THE RELATIONAL VIEW: COOPERATIVE STRATEGY AND …

® Academy o/ Management fleview1998, Vol. 23, No. 4, 660-679.

THE RELATIONAL VIEW: COOPERATIVESTRATEGY AND SOURCES OF

INTERORGANIZATIONAL COMPETITIVEADVANTAGE

JEFFREY H. DYERHARBIR SINGH

University of Pennsylvania

In this article we oifer a view that suggests that a firm's critical resources may spanfirm boundaries and may be embedded in inferiirm resources and routines. We arguethat an increasingly important unit of analysis for understanding competitive advan-tage is the relationship between firms and identify four potential sources of interor-ganizational competitive advantage: (1) relation-specific assets, (2) knowledge-sharing routines, (3) complementary resources/capabilities, and (4) effectivegovernance. We examine each of these potential sources of rent in detail, identifyingkey subprocesses, and also discuss the isolating mechanisms that serve to preserverelational rents. Finally, we discuss how the relational view may offer normativeprescriptions for firm-level strategies that contradict the prescriptions offered by thosewith a resource-based view or industry structure view.

Scholars in the strategy field are concernedfundamentally with explaining differential firmperformance (Rumelt, Schendel, & Teece, 1991).As strategy scholars have searched for sourcesof competitive advantage, two prominent viewshave emerged regarding the sources of super-normal returns. The first—the industry stTuctureview—associated with Porter (1980), suggeststhat supernormal returns are primarily a func-tion of a firm's membership in an industry withfavorable structural characteristics (e.g., rela-tive bargaining power, barriers to entry, and soon). Consequently, many researchers have fo-cused on the industry as the relevant unit ofanalysis. The second view—the resource-basedview (RBV) of the firm—argues that differentialiirm performance is fundamentally due to firmheterogeneity rather than industry structure(Barney, 1991; Rumelt, 1984, 1991; Wernerfelt,1984). Firms that are able to accumulate re-sources and capabilities that are rare, valuable,nonsubstitutable, and difficult to imitate willachieve a competitive advantage over compet-

We presented an earlier version of this article at theannual meeting of the Academy of Management in Cincin-nati, August 12, 1996, We thank Gautum Ahuja, Jay Barney,Ben Bensaou, Connie Helfat, Joanne Oxley, Lori Rosenkopf,Brian Silverman, Gabriel Szulanski, and John Lafkas fortheir valuable comments on earlier drafts.

ing firms (Barney, 1991; Dierickx & Cool, 1989;Rumelt, 1984). Thus, extant RBV theory views thefirm as the primary unit of analysis.'

Although these two perspectives have contrib-uted greatly to our understanding of how firmsachieve above-normal returns, they overlook theimportant fact that the (dis)advantages of anindividual firm are often linked to the (dis)ad-vantages of the network of relationships inwhich the firm is embedded. Proponents of theRBV have emphasized that competitive advan-tage results from those resources and capabili-ties that are owned and controlled by a singlefirm. Consequently, the search for competitiveadvantage has focused on those resources thatare housed within the Him. Competing firmspurchase standardized (nonunique) inputs thatcannot be sources of advantage, because theseinputs (factors) are either readily available to allcompeting firms or the cost of acquiring them isapproximately equal to the economic value theycreate (Barney, 1986). However, a firm's criticalresources may extend beyond firm boundaries.For example, the typical manufacturing firm inthe United States purchases 55 percent of thevalue of each product it produces (this figure is69 percent in Japan), and many of these inputs

' The dynamic capabilities approach {Teece, Pisano, &Shuen, 1997) also views the firm as the unit of analysis.

660

1998 Dyer and Singh 661

are highly customized by suppliers (Ministry ofInternational Trade and Industry, 1987). More-over, this percentage has been increasing dur-ing the past two decades (Bresnen & Fowler,1994; Nishiguchi, 1994). Recent studies suggestthat productivity gains in the value chain arepossible when trading partners are willing tomake relation-specific investments and com-bine resources in unique ways (Asanuma, 1989;Dyer, 1996a). This indicates that firms who com-bine resources in unique ways may realize anadvantage over competing firms who are un-able or unwilling to do so. Thus, idiosyncraticinterfirm linkages may be a source of relationalrents^ and competitive advantage.

This analysis suggests that a firm's criticalresources may span firm boundaries and maybe embedded in interfirm routines and process-es.^ Indeed, the "explosion in alliances" duringthe past decade suggests that a pair or networkof firms is an increasingly important unit ofanalysis and, therefore, deserves more study(Anderson, 1990; Gomes-Casseres, 1994; Smith,Carroll, & Ashford, 1995). Although there hasrecently been increased attention on interorgan-izational relationships in the strategic manage-ment literature, to date, no attempt has beenmade to integrate what we have learned andsystematically examine the interorganizationalrent-generating process. In instances where re-searchers have explicitly studied how firms col-laborate to generate economic rents, they havetended to focus on one particular benefit asso-ciated with collaboration, such as learning,lower transaction costs, or pooling of resources(DQre, 1983; Dyer, 1996a; Hamel, 1991; Larson,1992; Powell, Koput, & Smith-Doerr, 1996; Teece,1987).

^We use the term reJa(ionaJ rent, although, technicallyspeaking, trading partners generate quasi-rents, Peteraf de-fines quasi-rents as "returns that exceed a factor's short runopportunity cost ,,, [and] are an excess over the returns to afactor in its next best use" (1994: 155), The term quasi-ientssuggests that the rents are not permanent in nature,

^ The fact that a firm's valuable resources may extendbeyond a firm's boundaries is increasingly recognized, evenwithin the investment community. For example, Powellfound that industry investment analysts explicitly evaluateand assess the quality of a biotechnology firm's relation-ships with outside partners (1996: 206), Firms with more—and higher quality—partnerships receive higher market val-uations from the analysts who recognize that abiotechnology firm's critical resources extend beyond firmboundaries.

Our primary purpose in this article is to ex-amine how relational rents are earned and pre-served. We offer a relational view of competitiveadvantage that focuses on dyad/network rou-tines and processes as an important unit ofanalysis for understanding competitive advan-tage.^ This framework is valuable because itprovides a theoretical basis for cumulative ad-ditions to our understanding of the sources ofinferorganizational competitive advantage (Ol-iver, 1990). In the following sections we identifyand delineate the various sources of rents at theinterfirm unit of analysis. We also examine themechanisms that preserve the relational rentsthat dyads and networks jointly create. Finally,we discuss how the relational view may offernormative prescriptions for firm-level strategiesthat contradict the prescriptions offered by theRBV and industry structure view.

SOURCES OF RELATIONAL RENTS

Theoretical Discussion

By examining the relevant characteristics ofarm's-length market relationships, we find cluesthat guide our search for relational advantages.Arm's-length market relationships are charac-terized by

1, nonspecific asset investments,2, minimal information exchange (i.e., prices

act as coordinating devices by signaling allrelevant information to buyers and sellers),

3. separable technological and functional sys-tems within each firm that are character-ized by low levels of interdependence (i.e.,the two organizations have only a sales-to-purchasing interface and do not jointly cre-ate new products through multifunctionalinterfaces), and

4. low transaction costs and minimal invest-ment in governance mechanisms (William-son, 1985).

Under these conditions it is easy for firms toswitch trading partners with little penalty be-cause other sellers offer virtually identical prod-ucts. As Ghoshal notes, "Efficiency in the execu-tion of roufine tasks is the strength of markets"(1995: 16). Thus, arm's-length market relation-

•* For the convenience of exposition, we use two firms,rather than multiple firms, as the unit of analysis.

662 Academy of Management Review October

ships are incapable of generating relationalrents because theie is nothing idiosynciaticabout the exchange ielationship that enablesthe two paities to geneiate piofits above andbeyond what other sellei-buyei combinationscan geneiate. The relationships are not rare ordifficult to imitate. Buyers can only achieve adifferential advantage if they bring greater bar-gaining power to the table.

This analysis suggests that alliances gener-ate competitive advantages only as they movethe relationship away from the attributes ofmarket relationships. In other words, the com-petitive advantages of partnerships, as docu-mented in studies to date, seem to fall into fourcategories:

1. investments in relation-specific assets;2. substantial knowledge exchange, including

the exchange of knowledge that results injoint learning;

3. the combining of complementary, butscarce, resources or capabilities (typicallythrough multiple functional interfaces),which results in the joint creation of uniquenew products, services, or technologies; and

4. lower transaction costs than competitor al-liances, owing to more effective governancemechanisms.

We define a relational rent as a supernormalprofit jointly generated in an exchange relation-ship that cannot be generated by either firm inisolation and can only be created through thejoint idiosyncratic contributions of the specificalliance partners.

In summary, at a fundamental level, rela-tional rents are possible when alliance partnerscombine, exchange, or invest in idiosyncraticassets, knowledge, and resources/capabilities,and/or they employ effective governance mech-anisms that lower transaction costs or permitthe realization of rents through the synergisticcombination of assets, knowledge, or capabili-ties. In the sections that follow we examine indetail these four key sources of relational rents;in each section we develop a major propositionand a set of subpropositions, as summarized inFigure 1. After examining assets, knowledge,and resources, we examine governance, be-cause although governance may generate rela-tional rents by simply lowering transactioncosts, governance issues cut across each of theother sources of rents (e.g., influence what rela-tion-specific investments will be made, what

knowledge will be shared, and so on). We be-lieve it is easier to understand how governanceinfluences the ability to generate rents throughassets, knowledge, and capabilities if we havefirst examined these constructs.

Interfirm Relation-Specific Assets

Amit and Schoemaker argue that specializa-tion of assets is "a necessary condition for rent"and "strategic assets by their very nature arespecialized" (1993: 39). Thus, by definition, firmsmust do something specialized or unique to de-velop a competitive advantage. A firm maychoose to seek advantages by creating assetsthat are specialized in conjunction with the as-sets of an alliance paitner (Klein, Crawford, &Alchian, 1978; Teece, 1987). Productivity gains inthe value chain are possible when firms arewilling to make relation/transaction-specific in-vestments (Perry, 1989; Williamson, 1985).

Williamson (1985) identifies three types of as-set specificity: (1) site specificity, (2) physicalasset specificity, and (3) human asset specific-ity. Site specificity refers to the situationwhereby successive production stages that areimmobile in nature are located close to one an-other. Previous studies suggest that site-specificinvestments can substantially reduce inventoryand transportation costs and can lower the costsof coordinating activities (Dyer, 1996a). Physicalasset specificity refers to transaction-specificcapital investments (e.g., in customized machin-ery, tools, dies, and so on) that tailor processesto particular exchange partners. Physical assetspecialization has been found to allow for prod-uct differentiation and may improve quality byincreasing product integrity or fit (Clark & Fuji-moto, 1991; Nishiguchi, 1994). Human asset spec-ificity refers to transaction-specific know-howaccumulated by transactors through long-standing relationships (e.g., dedicated supplierengineers who learn the systems, procedures,and the individuals idiosyncratic to the buyer).Human cospecialization increases as alliancepartners develop experience working togetherand accumulate specialized information, lan-guage, and know-how. This allows them to com-municate efficiently and effectively, which re-duces communication errors, thereby enhancingquality and increasing speed to market(Asanuma, 1989; Dyer, 1996a).


FIGURE 1Determinants of Interorganizational Competitive Advantage

Determinants ofrelational rents

Subprocesses facilitatingrelational rents

1. Relation-specific assets

2. Knowledge-sharing routines

3. Complementaryresources and capabilities

4. Effective governance

la. Duration of safeguards

lb. Volume of interfirm transactions

2a. Partner-specific absorptive capacity

2b. Incentives to encourage transparencyand discourage free riding

3a. Ability to identify and evaluate potentialcomplementarities

3b. Role of organizational complementaritiesto access benefits of strategic resourcecomplementarity

4a. Ability to employ self-enforcementrather than third-party enforcementgovernance mechanisms

4b. Ability to employ informal versus formalself-enforcement governance mechanisms

Asanuma (1989) was among the first to doc-ument how the relation-specific skills devel-oped between Japanese suppliers and theirautomakers generated surplus profits andcompetitive advantages for collaboratingfirms. Similarly, Dyer (1996a) found a positiverelationship between relation-specific invest-ments and performance in a sample of auto-makers and their suppliers. Additionally, Sax-enian (1994) found that Hewlett Packard andother Silicon Valley firms greatly improvedperformance by developing long-term partner-ships with physically proximate suppliers.She claims that proximity in high-technology

industries "greatly facilitates the collabora-tion required for fast-changing and complextechnologies" (1990: 101). Indeed, severalscholars have shown that physical proximitycreated through site-specific investments fa-cilitates interfirm cooperation and coordina-tion, thereby enhancing performance (Dyer,1996a; Enright, 1995; Nishiguchi, 1994). Finally,Parkhe (1993) found that the commitment of"nonrecoverable investments" in a sample ofstrategic alliances was positively related toperformance. These studies indicate that rela-tional rents generated through relation-specific investments are realized through


lower total value chain costs, greater productdifferentiation, fewer defects, and faster prod-uct development cycles.

Proposition 1: The greater the alliancepartners' investment is in relation-specific assets, the greater the poten-tial will be for relational rents.

Regarding relation-specific assets, there aretwo key subprocesses that influence the abilityof partners to generate relational rents. First, thelength (i.e., in years) of the governance arrange-ment designed to safeguard against opportun-ism influences the ability of alliance partners toinvest in relation-specific assets. Since relation-specific investments create appropriable quasi-rents, transactors need to safeguard those in-vestments (Klein et al., 1978). Partners are morelikely to make investments in relation-specificassets when they have crafted effective safe-guards (Williamson, 1985). Moreover, there istypically a fixed, up-front cost associated withmaking a particular type of relation-specific in-vestment (such as in specialized equipment or adedicated plant). Some relation-specific invest-ments (e.g., a dedicated plant) are more durableand costly than others (e.g., a specialized tool or

jig)-Given the fixed-cost nature of some invest-

ments, alliance partners need to assess whetheror not they will make the necessary return on theinvestment during the payback period or lengthof the governance agreement (e.g., length of con-tract). For example. Dyer (1997) found that Japa-nese suppliers were more likely to make dura-ble and costly relation-specific investmentsbecause automakers provided safeguards onthose investments for at least 8 years or more. Incontrast, U.S. automakers offered average con-tracts of 2.3 years, and suppliers rationally re-fused to make relation-specific investmentswith a long payback period.

Proposition la: The greater the lengthof the safeguard is to protect againstopportunism, the greater the potentialwill be to generate relational rentsthrough relation-specific assets.

Second, the ability to substitute special-purposeassets for general-purpose assets is influencedby the total volume (scale) and breadth (scope)of transactions between the alliance partners.Just as firms that achieve production economies

of scale are able to increase productivity bysubstituting special-purpose assets for general-purpose assets, alliance partners are also ableto increase the efficiency associated with inter-firm exchanges as they increase the volume andscope of transactions between the alliance part-ners. A similar argument has been made byWilliamson (1985), who claims that transactorsengaging in frequent, recurring transactionscan afford to adopt more specialized and com-plex governance structures.

Proposition lb: The greater the volumeof exchange is between the alliancepartners, the greater the potential willbe to generate relational rents throughrelation-specific assets.

In summary, the length of the safeguard and thevolume of transactions are key subprocessesthat influence the ability of alliance partners togenerate rents through relation-specific assets.

Interfirm Knowledge-Sharing Routines

Various scholars have argued that interorgan-izational learning is critical to competitive suc-cess, noting that organizations often learn bycollaborating with other organizations (Levin-son & Asahi, 1996; March & Simon, 1958; Powellet al., 1996). For example. Von Hippel (1988)found that in some industries (e.g., scientific in-struments) more than two-thirds of the innova-tions he studied could be traced back to a cus-tomer's initial suggestions or ideas. In otherindustries (e.g., wire termination equipment) themajority of innovations could be traced back tosuppliers. Von Hippel argues that a productionnetwork with superior knowledge-transfermechanisms among users, suppliers, and man-ufacturers will be able to "out innovate" produc-tion networks with less effective knowledge-sharing routines. Similarly, Powell et al. (1996)found that the locus of innovation in the biotech-nology industry was the network—not the indi-vidual firm. Patents were typically filed by alarge number of individuals working for a num-ber of different organizations, including biotechfirms, pharmaceutical companies, and universi-ties. Powell et al. (1996) argue that biotech firmswho are unable to create (or position themselvesin) learning networks are at a competitive dis-advantage.

1998 Dyei and Singh 665

These studies suggest that a firm's alliancepartners are, in many cases, the most importantsource of new ideas and information that resultin performance-enhancing technology and inno-vations. Thus, alliance partners can generaterents by developing superior interfirm knowl-edge-sharing routines. We define an interfirmknowledge-sharing routine as a regular patternof interfirm interactions that permits the trans-fer, recombination, or creation of specializedknowledge (Grant, 1996). These are institution-alized interfirm processes that are purposefullydesigned to facilitate knowledge exchanges be-tween alliance partners.

Proposition 2: The greater the alli-ance partners' investment is in inter-tirm knowledge-sharing routines,the greater the potential will be forrelational rents.

Beyond simply arguing that alliance partnerscan generate relational rents through knowl-edge-sharing routines, it is important to under-stand how partners create knowledge-sharingroutines that result in competitive advantage.Many scholars divide knowledge into two types:(1) information and (2) know-how (Grant, 1996;Kogut & Zander, 1992; Ryle, 1984). We can defineinformation as easily codifiable knowledge thatcan be transmitted "without loss of integrityonce the syntactical rules required for decipher-ing it are known. Information includes facts, ax-iomatic propositions, and symbols" (Kogut &Zander, 1992: 386). By comparison, know-how in-volves knowledge that is tacit, "sticky," com-plex, and difficult to codify (Kogut & Zander,1992; Nelson & Winter, 1982; Szulanski, 1996).Since know-how is tacit, sticky, and difficult tocodify, it is difficult to imitate and transfer. How-ever, these properties also suggest that, com-pared to information, know-how is more likely toresult in advantages that are sustainable. As aresult, alliance partners that are particularly ef-fective at transferring know-how are likely tooutperform competitors who are not.

The ability to exploit outside sources ofknowledge is largely a function of prior relatedknowledge or the "absorptive capacity" of therecipient of knowledge. Cohen and Levinthaldefine absorptive capacity as "the ability of afirm to recognize the value of new, external in-formation, assimilate it, and apply it to commer-cial ends" (1990: 128). However, their definition

suggests that if a firm has absorptive capacity,it is equally capable of learning from all otherorganizations. Although Cohen and Levinthalfocus on the absolute absorptive capacity of in-dividual firms, the concept is particularly usefulin thinking about how alliance partners maysystematically engage in interorganizationallearning. Thus, partner-specific absorptive ca-pacity refers to the idea that a firm has devel-oped the ability to recognize and assimilatevaluable knowledge from a particular alliancepartner. This capacity would entail implement-ing a set of interorganizational processes thatallows collaborating firms to systematicallyidentify valuable know-how and then transfer itacross organizational boundaries. Partner-specific absorptive capacity is a function of(1) the extent to which partners have developedoverlapping knowledge bases and (2) the extentto which partners have developed interactionroutines that maximize the frequency and inten-sity of sociotechnical interactions. Previouswork suggests that the ability of a receiver ofknowledge to "unpackage" and assimilate it islargely a function of whether or not the firm hasoverlapping knowledge bases with the source(Mowery, Oxley, & Silverman, 1996; Szulanski,1996). Thus, this is a critical component of part-ner-specific absorptive capacity.

In addition, partner-specific absorptive capac-ity is enhanced as individuals within the alli-ance partners get to know each other wellenough to know who knows what and wherecritical expertise resides within each firm. Inmany cases this knowledge develops informallyover time through interfirm interactions. How-ever, it may be possible to codify at least someof this knowledge. For example, Fuji and Xeroxhave attempted to codify this knowledge by cre-ating a "communications matrix," which identi-fies a set of relevant issues (e.g., products, tech-nologies, markets, and so on) and then identifiesthe individuals (by function) within Fuji-Xerox,Fuji, and Xerox who have relevant expertise onthat particular issue. This matrix provides valu-able information regarding where relevant ex-pertise resides within the partnering firms.

This example illustrates that alliance part-ners can increase partner-specific absorptivecapacity by designing interfirm routines that fa-cilitate information-sharing and increase socio-technical interactions. These types of routinesare particularly important since know-how

666 Academy ol Management Review October

transfers typically involve an iterative processof exchange, and the success of such transfersdepends on whether personnel from the twofirms have direct, intimate, and extensive face-to-face interactions (Arrow, 1974; Badaraco, 1991;Daft & Lengel, 1986; Marsden, 1990).

Proposition 2a: The greater the part-ner-specific absorptive capacity is, thegreater the potential will be to gener-ate relational rents through knowl-edge sharing.

Finally, the ability of alliance partners to gen-erate rents through knowledge sharing is de-pendent on an alignment of incentives that en-courages the partners to be transparent, totransfer knowledge, and not to free ride on theknowledge acquired from the partner. In partic-ular, the transferring firm must have an incen-tive to devote the resources required to transferthe know-how since it typically incurs signifi-cant costs during the transfer—costs compara-ble to those incurred by the receiving firm (Szu-lanski, 1996). Thus, the mechanisms employed togovern the alliance relationship must create ap-propriate incentives for knowledge sharing.These may be formal financial incentives (e.g.,equity arrangements) or informal norms of reci-procity. In various studies scholars have foundthat equity arrangements are particularly effec-tive at aligning partner incentives and, there-fore, promote greater interfirm knowledge trans-fers than contractual arrangements (Kogut, 1988;Mowery et al., 1996).

Proposition 2b: The greater the align-ment of incentives by alliance part-ners is to encourage transparencyand reciprocity and to discouragefree riding, the greater the potentialwill be to generate relational rentsthrough knowledge sharing.

A comparison of Toyota's and GM's produc-tion networks illustrates how knowledge-sharing routines can create interorganizationalcompetitive advantage. Toyota has developed anumber of practices that facilitate knowledgetransfers to—and among—suppliers. For exam-ple, Toyota may transfer knowledge directly tosuppliers, through its "operations managementconsulting division" consultants, who will re-side at the supplier for days, weeks, or evenmonths to see' that the transfer takes place

(Nishiguchi, 1994; Womack, Jones, & Roos, 1990).Toyota also transfers its personnel to the sup-plier (on a temporary or permanent basis) toincrease the supplier's ability to assimilate andapply the new knowledge. These transfers resultin dense interfirm social networks that increasepartner-specific absorptive capacity. Conse-quently, Toyota personnel know what knowl-edge will be useful to the supplier, whom tocontact at the supplier, and where the absorp-tive capacity resides at the supplier.

In contrast, GM and its suppliers have a his-tory of keeping innovations proprietary. Thisstrategy is viewed, according to the RBV, as thebest way for an individual firm to generate rentsfrom a particular innovation. Of course, the de-cision not to share knowledge is the only ration-al one for suppliers, since GM has not cultivateda stable network of supplier companies thathave developed overlapping knowledge bases,dense social interactions, or a norm of reciproc-ity for knowledge sharing. GM does not have asupplier association to facilitate knowledgesharing, nor does GM transfer or lend personnelto suppliers to facilitate interfirm knowledgesharing. Consequently, suppliers rationallyrefuse to engage in costly knowledge-sharingactivities since they do not expect to receivesome benefit (i.e., knowledge) in return. It is notsurprising then that there is significantlygreater knowledge sharing between Toyota andits suppliers than between GM and its suppliers(Dyer, 1997).

Complementary Resource Endowments

Another way firms can generate relationalrents is by leveraging the complementary re-source endowments of an alliance partner. Insome instances a firm's ability to generate rentsfrom its resources may require that these re-sources be utilized in conjunction with the com-plementary resources of another firm. Comple-mentary resource endowments have been thefocus of much prior discussion on the formationand management of alliances and have beendiscussed widely as a key factor driving returnsfrom alliances (Hamel, 1991; Harrigan, 1985; Hill& Hellriegel, 1994; Shan, Walker, & Kogut, 1994;Teece, 1987). We define complementary resourceendowments as distinctive resources of alliancepartners that collectively generate greater rentsthan the sum of those obtained from the individ-


ual endowments of each paitnei. For these re-sources to generate rents through an alliance, itis necessarily the case that neither firm in thepartnership can purchase the relevant resourcesin a secondary market. Also, these resourcesmust be indivisible, thereby creating an incen-tive for each firm to form an alliance in order toaccess the complementary resources. As Oliverobserves, "Strategic alliances allow firms to pro-cure assets, competencies, or capabilities notreadily available in competitive factor markets,particularly specialized expertise and intangi-ble assets, such as reputation" (1997: 707).

The cooperative relationship between Nestleand Coca-Cola to distribute hot canned drinksthrough vending machines (a business largelyunknown outside of Japan) is an example of analliance in which complementary resource en-dowments are a source of relational rents. Thisalliance combines Nestle's brand names (Nes-cafe and Nestea) and competence in developingand producing soluble coffee and tea productswith Coca-Cola's powerful international distri-bution and vending machine network (Hamel &Prahalad, 1994:187). The alliance creates advan-tages over Japanese competitors (e.g., Suntory),who are better than Coca-Cola at soluble coffeeand tea and have a larger distribution andvending machine network than Nestle, but can-not match the Coca-Cola-Nestle combination ofcapabilities.

Shan and Hamilton (1991) offer another illus-tration. They found that complementarity of bothfirm- and country-specific resources betweendomestic and foreign firms was a key factor inthe formation of cross-border strategic alliancesin biotechnology. The complementarity in thecases they studied consisted of linkages be-tween the strong basic research capabilities ofU.S. firms with the unique local knowledge anddistribution capabilities of their partners inoverseas markets.

In the cases described above, the alliancepartners brought distinctive resources to the al-liance, which, when combined with the re-sources of the partner, resulted in a synergisticeffect whereby fhe combined lesouice endow-ments weie moie valuable, iare. and difficult toimitate than they had been before they werecombined. Consequently, these alliances pro-duced stronger competitive positions than thoseachievable by the firms operating individually.It is important to note, however, that not all of

the resources of a potential alliance partner willbe complementary. In assessing the extent towhich alliance partners can generate relationalrents by combining complementary resources, itis worthwhile to think about the proportion ofthe potential partner's strategic resources that issynergy sensitive with the firm's resources. Asthe proportion of synergy-sensitive resources inthe potential partners increases, so does the po-tential for earning relational rents by combiningthe complementary resources.

Pioposition 3: The gieatei the propor-tion is of syneigy-sensitive resourcesowned by alliance paitneis that,when combined, increase fhe degieeto which the resources are valuable,rare, and difficult to imitate, thegieatei the potential will be to genei-ate lelational lents.

There are several challenges faced by firmsattempting to generate relational rents withcomplementary resources. In particular, theymust find each other and recognize the potentialvalue of combining resources. If potential alli-ance partners possessed perfect information,they could easily calculate the value of differentpartner combinations and then rationally allywith the partner(s) who would generate thegreatest combined value. However, it is oftenvery costly and difficult (if not impossible) toplace a value on the complementary resourcesof potential partners. In fact, firms vary in theirability to identify potential partners and valuetheir complementary resources for three primaryreasons: (1) differences in prior alliance experi-ence, (2) differences in internal search and eval-uation capability, and (3) differences in theirability to acquire information about potentialpartners owing to different positions in their so-cial/economic network(s).

First, firms with higher levels of experience inalliance management may have a more preciseview on the kinds of partner/resource combina-tions that allow them to generate supernormalreturns. Previous research suggests that prioralliance experience results in more opportuni-ties to enter into future alliances, presumablybecause of the development of alliance capabil-ities and reputation (Gulati, 1995a; Mitchell &Singh, 1996; Walker, Kogut, & Shan, 1997).

Second, many organizations are developingways to accumulate knowledge on screening


potential partners by creating a "strategic alli-ance" function. For example, firms such asHewlett Packard, Xerox, and Microsoft have ap-pointed a Director of Strategic Alliances, withhis or her own staff and resources. The role ofthese individuals is to identify and evaluate po-tential alliance partners as well as to monitorand coordinate their firm's current alliances.The creation of these roles ensures some ac-countability for the selection and ongoing man-agement of alliance partners and also ensuresthat knowledge on successful partner combina-tions and on effective alliance managementpractices will be accumulated. An opportunityexists to codify some of this knowledge, as illus-trated by the fact that some firms, such asHewlett Packard, have created manuals that at-tempt to codify alliance-specific knowledge(Hewlett Packard's manual has more than 300pages). Research on acquisitions suggests thatcodification of knowledge is predictive of suc-cess in post-acquisition contexts (Singh & ZoUo,1997). Although alliances are a different contextthan acquisitions, a parallel argument can beapplied to the management of alliances.

Third, the ability of a firm to identify and eval-uate partners with complementary resources de-pends on the extent to which the firm has accessto accurate and timely information on potentialpartners. An investment in an internal alliancefunction likely will facilitate the acquisition ofthis information, but it also depends on the ex-tent to which the firm occupies an information-rich position within social/economic networks.Previous research suggests that firms occupyingcentral network positions with greater networkties have superior access to information and,thus, are more likely to increase the number oftheir alliances in the future (Gulati, 1995a;Mitchell & Singh, 1996; Walker et al., 1997). Whena firm is well positioned in networks, the firmhas' access to more reliable information aboutpotential partners because of trusted informantswithin the network who may have direct expe-rience with the potential partner (Burt, 1992;Chung, Singh, & Lee, in press; Granovetter, 1985;Nohria, 1992). An information-rich positionwithin a network, therefore, provides a firm withadditional information about the nature and de-gree of accessibility of the complementary re-sources of potential partners.

Proposition 3a: The ability of fiims togeneiate relational lents by combin-ing complementaiy resources in-cieases with the film's (1) prior aJii-ance experience, (2) investment ininternal search and evaluation capa-bility, and (3) ability to occupy an in-foimation-iich position in its social!economic netwoiks.

Thus far, our discussion has focused on thebenefits associated with combining resourceswith strategic complementarity. However, oncea firm has identified a potential partner with therequisite complementary strategic resources,another challenge is developing organizafionaicomplementaiity—the organizational mecha-nisms necessary to access the benefits fromcomplementary strategic resources. The abilityof alliance partners to realize the benefits fromcomplementary strategic resources is condi-tioned on compatibility in decision processes,information and control systems, and culture(Doz, 1996; Kanter, 1994). Although complemen-tarity of strategic resources creates the potentialfor relational rents, the rents can only be real-ized if the firms have systems and cultures thatare compatible enough to facilitate coordinatedaction. Previous research suggests that a pri-mary reason for failure of both acquisitions andalliances is not that the two firms do not possessstrategic complementarity of resources, butrather because they do not have compatible op-erating systems, decision-making processes,and cultures (Buono & Bowditch, 1989). Doz(1996), therefore, distinguishes between initialcomplementarity (strategic complementarity),based on potential combinations of resources,and revealed complementarities (organizationalcomplementarity), based on the realized resultsof cooperation between the firms involved in thepartnership.

Proposition 3b: The ability of alliancepartners to geneiate ielational ientsfiom complementaiy stiategic le-souices incieases with the degiee ofcompatibility in theii oiganizationalsystems, processes, and cultuies (oi-ganizational complementaiity).

In summary, both strategic and organizationalcomplementarity are critical for realizing the po-


tential benefits of combining complementarystrategic resources.

Effective Governance

Governance plays a key role in the creation ofrelational rents because it influences transac-tion costs, as well as the willingness of alliancepartners to engage in value-creation initiatives.For example, although alliance partners cangenerate relational rents through investmentsin relation-specific assets, their incentive tomake specialized investments is tempered bythe fact that the more specialized a resourcebecomes, the lower its value is in alternativeuses. The contingent value of a specialized re-source exposes its owner to a greater risk ofopportunism than does a generalized resource(Klein et al., 1978). An important objective fortransactors is to choose a governance structure(safeguard) that minimizes transaction costs,thereby enhancing efficiency (North, 1990; Wil-liamson, 1985).

We distinguish between two classes of gover-nance used by alliance partners: the first relieson third-party enforcement of agreements (e.g.,legal contracts), whereas the second relies onself-enforcing agreements, in which "no thirdparty intervenes to determine whether a viola-tion has taken place" (Telser, 1980: 27). Thetransaction cost economics perspective falls pri-marily within the first class, suggesting thatdispute resolution requires access to a third-party enforcer, whether it be the state (i.e.,through contracts) or a legitimate organizationauthority (Williamson, 1991b). In contrast, self-enforcing agreements (sometimes called "pri-vate ordering" in the economics literature or"trust/embeddedness" in the sociology litera-ture) involve safeguards that allow for self:enforcement. Within the self-enforcement classof governance mechanisms, we further distin-guish between "formal" safeguards, such as fi-nancial and investment hostages (Klein, 1980;Williamson, 1983), and "informal" safeguards,such as goodwill trust or embeddedness (Gulati,1995b; Powell, 1990; Sako, 1991; Uzzi, 1997) andreputation (Larson, 1992; Weigelt & Camerer,1988).

Formal self-enforcing safeguards are eco-nomic hostages created intentionally to controlopportunism by aligning the economic incen-tives of the transactors (Klein, 1980; Williamson,

1983). These hostages may be financial (e.g., eq-uity) or symmetric investments in specialized orcospecialized assets, which constitute a visiblecollateral bond that aligns the economic incen-tives of exchange partners. The fact that thevalue of the economic hostage will decrease invalue if a party is opportunistic provides anincentive for trading partners to behave in amore trustworthy fashion (Dyer & Ouchi, 1993;Pisano, 1989). Further, since these investmentsmay increase in value if the alliance partnerscooperate, there is an incentive for the alliancepartners to engage in value-creation initiatives.

Sociologists, anthropologists, and law and so-ciety scholars long have argued that informalsocial controls supplement—and often sup-plant—formal controls (Black, 1976; EUickson,1991; Granovetter, 1985; Macaulay, 1963). Thus,informal self-enforcing agreements may rely onpersonal trust relations (direct experience) orreputation (indirect experience) as governancemechanisms. A number of scholars have sug-gested that informal safeguards (e.g., goodwilltrust) are the most effective and least costlymeans of safeguarding specialized investmentsand facilitating complex exchange (Hill, 1995;Sako, 1991; Uzzi, 1997). For example, some schol-ars have argued that goodwill trust reducestransaction costs related to bargaining andmonitoring, thereby enhancing performance(Barney & Hansen, 1994; Sako, 1991).̂ Thus, self-enforcing safeguards result in transaction coststhat are lower than they are in situations wheretransactors must erect more elaborate gover-nance structures (e.g., contracts), which arecostly to write, monitor, and enforce.

The ability of exchange partners to matchgovernance structures with exchange attributesis viewed as critical to realizing "economizingadvantages."^ Williamson states, "The main hy-

^ Goodwill trust is defined as one party's confidence thatthe other party in the exchange relationship will not exploitits vulnerabilities (Ring & Van de Ven, 1992; Sako, 1991),Goodwill trust at the interfirm level refers to the extent towhich there is a collectively held trust orientation by organ-izational members toward a partner firm (Zaheer, McEvily, &Perrone, 1998),

^ Although the literature on choice of governance mecha-nisms has focused primarily on transaction costs, Gulati andSingh (in press) show that coordination costs stemming fromthe nature of the interdependence between partners (pooled,reciprocal, or sequential) are very important determinants ofalliance governance structures.


pothesis out of which transaction cost econom-ics works is this: align transactions, which differin their attributes, with governance structures,which differ in their costs and competencies, in adiscriminating (mainly transaction cost minimiz-ing) way" (1991a: 79; emphasis in original). Wil-liamson (1991a) argues that misalignments oc-cur frequently because of bounded rationalityand uncertainty. Thus, transactors who are ef-fective at aligning transactions with gover-nance structures will have an advantage overcompeting transactors who do not employ effi-cient governance mechanisms.

Proposition 4: The greater the alliancepartners' ability is to align transac-tions with governance structures in adiscriminating (transaction cost mini-mizing and value maximizing) way,the greater the potential will be forrelational rents.

We should emphasize that, although the dis-cussion thus far has followed a transaction costlogic with an emphasis on efficiency, we use theterm effective governance to suggest that gov-ernance mechanisms play an important role ingenerating relational rents that extends beyondefficiency arguments. More specifically, a smallbut growing body of literature on transactionvalue is emphasizing the influence of gover-nance on the value-creation initiatives of alli-ance partners (Dyer, 1997; Hansen, Hoskisson, &Barney, 1997; Madhok, 1997; Ring & Van de Ven,1992; Zajac & Olsen, 1993). Effective governancecan generate relational rents by either (1) low-ering transaction costs or (2) providing incen-tives for value-creation initiatives, such as in-vesting in relation-specific assets, sharingknowledge, or combining complementary stra-tegic resources.

In the first case transactors achieve an advan-tage by incurring lower transaction costs thancompetitors to achieve a given level of invest-ment in specialized assets. In the second caseeffective governance (e.g., trust) may allowtransactors to make greater investments in spe-cialized assets than competing transactors, whorefuse to make the relation-specific investmentsbecause of the high cost of safeguarding them.Similarly, alliance partners may be unwilling toshare valuable, proprietary knowledge withtrading partners if they are not credibly assuredthat this knowledge will not be readily shared

with competitors. The willingness of firms tocombine complementary strategic resourcesmay also hinge upon credible assurances thatthe trading partner will not attempt to duplicatethose same resources, thereby becoming a fu-ture competitor. Thus, effective governancemechanisms may generate rents by either low-ering transaction costs or by providing incen-tives for partners to engage in value-creationinitiatives.

In general, self-enforcing mechanisms aremore effective than third-party enforcementmechanisms at both minimizing transactioncosts and maximizing value-creation initiatives.Transaction costs are lower under self-enforcingagreements for four primary reasons.

First, contracting costs are avoided becausethe exchange partners trust that payoffs will bedivided fairly. Consequently, exchange partnersdo not have to bear the cost—or time—of spec-ifying every detail of the agreement in a con-tract. Further, contracts are likely to be less ef-fective than self-enforcing agreements atcontrolling opportunism because they fail to an-ticipate all forms of cheating that may occur.Second, monitoring costs are lower becauseself-enforcement relies on self-monitoringrather than external or third-party monitoring.Exchange partners do not need to invest incostly monitoring mechanisms to ensure con-tract fulfillment and to document infractions tothe satisfaction of a third party (e.g., court).Third, self-enforcing agreements lower the costsassociated with complex adaptation, thereby al-lowing exchange partners to adjust the agree-ment "on the fly" to respond to unforeseen mar-ket changes (Uzzi, 1997: 48). Fourth, self-enforcing agreements are superior to contractsat minimizing transaction costs over the longrun because they are not subject to the timelimitations of contracts. Contracts are typicallywritten for a fixed duration and, in effect, depre-ciate because they only provide protection dur-ing the designated length of the agreement. Atthe end of the contract duration, the alliancepartners need to write a new contract (or employa different safeguard). Exchange partners canavoid the costs of "recontracting" by employingself-enforcing agreements, which, over time,may in fact appreciate in the sense that trust orembeddedness increases with increased famil-iarity and interaction (Gulati, 1995b; Larson,1992).

1998 Dyei and Singh 671

Self-enforcing agreements also call forthgreater value-creation initiatives on the part ofthe exchange partners. For example, it is diffi-cult (if not impossible) to explicitly contract forvalue-creation initiatives, such as sharing fine-grained tacit knowledge, exchanging resourcesthat are difficult to price, or offering innovationsor responsiveness not explicitly called for in thecontract. Under self-enforcing agreements, ex-change partners are more likely to engage inthese activities because they have credible as-surances that they will be rewarded for them.Finally, contractual agreements are relativelyeasy to imitate as a form of governance and,therefore, are unlikely to create sustainable ad-vantages. Competing firms are likely to haveequal access to lawyers (to write the agree-ments) and the state (to enforce the agreements).

Proposition 4a: The greater the alli-ance partners' ability is to employself-enforcing safeguards (e.g., trust orhostages) rather than third-party safe-guards (e.g., legal contracts), thegreater the potential will be for rela-tional rents, owing to (1) lower con-tracting cosfs, (2) lower monitoringcosts, (3) lower adaptation costs, (4)lower recontracting costs, and (5) su-perior incentives for value-creation in-itiatives.

Likewise, within the self-enforcement mecha-nism category, informal safeguards are morelikely to generate relational rents than are for-mal safeguards, for two primary reasons. First,the marginal cost associated with formal hos-tages typically is higher than for informal safe-guards because formal hostages involve capitaloutlays for equity or other types of collateralbonds. Furthermore, formal safeguards aremuch easier for competitors to imitate. If the keyto minimizing transaction costs and encourag-ing value-creation initiatives by partners is sim-ply swapping stock, creating a joint venture, orhaving a partner (e.g., franchisee) post a bond,then competitors can imitate this governancemechanism with relative ease. Informal safe-guards (goodwill trust or reputation) are muchmore difficult to imitate because they are so-cially complex and idiosyncratic to the ex-change relationship.

Proposition 4b: The greater the alli-ance partners' ability is to employ in-formal self-enforcing safeguards (e.g.,trust) rather than formal self-enforcingsafeguards (e.g., financial hostages),the greater the potential will be forrelational rent, owing to (1) lower mar-ginal costs and (2) difficulty of imita-tion.

Although informal safeguards have the great-est potential to generate relational rents, theyare subject to two key liabilities: (1) they requiresubstantial time to develop, because they re-quire a history of interactions and personal ties,and (2) they are subject to the "paradox of trust,"which means that although trust establishesnorms and expectations about appropriate be-havior, lowering the perception of risk in theexchange, it provides the opportunity for abusethrough opportunism (Granovetter, 1985). Inpractice, it appears that many effective alli-ances use multiple governance mechanisms si-multaneously (Borch, 1994). Many alliances be-gin with the use of formal mechanisms and then,over time, employ more informal ones (Gulati,1995b).

Recent empirical studies support the argu-ment that effective governance, in the form oflower transaction costs, may be a source of re-lational rents. For example. Dyer (1997) foundthat General Motors' procurement (transaction)costs were more than twice those of Chrysler'sand six times higher than Toyota's. GM's trans-action costs are persistently higher than Toyo-ta's and Chrysler's primarily because suppliersview GM as a much less trustworthy organiza-tion. Similarly, Zaheer et al. (1998) found that, inthe electrical equipment industry, interorganiza-tional trust reduced negotiation costs and con-flict and had a positive effect on performance.

MECHANISMS THAT PRESERVE RELATIONALRENTS

An explanation of how firms generate rela-tional rents necessarily requires an explanationof why competing firms do not simply imitatethe partnering behavior, thereby eliminatingany competitive advantages that might begained through collaboration. There are a vari-ety of isolating mechanisms that preserve therents generated by alliance partners. First, it is


important to recognize that some of the mecha-nisms already described in the literature on thesustainability of rents within the RBV of the firmapply at the dyadic level. These include causalambiguity and time compression diseconomies(see Barney, 1991; Dierickx & Cool, 1989; Lipp-man & Rumelt, 1982; Reed & DeFillippi, 1990). Forexample, the development of goodwill trust issubject to considerable causal ambiguity be-cause it is a highly complex and situation-specific process (Butler, 1991; Larzelere & Hus-ton, 1980). Moreover, the development of trust orpartner-specific absorptive capacity is subjectto time compression diseconomies because itcannot be developed quickly, nor can it bebought or sold in the marketplace (Arrow, 1974;Sako, 1991).

However, in addition to these mechanisms,relational rent s may be preserved through in (eror-ganizafionaJ asset interconnectedness; partnerscarcity (rareness); resource indivisibility (coevo-lution of capabilities); or a socially complex, andtherefore difficult to imitate, institutional envi-ronment (e.g., country specific). We do not dis-cuss causal ambiguity and time compressiondiseconomies, since these rent-preservationmechanisms have been discussed in detail else-where.

Interorganizational Asset Interconnectedness

Our concept of relational advantage takes theidea of asset interconnectedness across organi-zational boundaries. We submit that interorgan-izational asset interconnectedness will occur incumulative increments on an existing stock ofassets held by a firm or its alliance partner. Toillustrate, a Nissan seat supplier built its planton the property adjacent to a Nissan assemblyplant. The supplier was willing to make thissite-specific investment because Nissan had aminority equity position in the supplier and be-cause the two parties had developed a highlevel of trust. Once this site-specific investmentwas made, the two parties discovered thatrather than transport the seats by truck (a gen-eral-purpose asset), it would be more economi-cal to build a conveyor belt (a highly specializedasset). Consequently, they jointly invested inbuilding the conveyor belt.

This example demonstrates how initial rela-tion-specific investments (i.e., a site-specificplant) create conditions that make subsequent

specialized investments (i.e., customized equip-ment) economically viable. Thus, there is a cu-mulative (snowball) effect that is due to the in-terconnectedness of current relation-specificinvestments with previous relation-specific in-vestments. In contrast, GM's suppliers have notmade the initial site-specific investment; there-fore, it is not economically feasible for them tomake other subsequent specialized invest-ments. The key strategic implication of this iso-lating mechanism is that alliance partners mayneed to make "bundles" of related relation-specific investments in order to realize the fullpotential of those investments in an alliancerelationship.

Partner Scarcity

The creation of relational rents is often con-tingent on a firm's ability to find a partner with(1) complementary strategic resources and (2) arelational capability (i.e., a firm's willingnessand ability to partner). In some cases a late-comer to the partner scene may find that allpotential partners with the necessary comple-mentary strategic resources have already en-tered into alliances with other firms. This is aparticular problem for late movers into foreignmarkets, where there may be few local firmswith the local market knowledge, contacts, anddistribution network needed to facilitate marketentry. In other instances potential partners maysimply lack the relational capability or the rela-tion-building skills and process skills necessaryto employ effective governance mechanisms,make relation-specific investments, or developknowledge-sharing routines (Eisenhardt &Schoonhoven, 1996; Larson, 1992). Firms with col-laboration experience have been found to bemore desirable as partners and more likely togenerate value through partnerships (Gulati,1995a; Mitchell & Singh, 1996).

To illustrate the importance of relational ca-pability, Koichiro Noguchi, Toyota's Interna-tional Purchasing Chief, told the first author thatone of the difficulties Toyota faced in enteringthe U.S. market was finding U.S. suppliers whowere willing to work in partnership fashion.Stated Noguchi:

Many U.S, suppliers do not understand our way ofdoing business. They do not want us to visit theirplants and they are unwilling to share the infor-mation we require. This makes it very difficult for


US to work with them effectively; we also can'thelp them to improve (author interview, July 22,1992).

Thus, even though Toyota had developed a re-lational capability and was effective at partner-ing, it found that it was unable to effectivelygenerate relational rents with U.S. supplierswho had not developed a relational capability.Thus, relational rents may be difficult to imitatebecause potential alliance partners with thenecessary complementary resources and rela-tional capability are rare. The key strategic im-plication of this isolating mechanism is thatthere are strong first mover advantages forthose firms that develop a capability of quicklyidentifying and allying with partners possess-ing complementary strategic resources and/or arelational capability.

Resource Indivisibility

Partners may combine resources or jointly de-velop capabilities in such a way that the result-ing resources are both idiosyncratic and indivis-ible. The VISA organization is an example ofalliance partners (23,000 banks) jointly creatingindivisible assets that help generate returns forthe alliance partners. In particular, the VISAbrand name and distribution network are idio-syncratic and indivisible assets that are collec-tively owned by the participating banks in alarge multifirm alliance. Individual banks canonly access the brand name and distributionnetwork through the alliance.

In other settings, such as with Fuji and Xerox,alliance partners combine resources and capa-bilities, which then coevolve over time. Underthese conditions the mutual coevolution of ca-pabilities of the partner firms can serve as apreserver of rents from the partnership. As thepartners engage in a long-term relationship,they develop dedicated linkages that enhancethe benefits from engaging in the joint relation-ship. Over time, these coevolved capabilitiesare increasingly difficult to imitate, owing topath dependence and resource indivisibility.

A key strategic implication is that the part-ners' resources and capabilities may coevolveand change over time, thereby restricting eachfirm's ability to control and redeploy the re-sources. Although value may be generatedthrough the partnership, there is the potential

for a loss of flexibility, which should be consid-ered at the outset.

Institutional Environment

An institutional environment that encouragesor fosters trust among trading partners (i.e., haseffective institutional "rules" or social controlsfor enforcing agreements) may facilitate the cre-ation of relational rents (North, 1990). Indeed, ata broader level, arguments regarding relationaladvantage can be extended to consider the is-sue of national or country advantage (Casson,1991; Fukuyama, 1995; Hill, 1995). For example,numerous scholars suggest that Japanese trans-actors incur lower transaction costs than U.S.transactors, thereby generating relational rents(Dore, 1983; Dyer, 1996b; Hill, 1995; Sako, 1991;Smitka, 1991). Japanese firms appear to havebeen successful at generating relational rents inpart because of a country-specific institutionalenvironment that fosters goodwill trust and co-operation (Dore, 1983; Hill, 1995; Sako, 1991;Smitka, 1991).

Borys and Jemison (1989) refer to these types ofenvironmentally embedded mechanisms thatcontrol opportunism as "extrahybrid institu-tions." Collaborating firms in other countries(e.g., the United States and Russia) may not beable to replicate the low transaction costs ofJapanese alliance partners because of an in-ability to replicate the socially complex extrahy-brid institutions embedded in the Japanese in-stitutional environment. Thus, following North(1990), one can argue that the institutional envi-ronment can either raise or lower the transac-tion costs that must be borne to achieve a givenlevel of specialization and cooperation. Thestrategic implication of this isolating mecha-nism is that firms may need to locate operationsin particular institutional environments in orderto realize the benefits associated with extrahy-brid institutions.

In summary, the relational rents generated byalliance partners are preserved because com-peting firms

1. cannot ascertain what generates the re-turns because of causal ambiguity;

2. can figure out what generates the returnsbut cannot quickly replicate the resourcesbecause of time compression diseconomies;

3. cannot imitate practices or investments be-cause of asset stock interconnectedness

674 Academy ol Management Review October

(they have not made the previous invest-ments that make subsequent investmentseconomically viable) and because the costsassociated with making the previous in-vestments are prohibitive;

4. cannot find a partner with the requisitecomplementary strategic resources or rela-tional capability;

5. cannot access the capabilities of a potentialpartner because these capabilities are indi-visible, perhaps having coevolved with an-other firm; and

6. cannot replicate a distinctive, socially com-plex institutional environment that has thenecessary formal rules (legal controls) orinformal rules (social controls) controllingopportunism/encourage cooperative behav-ior.

COMPARING THE RELATIONAL. RBV. ANDINDUSTRY STRUCTURE VIEWS

Although an individual firm's ability to workeffectively with other firms may be classified asa firm-specific capability (which may generaterelational rents), there is value in distinguishinga relational view, which offers a distinct, but

complementary, view on how firms generaterents. A relational view considers the dyad/network as the unit of analysis and the rentsthat are generated to be associated with thedyad/network. Although complementary to theRBV, this view differs somewhat in terms of unitof analysis and sources of rent, as well as con-trol and ownership of the rent-generating re-sources (see Table 1).

To illustrate, a Toyota supplier may generaterents by actively participating in the knowl-edge-sharing processes in Toyota's supplier as-sociation. However, the supplier will be unableto generate the knowledge rents if the othermembers decide to exclude it from the network.Similarly, the 23,000 member banks of the VISAorganization have achieved an advantage overAmerican Express and Discover by pooling theirenormous distribution power, which allows foruse of the card at more locations than its com-petitors. Individual banks generate profits withVISA, owing to the jointly created brand nameand distribution network. In both of these cases,the resources that create the relational rents are

TABLE 1Comparing the Industry Structure. Resource-Based, and Relational Views of Competitive

Advantage

Dimensions Industry Structure View Resource-Based View Relational View

Unit oi analysis

Primary sources ofsupernormal profitreturns

Mechanisms thatpreserve profits

Ownership/control ofrent-generatingprocess/resources

Industry

Relative bargaining power

Collusion

Industry barriers to entry

• Government regulations• Production economies/

sunk costs

Firm

Scarce physical resources(e.g., land, raw materialinputs)

Human resources/know-how(e.g., managerial talent)

Technological resources(e.g., process technology)

Financial resourcesIntangible resources (e.g.,

reputation)

Firm-level barriers to imitation

• Resource scarcity/propertyrights

• Causal ambiguity• Time compression

diseconomies• Asset stock

interconnectedness

Collective (with competitors) Individual firm

Pair or network of firms

Relation-specific investments

Interfirm knowledge-sharingroutines

Complementary resourceendowments

Effective governance

Dyadic/network barriers toimitation

• Causal ambiguity• Time compression

diseconomies• Interorganizational asset

stock interconnectedness• Partner scarcity• Resource indivisibility• Institutional environment

Collective (with tradingpartners)


essentially beyond the control of the individualfirm.

In summary, the RBV focuses on how individ-ual firms generate supernormal returns basedupon resources, assets, and capabilities that arehoused within the firm. However, according to arelational perspective, rents are jointly gener-ated and owned by partnering firms.̂ Thus, re-lational rents are a property of the dyad or net-work. A firm in isolation, irrespective of itscapabilities or resources, cannot enjoy theserents. Thus, a relational capability is not a suf-ficient condition for realizing relational rents. AsZajac and Olsen argue, "[B]oth parties use theinterorganizational strategy to establish an on-going relationship that can create value thatcould otherwise not be created by either firmindependently" (1993: 137).

A relational view may offer different norma-tive implications for the strategies firms shoulduse to achieve high profits. For example, accord-ing to the RBV, an individual firm should at-tempt to protect, rather than share, valuable pro-prietary know-how to prevent knowledgespillovers, which could erode or eliminate itscompetitive advantage. However, an effectivestrategy from a relational view may be for firmsto systematically share valuable know-howwith alliance partners (and willingly acceptsome spillover to competitors) in return for ac-cess to the stock of valuable knowledge residingwithin its alliance partners. Of course, this strat-egy makes sense only when the expected valueof the combined inflows of knowledge from part-ners exceeds the expected loss/erosion of ad-vantages due to knowledge spillovers to com-petitors.

Similarly, the relational view and industrystructure view may offer different prescriptionsfor firm-level strategies. For example, according

' We expect the distribution of the relational rents to beconsistent with a resource-dependency perspective (Pfeffer& Salancik, 1978). Partners that bring the more critical (i.e.,scarce) resources to the relationship will be able to appro-priate a higher percentage of the rents (see Asanuma, 1989,and Dyer, 1996a). For example, Toyota made higher profits(return on assets [ROA] = 13.0 percent) than its suppliers(average ROA = 7.1 percent) from 1982-1992, owing to itsgreater relative bargaining power and control over morecritical resources. However, some suppliers, like Denso—asupplier of key electronic components, which brings criticaland scarce resources to the relationship—made profit re-turns (ROA = 12.8 percent) similar to those of Toyota.

to the industry structure view, firms should beeager to increase the number of their suppliers,thereby maximizing bargaining power and prof-its. Porter states, "In purchasing, then, the goalis to find mechanisms to offset or surmountthese sources of suppliers' power. . . . Purchasesof an item can be spread among alternate sup-pliers in such a way as to improve the firm'sbargaining power" (1980: 123).

This strategy is in direct contrast to a rela-tional perspective, which holds that firms canincrease profits by increasing their dependenceon a smaller number of suppliers, thereby in-creasing the incentives of suppliers to shareknowledge and make performance-enhancinginvestments in relation-specific assets. State Ba-kos and Brynjolfsson:

By committing to a small number of suppliers, thebuyer firm can guarantee them greater ex postbargaining power and therefore greater ex anteincentives to make noncontractible investments,such as investments in innovation, responsive-ness, and information sharing; the buyer ends upbeing better off by keeping a smaller piece of abigger pie (1993: 43).

Thus, a relational view may differ from existingviews in the normative prescriptions offered topracticing managers. The fact that there areclear contradictions between these views sug-gests that existing theories of advantage are notadequate to explain interorganizational com-petitive advantage.

CONCLUSION

The central thesis of this article is that a pairor network of firms can develop relationshipsthat result in sustained competitive advantage.Competition between single firms, while per-haps still the rule, is becoming less universal,as pairs and networks of allied firms have be-gun to compete against each other. Our analysissuggests that although looking for competitiveadvantage within firrhs and industries has been(and is still) important, a singular focus on theseunits of analysis may limit the explanatorypower of the models we develop to explain firm-level profitability.

The view we offer here extends the existingliterature on alliances and networks in a num-ber of ways. First, we have attempted to inte-grate what is known regarding the benefits ofcollaboration by examining the interorganiza-


tional rent-generating process. We have arguedthat collaborating firms can generate relationalrents through relation-specific assets, knowl-edge-sharing routines, complementary resource.endowments, and "effective governance." Sec-ond, we have identified the isolating mecha-nisms that preserve the relational rents gener-ated through effective interfirm collaboration.Moreover, we have introduced mechanisms notdiscussed previously in the literature on sus-tainability of rents: interorganizational assetconnectedness, partner scarcity, resource indi-visibility (coevolution of capabilities), and theinstitutional environment. Third, we have ar-gued that a relational perspective may offer nor-mative prescriptions for practicing managersthat contradict the prescriptions offered by theRBV and industry structure view.

In future research scholars might explicitlyexamine these differences in greater detail. An-other important avenue for future researchwould be to examine how relational rents aredistributed among alliance partners. Finally,given the poor track record of many alliances,researchers might examine, in detail, the factorsthat impede the realization of relational rents.

In conclusion, we reemphasize the primaryobjective of this article, which is to propose thatrelationships between firms are an increasinglyimportant unit of analysis for explaining super-normal profit returns. The relational view offersa useful theoretical lens through which re-searchers can examine and explore value-creating linkages between organizations.

REFERENCES

Amit, R., & Schoemaker, P. 1993. Strategic assets and organ-izational rent. Strategic Management Journal, 14: 33-46.

Anderson, E. 1990. Two firms, one frontier: On assessing jointventure performance. Sloan Management Review, 31(2):19-30.

Arrow, K. J. 1974. The limits o/ organization. New York: Nor-ton.

Asanuma, B. 1989. Manufacturer-supplier relationships inJapan and the concept of relation-specific skill. Journalof the Japanese and International Economies, 3: 1-30.

Badaraco, J. L., Jr. 1991. The knowledge link. Boston: HarvardBusiness School Press.

Bakos, J. Y., & Brynjolfsson, E. 1993. Information technology,incentives, and the optimal number of suppliers. Journalol Management Information Systems, 10(2): 37-53.

Barney, J. B. 1986. Strategic factor markets: Expectations,

luck, and business strategy. Managemenf Science, 32:1231-1241.

Barney, J. B. 1991. Firm resources and sustained competitiveadvantage. Journal of Management, 17: 99-120.

Barney, J. B., & Hansen, M. H. 1995. Trustworthiness as asource of competitive advantage. Strategic Manage-ment Journal, 15: 175-190.

Black, D. 1976. The behavior of law. New York: AcademicPress.

Borch, 0 . J. 1994. The process of relational contracting; De-veloping trust-based strategic alliances among smallbusiness enterprises. Advances in Strategic Manage-ment, lOB: 113-135.

Borys, B., & Jemison, D. B. 1989. Hybrid arrangements asstrategic alliances: Theoretical issues in organizationalcombinations. Academy of Management Review, 14:234-249.

Bresnen, M., & Fowler, C. 1994. The organizational correlatesand consequences of subcontracting: Evidence from asurvey of South Wales businesses. Journal of Manage-ment Studies, 31: 847-864.

Buono, A. F., & Bowditch, J. L. 1989. Tiie human side of merg-ers and acquisitions. San Francisco: Jossey-Bass.

Burt, R. 1992. Structural holes: The social structure of compe-tition. Cambridge, MA: Harvard University Press.

Butler, J. K. 1991. Toward understanding and measuring con-ditions of trust: Evolution oi a conditions of trust inven-tory. Journal of Management, 17: 643-663.

Casson, M. 1991. The economics of business culture. Oxford,England: Clarendon Press.

Chung, S., Singh, H., & Lee, K. In press. Complementarity,status similarity, and social capital as drivers of alli-ance formation. Strategic Management Journal.

Clark, K. B., & Fujimoto, T. 1991. Producf deve/opmenf perfor-mance. Boston: Harvard Business School Press.

Cohen, W. M., & Levinthal, D. A. 1990. Absorptive capacity: Anew perspective on learning and innovation. Adminis-trative Science Quarterly, 35: 128-152.

Daft, R., & Lengl, R. 1986. Organizational information re-quirements, media richness and structural design. Man-agement Science, 32: 554-571.

Dierickx, I., & Cool, K. 1989. Asset stock accumulation andsustainability of competitive advantage. ManagementScience, 35: 1504-1513.

Dore, R. 1983. Goodwill and the spirit of market capitalism.British Journal of Sociology, XXXIV(4): 459-482.

Doz, Y. 1996. The evolution of cooperation in strategic alli-ances: Initial conditions or learning processes. StrategicManagement Journal, 17: 55-83.

Dyer, J. H. 1996a. Specialized supplier networks as a sourceof competitive advantage: Evidence from the auto indus-try. Strategic Management Journal, 17: 271-292.

Dyer, J. H. 1996b. Does governance matter? Keiretsu alliancesand asset specificity as sources of Japanese competitiveadvantage. Organization Science, 7: 649-666.


Dyer, J. H. 1997. Effective interfirm collaboration: How firmsminimize transaction costs and maximize transactionvalue. Strategic Management Journal, 18: 553-556.

Dyer, J. H., & Ouchi, W. G. 1993. Japanese style businesspartnerships: Giving companies a competitive edge.Sloan Management Review, 35(1): 51-63.

Eisenhardt, K., & Schoonhoven, C. B. 1996. Resource-basedview of strategic alliance formation: Strategic and so-cial effects in entrepreneurial firms. Organization Sci-ence, 7: 136-150.

Ellicksoh, R. C. 1991. Order without law. Cambridge, MA:Harvard University Press.

Enright, M. J. 1995. Organization and coordination in geo-graphically concentrated industries. In N. Lamoreaux &D. Raff (Eds.), Coordination and information: Historicalperspectives on the organization of enterprise: 103-142.Chicago: University of Chicago Press.

Fukuyama, F. 1995. Trust; The social virtues and the creationof prosperity. New York: Free Press.

Gerlach, M. L. 1992. Alliance capitalism. Berkeley, CA: Uni-versity of California Press.

Ghoshal, S. 1995. Bad for practice: A critique of the transac-tion cost theory. Proceedings of the Academy of Manage-ment: 12-16.

Gomes-Casseres, B. 1994. Group versus group: How alliancenetworks compete. Harvard Business Review, 4(July-August): 4-11.

Granovetter, M. 1985. Economic action and social structure:The problem of embeddedness. American Journal of So-ciology, 91: 481-510.

Grant, R. 1996. Prospering in dynamically-competitive envi-ronments: Organizational capability as knowledge inte-gration. Organization Science, 7: 375-387.

Gulati, R. 1995a. Social structure and alliance formationpatterns: A longitudinal analysis. Administrative Sci-ence Quarterly, 40: 619-652.

Gulati, R. 1995b. Does familiarity breed trust? The implica-tions of repeated ties for contractual choice in alliances.Academy of Management Journal, 38: 85-112.

Gulati, R., & Singh, H. In press. The architecture of coopera-tion: Managing coordination costs and appropriationconcerns in strategic alliances. Administrative ScienceQuarterly.

Hamel, G. 1991. Competition for competence and inter-partner learning within international strategic alli-ances. Strategic Management Journal, 12(Winter SpecialIssue): 83-104.

Hamel, G., & Prahalad, C. K. 1994. Competing for the future.Boston: Harvard Business School Press.

Hansen, M., Hoskisson, B., & Barney, J. 1997. Trustworthiness:A cooperative resource. Working paper, Brigham YoungUniversity, Provo, UT.

Harrigan, K. 1985. Strategic flexibility. Lexington, MA: Lex-ington Books.

Hill, C. W. L. 1995. National institutional structures, transac-

tion cost economizing, and competitive advantage: Thecase of Japan. Organization Science, 6: 119-131.

Hill, R. C, & Hellriegel, D. 1994. Critical contingencies injoint venture management: Some lessons from manag-ers. Organization Science, 5: 594-607.

Kanter, R. M. 1994. Collaborative advantage: The art of alli-ances. Harvard Business Review, 4(July-August): 96-108.

Klein, B. 1980. Transaction cost determinants of "unfair" con-tractual arrangements. American Economic Review,70(2): 56-62.

Klein, B., Crawford, R. G., & Alchian, A. A. 1978. Verticalintegration, appropriable rents, and the competitivecontracting process. Journal of Law and Economics, 21:297-326.

Kogut, B. 1988. Joint ventures: Theoretical and empirical per-spectives. Strategic Management Journal, 9: 319-332.

Kogut, B., & Zander, U. 1992. Knowledge of the firm, combi-native capabilities, and the replication of technology.Organization Science, 3: 383-397.

Larson, A. 1992. Network dyads in entrepreneurial settings: Astudy of the governance of exchange relationships. Ad-ministrative Science Quarterly, 37: 76-104.

Larzelere, R., & Huston, T. 1980. The dyadic trust scale: To-ward understanding interpersonal trust in close rela-tionships. Journal of Marriage and the Family, 42: 595-604.

Levinson, N. S., & Asahi, M. 1996. Cross-national alliancesand interorganizational learning. Organizational Dy-namics, 24: 51-63.

Lippman, S. A., & Rumelt, R. P. 1982. Uncertain imitability: Ananalysis of interfirm differences in efficiency undercompetition. The Bell Journal of Economics, 13(Autumn):418-438.

Macaulay, S. 1963. Non-contractual relations in business: Apreliminary study. American SocioJogical fleWew, 28:55-69.

Madhok, A. 1997. Cost, value and foreign market entry mode:The transaction and the firm. Strategic ManagementJournal, 18: 39-61.

March, J. G., & Simon, H. A. 1958. Organizations. New York:Wiley.

Marsden, P. V. 1990. Network data and measurement. AnnualReview of Socioiogy, 16: 435-463.

Ministry of International Trade and Industry. 1987. Whitepaper on small and medium enterprises in Japan. Tokyo:MITI.

Mitchell, W., & Singh, K. 1996. Entrenched success: The re-ciprocal relationship between alliances and businesssales. Proceedings of the Academy of Management: 31-35.

Mowery, D. C, Oxley, J. E., & Silverman, B. S. 1996. Strategicalliances and interfirm knowledge transfer. StrategicManagement Journal, 17: 77-91.

Nelson, R., & Winter, S. 1982. An evolutionary theory of eco-nomic change. Cambridge, MA: Belknap Press of Har-vard University Press.


Nishiguchi, T. 1994. Strategic industrial sourcing. New York:Oxford University Press.

Nohria, N. 1992. Is a network perspective a useful way ofstudying organizations? In N. Nohria & R. G. Eccles(Eds.), Networks and organizations: 1-21. Boston: Har-vard Business School Press.

North, D. C. 1990. /nstitutions, institutional change and eco-nomic performance. Cambridge, England: CambridgeUniversity Press.

Oliver, C. 1990. Determinants of interorganizational relation-ships: Integration and future directions. Academy ofManagement Review, 15: 241-265.

Oliver, C. 1997. Sustainable competitive advantage: Com-bining institutional and resource-based views. StrategicManagement Journal, 18: 697-714.

Parkhe, A. 1993. Strategic alliance structuring: A game the-oretic and transaction cost examination of interfirm co-operation. Academy of Management Journal, 36: 794-829.

Perry, M. K. 1989. Vertical integration. In R. Schmalensee &R. Willig (Eds.), Handbook of industrial organization:185-255. Amsterdam: North Holland.

Peteraf, M. 1994. Commentary. In P. Shrivastava, A. Huff, &J. Dutton (Eds.), Advances in strategic management, vol.lOB: 153-158. Greenwich, CT: JAI Press.

Pfeffer, J., & Salancik, G. R. 1978. The external control oforganizations: A resource dependence perspective. NewYork: Harper & Row.

Pisano, G. P. 1989. Using equity participation to supportexchange: Evidence from the biotechnology industry.Journal of Law, Economics and Organization, 5: 109-126.

Porter, M. E. 1980. Competitive strategy. New York: FreePress.

Powell, W. W. 1990. Neither market nor hierarchy: Networkforms of organization. In B. M. Staw & L. L. Cummings(Eds.), flesearcli in organizational behavior, vol. 12: 295-336. Greenwich, CT: JAI Press.

Powell, W. W. 1996. Inter-organizational collaboration in thebiotechnology industry. Journal of Institutional and The-oretical Economics, 152: 197-225.

Powell, W. W., Koput, K. W., & Smith-Doerr, L. 1996. Interor-ganizational collaboration and the locus of innovation:Networks of learning in biotechnology. AdministrativeScience Quarterly, 41: 116-145.

Reed, R., & DeFillippi, R. J. 1990. Causal ambiguity, barriersto imitation, and sustainable competitive advantage.Academy of Management Review, 15: 88-102.

Ring, P., & Van de Ven, A. 1992. Structuring cooperativerelationships between organizations. Strategic Manage-ment Journal 13: 483-498.

Rumelt, R. P. 1984. Towards a strategic theory of the firm. InR. B. Lamb (Ed.), Competitive strategic management:556-571. Englewood Cliffs, NJ: Prentice-Hall.

Rumelt, R. P. 1991. How much does industry matter? StrategicManagement Journal, 12: 167-185.

Rumelt, R. P., Schendel, D., & Teece, D. J. 1991. Strategic

management and economics. Strategic ManagementJournal, 12: 5-29.

Ryle, G. 1984. The concept of mind. Chicago: University ofChicago Press.

Sako, M. 1991. The role of "trust" in Japanese buyer-supplierrelationships, flicerche Economiche, XLV: 449-474.

Saxenian, A. 1994. Regional advantage. Cambridge, MA:Harvard University Press.

Shan, W., & Hamilton, W. 1991. Country-specific advantageand international cooperation. Strategic ManagementJournal, 12: 419-432.

Shan, W., Walker, G., & Kogut, B. 1994. Interfirm cooperationand startup innovation in the biotechnology industry.Strategic Management Journal, 15: 387-394.

Singh, H., & Zollo, M. 1997. Learning to acquire: Knowledgeaccumulation mechanisms and the evolution of post-acquisition integration strategies. Working paper. TheWharton School, University of Pennsylvania, Philadel-phia.

Smith, K. G., Carroll, S. J., & Ashford, S. I. 1995. Intra- andinterorganizational cooperation: Toward a researchagenda. Academy of Management Journal, 38: 7-23.

Smitka, M. J. 1991. Competitive ties: Subcontracting in theJapanese automotive industry. New York: Columbia Uni-versity Press.

Szulanski, G. 1996. Exploring internal stickiness: Impedi-ments to the transfer of best practice within the firm.Strategic Management Journal, 17: 27-43.

Teece, D. J. 1987. Profiting from technological innovation:Implications for integration, collaboration, licensingand public policy. In D. J. Teece (Ed.), The competitivechallenge: Strategies for industrial innovation and re-newaJ: 185-219. Cambridge, MA: Ballinger.

Teece, D. J., Pisano, G., & Shuen, A. 1997. Dynamic capabili-ties and strategic management. Strategic ManagementJournal, 18: 509-533.

Telser, L. G. 1980. A theory of self-enforcing agreements.Journal of Business, 53: 27-44.

Uzzi, B. 1997. Social structure and competition in interfirmnetworks: The paradox of embeddedness. Administra-tive Science Quarterly, 42: 35-67.

von Hippel, E. 1988. The sources of innovation. New York:Oxford University Press.

Walker, G., Kogut, B., & Shan, W. 1997. Social capital, struc-tural holes and the formation of an industry network.Organization Science, 8: 109-125.

Weigelt, K., & Camerer, C. 1988. Reputation and corporatestrategy: A review of recent theory and applications.Strategic Management Journal, 9: 443-454.

Wernerfelt, B. 1984. A resource based view of the firm. Stra-tegic Management Journal, 5: 171-180.

Williamson, O. E. 1983. Credible commitments: Using hos-tages to support exchange. American Economic Review,73: 519-535.

Williamson, O. E. 1985. The economic institutions of capital-ism. New York: Free Press.


Williamson, O. E. 1991a. Comparative economic organiza- Zaheer, A., McEvily, B., & Perrone, V. 1998. Does trust matter?tion: The analysis of discrete structural alternatives. Exploring the effects of interorganizational and inter-Administrative Science Quarterly, 36: 269-296. personal trust on performance. Organization Science, 9:

Williamson, O. E. 1991b. Strategizing, economizing, and eco- 141-159.nomic organization. Strategic Management Journal, 12: Zajac, E. J., & Olsen, C. P. 1993. From transaction cost to^ ^ transactional value analysis: Implications for the study

Womack, J. P., Jones, D. T., & Roos, D. 1990. The machine that of interorganizational strategies. Journal of Manage-changed the world. New York: Harper Perennial. ment Studies, 30: 131-145.

Jeffrey H. Dyer is the Stanley Goldstein Assistant Professor of Management at TheWharton School, University of Pennsylvania. He received his Ph.D. in strategy andorganization from the Anderson Graduate School of Management, University of Cal-ifornia at Los Angeles. His research focuses on creating competitive advantagesthrough interfirm collaboration, such as strategic alliances and networks.

Harblr Singh is a professor of management and the Chair of the Management Depart-ment at The Wharton School, University of Pennsylvania. He received his Ph.D. fromthe University of Michigan at Ann Arbor. He has conducted research and publishedarticles on the management of acquisitions, alliances, and corporate restructuring.

THE RELATIONAL VIEW: COOPERATIVE STRATEGY AND …

Documents