interfirm rivalry and managerial complexity

17
1047-7039/01/1201/0037/$05.00 1526-5455 electronic ISSN ORGANIZATION SCIENCE, 2001 INFORMS Vol. 12, No. 1, January–February 2001, pp. 37–53 Interfirm Rivalry and Managerial Complexity: A Conceptual Framework of Alliance Failure Seung Ho Park • Gerardo R. Ungson Department of Organization Management, Rutgers University, New Brunswick, New Jersey 08840 Department of Management, Graduate School of Management, University of Oregon, Eugene, Oregon 97403 [email protected][email protected] Abstract Empirical research indicates that more than half of strategic alliances fail, and the outcomes of alliance failure can be dev- astating. Despite the increased concern about managing strate- gic alliances, the field still lacks a theoretical framework to de- scribe the conditions and dynamics leading to the failure of strategic alliances. This paper attempts to distill, derive, and integrate theories across different disciplines into a unified framework that offers a better understanding of alliance failure. The conceptual framework focuses on two primary sources of alliance failure: interfirm rivalry and managerial complexity. We propose that strategic alliances fail because of the oppor- tunistic hazards as each partner tries to maximize its own in- dividual interests instead of collaborative interests. Also, stra- tegic alliances fail because of the difficulties in coordinating two independent firms (i.e., coordination costs), and in aligning operations at the alliance level with parent firms’ long-term goals (i.e., agency costs). The paper extends the theoretical framework by looking into a process model of alliance devel- opment and failure. (Alliance Failure; Interfirm Rivalry; Managerial Com- plexity) When two parties in an alliance cooperate at an early stage, they realize that they might compete with each other at some later stage. Nevertheless, they are willing to invest time and effort in anticipation of specific out- comes that would benefit them, even if the period of col- laboration is temporary. However, both parties know that the current intention by itself is insufficient, for, once completed, they could act opportunistically by withhold- ing important information, providing false information, or simply cheating the other. Moreover, the sheer com- plexity of the alliance might preclude the partners from evaluating their contributions over time, leading to per- ceptions that their contributions are unbalanced and asymmetric. As one partner learns faster about the other, dependencies escalate, creating even more asymmetry. If the partners trust each other (confidence on the goodwill of the other), and if significant resources and commitment are expended on the alliance, these problems would be minimized. But trust and commitment are tempered by perceptions of gains and losses, equity considerations, goal conflicts, and role ambiguities between partners. Even in a highly complementary alliance, it is a daunting task to manage conflicts arising from managerial and or- ganizational dissimilarities between the partners. Within the context of our opening vignette, several conditions affect stabilizing and destabilizing character- istics of alliances: need for specialization based on partner complementarity, problems of opportunism, complexity in monitoring behaviors, difficulty in coordinating part- ners, conflict in strategic directions, and manner in which trust can attenuate rivalry. These characteristics are em- bedded in the exchange of skills and capabilities, inten- tionally or unintentionally, between partners, which affect the process and outcome of the cooperative relationship. Hence, the basic concept of an alliance, particularly a partnership between competitors, juxtaposes two coun- tervailing tendencies: cooperative activities leading to the attainment of goals that advance the interests of both part- ners and competitive behaviors by one or both partners in pursuing their self-interests (Kogut 1988, Hamel 1991, Parkhe 1993a, Gulati 1999). Cast in this context, a stra- tegic alliance represents a temporal structure of exchange relationships that generate cooperative or competitive be- haviors between partners, depending on their private in- centives. In recent years, alliances between erstwhile competi- tors have increased significantly (Gulati et al. 1994, Park and Russo 1996). Competitors are motivated to form stra- tegic alliances with one another not only to improve their market positioning, but also with the expectation of re- ducing rivalry or attenuating contractual hazards. Even so, competitive factors that motivate the alliance can be

Upload: nguyen-hung-hai

Post on 16-Oct-2014

52 views

Category:

Documents


1 download

TRANSCRIPT

Page 1: Interfirm Rivalry and Managerial Complexity

1047-7039/01/1201/0037/$05.001526-5455 electronic ISSN

ORGANIZATION SCIENCE, � 2001 INFORMSVol. 12, No. 1, January–February 2001, pp. 37–53

Interfirm Rivalry and Managerial Complexity:A Conceptual Framework of Alliance Failure

Seung Ho Park • Gerardo R. UngsonDepartment of Organization Management, Rutgers University, New Brunswick, New Jersey 08840

Department of Management, Graduate School of Management, University of Oregon, Eugene, Oregon [email protected][email protected]

AbstractEmpirical research indicates that more than half of strategicalliances fail, and the outcomes of alliance failure can be dev-astating. Despite the increased concern about managing strate-gic alliances, the field still lacks a theoretical framework to de-scribe the conditions and dynamics leading to the failure ofstrategic alliances. This paper attempts to distill, derive, andintegrate theories across different disciplines into a unifiedframework that offers a better understanding of alliance failure.The conceptual framework focuses on two primary sources ofalliance failure: interfirm rivalry and managerial complexity.We propose that strategic alliances fail because of the oppor-tunistic hazards as each partner tries to maximize its own in-dividual interests instead of collaborative interests. Also, stra-tegic alliances fail because of the difficulties in coordinatingtwo independent firms (i.e., coordination costs), and in aligningoperations at the alliance level with parent firms’ long-termgoals (i.e., agency costs). The paper extends the theoreticalframework by looking into a process model of alliance devel-opment and failure.(Alliance Failure; Interfirm Rivalry; Managerial Com-plexity)

When two parties in an alliance cooperate at an earlystage, they realize that they might compete with eachother at some later stage. Nevertheless, they are willingto invest time and effort in anticipation of specific out-comes that would benefit them, even if the period of col-laboration is temporary. However, both parties know thatthe current intention by itself is insufficient, for, oncecompleted, they could act opportunistically by withhold-ing important information, providing false information,or simply cheating the other. Moreover, the sheer com-plexity of the alliance might preclude the partners fromevaluating their contributions over time, leading to per-ceptions that their contributions are unbalanced andasymmetric. As one partner learns faster about the other,

dependencies escalate, creating even more asymmetry. Ifthe partners trust each other (confidence on the goodwillof the other), and if significant resources and commitmentare expended on the alliance, these problems would beminimized. But trust and commitment are tempered byperceptions of gains and losses, equity considerations,goal conflicts, and role ambiguities between partners.Even in a highly complementary alliance, it is a dauntingtask to manage conflicts arising from managerial and or-ganizational dissimilarities between the partners.

Within the context of our opening vignette, severalconditions affect stabilizing and destabilizing character-istics of alliances: need for specialization based on partnercomplementarity, problems of opportunism, complexityin monitoring behaviors, difficulty in coordinating part-ners, conflict in strategic directions, and manner in whichtrust can attenuate rivalry. These characteristics are em-bedded in the exchange of skills and capabilities, inten-tionally or unintentionally, between partners, which affectthe process and outcome of the cooperative relationship.Hence, the basic concept of an alliance, particularly apartnership between competitors, juxtaposes two coun-tervailing tendencies:cooperative activitiesleading to theattainment of goals that advance the interests of both part-ners andcompetitive behaviorsby one or both partnersin pursuing their self-interests (Kogut 1988, Hamel 1991,Parkhe 1993a, Gulati 1999). Cast in this context, a stra-tegic alliance represents a temporal structure of exchangerelationships that generate cooperative or competitive be-haviors between partners, depending on their private in-centives.

In recent years, alliances between erstwhile competi-tors have increased significantly (Gulati et al. 1994, Parkand Russo 1996). Competitors are motivated to form stra-tegic alliances with one another not only to improve theirmarket positioning, but also with the expectation of re-ducing rivalry or attenuating contractual hazards. Evenso, competitive factors that motivate the alliance can be

Page 2: Interfirm Rivalry and Managerial Complexity

SEUNG HO PARK AND GERARDO R. UNGSON Interfirm Rivalry and Managerial Complexity

38 ORGANIZATION SCIENCE/Vol. 12, No. 1, January–February 2001

only imperfectly redressed and can cause future problemsand instability. In a broad sense, failure occurs when ex-cessive rivalry eclipses cooperative tendencies.

Despite the popularity of strategic alliances and thepurported benefits for participating firms, failures abound,with the failure rate reported at or higher than 50% (Porter1987, Harrigan 1988, Parkhe, 1993a). From a study oflarge U.S. corporations during the 1950–1986 period,Porter (1987) reported the failure rate of alliances washigher than 50%, which was much higher than the failurerate of internal venturing or corporate buyouts. Alliancefailure often causes several other adverse effects to par-ticipants. Studies indicate that in cross-border alliances,U.S. firms incurred more serious losses than their coun-terparts because of an involuntary loss of potential reve-nues and uncompensated transfers of rent-generating re-sources, such as technology (Hamel et al. 1989). Otheradverse effects include operational difficulties and prob-lems, disagreements, and anxieties over the loss of pro-prietary information. Alliance failure is also associatedwith more intangible, adverse outcomes such as the lossof reputation.

Despite overwhelming empirical evidence that alli-ances are unstable, there are still very few studies, if any,that provide comprehensive discourses on why alliancesfail (Parkhe 1993a). On the surface, there is no dearth ofexplanations as to why strategic alliances fail: poor man-agement, poor communications, lack of trust, competitiverivalry, lack of top management commitment, culturaldifferences, etc. While these reasons appear to be intui-tively clear, they also tend to be anecdotal in origin, adhoc in content, and fragmented in their development.Parkhe (1993b) criticized researchers for having noncu-mulatively focused on different dimensions of the stra-tegic alliance, whose results have not been coalesced intoa collectively coherent body of work. Smith et al. (1995)implore future researchers to be more systematic in theirexamination of the theoretical variables that promote co-operation (and competition). The lack of a comprehensiveframework has led to empirical fragmentation and hasencouraged static representations of alliance failure.Without a general theory of alliance failure for whichboundary conditions can be defined and tested, future em-pirical tests may remain fragmented, impeding furthertheoretical development.

The objective of this paper is to develop a theoreticalframework of alliance failure. In developing this frame-work, two tracks were pursued. One strategy was to findsystemic relationships among the different variables thataccount for failure in a strategic alliance. Of particularimportance, however, was distinguishing between failurein a strategic alliance and that in any generic partnership

within an organization. The relevant context is:What isdistinctive within a strategic alliance, as opposed to ageneric partnership within an organization, that leads toparticular patterns of failure?Our focus in this study isalso limited to the failure of a strategic alliance as analternative form of governance structure for interfirmtransactions, not the failure of the alliance from each part-ner’s point of view. Within a broadened transactional costframework, we postulate that alliance failure results fromcompetitive rivalry and managerial complexity that act asdestabilizing characteristics within an alliance. We alsointroduce some discussion of specific conditions that af-fect these two dimensions along the processes of man-aging alliances. This is undertaken by postulating ex ante,in situ, and ex post variables and conditions underlyingthe stage-of-growth in alliances. Using this track, coun-tervailing tendencies are explored. The paper concludeswith suggestions for future studies of alliance failure.

A Conceptual Framework of AllianceFailureIn our view, two continuing developments have impededthe introduction of a systematic and comprehensiveframework of alliance failure: the contestability of dif-ferent interpretations (measures) of alliance failure, andthe lack of cross-fertilization between numerous theo-retical schemes. While helpful in addressing specific is-sues on alliance management at the midrange level ofanalysis, they have also retarded integration and subse-quent advancement at a broader level of analysis.

Contestable MeasuresA unique nature of the alliance is the involvement of twoindependent firms in one frontier, which makes it elusiveto assess performance of an alliance (Anderson 1990).Studies have utilized different measures, including bothobjective (e.g., survival, termination, duration, financialgains, etc.) and subjective or process-oriented measures(e.g., goal attainment, satisfaction, learning, competencebuilding, etc.) to assess alliance outcomes (see Table 1for summary). These measures may reflect the perfor-mance of strategic alliances, but in different theoreticalquestions. They also focus on alliance activities at differ-ent levels, i.e., either the parent firm or the alliance levels,with different understanding of alliance goals. At the par-ent level, partners are often contradictory in their assess-ment of an alliance given the asymmetry in their goalsand interests in relation to the alliance. It is possible thatone partner’s success means complete failure to the otherpartner. Local management at the alliance level also dif-fers from their parent firms in terms of the direction of

Page 3: Interfirm Rivalry and Managerial Complexity

SEUNG HO PARK AND GERARDO R. UNGSON Interfirm Rivalry and Managerial Complexity

ORGANIZATION SCIENCE/Vol. 12, No. 1, January–February 2001 39

Table 1 Summary of Performance Measures in Alliance Studies

Performance Measures Authors Sample Size (Failure Rate) Operational Definition

Alliance LevelSurvival Franko (71) 1100 (24.1%) Liquidation/Ownership Change(Stability) Tomlinson (70) 107 (50%) Termination

Holton (81) Conceptual Complete WithdrawalKilling (83) 36 (30%) Liquidation/ReorganizationKogut (88) 149 (46.3%) TerminationBeamish (87) 66 (45%)Porter (87) 300 (50.3%) DivestitureHarrigan (88) 895 (54.8%)Park & Russo (96) 204 (67.5%) Dissolution/Acquisition

Duration Harrigan (88) Alliance AgeKogut (88)Park & Ungson (97)

Financial Tomlinson (70) ROIPerformance Good (72) Growth/Capital Expenditure

Raffii (78) Alliance Returns

Subjective Holton (81) Conceptual Chronic DissatisfactionIndex Killing (83) 36 (36%) Poor—Good

Beamish (87) 66 (61%) Mutual AgreementHarrigan (88) 895 (54.7%) Mutual Agreement

Composite Index Parkhe (93a)

Partner LevelAchievement of Hamel et al. (89) ConceptualIndividual Goals Koot (88)and Learning Hamel (91) Case studies

the alliance. While parent firms try to fit alliance opera-tions with their long-term strategic goals, alliance man-agers are more concerned with local interests as they tryto adapt to market changes. Another problem in the lit-erature has been the difficulty in cross-validating findingsfrom various studies because of their use of different con-cepts and measures of alliance outcomes (Parkhe 1993b).

Despite the proliferation of measures, there is anemerging preference for dissolution as the appropriatemeasure of alliance instability, if not failure. A funda-mental premise is that a strategic alliance is not expectedto last indefinitely and, therefore, the instability of anyalliance may be measured in terms of its expected dura-tion. It is generally believed that a strategic alliance per-ceived as performing successfully by the partners is morelikely to remain in operation than those viewed as lesssuccessful (Porter 1987). Hence, instability may be sig-naled byunexpectedtermination, i.e., dissolution of thecollaborative relationship.

Even though disagreement exists regarding whether

dissolution implies failure of a strategic alliance, there hasbeen substantial theoretical and empirical support for dis-solution as a sign of downside in interfirm collaboration.Kogut (1988) argued that dissolution typically reflects abusiness failure or irresolvable conflict among the part-ners. Porter (1987) contended that dissolution is signifi-cant because companies generally do not divest or shutdown a successful alliance; dissolution happens onlywhen the alliance is not financially viable. Not surpris-ingly, he found dissolution to reflect poor financial per-formance in one, if not both, parties. In addition to eco-nomic considerations, dissolving an ongoing cooperativestructure is significant considering powerful social-psychological motivations to preserve the cooperative re-lationships that entail transaction-specific investments.Investments in interfirm cooperation include not onlyeconomic and technological resources, but also socialcommitments and entanglements of individual agents. Itis not only in the economic but also in the psychologicalbest interests of the organizational parties to find ways to

Page 4: Interfirm Rivalry and Managerial Complexity

SEUNG HO PARK AND GERARDO R. UNGSON Interfirm Rivalry and Managerial Complexity

40 ORGANIZATION SCIENCE/Vol. 12, No. 1, January–February 2001

preserve their socially embedded relationship as they faceadverse situations (Ring and Van de Ven 1994). In inter-national alliances, studies often define successful col-laboration as a stable business relationship that meets theneeds of both partners over the long term while dissolu-tion creates long-term political tensions (Franko 1971).Studies also show that dissolution is highly correlatedwith parent firms’ reported dissatisfaction with the ven-ture and perceptions of how the ventures performed rela-tive to their initial objectives (e.g., Geringer and Herbert1989).

Given this confusion regarding the concept of alliancefailure, the theoretical framework of alliance failure de-veloped below is limited to the alliance level. In partic-ular, our model focuses on the assessment of strategicalliances as a governance mechanism that is formed topursue collaborative interests between two or more in-dependent firms. Accordingly, our model does not fullyaddress issues on alliance failure at other levels, such asparent firms, or theoretical constructs than the failure ofalliances as an alternative governance mechanism for in-terfirm transactions. Because our focus lies on the failureof a strategic alliance, not necessarily the failure of part-ners, dissolution may best reflect the concept of alliancefailure as featured in our conceptual model of alliancefailure.

Lack of Cross-Disciplinary IntegrationAnother problem in studying alliance failure is the pro-liferation of theories on the subject, with little cross-fertilization. The study of cooperation and competitiontends to be multifaceted and multidisciplinary, includingdisciplines from economics, political science, sociology,organization science, business strategy, etc. While the as-sumptions and approaches vary, these studies suggest thatinterfirm cooperation and competition are embedded insocial conditions or contexts that structure the ‘‘rules ofconduct’’ to govern cooperative and competitive behav-iors. These conditions influence how such behaviors aredefined, chosen, and implemented.

Not surprisingly, scholars have used a wide range ofdisciplines to explain the failure of strategic alliances: forexample, transaction-cost economics (Buckley andCasson 1988, Parkhe 1993a), strategic behavior theory(Kogut 1988), game theory (Parkhe 1993a, Gulati et al.1994), agency theory (Geringer and Hebert 1989),resource-based view (Eisenhard and Schoonhoven 1996),learning theory (Hamel 1991), and investment option the-ory (Kogut 1991) have been used to relate alliance failureto opportunistic hazards of competitive collaboration.Also, along with transaction-cost theory and agency the-ory (Geringer and Hebert 1989), various organization the-ories, such as resource dependence (Cummings 1984),

workflow interdependence theory (Cummings 1984),congruence model (Barkema et al. 1996), information-processing model (Mohr and Spekman 1994), evolution-ary model (Koza and Lewin 1998), and strategy-structurefit model (Holton 1981) have focused on managerial andorganizational problems, such as coordination costs andagency costs, to explain alliance failure. Table 2 sum-marizes these theoretical perspectives, which vary in sub-stance and underlying contexts, and key metaphors andgenerating forces that have been used to explain the fail-ure of strategic alliances.

Although these theories were unique in their contri-bution to explaining alliance failure, they mostly adoptedcontingency approaches, with each theory addressing asmall part of the relationships leading to alliance failure.So far, there has been little effort to integrate these mul-tiple views and to look for the underlying theoretical con-structs that could tie these eclectic explanations together.This explains why the field still remains so fragmentedand is lacking a systematic framework to explain alliancefailure. As the summary in Table 2 indicates, however,these multiple perspectives can be complementary; it ap-pears that they converge into two primary dimensions ofalliance failure: rivalry (or appropriation hazards) andmanagerial complexity. Below we develop a conceptualframework of alliance failure by focusing on these twodimensions, which cover most key issues as discussed inprevious studies. Accordingly, multiple perspectives in-troduced in earlier studies lay the theoretical foundationin our attempt to develop a comprehensive theoreticalframework of alliance failure, which highlights two of thedimensions most commonly addressed by these perspec-tives. Although not complete in integrating these multipleperspectives, we hope our effort lays the foundation forfuture research to develop a general theory of alliancefailure.

Toward a SynthesisEarlier we posited a strategic alliance as a temporal struc-ture of exchange relationship that generates cooperativeor competitive behaviors between partners, depending ontheir private incentives. Our discussion of different theo-retical approaches in the preceding section suggests thatalliance failure can arise from rivalry between partners,and managerial complexity. The contractual stipulationof strategic alliances, that is a hybrid of market and hi-erarchy, explains this dual nature of alliance failure. Prin-ciples of market and hierarchy are still operative in analliance although the alliance is a governance mechanismmeant to remedy failures of these governance structures.Strategic alliances purportedly protect against the hazardsof market and hierarchy through partners’ investment of

Page 5: Interfirm Rivalry and Managerial Complexity

SEUNG HO PARK AND GERARDO R. UNGSON Interfirm Rivalry and Managerial Complexity

ORGANIZATION SCIENCE/Vol. 12, No. 1, January–February 2001 41

Table 2 Theoretical Perspectives on Alliance Failure

Theoretical Perspectives Metaphors General Forces

Interfirm Rivalry

Transaction cost theory (Kogut 1988, Parkhe 1993a,Buckley and Casson 1988, Defillipi and Reed 1991,Park and Russo 1996)

Strategic behavior theory (Kogut 1988)Game theory (Parkhe 1993a, Gulati et al. 1994)Resource dependency theory (Cummings 1984)Transactional value approach (Zajac and Olsen

1993)Resource-based view (Eisenhardt and Schoonhoven

1996)Investment option theory (Kogut 1991)Learning theory (Kogut 1988, Hamel 1991)Workflow interdependence view (Cummings 1984,

Park and Russo 1996)

Mutual forbearanceOpportunismTrustBehavioral transparencyReciprocityReputation

Incentives among thepartners

Nature of competitiverivalry

Goal divergenceOperational overlapAsymmetric dependence

& learning

Managerial Complexity &Uncertainty

Information-processing model (Mohr and Spekman1994)

Organizational control (Yan and Gray 1994, Killing1983, Geringer and Hebert 1989, Holton 1981)

Organization (Cultural) congruence (Park andUngson 1997, Barkema et al. 1996, Brown et al.1988)

Role conflict and formalization (Shenkar and Zeira1992, Ring and Van de Ven 1994)

Agency theory (Geringer and Hebert 1989)Transaction cost theory (Park and Ungson 1997,

Geringer and Hebert 1989)

ConflictDissimilarityAmbiguityControl

Uncertainty andequivocality

Cultural dissimilarityOrganizational misfitExcessive formalizationAdaptive capacityInterpersonal dynamics

Figure 1 An Integrative Model of Alliance Failuretime and resources, which in return drive strong incen-tives for them to honor contractual terms of the alliance.Unlike this purported belief, however, studies have in-dicated that there still remain competitive threats to thepartners, such as an involuntary loss of potential revenuesand uncompensated transfers of rent-generating re-sources, e.g., technology (Hamel 1991). Even with min-imal opportunism, partners also face difficulties in coor-dinating their activities toward collective operations andin linking local operations at the alliance level with thepartners’ long-term strategic goals. Figure 1 presents amodel that explains alliance failure according to thesetwo sources of problems: opportunistic hazards due torivalry between partners, and coordination and agencycosts resulting from complexity and uncertainty in man-aging a cooperative relationship.

Previous studies related alliance failure mostly to com-petitive issues between partners (Kogut 1988, Parkhe1993a). However, we believe managerial problems areequally—if not more—important in explaining the failure

of strategic alliances, especially in the case of cross-border alliances. From an in-depth case study, Larson(1992) concluded that economic transactions cannot beseparated from the social context in which they takeplace. An idealized economic exchange is neither over-socialized, whose primary governance mechanism is val-ues and norms, nor undersocialized, where partners re-main as isolated and rational economic units (Larson1992). In a cooperative relationship, partners face two

Page 6: Interfirm Rivalry and Managerial Complexity

SEUNG HO PARK AND GERARDO R. UNGSON Interfirm Rivalry and Managerial Complexity

42 ORGANIZATION SCIENCE/Vol. 12, No. 1, January–February 2001

prime uncertainties: uncertainties about the people andorganization, and about their capabilities. The formertype of uncertainty refers to a social-relations condition,the latter to an economic condition, affecting personal andeconomic trusts, respectively. Our model in Figure 1 ad-dresses both economic and transaction-related issues andsocial-psychological, organizational, and managerial is-sues of alliance failure through the two dimensions ofcompetitive rivalry and managerial complexity, respec-tively. The model also indicates that trust and commit-ment work as countervailing tendencies in these relation-ships by narrowing the disparity in the partners’competitive incentives and managerial practices.

Interfirm Rivalry. In 1984, GE and Rolls Royceformed a strategic alliance to manufacture jet engines forcommercial airliners. The GE-Rolls alliance seemed tobe an ideal way of getting the two partners into marketswhere they had been weak, while enabling them to avoidsome of the development costs required to design a newengine from scratch. However, this ideal fit was scrappedin two years because the two companies’ cooperation inmaking jet engines turned into competition. Both sidesrealized that cooperation would not be possible given thedirect competition and the lack of trust between them(Wall Street Journal1986).

The GE-Rolls alliance illustrates that even when stra-tegic alliances between competitors are formed to im-prove their competitive positioning, this is not necessarilycongruent to their individual goals. The same competitivemotives to form an alliance accounts for the eventual de-mise of the alliance (Kogut 1988). Bleeke and Ernst(1993) reported that the success rate for alliances betweencompetitors with similar core businesses, geographicmarkets, and functional skills is only about one in three.They concluded that managers typically choose directcompetitors as partners in pursuit of short-term synergythrough consolidation of overlapping product and marketpositions. However, such competitive alliances tend tofail because individual partner goals become misalignedwith these collective goals, as illustrated by the GE-Rollsalliance. For so long as interfirm transactions are madewithin a strategic alliance, competition will motivate part-ners to pursue individual interests at the expense of theother partner (Parkhe 1993b). Kogut (1988) noted thateven equity-based alliances, short of full ownership, can-not fully resolve the potential for competitive conflict.Empirical evidence also supports that strategic alliancesbetween direct competitors are most likely to fail (Parkand Russo 1996). Park and Russo concluded that the fail-ure of strategic alliances is primarily attributable to causesassociated with exchange between distinct parties, and

that once an exchange is moved away from a hierarchicaloperation, a major transactional threshold has beenbreached.

The presence of competition within a cooperative ar-rangement causes a high level of transaction costs as part-ners adopt various contractual stipulations to avoid op-portunistic hazards. According to a transaction costperspective, these stipulations and costs involve those inacquiring and processing information, monitoring and ad-ministering contractual promises, and enforcing contrac-tual promises. The potential transaction costs also involvelegal and organizational costs in coordinating activity andenforcing conventional behaviors. Formal contracts be-tween parties typically codify only a minimal set of ob-ligations while a broader set of obligations is determinedinformally (Buckley and Casson 1988). It is, however,essential that parties meet both formal and informal ob-ligations to sustain a cooperative relationship. Becauseof the need for both formal and informal safeguards,Williamson (1991) described strategic alliances as inher-ently temporal, unstable, and disfavored, particularly inthat changes in an alliance cannot be made unilaterally(as with market consent), or by fiat (as with hierarchy).The success of an alliance depends on mutual consentfrom partners, but this takes time, particularly when ex-ternal conditions change over time. As long as an allianceremains as an alternative mode of competition, defectionfrom the spirit of collaboration is inevitable, especiallywhen the lawful gains from insisting upon literal enforce-ment exceed the discounted value of continuing the co-operative relationship.

Collective benefits from an alliance are typicallyfuture-oriented and uncertain, while opportunity costsfrom cheating are more immediate and often tangible,which further aggravates opportunistic tendencies in analliance. A partner’s opportunistic behavior brings im-mediate realization of individual goals without facing theuncertainty of long-term returns. Interfirm rivalry in analliance also makes mutual forbearance less appealing tothe partners as they lack a long-term view. Buckley andCasson (1988, pp. 34–35) concurred that ‘‘a short-termview is likely to prevail when the agent expects the ven-ture to fail because of cheating by others. The risk ofprejudicing the venture through its own cheating is cor-respondingly low, and there may be considerable advan-tages in being the first to cheat because the richest pick-ings are available at this stage.’’

The prisoner’s dilemma also provides an insight abouthow competitive rivalry emerges within an alliance(Parkhe 1993a). The prisoner’s dilemma problem depictsconditions under which, at each iteration, there is an in-centive to deviate from an institutional rule. In a com-petitive alliance, application of the game theory suggests

Page 7: Interfirm Rivalry and Managerial Complexity

SEUNG HO PARK AND GERARDO R. UNGSON Interfirm Rivalry and Managerial Complexity

ORGANIZATION SCIENCE/Vol. 12, No. 1, January–February 2001 43

that although each partner does better when all act co-operatively than when all act noncooperatively, any onepartner can do better by acting noncooperatively when allother partners act cooperatively. Partners face constanttemptations to renege on an agreement because a partnerthat does not renege, while its partner does, will do worsethan when all partners renege. Therefore, partners sacri-fice collective goals in pursuit of private self-interests,which, in turn, can result in poorer performance for bothparties and failure of the alliance. Both parties in an al-liance can gain considerably by divulging meaningfulproprietary information. However, one party stands tolose a lot if it provides this information while the otherparty does not reciprocate it. In fact, the party gains theadvantage by holding out. Both parties stand to lose orgain, depending on their propensity to compete or coop-erate.

As customarily thought, in the absence of binding con-tracts, adverse consequences will result in the event thatboth parties do not cooperate. The game theory modelimplies that strategic alliances fail not because of poordecision making by managers, but because of the natureof incentive structure in a prisoner’s dilemma (Parkhe1983a). This is true unless the pattern of payoff is ad-justed to encourage mutual coordination of trust throughcontinuing interaction between the partners (Zajac andOlsen 1993). Gulati et al. (1994) suggested that partnerscan avoid the prisoner’s dilemma by restructuring the na-ture of cooperative decision making. Because partners ex-pect better payoffs from the alliance than they wouldwithout the alliance, they may restructure the decisionmaking into a sequential process instead of a simultane-ous one, which creates the prisoner’s dilemma. Partnersmay initiate unilateral commitment by committing to oneof the choices already available to them, creating a situ-ation similar to mutual hostage and credible commitment.

Still another perspective for understanding rivalrycomes from the resource-based view of the firm. Withinthis perspective, firms are heterogeneous with respect totheir resources/capabilities endowments. Resource en-dowments are sticky over any strategically relevant timeframe, and absent some linkages—like alliances—firmsare stuck with what they have because of imperfect mo-bility and complex development process of resources/capabilities. It is easier to acquire existing resources/capabilities from alliance partners than to develop theminternally, which has been the main cause for prolifera-tion of technological alliances in recent years (Eisenhardtand Schoonhoven 1996). It is likely in an alliance thatthe resource-advantaged partners will appropriate theeconomic rents of their less-advantaged partners. Hamel(1991) reported that several partners in his study regarded

alliances as transitional devices where the primary goalwas to internalize partner skills. Reich and Mankin (1986)criticized that unintended unilateral transfer of technol-ogy competencies from U.S. firms to their Japanese part-ners has caused the erosion of U.S. firms’ resource basesand competitive advantages. Resource appropriation ismore probable in the direction of partners with more am-biguous, immobile resources appropriating the less am-biguous resources of their partners. The partners withtransparent resources and capabilities are more vulnerableto resource appropriation by the other partner. The ap-propriation hazard is further heightened as the partnersdiffer in their capabilities to learn and internalize externalcapabilities (Hamel 1991).

Resource appropriation is also affected by the patternof interdependence between the partners (Park and Russo1996). Integrative alliances, in which the partners’ con-tributions represent a pooling of their talents (e.g., whenpartners jointly manufacture a new good), require morefrequent and intensive information flow. As the interac-tion continues, it is possible that one of the partners couldhave opportunities to gradually expose and appropriatethe other’s key firm-specific assets. Studies indicate thatintegrative alliances face a stronger threat of appropria-tion hazards where partners may lose important know-how and its competitive basis, than sequential allianceswhere each partner’s contribution lies on a sequential pathwith clear boundaries between partners.

In summary, various studies have attributed the failureof strategic alliances to interfirm rivalry by applying mul-tiple theoretical perspectives, such as transaction-costtheory, game theory, resource-based view, learning the-ory, and resource dependence theory. This line of studiesfocuses on rivalry between partners, whose primary em-phasis is on pursuing self-interests over long-term collec-tive goals. Opportunistic hazards are inevitable in stra-tegic alliances because of this competitive rivalrybetween partners. There exists a threat of each partnerpursuing short-term and tangible gains by appropriatingits partner and reneging on the contracts. In a competitivealliance, it is difficult to develop trust between partners,and once each party begins to doubt the other there isoften no end to it. Subsequently, strategic alliances be-come a costly governance structure to arrange interfirmtransactions as partners try to employ safeguards againstthese potential hazards. These appropriation hazardseventually lead to the failure of the alliance.

Interfirm rivalry is an issue that resides in the nature ofcooperation and partners’ role and disposition toward analliance. This dimension of alliance failure is thus closelytied with partners’ incentive and motivation in a coop-erative relationship. Underlying rivalry between partners

Page 8: Interfirm Rivalry and Managerial Complexity

SEUNG HO PARK AND GERARDO R. UNGSON Interfirm Rivalry and Managerial Complexity

44 ORGANIZATION SCIENCE/Vol. 12, No. 1, January–February 2001

presents a situation in which each partner weighs gainsfrom the collective operation against individual interestsas new opportunities emerge in the market. As long asthis competitive uncertainty continues, it is difficult todevelop trust between partners and the alliance loses itsadvantage as a transactional mechanism over market andhierarchy. Interfirm rivalry, however, explains only partof the reasons for alliance failure, primarily related toeconomic, strategic, and transactional issues at the inter-organizational level. Managerial complexity and uncer-tainty also poses another major problem in managing analliance. Implementing an alliance is a difficult task be-cause of differences in partners’ managerial and organi-zational practices.

Managerial Complexity. The AT&T-Olivetti alliancein the early 1980s, which combined Olivetti’s compe-tence in global marketing and AT&T’s technological su-periority, was considered a perfect match, at least in eco-nomic terms. Despite such strong economic and strategiccompatibility, however, this alliance was troubled fromthe beginning, primarily due to managerial and organi-zational incompatibility. Olivetti was an extremely ag-gressive, fast-moving, entrepreneurial organization whileAT&T was regarded as the ultimate example of profes-sional, technocratic management culture. As this case il-lustrates, strong economic and strategic complementarityand the absence of rivalry and asymmetry between part-ners do not necessarily lead to success. Strategic alliancesalso fail because they require excessive effort to coordi-nate and integrate two independent organizations whichresult in a high level of managerial complexity and un-certainty. Along with coordination costs as partners striveto work together, they also experience agency problemsin aligning their interests with alliance managers who donot necessarily work for their parent firms’ interests.

In a most basic sense, alliances are formed because thegains from specialization override the costs of coordina-tion. Specialization by discrete parties creates the needfor collaboration when further specialization provescostly. For example, if Partner A has specialized in a par-ticular skill, such as research and development, and Part-ner B has a distinctive edge in manufacturing, forging analliance between them to meet predetermined goalswould be more efficient than if they had to learn eachother’s specialties.

Coordination problems within the alliance stem fromthe complexity of activities that arise from specialization.Even when the overriding goal is to cooperate, the lackof information that parties need to have to enable themto coordinate their specialized activities may impede co-operation. Without the exchange of proper information

over time, partnerships can erode into adverse, competi-tive situations. The complexity of activities that resultfrom the difficulty in accessing specialized or idiosyn-cratic knowledge creates difficulties in communicationand may prevent effective cooperation. Early writers,such as Simon (1947) and Barnard (1968), have arguedthat this uncertainty and complexity in information pro-cessing is why complete cooperation is not possible. Inthe context of alliances, greater coordination betweenmore specialized partners is needed. The logic for col-laboration is that each partner derives value from coor-dinating specialized activities that far exceed the costs ofcoordination. However, it is difficult to sustain this po-sition over time. The dynamics of the exchange changesthe balance of contributions, creating asymmetric depen-dence (Hamel et al. 1989). Unless this is corrected, co-ordination difficulties will overwhelm cooperative activ-ities, leading to the failure of the alliance.

Successful alliances require nurturing the collaborativerelationship and creating new value together, which goesfar beyond simple exchange of competencies. Top man-agers, however, often devote more time to screening po-tential partners in financial terms and controlling the re-lationship than to managing the partnership in humanterms. People-related issues are often not addressed untilan alliance is formed, which may be due to the belief thatthese can be worked out later, or perhaps to maintainsecrecy in negotiating and planning such ventures (Cascioand Serapio 1991).

Strategic alliances often involve managers from differ-ent parent companies with different national, cultural, so-cial, political, and economic backgrounds. In the case ofcross-border alliances, the social context in which theyoperate is partly defined by the cultural and institutionalbackground of the nationalities the partners represent. Na-tional culture then affects managerial behavior and mod-erates the relationship between structural and economicvariables and performance of the alliances. It has beenproposed that the similarity of cultural values may reducemisunderstanding between the partners, and that cultur-ally distant partners experience greater difficulty in theirinteractions. Lane and Beamish (1990) argued that cul-tural compatibility between partners is the most importantfactor in the survival of a global alliance, because thedefects in alliances often stem from the unobtrusive in-fluence of national culture on behavior and managementsystems, which may create unresolved conflicts. Partnersfrom different national cultures tend to experience a lackof fit in organizational and strategic practices, makingcoordination more difficult (Park and Ungson 1997).Lack of fit with a partner’s culture also leads to poorcommunication and mutual distrust.

Page 9: Interfirm Rivalry and Managerial Complexity

SEUNG HO PARK AND GERARDO R. UNGSON Interfirm Rivalry and Managerial Complexity

ORGANIZATION SCIENCE/Vol. 12, No. 1, January–February 2001 45

In addition to the difference in national culture, thefailure of a strategic alliance is also caused by how com-patible the partners are in regard to specific organizationalcharacteristics (Doz 1988, Yan and Gray 1994, Park andUngson 1997). Because an alliance is a hybrid of morethan two independent firms, dissimilarities in organiza-tional structures and processes can create problems in co-ordination. Brown et al. (1988) conjectured that organi-zational compatibility is a more significant factor thannational cultural differences in explaining the outcome ofan alliance. As dissimilar partners may expend time andenergy to establish standard managerial routines to fa-cilitate communication, they may incur higher costs andmistrust relative to a strategic alliance with similar part-ners. Disagreements over operating strategies, policies,and methods often arise even when the alliances are quiteself-contained and have a minimum of interaction withother units within the foreign-based multinational part-ner’s operations. Thus, differences of opinion about div-idend payout policies, debt-equity ratios, marketing pol-icies, quality control methods, and the like can becomeso great that one or more of the partners either wishes towithdraw from the alliance or at least is chronically un-happy with the operation of the enterprise.

There is also a high degree of uncertainty and equi-vocality in managing an alliance. With the inherent sys-tem multiplicity, serving two bosses and two sets of ex-pectations, alliance managers are prone to role conflictsand ambiguities (Shenkar and Zeira 1992). Alliance man-agers often lack information about the specific expecta-tions of different managers in each parent firm, the dif-ferent expectations of different employee groups in thealliance, and the different expectations of host countryorganizations with which they must interact. Alliancemanagement thus needs to exert costly information-search effort to reduce these uncertainties and promote astrategic fit of allying firms with the alliance.

Strategic alliances also face agency problems in inte-grating alliance activities with the parent firms. The em-ployment relationship between the parent organizationand alliance managers reflects the basic agency structureof a principal and an agent who have different goals andattitudes toward risk (Holton 1981). Doz (1988, p. 335)described this conflict in the following terms: ‘‘Most ofthem (alliance managers) had little to lose, in careerterms, either because they were already marginal in theirown company or, on the contrary, because their positionwas so secure as to be unassailable. In both cases, thisallowed them to accept unusual levels of personal risksand to deviate from corporate norms in making the part-nership work.’’

Agency problems are exacerbated by strong interde-pendence between the alliance and the parent organiza-tion. This is because decisions at the alliance level havedirect effects on the parents’ overall costs and revenues.Partly as a consequence, there should be tight and effec-tive monitoring of the alliance in order for it to achieveits objectives (Franko 1971). Even so, it is difficult andcostly to verify what alliance managers do. Alliances aretypically undertaken in highly uncertain and risky set-tings, creating a high level of outcome uncertainty andaggravating the problems of monitoring managers’ activ-ities, much less deciding their appropriate incentives. Ithas been reported that most alliance managers often dis-regarded the impact of alliance activities on parent firms’operations (Holton 1981). The agency problem was ob-vious in the Yamatake-Honeywell alliance as Honeywelland the alliance both entered China and ended up in directcompetition (Gomes-Casseres 1996).

This conflict between parent firms and the alliance iswhy control has been a variable of interest in prior re-search (Geringer and Hebert 1989). Even though empir-ical findings are conflicting, the emerging pattern directlyrelates management control to alliance performance (Yanand Gray 1994). Findings also suggest that problems inpiloting a strategic alliance most likely occur when con-trol is equally shared. In an unbalanced alliance, the non-dominant partner has an a priori notice that their pursuitof individual interests may be difficult (Park and Russo1996). Accordingly, they expect at least some ex poststrategic choices to diverge from what they would havechosen. Dominant ownership can help avoid agencyproblems by simplifying the control process and main-taining easier integration with the parent firm’s strategicdirection, although incentives to cheat are greater in adominant alliance.

Yan and Gray (1994) argued that when the control iseven between partners, each partner’s performance, atleast as assessed by its own perspective, is equal. Controlhelps the parent firm to achieve the full potential of theirstrategies and to attain their objectives by reducingagency costs. However, control may also lead to in-creased overhead costs because it implies more respon-sibility and resources for the partner (Geringer and Hebert1989). Moreover, the exercise of extensive control (andformalization to manage the control process) over an al-liance’s activities and decisions may incur a high level ofbureaucratic costs and limit the efficiency of the alliance.Excessive control by a foreign partner over a local alli-ance often hinders the autonomy and flexibility of thealliance and its management, which weakens potentialbenefits of the alliance. Therefore, it is critical to find thestrategy-structure fit at which the margin between benefits

Page 10: Interfirm Rivalry and Managerial Complexity

SEUNG HO PARK AND GERARDO R. UNGSON Interfirm Rivalry and Managerial Complexity

46 ORGANIZATION SCIENCE/Vol. 12, No. 1, January–February 2001

and costs from an alliance becomes optimal. This willminimize potential agency problems, achieve tight inte-gration between the parent firm and the alliance, and al-low enough freedom for alliance managers to flexibly re-spond and adapt to the needs in local markets (Bleekeand Ernst 1993). The trouble in the Yamatake-Honeywellalliance illustrates the importance of maintaining balancein controlling an alliance. Honeywell’s hands-off ap-proach explains the initial success of the alliance, but theeventual conflict between Honeywell and the alliance isattributed to the weak control over the alliance.

As strategic alliances fail because of economic andstrategic incompatibility, alliances also fail because ofdifficulties in implementing the collaborative efforts andintegrating them with the parent’s strategic objectives.These managerial and organizational concerns in strategicalliances are rooted in various theoretical perspectives,such as cultural complementarity, information-processingmodel, organizational congruence, strategy-structure fit,and agency theory. These theories focus on managerialcomplexity and uncertainty in coordinating the relation-ships between partners, and between partners and the al-liance, that become the sources of coordination andagency costs and organizational rigidity of the alliance.Along with appropriation hazards, these managerial con-cerns incur a high level of transaction costs, especiallybureaucratic costs and agency costs, in the process of co-ordination. Alliance partners often end up wasting valu-able time and resources to understand each other, evenbefore the actual collaboration starts. This is what hap-pened in the proposed alliance between Ford and Fiat in1995. After more than a year of discussing a wide rangeof proposals, they realized that they could not agree onthe management structure necessary to launch the alli-ance. The differences in management and operationalstructures were so severe that they decided it was im-practical to enter into a cooperative relationship. It is im-portant to maintain flexibility at the alliance level becauseit is inevitable that internal and external conditions of thepartners and the alliance gradually change over time. Ex-cessive flexibility at the alliance level, however, often be-comes a serious problem in working with parent firms,causing them to withdraw their commitment to the alli-ance, as illustrated by the trouble in the Yamatake-Honeywell alliance.

Interaction Between Interfirm Rivalryand Managerial ComplexityIn the foregoing sections, we have illustrated the sourcesof alliance failure in terms of interfirm rivalry and man-agerial complexity. While they are conceptually distinc-tive, these two components are inseparable in practice.

Strategic alliances fail when there is strong rivalry or ahigh level of managerial complexity, but failure often re-sults from the interaction between these components.These two dimensions represent strategic and managerialissues that affect partners’ motivation and ability to con-tinue the cooperative relationship.

The most unstable type of alliance is one featured witha high level of rivalry and managerial complexity. Thereis little strategic and organizational complementarity inthis type of alliance; participants put private interests overcollective interests and have difficulty in coordinatingmanagerial and organizational differences between them.This type of alliance is likely to be between direct com-petitors, as indicated by strong rivalry. Accordingly, it isdifficult to develop a trust-based relationship, and part-ners become hesitant to make a strong commitment to thecollaborative operation. Such an alliance also becomes ahighly inefficient governance structure to pursue interfirmtransactions. Coordination costs are high because it is dif-ficult to resolve organizational dissimilarities and to man-age cooperative processes. The Ford-Volkswagen alli-ance set up in the 1980s to expand their market basis inLatin America represents a typical case of this type ofalliance. Although the partners had a common goal ofestablishing a critical mass in Latin America’s emergingmarkets, the alliance was an uncomfortable marriagefrom the beginning. Because the partners were directcompetitors in most other markets, they were not willingto share their own design skills and marketing strategieswith their partner. Differences in cultural values and or-ganizational practices also made it difficult to develop acoherent strategy to challenge GM, who was a majorcompetitor in the market.

By extension, we propose that the alliances with weakrivalry and low managerial complexity are least likely tofail. Previous studies point out the importance of com-plementary needs for the viability and potential of an al-liance (Porter and Fuller 1985). These types of alliancesare based on highly complementary strategic interests be-tween partners that can easily converge toward eachother. There is thus little rivalry in the alliance and co-operative behaviors become more compelling. Thiswould be an ideal setting for interfirm collaboration with-out any serious transaction and coordination costs; thepartners are easily able to understand and trust each otherbecause of similar cultural and organizational back-grounds and strong economic complementarity (or weakrivalry). They also become easily committed to collectiveinterests that can be carried out even at a low level ofsocialization between partners, which further contributesto improving transactional efficiency. We can also expectsimilar economic mindsets and strong organizational

Page 11: Interfirm Rivalry and Managerial Complexity

SEUNG HO PARK AND GERARDO R. UNGSON Interfirm Rivalry and Managerial Complexity

ORGANIZATION SCIENCE/Vol. 12, No. 1, January–February 2001 47

transparency between the partners, which help them un-derstand each other better.

Unlike the previous two types of alliances, there ismuch ambiguity in predicting the outcome of the alli-ances with either strong rivalry or high managerial com-plexity. The likelihood of failure in these alliances de-pends on how the partners manage the problems causedby either their competitive intentions or organizationaland managerial dissimilarities. Firms enter strategic alli-ances in pursuit of their self-interests through collabora-tion with other firms. The first-order condition for suc-cessful alliances is the presence of complementarybenefits for both partners, which cannot be acquired ef-ficiently through individual operations. Once an allianceis formed based on partners’ economic interests, its suc-cess subsequently depends on the dynamic change ofeach partner’s economic interests and the effective im-plementation of alliance activities. We thus view the man-agement of managerial and organizational dissimilaritiesas a second-order condition for managing successful al-liances. Strategic alliances fail when either or both ofthese first-order and second-order conditions are not met.An alliance is able to sustain its structure and remains asan efficient mechanism for interfirm transactions as longas partners’ economic benefits overshadow potential costsin managing the alliance. Alliance partners would bemore eager to work out conflicts from managerial andorganizational differences as long as potential economicbenefits remain strong. Because they have strong eco-nomic incentives and are more willing to cooperate, it isalso much easier to develop mutual trust. Once partnersdevelop trust toward each other, the alliance becomesmore sustainable as they can better coordinate their ef-forts to overcome managerial complexity and reduce co-ordination costs. The strong economic incentive also al-lows effective socialization and governance in managingthe cooperative structure.

Governance structures that are closer to a market mech-anism, such as licensing and other types of nonequityrelationships, would work better in an alliance with weakrivalry but high managerial complexity. Hierarchy-likegovernance structures (e.g., joint ventures), involve ahigh degree of interdependence and coordination betweenpartners, and interfirm coordination becomes a highly de-manding and difficult task. On the other hand, market-like governance structures require (relatively) a muchlower level of interdependence and a simpler coordina-tion process than hierarchy-like governance structures. Itis thus possible that alliances can overcome problems re-sulting from organizational dissimilarities by adopting aproper governance mechanism, as long as there are suf-ficient economic gains. This does not necessarily imply

that alliances with high managerial complexity and lowrivalry do not fail. Alliances often face a situation inwhich coordination difficulties offset even strong eco-nomic gains. Firms would be better off adopting differenttransaction mechanisms for similar economic interestsrather than entering into an inefficient alliance structure.The alliance among IBM, Toshiba, and Siemens illus-trates a failure resulting from high coordination costs, de-spite strong economic interests for the partners. This al-liance was set up to design and manufacture 256-megabitchips, which required a high level of interaction amongthe partners. However, the economic potential in this al-liance was soon overshadowed by difficulties in coordi-nating the three culturally very distinctive firms; IBM per-ceived Toshiba as excessively group-oriented andSiemens as overly concerned with costs and details offinancial planning.

Direct rivalry between partners, however, poses a se-rious threat to alliance success in that it is difficult to builda trust-based relationship. This lack of trust impedes theexchange of firm-specific assets to accomplish collectivegoals. Especially when alliances involve technologytransfer, there is reluctance to provide extensive infor-mation about the technology. Expecting to be cheated,each partner tends to disguise or limit its contribution. Asa result, only low-quality or insufficient information willbe transferred in competitive alliances, and eventuallydishonest activities will drive honest activities out of thealliances, causing the failure of the alliance (Akerlof1970). Presumably, alliance partners can prevent theseproblems by adopting a tightly specified governancestructure, such as hierarchy-like governance structure(e.g., equity alliances). However, these alliances cannotavoid possible transaction hazards because each party fo-cuses more on its self-interest, given the competitivenessbetween the partners. These are similar to the type ofintermediate governance structures Williamson (1991)referred to as a highly unstable form of governance struc-ture. Williamson argued that intermediate types of gov-ernance structures are not impervious to opportunisticthreats, and they are nothing more than a temporary ar-rangement for interfirm transactions. Even if the partnersin a competitive alliance can design a mechanism to gov-ern possible opportunistic hazards, this only compoundsmanagerial complexity as tighter monitoring is needed.In the following section, we discuss the underlying con-ditions that affect interfirm rivalry and managerial com-plexity in an alliance. In particular, it focuses on some ofthe process-related conditions that affect rivalry and man-agerial complexity differentially over time as the coop-eration continues between partners.

Page 12: Interfirm Rivalry and Managerial Complexity

SEUNG HO PARK AND GERARDO R. UNGSON Interfirm Rivalry and Managerial Complexity

48 ORGANIZATION SCIENCE/Vol. 12, No. 1, January–February 2001

Figure 2 The Dynamics of Erosion and Failure of StrategicAlliances

Toward a Process Theory of AllianceErosion and FailureWhat would make a cooperative alliance more competi-tive (and vice versa)? In our view, there are a number ofconditions,ex ante, in situ,and ex post, that influencecooperative and competitive behaviors within an alliance,and that explain how alliances might transform over timeinto different types discussed earlier. These conditions areillustrated in Figure 2.

Ex Ante ConditionsEx ante conditions refer to prevailing conditions at thetime of the formation of an alliance. For the most part,these conditions influence the structure of the alliance andcan predispose partners toward competitive or coopera-tive behavior. As developed in earlier sections, these con-ditions include the nature of competition (direct compet-itors are more likely to engage in future competitivebehavior), complementarity (partners that are wellmatched in terms of their contributions are more likely toengage in cooperative behavior), and the perceptions ofendgames (partners who recognize the downfalls of

cheating will be less likely to engage in opportunistic be-havior).

In addition to rivalry, ex ante conditions cover struc-turing agreements that arise from the need for coordina-tion, particularly in the context of parent-firm relation-ships. We noted two conditions that influence competitiveand cooperative behavior: cultural incongruities (differ-ences in corporate culture and internal politics create highcoordination costs in an alliance) and agency issues(agency problems also occur because alliance managershave different attitudes on risk than those of the parentorganization), under which coordination costs occur.

In Situ ConditionsWhat are the dynamics that characterize possible move-ments among the different types of alliances?In situcon-ditions relate to those that affect rivalry and managerialcomplexity within the course of an alliance. In a funda-mental sense, these conditions are process related andstrongly influence development of the alliance. Eachstage of cooperation provides a receptive context for theinitiation and evolution of economic exchanges in the fol-lowing stage. Cooperative dynamics in a stage builds ahistorical context that affects the nature of cooperation insubsequent stages. Each stage of alliance evolution alsoruns through a cyclical process of negotiation, commit-ment, and execution according to the partners’ evaluationof the efficiency and effectiveness of the alliances (Ringand Van de Ven 1994). The extent to which strategicalliances are affected by competitive rivalry and mana-gerial complexity changes over time as resource ex-changes and social interaction continue between partners.At each stage, alliance partners begin to reassess the over-all values of the alliance in comparison with other typesof governance arrangements, e.g., market and hierarchy,because their strategic interests and cooperative roleschange over time. Empirical evidence indicates that thefailure rate of equity alliances is nonmonotonic, rising inearly years, but then declining after reaching a peak threeor four years into the operation (Park and Russo 1996).Strategic alliances are formed with expectations of suc-cess and complementary economic benefits. The survivalof strategic alliances hinges on their adaptive capabilitiesto such dynamic changes over time. In our view, fourinterrelated factors are salient at this stage: changes in theexternal environment, perceptions of balanced contribu-tions, the pace of bilateral learning, and asymmetric de-pendencies.

Changes in the External Environment.Volatile con-ditions also lead to vulnerabilities within an alliance, asthese affect the partners’ competitive positions. This isconsistent with others who have emphasized industry and

Page 13: Interfirm Rivalry and Managerial Complexity

SEUNG HO PARK AND GERARDO R. UNGSON Interfirm Rivalry and Managerial Complexity

ORGANIZATION SCIENCE/Vol. 12, No. 1, January–February 2001 49

environmental factors as the major reasons why organi-zations fail (e.g., Hannan and Freeman 1977), or whyspecific alliances are prone to failure (e.g., Kogut 1988).However, there is mounting evidence that alliance failureis not so much due to external factors, such as changingeconomic conditions, slackened demand for the product,opposition of government agencies, or changes in tech-nological structure, but to internal factors, such as behav-ioral and managerial problems (Parkhe 1993b). An ex-planation is that once firms invest dedicated assets in analliance, changes in the external environment become aless significant factor for alliance failure unless changesare adverse and revolutionary.

One serious concern for students in economics wouldbe the anticompetitive issue of interfirm collaboration(Berg and Friedman 1980). Some alliances, particularlyjoint production ventures, can in fact be anticompetitivein terms of price fixing and bid rigging. Such illegitimateventures are merely shams to mask cartel behavior andwould face legal constraints as a serious external threatto their long-term survival. The changes in the U.S. leg-islation have reduced the likelihood of antitrust liabilityin recent years. The National Cooperative Research Actof 1984 allows joint R&D to obtain partial or full im-munity from triple damage suits threatening the mainte-nance of the ventures. The U.S. government also does notintend to condemn presumptively joint production ven-tures, including ventures among competitors, and willtake full account of foreign competition and potential pro-competitive efficiencies. Unless a venture is a disguise bya cartel to exercise market power and raise prices to con-sumers, it is safe from legal challenges, and its survivaldepends mainly on the partners’ behavioral and mana-gerial dispositions. Furthermore, if an alliance is anticom-petitive, the venture fails even before its operation getsoff the ground. In other words, in the process of negoti-ation, anticompetitive ventures are normally blocked bygovernment agencies as illegal operations. In short,changes in the external environment can change the mar-ket positions of alliance partners, thereby triggering morecompetitive or cooperative activities aimed at restoringor improving prealliance market positions.

The Perceptions of Balanced Contributions.Therai-son d’etatfor alliances are complementary benefits thataccrue to participating partners. Alliances provide accessto the other firms’ proprietary know-how, assets, humanresource needs, market access needs, government and po-litical needs, and knowledge needs. Porter and Fuller(1985) argued that an alliance is stable for as long ascontributions by each partner are perceived to be bal-anced and equitable.

The investment of both time and resources providesincentives for parties to honor the terms of the contract.However, the balance of contributions is perceived to beequitable provided that there are sufficient monitors forpartners to make these assessments. But even with thebest intentions, the complexity of the alliance allows theparties to pursue disparate goals and creates the need formonitoring activities. However, because monitoring isalso costly and difficult, partners may attempt to reducetheir work effort.

The dynamic changes in the compatibility of partners’contributions also affect the partners’ bargaining posi-tions during the course of collaboration. Dymsza (1988)illustrated that international alliances between U.S.MNCs and local partners reached a stalement and brokeup once the local partners developed the capabilities totake charge of the alliance operation. Many foreign firmsalso start out in Japan through joint ventures, but theysoon terminate these ventures once they feel confidentenough to set up their own sales networks. Maintainingan alliance that does not offer significant complementarybenefits to the partners generates unnecessary costs. Theperceptions of equity in a cooperative relationship are acritical factor to stabilize the cooperative structure.

Pace of Bilateral Learning. Perhaps the key processvariable in managing an alliance is the difference in learn-ing capability and strategic intent for learning betweenthe partners. Strategic alliances provide an importantlearning mechanism for other firms’ firm-specific skills.However, the asymmetry in learning and the subsequentasymmetric dependence between partners lead to alliancefailure. This asymmetry is often cited as a primary reasonfor alliance failure between U.S. and Japanese firms(Hamel et al. 1989). It is also possible that changes in apartner’s competitive motives and strategic positionscause the other party’s contribution less significant overtime, if not obsolete. Based on in-depth case studies,Hamel (1991, p. 88) concluded that ‘‘in a narrow sensemanagers saw collaboration as a race to learn, but in abroader sense they saw it as a race to remain ’attractive’to their partners.’’

Failed joint ventures, particularly in the case of Japa-nese and American firms, can be traced to the ability ofone of the partners, e.g., Japanese partners, to learn morequickly the core skills of its partner, e.g., American part-ners. Learning is determined by the propensity or desirefor one partner (firm) to learn the core skills of anotherand the degree to which partners are ‘‘transparent’’, thatis, easy to learn about (see Hamel et al., 1989). Americanfirms have been known to be more ‘‘transparent’’ abouttheir work and attitudes compared to their Japanese orKorean counterparts.

Page 14: Interfirm Rivalry and Managerial Complexity

SEUNG HO PARK AND GERARDO R. UNGSON Interfirm Rivalry and Managerial Complexity

50 ORGANIZATION SCIENCE/Vol. 12, No. 1, January–February 2001

Asymmetry. To the extent that asymmetries in contri-butions occur, dependencies between alliance partnersdevelop. Over time, one partner may be more dependenton the other (Hamel et al. 1989). Partly as a response, thedisadvantaged party may reveal more and more of its pro-prietary know-how to maintain its initial attractiveness.Vulnerabilities created by this situation lead to ‘‘hostagesituations’’: The more dependent partner may face a smallnumber bargaining situation that is held up by the otherparty. This may in turn engender opportunistic behaviorsfrom the advantaged party. Because the survival of thedisadvantaged partner depends on overall alliance per-formance, it may become even more vulnerable to actionstaken by the dominant partner. The disadvantaged partnermay initiate renegotiation of contractual terms to main-tain equal or more control over the alliance, or leave thealliance altogether.

Strategic alliances are also vulnerable to shifts in in-ternal politics, as well as to wavering support by parentorganizations. For example, while Honeywell was goingthrough corporate restructuring in 1986, the alliances withMachines Bull of France, NEC of Japan, and EricssonTelephone of Sweden were all phased out. Doz (1988)argued that parochial subunit goals and corporate politicswithin top management could erode any teamwork, un-derstanding, and tolerance of rivalry. Difficulties withinthe top management team, in turn, can create wide cul-tural gaps and poor communications at the level of middlemanagement, where the need for collaboration is moreacute. Accordingly, middle managers and staff in an al-liance are often divided into these rival camps.

Ex Post ConditionsEx post conditions relate to outcomes arising from part-ners’ assessments of efficiency, equity, and goal attain-ment. Because purposes, values, and expected conse-quences are grounds for human choices, it is expectedthat continuation of an alliance will depend on such as-sessments. At the end of each collaborative cycle, eachpartner engages in redefinition of the alliance and initiatesa new loop of negotiation, commitment, and execution ofcooperation.

In claiming and distributing the value created by thestrategic alliance, partners may experience conflict whenthey perceive divergence of interests. However, partners’perceived level of the alliance’s and the partners’ value,relative to other types of governance structures like mar-ket or hierarchy and other potential partners, may changeat this stage. When partners realize a gap in performancefrom their historical comparison of actual to expectedvalue creation, they decide to terminate the cooperativerelationships unless they emphasize joint value maximi-zation supported by strong relational norms that might

have developed over time. If partners manage to buildsocial norms for joint value maximization, partners pur-sue joint searches for satisfactory outcomes to conflictingsituations and unsatisfactory results of the alliances (Za-jac and Olsen 1993).

Each of the looping sequences of negotiation, com-mitment, and execution stages, or the sequence oflearning-reevaluation-readjustment in cooperative rela-tionships, is assessed in terms of efficiency and equity.Sometimes what counts goes beyond the actual contri-butions made by each partner and the profits and otherbenefits obtained by each one. Even more important iswhat the partners perceive over the whole life of theoperation. Strategic alliances fail when one or both part-ners perceive unfair treatment or an unsatisfactory ratiobetween compensation and contribution from the alli-ances. It is imperative to maintain mutually beneficial re-wards for the survival of an alliance.

Earlier we postulated that trust and commitment couldmitigate competitive rivalry and managerial complexity.In this section, we discuss how trust emerges as a resultof ex ante, in situ, and ex post conditions. As partnersdevelop mutual understanding of each other over time, astronger form of trust, i.e., process-based trust, starts toreplace the initial characteristic-based trust (Zucker1986), and the alliance then becomes more dependent oninformal measures than it was in the earlier stages. Stra-tegic alliances become institutionalized and take on morehierarchical characteristics (Parkhe 1993b), substantiallylowering managerial complexity and coordination costs.There is a higher level of integration between partners,and social control appears as a primary governance mech-anism. However, even though the alliance becomes lessmarketlike in its design and governance, it is still affectedby competitive spirit (like market transactions) betweenpartners. As the interaction between partners continues,the alliance improves its repository of knowledge assets,i.e., interpartner relationships and skills, and the attach-ment between partners. The initial role relationships andexpectations become socially embedded through repeatedinteractions over time (Ring and Van de Ven 1994).These enhanced interfirm ties, however, become offset aschanges occur in resource fit between partners over time.

Seabright et al. (1992) maintain that resource fit dete-riorates over time when there is a reduction in partners’ability to satisfy resource requirements because of theincreases in a partner’s resource requirements, the de-creases in its exchange partner’s provisions, or the in-creases in the opportunity set. Strategic alliances are sub-ject to environmental as well as internal changes that arederivative changes of parent firms’ management structureor strategic interests. The payoffs from strategic

Page 15: Interfirm Rivalry and Managerial Complexity

SEUNG HO PARK AND GERARDO R. UNGSON Interfirm Rivalry and Managerial Complexity

ORGANIZATION SCIENCE/Vol. 12, No. 1, January–February 2001 51

alliances may change over time, altering incentive struc-tures for partners. Studies show that even though no vi-olations of trust were found, strategic alliances were de-mised to decline or dissolve due to changes in competitiveconditions and strategic direction; in particular, heavy re-liance on a partner and the exchange of proprietary in-formation represented major risks (Larson 1992).

While trust and commitment provide countervailing ef-fects for alliance stability, they do not evolve overnightand seldom come easily; they can be built only over time.As one popular resolution to the prisoner’s dilemmagame, authors suggested ‘‘tit-for-tat’’ and ‘‘mutual for-bearance’’ as strategies to develop trust and commitment(Buckley and Casson 1988). These strategies match for-bearance with forbearance but punish cheating. There-fore, one partner’s forbearance or credible commitmentat a stage gives rise to a positive feedback over time,while the perception of injustice initiates a negative feed-back, leading to eventual failure of the alliance. Doz(1996) suggests that partners in successful alliances werewilling to make irreversible commitments first, expectingother partners to reciprocate, but using the opportunity tobuild and test trust. Generally, trust emerges based on‘‘norms of equity,’’ and it can be extended to an orga-nization with a reputation for trustworthiness (Ring andVan de Ven 1994). Transaction parties become aware ofthe relevant party’s reputation on trust through consultingtheir own experience and/or from the experience of others(Gulati 1999). Taken collectively, opportunism leads toan asymmetric bargaining position and the dissipation offirm-specific assets. When these costs become insur-mountable, the alliance may not succeed. Over time, part-ner behaviors become path-dependent; there will be es-calating commitment as the alliances meet the partners’individual goals and outcomes are perceived to be fairlydistributed, but there will be decelerating commitment asthey begin to note any sign of mistrust or lack of com-mitment from the other parties.

The best and least costly mode of governance for analliance would be trust and mutual commitment. A Jap-anese executive involved in an alliance commented thata partnership works on the basis of trust and commitmentor not at all (Business Week1986). American managersnormally refer to formal contracts to avoid conflicts anduncertainties, while Japanese managers prefer mutual dis-cussion and mediation to formal contracts. Japanese man-agers view formality and the use of third-party manage-ment to control an alliance as signs of mutual distrust.Mutual trust in an alliance reduces interaction and mini-mizes bureaucratic complexity. It also helps partnersovercome conflicting situations and unexpected difficul-ties.

ConclusionsStrategic alliances have become an essential element inrecent corporate strategy. A review of the alliance liter-ature reveals an extensive list of reasons for entering analliance. Strategic alliances help firms gain market powerand access, institutional legitimacy, and new competen-cies, as well as to exploit firm-specific competencies orreduce environmental uncertainty (Eisenhardt andSchoonhoven 1996). Strategic alliances also have becomea powerful force in shaping a firm’s global strategy.

Despite these purported benefits, strategic alliances en-tail serious competitive risks, and managerial difficultiesin implementing the cooperative relationships. The highfailure rates documented in the literature indicate thatstrategic alliances are more likely to fail than to succeed.In fact, success in an alliance is now regarded as an ex-ception rather than a rule. Although several studies at-tempted to explore the potential sources of alliance fail-ure, the field still remains without a comprehensivetheoretical framework. Previous studies, whether empir-ical or theoretical, seemed to have adopted contingencyapproaches focusing on a series of interactive and struc-tural variables to explain alliance failure. This paper pres-ents a theoretical framework that integrates these multipleexplanations of alliance failure by focusing on interfirmrivalry and managerial complexity. We believe that thesetwo dimensions could be the logical foundations to un-derstand alliance failure that have been addressed so farin the literature. We also believe that the recent slowdownof studies on alliance failure, or alliance performance ingeneral, is partially the result of a lack of theoreticalframework that could offer a direction for further studies.In this regard, we contend that this paper serves as a firststep toward deepening our understanding of alliance fail-ure and developing a general theory of alliance failurewith well-defined and tested boundary conditions.

To conclude, this study presents a conceptual model ofalliance failure based on rivalry and managerial com-plexity in an alliance that affects partners’ strategic in-terests and motivation to cooperate (or compete), andtheir ability to implement the cooperative structure. Wefurther illustrate that the operation of rivalry and mana-gerial complexity in an alliance is a dynamic function ofthe interplay among ex ante, in situ, and ex post condi-tions. These stages of development occur as a repetitivesequence in assessing each partner’s contributions.Changes in external environment, perceptions of bal-anced contributions, the pace of bilateral learning, andasymmetric dependencies are key process variables thatinfluence whether alliances remain highly competitiveand complex, more complex than competitive, more com-petitive than complex, or relatively noncompetitive and

Page 16: Interfirm Rivalry and Managerial Complexity

SEUNG HO PARK AND GERARDO R. UNGSON Interfirm Rivalry and Managerial Complexity

52 ORGANIZATION SCIENCE/Vol. 12, No. 1, January–February 2001

complex. These variables also lead to pathological ten-dencies; more learning about the other may help promotetrust, but it creates greater dependency. Underlying thesestages, therefore, are social-psychological processes ofsensemaking and enactment by both parties (Ring andVan de Ven 1994).

A critical issue future studies should keep in mind isto clarify the concept of failure in a strategic alliance. Asillustrated earlier, the outcomes and their predictors in analliance vary according to the operational definition ofalliance failure. These multiple indicators of outcomemeasures that have been used in previous studies are notnecessarily accommodating to each other. In fact, someof the predictors are contradictory, although they mayexplain some form of alliance failure. For example, pre-dictors of one party’s success often indicate failure of thealliance as a governance structure. For this reason, itwould be a daunting task to integrate multiple indicatorsof alliance failure and test them simultaneously in a singlestudy. Rather, it would be more desirable to focus on aclearly defined concept of alliance failure to develop atheoretically and methodologically coherent study; or,studies may focus on multiple measures as long as theyrefer to the same level of analysis, i.e., alliance vs. partnerlevels.

Another critical issue in future research is to followdynamic, process-oriented approaches to study alliancefailure (Koza and Lewin 1998, Yan and Zeng 1999). Re-cently, Koza and Lewin advocated the importance ofadopting a coevolutionary perspective to study strategicalliances. Strategic alliances, like other organizationalforms, emerge as an adaptive mechanism to market un-certainty, and their developments over time reflect thecoevolution of distinctive firm capabilities and of industryand market activities. Studies have so far been limited tostatic explanations of contingency factors of alliance fail-ure. Given the change of in situ conditions over time,partners and alliances go through similar coevolutionarycycles in terms of their motives and capabilities towardthe cooperative relationships. Alliance failure is an out-come of this coevolutionary adjustment to changes in themarket, the competitive dynamics between partners, andassessment of efficiency of the alliance as an alternativegovernance structure. It is thus critical to adopt a dynamicperspective and historical observations of cooperativeprocesses. Because earlier studies followed a static ap-proach, they could explore only the beginning and theend of a cooperative alliance, but not the cooperative pro-cess in the mid-life of alliances (Yan and Zeng 1999). Itwould be interesting to study how partners’ opportunistictendencies evolve over time, how they react to divergenteconomic interests to avoid potential appropriation haz-ards, and how they manage ongoing conflicts due to man-

agerial and organizational dissimilarities. Another prom-ising area for future studies is to dissect and makecomparisons of determinants of alliance failure acrossvarious stages of alliance life cycle.

Harrigan (1988) admonishes firms to understand thestrategic impact of alliances, whether they join an allianceor not, because they will undoubtedly face competitorsthat cooperate with each other. By delineating the struc-ture of alliances, as well as the dynamics that influencetheir development, the proposed framework illuminatesthe factors that can improve management of strategic al-liances. It is hoped that this may be used as one guide inredirecting future studies of alliances. We also would liketo think that this might redress the imbalance created bythe prevailing attention to alliance success, as opposed toalliance failure.

AcknowledgmentsThe authors are grateful to Arvind Parkhe, Michael Gordon, Mike Russo,Jay Barney, Michael Lawless, Claudia Schoonhoven, and three anonymousreviewers for their helpful comments on earlier versions of this paper.

ReferencesAkerlof, G. A. 1970. The market for ‘‘lemons’’: Quality uncertainty

and the market mechanism.Quart. J. Econom.84 488–500.Anderson, E. 1990. Two firms, one frontier: On assessing joint venture

performance.Sloan Management Rev.(Winter) 19–30.Barkema, H., J. Bell, J. Pennings. 1996. Foreign entry, cultural barriers,

and learning.Strategic Management J.17 151–166.Barnard, C. I. 1968.The Functions of the Executive.Harvard Univ.

Press, Cambridge, MA.Beamish, P. W. 1987. Joint ventures in LDCs: Partner selection and

performance.Management Internat. Rev.27 23–37.Berg, S. V., P. Friedman. 1980. Causes and effects of joint venture

activity: Knowledge acquisition vs. parent horizontality.The Anti-trust Bull. (Spring)25 143–168.

Bleeke, J., D. Ernst. 1993.Collaborating to Compete. John Wiley andSons, New York.

Brown, L., A. Rugman, A. Verbeke. 1988. Japanese joint ventures withwestern multinationals: Synthesizing the economic and culturalexplanations of failure.Asia Pacific J. Management6 225–242.

Buckley, P., M. Casson. 1988. A theory of cooperation in internationalbusiness. InCooperative Strategies in International Business. F.Contractor and P. Lorange, eds., Lexington Books, Lexington,MA.

Business Week. 1986. Odd couples. (July 21) 100–106.Cascio, W. F., M. G. Serapio. 1991. Human resources systems in an

international alliance: The undoing of a done deal?Organ. Dy-nam.(Winter) 19 63–74.

Cummings, T. 1984. Transorganizational development.Res. Organ.Behavior6 367–422.

Defillipi, R., R. Reed. 1991. Three perspectives on appropriation haz-ards in cooperative agreements. Working paper, Suffolk Univer-sity.

Doz, Y. L. 1988. Technology partnerships between larger and smallerfirms: Some critical issues. InCooperative Strategies in Inter-

Page 17: Interfirm Rivalry and Managerial Complexity

SEUNG HO PARK AND GERARDO R. UNGSON Interfirm Rivalry and Managerial Complexity

ORGANIZATION SCIENCE/Vol. 12, No. 1, January–February 2001 53

national Business.F. Contractor, P. Lorange, eds., LexingtonBooks, Lexington, MA.

——. 1996. The evolution of cooperation in strategic alliances: Initialconditions or learning processes?Strategic Management J.17 55–83.

Dymsza, W. A. 1988. Successes and failures of joint ventures in de-veloping countries: Lessons from experience.Cooperative Strat-egies in International Business.F. Contractor, P. Lorange, eds.Lexington Books, Lexington, MA.

Eisenhardt, K. M., C. B. Schoonhoven. 1996. Resource-based view ofstrategic alliance formation: Strategic and social effects in entre-preneurial firms.Organ. Sci.7 136–150.

Franko, L. G. 1971.Joint Venture Survival in Multinational Corpo-rations. Praeger Publishers, New York.

Geringer, J. M., L. Hebert. 1989. Control and performance of inter-national joint ventures.J. Internat. Bus. Stud.18 235–254.

Gomes-Casserres, B. 1996.The Alliance Capitalism: The New Shapeof Business Rivalry.Harvard Univ. Press, Cambridge, MA.

Good, L. 1972. United States Joint Ventures and Manufacturing Firmsin Monterrey, Mexico: Comparative styles of management. Un-published doctoral dissertation, Cornell University, Ithaca, NY.

Gulati, R. 1999. Network location and learning: The influence of net-work resources and firm capabilities on alliance formation.Stra-tegic Management J.20 397–420.

——, K. Tarun, N. Nohria. 1994. Unilateral commitments and the im-portance of process in alliances.Sloan Management Rev.35 61–69.

Hamel, G. 1991. Competition for competence and interpartner learningwithin international strategic alliances.Strategic Management J.(Summer)12 83–103.

——, Y. L. Doz, C. K. Prahalad. 1989. Collaborate with your com-petitors—and win.Harvard Bus. Rev.(January–February) 133–139.

Hannan, M. T., J. Freeman. 1977. The population ecology of organi-zations.Amer. J. Soc.83 929–964.

Harrigan, K. R. 1988. Strategic alliances and partner asymmetries. InCooperative Strategies in International Business. F. Contractor,P. Lorange, eds. Lexington Books, Lexington, MA.

Holton, R. E. 1981. Making international JVs work.Management ofHeadquarters-Subsidiary Relationships in Multinational Corpo-rations. L. Otterbeck, ed. St. Martins Press, New York.

Killing, J. P. 1983.Strategies for Joint Venture Success. Praeger, NewYork.

Kogut, B. 1988. Joint ventures: Theoretical and empirical perspectives.Strategic Management J.9 319–332.

——. 1991. Joint ventures and the option to expand and acquire.Man-agement Sci.37 19–33.

——, H. Singh. 1988. The effect of national culture on the choice ofentry mode.J. Internat. Bus. Stud.(Fall) 15 411–432.

Koot, M. 1988. Underlying dilemmas in the management of interna-tional joint ventures. InCooperative Strategies in InternationalBusiness. F. Contractor, P. Lorange, eds. Lexington Books, Lex-ington, MA.

Koza, M. P., A. Y. Lewin. 1998. The co-evolution of strategic alliances.Organ. Sci.9 255–263.

Lane, H. W., P. W. Beamish. 1990. Cross-cultural cooperative behaviorin joint ventures in LDCs.Management Internat. Rev.30 87–102.

Larson, A. 1992. Network dyads in entrepreneurial settings: A study

of the governance of exchange relationships.Admin. Sci. Quart.37 76–104.

Mohr, J., R. Spekman. 1994. Characteristics of partnership success:Partnership attributes, communication behavior, and conflict res-olution. Strategic Management J.15 135–152.

Park, S. H., M. Russo. 1996. When competition eclipses cooperation:An event history analysis of alliance failure.Management Sci.42875–890.

——, G. Ungson. 1997. The effect of partner nationality, organiza-tional dissimilarity, and economic motivation on the dissolutionof joint ventures.Acad. Management J.39 279–307.

Parkhe, A. 1993a. The structuring of strategic alliances: A game-theoretic and transaction-cost examination of interfirm coopera-tion. Acad. Management J.36 794–829.

——. 1993b. ‘‘Messy’’ research, methodological predispositions, andtheory development in international joint ventures.Acad. Man-agement Rev.18 227–268.

Porter, M. 1987. From competitive advantage to corporate strategy.Harvard Bus. Rev.65 43–59.

——, M. Fuller. 1985. Coalitions and global strategy. InCompetitionin Global Industries. M. Porter, ed. Harvard Business SchoolPress, Cambridge, MA.

Rafii, F. 1978. Joint ventures and transfer of technology to Iran: Theimpact of foreign control. Unpublished doctoral dissertation, Har-vard University. Cambridge, MA.

Reich, R. B., E. D. Mankin. 1986. Joint ventures with Japan give awayour future.Harvard Bus. Rev.64 78–139.

Ring, P. S., A. H. Van De Ven. 1994. Developmental processes ofcooperative interorganizational relationships.Acad. ManagementRev.19 90–118.

Seabright, M. A., D. A. Levinthal, M. Fichman. 1992. Role of individ-ual attachments in the interorganizational relationships.Acad.Management J.35 122–160.

Shenkar, O., Y. Zeira. 1992. Role conflict and role ambiguity of chiefexecutive officers in international joint ventures.J. Internat. Bus.Stud.23 55–75.

Simon, H. 1947.Administrative Behavior. MacMillan, London, U.K.Smith, K. G., S. J. Carroll, S. J. Ashford. 1995. Intra- and interorga-

nizational cooperation: Toward a research agenda.Acad. Man-agement J.38 (1) 7–23.

Tomlinson, J. W. C. 1970.The Joint Venture Process in InternationalBusiness: India and Pakistan. MIT Press, Cambridge, MA.

Wall Street Journal.1986. Jet engine pact of Rolls-Royce, GE isscrapped. (November 20).

Williamson, O. E. 1991. Comparative economic organization: Theanalysis of discrete structural alternatives.Admin. Sci. Quart.36269–296.

Yan, A., B. Gray. 1994. Bargaining power, management control, andperformance in United States–China joint ventures.Acad. Man-agement J.37 1478–1517.

——, M. Zeng. 1999. International jv instability: A critique of previousresearch, a reconceptualization, and directions for future research.J. Internat. Bus. Stud.30 397–414.

Zajac, E., G. Olsen. 1993. From transaction cost to transactional valueanalysis: Implications for the study of interorganizational strate-gies.J. Management Stud.30 131–145.

Zucker, L. G. 1986. Production of trust.Res. Organ. Behavior8 53–111.

Accepted by Claudia Bird Schoonhoven; received May 15, 2000.