strategy of Conflict . HarvardUniversity Press: Cambridge, MA.

Schmidt JA. 1999. Realizing the Full Value of the Mergeror Acquisition: A Risk Management Perspective.Towers Perrin: Stamford, CT.

Schweiger DM, DeNisi AS. 1991. Communication withemployees following a merger: a longitudinal field

experiment. Academy of Management Journal 34(1):110–135.

Schweiger DM, Goulet PK. 2000. Integrating mergersand acquisitions: an international research review.In Advances in Mergers and Acquisitions , Cooper C,Gregory A (eds). JAI Press: Greenwich, CT; 61–91.

Shanley MT. 1988. Reconciling the rock and thehard place: management control versus humanresource accommodation in acquisition integration.Unpublished paper, University of Chicago.

Simon HA. 1947. Administrative Behavior . Macmillan:New York.

Smith V. 1982. Microeconomic systems as an experi-mental science. American Economic Review 72(5):923–955.

Smith V. 2000. Bargaining and Market Behavior: Essaysin Experimental Economics . Cambridge UniversityPress: New York.

Smith-Doerr L, Powell WW. 2005. Networks andeconomic life. In The Handbook of EconomicSociology , Smelser NJ, Swedberg R (eds). PrincetonUniversity Press: Princeton, NJ.

Song M, Calantone RJ, Di Benedetto CA. 2002. Compet-itive forces and strategic choice decisions: an exper-imental investigation in the United States and Japan.Strategic Management Journal 23(10): 969–978.

Tenbrunsel AE. 1998. Misrepresentation and expectationsof misrepresentation in an ethical dilemma: the role ofincentives and temptation. Academy of ManagementJournal 41(3): 330–339.

Thomas-Hunt MC, Ogden TY, Neale MA. 2003. Who’sreally sharing? Effects of social and expert statuson knowledge exchange within groups. ManagementScience 49(4): 464–477.

Thompson JD. 1967. Organizations in Action . McGraw-Hill: New York.

Tosi HL, Katz JP, Gomez-Mejia LR. 1997. Disaggregat-ing the agency contract: the effects of monitoring,incentive alignment, and term in office on agent deci-sion making. Academy of Management Journal 40(3):584–602.

Tushman M, Nadler D. 1978. Information processingas an integrating concept in organization design.Academy of Management Review 3(3): 613–624.

Vance A. 2009. Cisco is buying Starent for $2.9 billion tobolster its wireless networks business. The New YorkTimes , 14 October.

Vassolo RS, Anand J, Folta TB. 2004. Non-additivity inportfolios of exploration activities: a real options-based analysis of equity alliances in biotechnology.Strategic Management Journal 25(11): 1045–1061.

Weber RA, Camerer CF. 2003. Cultural conflict andmerger failure: an experimental approach. Manage-ment Science 49(4): 400–415.

Wieland K. 2005. More than money: when vendorsventure, their priorities can differ from the traditionalVC. Telecommunications Magazine, 6 April.

Williamson OE. 1975. Markets and Hierarchies . FreePress: New York.

Williamson OE. 1985. The Economic Institutions of Cap-italism: Firms, Markets and Relational Contracting .Free Press: New York.



Williamson OE. 1991. Comparative economic organiza-tion: the analysis of discrete structural alternatives.Administrative Science Quarterly 36: 269–296.

Zaheer A, Hernandez E, Banerjee S. 2010. Prior allianceswith targets and acquisition performance inknowledge-intensive industries. Organization Science21(5): 1072–1091.

Zaheer A, Venkatraman N. 1995. Relational governanceas an interorganizational strategy: an empirical testof the role of trust in economic exchange. StrategicManagement Journal 16(5): 373–392.

Zelditch M. 1969. Can you really study an army inthe laboratory? In Complex Organizations (2nd edn),Etzioni A (ed). Holt, Rinehart, Winston: New York;528–539.

Zeng M, Chen XP. 2003. Achieving cooperation inmulti-party alliances: a social dilemma approach topartnership management. Academy of ManagementReview 28(4): 587–605.

Zhao ZJ, Anand J. 2009. The role of collective teachingand collective absorptive capacity in transferring firm

capabilities: evidence from the Chinese automotiveindustry. Strategic Management Journal 30(9):959–983.

SUPPORTING INFORMATION

Additional supporting information may be foundin the online version of this article:

APPENDIX 1: Parameters and resulting equilibria

APPENDIX 2: Composite instructions

Please note: Wiley-Blackwell is not responsiblefor the content of functionality of any supportingmaterials supplied by the authors. Any queries(other than missing material) should be directedto the corresponding author for the article.


spillovers across organizational architectures: the role...

Documents