Published online EarlyView 18 August 2014 in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/smj.2300Received 14 May 2013; Final revision received 27 May 2014
THE IMPACT OF CULTURE ON THE RELATIONSHIPBETWEEN GOVERNANCE AND OPPORTUNISMIN OUTSOURCING RELATIONSHIPS
SEAN M. HANDLEY* and COREY M. ANGSTDepartment of Management, University of Notre Dame, Notre Dame, Indiana, U.S.A.
The goal of our research is to understand how cul-ture impacts the relationship between governancemechanisms and opportunism in outsourcing rela-tionships. The outsourcing of noncore business pro-cesses has been acknowledged as potentially offer-ing substantive organizational benefits. As such,outsourcing of business processes and services con-tinues to grow as a prevalent business model (e.g.,KPMG, 2012; NASSCOM, 2009). Notwithstandingthe popularity of outsourcing, there are salient risksassociated with outsourcing including the prospectthat the outsourcing service provider may act in adeceitful, self-serving (i.e., opportunistic) mannerto the detriment of the customer. Indeed, oppor-tunism is one of the key behavioral assumptions
Keywords: outsourcing; opportunism; contractual gover-nance; relational governance; culture*Correspondence to: Sean M. Handley, 345 Mendoza College ofBusiness, Notre Dame, IN 46556. E-mail: [email protected]
of transaction cost theory (TCT) and is heraldedas a central consideration in outsourcing decisions(Ghoshal and Moran, 1996; Williamson, 1979).However, Provan and Skinner (1989) suggest thatof the conditions of TCT that impact transactioncosts, opportunism is the least understood. While asignificant literature on outsourcing exists, researchexamining the unique challenges of outsourcingacross national boundaries is surprisingly sparse.A persistent challenge in managing internationalrelationships is understanding the implications ofculture (Couto et al., 2006). Considering that a sub-stantial portion of outsourced relationships spancountry boundaries, it is somewhat surprising thatmore research has not investigated the impact ofculture. Our study departs from others in thatwe investigate a context under which domesticfirms outsource to a wide variety of countriesthat inherently possess variance in culture. This istheoretically and practically interesting because itprovides broader insights about the role that cul-ture plays in transactions, which can be applied toother phenomena.
Culture and Governance in Outsourcing Relationships 1413
To address concerns of opportunism, outsourc-ing organizations need to decide the extent towhich governance (or control) mechanisms will bedeployed (Jap and Ganesan, 2000). The appropri-ate alignment of governance structures with thecharacteristics of an exchange is a crucial con-sideration in the establishment and managementof interfirm relationships (Dyer and Singh, 1998).Two general modes of interorganizational gover-nance mechanisms recognized in the literature areformal mechanisms (i.e., principally formal con-tracts) (Jayaraman, Narayanan, and Luo, 2013;Kalnins and Mayer, 2004; Kirsch et al., 2002;Lafontaine and Shaw, 1999; Masten and Crocker,1985; Poppo and Zenger, 2002) and informal mech-anisms (i.e., social or relational) (Bradach andEccles, 1989; Jayaraman et al., 2013; Kirsch et al.,2002; Macneil, 1985; Poppo and Zenger, 2002;Ring and Van De Ven, 1994). Agency theory andTCT emphasize the prominent role of formal con-tracts as incentive alignment mechanisms aimedat controlling the behavior of an exchange part-ner (Eisenhardt, 1989; Williamson, 1979). Therelational view of interorganizational competitiveadvantage highlights committed and cooperativerelationships (informal mechanism) as prominentgovernance mechanisms (Dyer and Singh, 1998).Contractual governance involves the extent to whichformal incentives/disincentives (typically financial)are used clearly to specify and control service activ-ities between a buyer (i.e., customer firm) and sup-plier (i.e., service provider). Implicit within ourconceptualization is that with contractual gover-nance, the service provider has some variable por-tion of its revenue or compensation explicitly tiedto performance against contractually specified out-comes. Relational governance, on the other hand,assesses the extent to which a business relationshipinvolves commitment and cooperation from bothparties, and performance targets that are less clearlyspecified. Both mechanisms are consistently arguedto be effective at bringing the service provider’sbehaviors into better alignment with the interestsof the customer, and improving the performance ofinterorganizational relationships (Goo et al., 2009;Kirsch et al., 2002; Poppo and Zenger, 2002). How-ever, the literature is not conclusive on whethercontractual and relational governance are substi-tutes (i.e., contractual can take the place of rela-tional governance and vice versa) or complements(one adds to the effectiveness of the other) in theireffect on opportunism (Lumineau and Malhotra,
2011; Rigdon, 2009). In light of the mixed resultsand arguments put forth, the relationship betweencontractual and relational governance is increas-ingly thought to be context-specific (Li et al., 2010;Lumineau and Henderson, 2012; Zhou and Poppo,2010). In fact, some have asserted that extantresearch findings in purely domestic contexts (i.e.,U.S.-based) may not hold in certain internationalenvironments with the complicating influence ofculture being specifically cited (Ellram, Tate, andBillington, 2008; Li et al., 2010; Poppo and Zenger,2002; Zhou, Poppo, and Yang, 2008).
Our study is most closely related to the liter-ature that concurrently evaluates the influence ofcontractual and relational mechanisms (e.g., Jayara-man et al., 2013; Li et al., 2010; Liu, Luo, andLiu, 2009; Poppo and Zenger, 2002; Wuyts andGeyskens, 2005). Simultaneous consideration ofboth contractual and relational governance enablesa cleaner view of the incremental effect of eachgovernance mechanism while controlling for theeffect of the other. For example, Liu et al. (2009)argue that transactional (contractual) mechanismsare more effective than relational in deterring oppor-tunism, however, culture is not considered in theirstudy. Similarly, Jap and Anderson (2003) con-sider how opportunism moderates the relationshipbetween safeguards and exchange performance, butnational culture is not within the scope of theirstudy. Lumineau and Henderson (2012) considerhow and when these governance mechanisms areeffective in mitigating conflict in buyer–supplierrelationships, but again, the role of culture is absent.Of the prior contributions that do concurrentlyexamine the effectiveness of contractual and rela-tional mechanisms, the one most similar to ours isLiu et al. (2009). They utilize dyadic data to evalu-ate the impact of contractual and relational mecha-nisms on opportunism in manufacturer–distributorrelationships. The key distinction between our studyand the contribution of Liu et al. (2009) is that wedirectly evaluate the moderating influence of culturewhereas Liu et al. (2009) do not explicitly considerany location-related moderators.
Williamson (1991) provides the lens throughwhich we examine our phenomenon of interest.He acknowledges the interactive role of gover-nance and institutional environments, arguingthat changes in the institutional environmentconstitute “shift parameters” that alter the relativeeffectiveness of governance mechanisms. In short,Williamson (1991) states that the institutional
environment is made up of a “locus of parameters”(e.g., political, social, legal, reputational, etc.)and changes in this set of parameters can change(shift) relative governance costs at the exchangerelationship level (we discuss this in greater detailduring our theoretical development below). Withfew exceptions (Oxley, 1999), empirical researchexamining the role of shift parameters vis-à-vistheir impact on governing interfirm exchangesis lacking. A likely reason for the sparse use ofshift parameters is that they evaluate the impactof changes or differences in the institutionalenvironment, and thus requires “heterogeneityin institutional environments” (i.e., culture in thepresent study) (Oxley, 1999: 284). For this reason, amulticountry study such as ours is both empiricallyand theoretically important to achieving the neces-sary cultural heterogeneity. Those studies that doexist tend to focus on the impact of formal institu-tions (e.g., legal and regulatory) on governance (Liet al., 2010; Zhou and Poppo, 2010) rather than therole of informal institutions such as culture. Whilesome authors have drawn connections betweennational institutions and organizational decisionmaking and structure (Crossland and Hambrick,2011; Oxley, 1999; Siegel, Licht, and Schwartz,2011), the interactions between national culture,contractual governance, and relational governancehas not been investigated in a multicountry study,despite the multiple calls for research (Poppo andZenger, 2002; Wuyts and Geyskens, 2005).
In accordance with these observations, thecurrent study was undertaken to examine therole that two distinct aspects of culture (i.e.,individualism–collectivism and uncertainty avoid-ance) play in moderating the relationship betweencontractual and relational governance and theireffect on opportunism in outsourcing relationships.In testing this set of relationships, we aim not onlyto illuminate the moderating influence of culture oneach governance mechanism individually, but alsoto use national culture as a means of contributing tothe reconciliation of the aforementioned divergentperspectives in the literature regarding the relation-ship between contractual and relational governance(Liu et al., 2009; Lumineau and Henderson, 2012;Zhou and Xu, 2012). Our objectives require thatwe draw from a sample that incorporates multiplecultural distinctions found in international contexts.We use dyadic data (i.e., from both customer andservice provider organizations) collected on 102outsourcing relationships to test these relationships.
In our sample, all customer firms are based in theU.S., whereas service providers had operations in41 different countries. This research setting allowsfor findings with greater external validity thanpreviously offered in the literature.
The results of the study are modestly support-ive of a complementary relationship between con-tractual and relational governance as they relateto deterring opportunism. The results related tothe moderating impact of culture suggest that theeffectiveness of both contractual and relationalgovernance mechanisms in mitigating opportunis-tic behavior is context-specific. Specifically, con-tractual governance is more effective at mitigat-ing service provider opportunism in more indi-vidualistic and low uncertainty avoidance cultures.Conversely, relational governance is more effec-tive in collectivist and high uncertainty avoidancecultures. These findings suggest that the effec-tiveness of these governance mechanisms is morecontext-specific than previously thought. Further-more, culture appears to be one variable (of poten-tially many) that might explain the inconsistentfindings related to the complementary/substitutiverole that the two governance mechanisms play intheir impact on opportunism. These findings makean important contribution to the nascent empiri-cal literature on the management of internationaloutsourcing as well as to the literature investigat-ing the independent and joint effects of contrac-tual and relational governance on interfirm oppor-tunism. Finally, our work provides further nuance tothe voluminous transaction cost literature in that itlends empirical support to the notion of shift param-eters in interfirm, multicultural relationships.
THEORETICAL DEVELOPMENT
Opportunism and governance
Opportunism is defined as “self-interest seekingwith guile” (Williamson, 1975: 9), with the term“guile” referring to “lying, stealing, cheating, andcalculated efforts to mislead, distort, disguise,obfuscate, or otherwise confuse” (Williamson,1975: 47). Transaction cost theorists argue thatexchanges characterized by a high risk of oppor-tunism require more pronounced expenditures onmechanisms aimed at controlling the behaviorof the other party; in our case, the outsourcingservice provider. In this study, opportunism is
Culture and Governance in Outsourcing Relationships 1415
reflected as a higher order construct comprisedof shirking and poaching; two germane forms ofopportunism in outsourcing relationships (Aron,Clemons, and Reddi, 2005; Wathne and Heide,2000). Shirking is the extent to which the providerdeliberately underperforms or withholds resourceswhen the customer is unable to detect the action(Handley and Benton, 2012) and is related to themoral hazard concern in agency theory (Eisenhardt,1989). Poaching, also referred to as diffusion risk(Walker, 1988), represents the degree to which theprovider is inclined to utilize information obtainedthrough the relationship for its own unauthorizedbenefit (Handley and Benton, 2012).
Proponents of TCT argue that contracts are “theprimary alternative to vertical integration as a solu-tion to the general problem of opportunistic behav-ior,” (Klein, Crawford, and Alchian, 1978: 302).Properly designed contractual incentives reducegoal misalignment and thus mitigate the risk ofself-interest seeking behavior. Accordingly, it issuggested that outsourcing contracts incorporatepenalty and reward clauses explicitly tied to perfor-mance against service level objectives (Eastwood,2006; Landis, Mishra, and Porrello, 2005; Robin-son et al., 2008). In keeping with these perspectives,the construct of contractual governance here rep-resents the degree to which the service provider’snear-term compensation and longer-term businessopportunities are impacted by their performanceagainst contractual service level agreements. Thesecontractual mechanisms should, ceteris paribus,reduce the risk of service provider opportunism.This relationship has been extensively supportedin prior research—particularly in the economicsfield (Kalnins and Mayer, 2004; Lafontaine andShaw, 1999; Masten and Crocker, 1985), but alsoin strategic management (Barthélemy and Quélin,2006; Dyer and Singh, 1998; Lumineau and Mal-hotra, 2011) and operations management (Li et al.,2010)—therefore, we do not hypothesize a specificlink but instead assume provider opportunism to benegatively related to contractual governance.
Unlike TCT’s emphasis on formal control, therelational view of interorganizational competitiveadvantage (Dyer and Singh, 1998) centers on infor-mal mechanisms to achieve control. This per-spective asserts that relationships characterized bycooperative practices such as open communica-tion, extensive information sharing, and joint prob-lem resolution result in shared norms and mutualcommitment (Bradach and Eccles, 1989; Li et al.,
2010; Poppo and Zenger, 2002; Uzzi, 1997). Insuch an environment, actions are governed by adesire for solidarity (Dyer and Singh, 1998; Heideand John, 1992; Liu et al., 2009; Ring and Vande Ven, 1992) and goal alignment is driven byinformal social pressures to behave in accordancewith the shared value system. The negative asso-ciation between committed and cooperative rela-tionships and the level of interfirm opportunism isalso widely recognized in the extant literature (e.g.,Gulati, 1995; Liu et al., 2009); therefore we do notformally present an associated hypothesis. Rather,our intent is to build upon these established rela-tionships between governance and opportunism andhighlight the important role that culture plays as amoderator. However, we first consider the expectedcomplementary role that contractual and relationalgovernance play in influencing opportunism.
The manner by which contractual and relationalgovernance interact with one another continues tobe subject to significant debate (e.g., Lumineauand Malhotra, 2011; Rigdon, 2009). Some arguethat more extensive use of either contractual orrelational governance reduces the need for the other(Dyer and Singh, 1998; Gulati, 1995; Macaulay,1963). That is, they are substitutes. Many suggestthat committed and cooperative partnerships withshared norms and values reduce the risk of oppor-tunism in ways that are more flexible and costeffective than can be achieved through formal con-tracts (Dyer and Singh, 1998; Gulati, 1995; Uzzi,1997). Thus, as the relationship becomes moreembedded, the need for costly formal control sub-sides (Ring and Van De Ven, 1994). Studies in theeconomics literature argue that coupling economicincentives to performance outcomes underminesthe intrinsic motivation to cooperate (Kessler andLeider, 2012) and signals a weakening relationshipleading to a reduction in cooperative exchangebehaviors (Rigdon, 2009). Despite these argumentsfor contractual and relational governance having asubstitutive relationship, others posit the opposite.Due to bounded rationality, the development ofa fully specified contract is often not possible orpractical (Klein et al., 1978; Williamson, 1979).This incompleteness necessarily results in a resid-ual opportunism risk that can be further addressedthrough informal, relational means (Li et al.,2010). Further, adaptations to the formal agreementare often necessary through the course of therelationship. In exchanges with greater relationalgovernance, contractual modifications can be
achieved more effectively due to the cooperativeand committed orientation that exists (Goo et al.,2009; Poppo and Zenger, 2002). Finally, contrac-tual agreements can improve clarity about the roles,responsibilities, and performance expectations ofeach party. This clarity reduces uncertainty whileenhancing coordination and commitment (Japand Ganesan, 2000). Thus, interfirm relationshipsgoverned jointly by economic and social incentivemechanisms should result in a lower risk of oppor-tunism than relationships governed by only one,suggesting that they serve as complements. Whilewe acknowledge that some studies have arguedand found support for contractual and relationalgovernance being substitutes (Dyer and Singh,1998; Gulati, 1995; Macaulay, 1963; Malhotra andMurnighan, 2002), the majority of recent empiricalevidence suggests the relationship is complemen-tary (Goo et al., 2009; Jayaraman et al., 2013; Liuet al., 2009; Poppo and Zenger, 2002). We proceedfrom the premise that they will be complementaryand focus our attention on the moderating role ofnational culture.
National culture as a shift parameter
There is some debate as to the impact of nationalculture on management practices. The convergenceview holds that as countries become more inte-grated into the global economy, their work practicesand organizational cultures become more similar(Naor, Linderman, and Schroeder, 2010; Shenkarand Ronen, 1987). On the contrary, the divergenceview suggests that due to deep-rooted differencesin norms and value systems, organizations indifferent countries will adopt different practicesand form heterogeneous organizational cultures(Ralston et al., 1997). The influence of nationalculture has been studied with regard to a widearray of managerial issues including some relatedto the current study. For instance, prior researchon interorganizational management has examinedthe impact of culture on the buyer’s long-termorientation toward supplier relationships (Cannonet al., 2010), the personal attachment of boundaryspanning personnel (Luo, 2001), internationalnegotiation schema (Brett and Okumura, 1998),joint-venture dissolutions (Park and Ungson, 1997),financial contracts between lenders and borrowers(Giannetti and Yafeh, 2011), and the propen-sity to enter international marketing alliances(Yeniyurt et al., 2009).
Following the divergence perspective and theWilliamson (1991) shift parameter framework, theoverriding thesis adopted here is that the culture(in provider countries) impacts the degree to whichcontractual and relational governance are effectiveat mitigating opportunism. Our intent is to exam-ine how governance decisions (contractual and rela-tional) interact with the cultural environment toaffect opportunistic behavior. The theoretical fram-ing of our model, as well as our intended empiricalanalysis, implicitly assumes that governance mis-alignments (i.e., mistakes) do occur, and that thesemisalignments have important performance impli-cations, with regard to opportunism. Such a fram-ing begs the question of “why governance mistakesare made.” Alternatively, “why would either partyagree to a governance framework that does not suf-ficiently protect its interests?” Extant research pro-vides several explanations (Argyres and Bigelow,2007; Nickerson and Silverman, 2003; Sampson,2004). One such explanation is that firms perceivethe hazards in their contracting environment withsome degree of error (Masten, 1993; Sampson,2004). Additionally, due to anchoring and insuf-ficient adjustment biases, some firms may erro-neously perceive contractual and/or relational gov-ernance to be similarly effective across cultural con-texts, when in fact they are likely to be different(Bowman, 1963; Tversky and Kahneman, 1974).In short, we argue that outsourcing across nationalboundaries represents a complex contracting envi-ronment subject to governance mistakes. These mis-takes place some firms in the unfortunate situationof operating under a governance framework thatis ineffective in mitigating opportunistic behavior.Thus, while we have no reason to expect that eitherparty will systematically and knowingly agree to anegregiously one-sided relationship, there is a possi-bility they will agree to a governance system that ismisaligned with their cultural norms to some extent.While it is difficult to perfectly disentangle gov-ernance decisions from their impact, the focus inthis study is not specifically how or why the firmsarrived at the agreed upon-governance system, butrather, once in place, how effective it is at deterringprovider opportunism.
In keeping with the concept of shift parame-ters (Oxley, 1999; Williamson, 1991), we posit thatcultural norms in the countries where the serviceproviders operate “shift” the effectiveness of gover-nance decisions. Because governance mechanismsoperate at the relationship level and culture operates
Culture and Governance in Outsourcing Relationships 1417
at a higher level of social analysis (see Williamson,2000), culture is posited to constrain or enhancethe role of contractual and relational governance.While cultural distance between partners has beenshown to be an important determinant of the qual-ity of interorganizational relationships, knowledgesharing, and other relational issues (Griffith, Myers,and Harvey, 2006), we focus on the provider’s cul-ture rather than the difference between the buyer’sand provider’s culture. Our theoretical argumentis that the culture of the provider firm influencestheir opportunistic behavior and the effectiveness ofmechanisms aimed at altering this behavior. Thislogic is similar to the theoretical arguments madeby Shane (1994) that a foreign country’s cultureinfluences American firms’ tendency to use licens-ing agreements rather than direct foreign investmentand the Hewett and Bearden (2001) hypothesis thatthe culture in a foreign country moderates the extentto which managers in that country will cooperatewith headquarters in the U.S. Yet, we acknowledgethat cross-cultural relationships may respond differ-ently from those that are intracultural, therefore wecontrol for this in our model. We do not use cul-tural distance because it suggests that providers willpotentially respond differently depending upon theculture of their customer. This is a different, albeitvery interesting, research question but one that isbeyond the scope of our study.
Drawing on the commonly invoked frameworkof core cultural dimensions, based principallyon the work of Geert Hofstede (Hofstede, 1980,1985), we hypothesize that the effectiveness ofcontractual and relational governance mechanismsis moderated by two dimensions of national culture:individualism–collectivism (IDV) and uncertaintyavoidance (UAI). We chose these two dimensionsfor multiple reasons. First, both IDV and UAIhave been found to be more predictive than othercultural dimensions in managerial decision making(Crossland and Hambrick, 2011). Second, althoughmultiple cultural value typologies have been pre-sented, the IDV dimension (or something similar)has appeared in several recognized frameworksand is consistently considered a core dimensiondistinguishing different cultures (Cannon et al.,2010; Crossland and Hambrick, 2011). The UAIdimension has previously been associated withthe design of international control and governancesystems (Homburg et al., 2009; Ueno and Sekaran,1992). We opt here to focus on the theoreticallysupported relationships regarding IDV and UAI
and leave the other dimensions—which would bemore exploratory—for future research.
As shown in Figure 1, extant research suggests anegative relationship between the extent of use ofgovernance mechanisms and provider opportunism(see a). The combined effect of relational andcontractual governance is said to further decreaseopportunism (see line CG+RG). In Figure 1(b–d),we highlight the role that one dimension of nationalculture—individualism/collectivism—plays inshifting the effectiveness of governance on oppor-tunism. Although not depicted, the shifting effectsassociated with uncertainty avoidance would besimilar in nature. Consistent with Williamson’s(2000) theorizing that higher levels of social analy-sis (e.g., culture) constrain or enhance lower levelgovernance structures, we argue that a tighter align-ment between the characteristics of the governancemechanism and the provider’s cultural values willmanifest in a more pronounced effect on mitigatingprovider opportunism, whereas misalignment isexpected to detract from the effectiveness.
The impact of individualism–collectivism
The individualism–collectivism dimension reflectshow tightly knit the social framework is in a society.In high individualism cultures, there is a preferencefor a loosely knit social framework in which “indi-viduals are supposed to take care of themselves andtheir immediate families only” (Hofstede, 1985:348). In more collectivist cultures, individualsexpect “their relatives, clan, or other in-group tolook after them” (Hofstede, 1985: 348). Individual-ism and collectivism represent opposite ends of thesame dimension. We argue that contractual gov-ernance is more suited to individualistic cultures;therefore increased contractual governance willbe more effective at deterring opportunism (seeFigure 1b; slope of CGI < slope of CGC). The basicintuition underlying our expectation that contrac-tual governance is more effective in individualisticcultures is that members of these societies are morerational, calculative in their decision making, andare therefore more responsive to extrinsic economicincentives (Cannon et al., 2010; Davis, Schoorman,and Donaldson, 1997; Doney, Cannon, and Mullen,1998). In more individualistic cultures, people andorganizations are more receptive to compensationsystems that are contingent on their performance(Gelfand et al., 2004; Redding, Norman, and Sch-lander, 1994), including incentive-based contracts
Culture and Governance in Outsourcing Relationships 1419
with compensation and future opportunities tied toperformance against service level agreements. Con-sistent with this logic, empirical research on CEOcompensation has found that CEOs in individual-istic cultures have a higher proportion of their totalcompensation contingent upon performance than dothose in more collectivist cultures (Tosi and Greck-hamer, 2004). Building on this extant research, weposit that contractual governance, which makesthe service provider’s financial prospects morecontingent upon performance, is more aligned withthe cultural norms in individualistic as opposedto collectivist societies. This tighter alignment isexpected to result in contractual governance beingmore effective at deterring opportunism. Thereforewe test
Hypothesis 1 (H1): The negative influence ofcontractual governance on opportunism isstronger in individualistic cultures than it is incollectivist cultures.
It is widely acknowledged that there are asym-metric conceptions of relational quality in businessrelationships that cross international borders andthus cultures (Zaheer and Zaheer, 2006). Whereasprior research associates individualistic cultureswith calculative, rational decision making moti-vated predominantly by economic concerns, otherresearch observes that members of collectivist soci-eties are more driven by social factors, potentiallyat the expense of material considerations (Daviset al., 1997; Gelfand et al., 2004). Cannon et al.(2010: 508–509) concisely summarize these obser-vations by stating that in collectivist cultures “peo-ple are valued over performance criteria” and “busi-ness exchange relationships are primarily socialrather than instrumental.” Because relational gov-ernance is often referred to as social or clan control(Eisenhardt, 1985; Li et al., 2010), we expect ser-vice providers in collectivist societies to be morereceptive to relational governance mechanisms. Associeties become less individualistic, cultural normsemphasize long-term relationships, harmony, andcooperation (Davis et al., 1997; Doney et al., 1998),hallmarks of strong relational governance (Dyer andSingh, 1998). The mismatch between individualis-tic cultures and relational governance is likely tominimize the effectiveness of this mechanism atattenuating opportunism. That is, we posit relationalgovernance to be more effective in collectivist soci-eties (see Figure 1c; slope of RGC < slope of RGI).
Hypothesis 2 (H2): The negative influence ofrelational governance on opportunism is weakerin individualistic cultures than it is in collectivistcultures.
The impact of uncertainty avoidance
Uncertainty avoidance represents the extent towhich a society feels “uncomfortable with uncer-tainty and ambiguity” and a preference for “beliefspromising certainty” (Hofstede, 1985: 347–348).Prior research has characterized members of highuncertainty avoidance cultures as being risk averse(Luque and Javidan, 2004). Established theory rec-ognizes that incentive or outcome-based contractsshift risk to the agent (i.e., the service provider inour context) (Beatty and Zajac, 1994; Eisenhardt,1989). Therefore, contracts that shift risk to theservice provider are misaligned with the normsin high uncertainty avoidance (i.e., risk averse)cultures. Consequently, we would not expectindividuals or organizations in high uncertaintyavoidance cultures to respond favorably to contrac-tual arrangements that place their compensation atrisk. Although in different contexts, prior empiricalwork has supported these theoretical argumentsby demonstrating a negative association betweenhigh uncertainty avoidance cultures and the useof incentive or performance-based compensationmechanisms (Brown Johnson and Droege, 2004;Tosi and Greckhamer, 2004). Our conceptualizationof contractual governance centers on the extentto which the service provider’s compensationand future commercial opportunities are tied toperformance against service level objectives. Thus,while some contract forms may be argued toserve as uncertainty reducing mechanisms throughthe clarification of roles and responsibilities, thecontingent nature of the contractual provisions weexamine introduces financial uncertainty and ambi-guity that members of high uncertainty avoidancesocieties seek to avoid. Again, this misalignmentof cultural norms with governance is proposed tomoderate negatively the effectiveness of contractualgovernance in discouraging opportunistic behavior.While we did not graphically depict the UAI shiftparameters, we expect them to function similarlyto the relationships represented in Figure 1(b–d).
Hypothesis 3 (H3): The negative influence ofcontractual governance on opportunism is
weaker in high uncertainty avoidance culturesthan it is in low uncertainty avoidance cultures.
We know from prior literature that uncertaintyavoidance is essentially a proxy for the extentof unpredictability and risk a firm is willing toundertake (Crossland and Hambrick, 2007). Onestrategic action that can serve as a source of unpre-dictability and ambiguity is discontinuing an exist-ing interfirm relationship. Given that members ofhigh uncertainty avoidance cultures seek to avoidrisk, it is expected that they would be reticent tomove in and out of business relationships. Indeed,existing research observes that people and organi-zations in low uncertainty avoidance cultures severexisting relationships and enter new relationshipsmore freely (Doney et al., 1998; Kale and Barnes,1992). Alternatively, high uncertainty avoidancesocieties favor permanence and stability in rela-tionships and nurture bonds for longer-term bene-fits (Luque and Javidan, 2004). Interorganizationalrelationships characterized by high levels of rela-tional governance also favor a long-term, partner-ship orientation where the parties are committedto the relationship (Cannon et al., 2010; Dyer andSingh, 1998; Jap and Ganesan, 2000). Thus, rela-tional governance is well aligned with the normsin high uncertainty avoidance cultures. Due tothis strong alignment, it is expected that relationalgovernance will be more effective at controllingprovider behavior in high uncertainty avoidancecultures. Therefore, it is posited
Hypothesis 4 (H4): The negative influenceof relational governance on opportunism isstronger in high uncertainty avoidance culturesthan it is in low uncertainty avoidance cultures.
Governance, culture and opportunism
Earlier we highlighted the debate in the literaturerelated to the joint effect of contractual and rela-tional governance on opportunism. We also arguedthat culture moderates the unique effect of eachof these governance mechanisms on opportunism.Since we do not hypothesize main effects for con-tractual and relational governance and instead focuson their complementary effect on opportunism, wefeel it is critical to examine the moderating effect ofculture on the two governance mechanisms jointlyas well. As we expect culture to influence the extent
to which both contractual and relational governanceare effective deterrents of opportunism, we wouldalso anticipate that culture will impact how the twomechanisms interact with one another in their effecton opportunism. From a pragmatic perspective, pro-viding theoretical support for complex three-wayinteractions is challenging. Some authors have cho-sen to simply posit a significant effect withoutcommitting to directionality (Halford et al., 2005)and acknowledge that the important insights comewhen graphing the interactions and post-hoc analy-ses (Angst and Agarwal, 2009; Dawson and Richter,2006). Ultimately this is somewhat exploratory, butit can provide guidance for future theoretical contri-butions. Thus, we simply hypothesize that both IDVand UAI will significantly moderate the two-wayinteraction between contractual and relational gov-ernance in its effect on opportunism (see Figure 1d)and in a post-hoc analysis we will discuss specificaspects of the findings. Therefore, we hypothesize:
Hypothesis 5 (H5): The individualism–collectivism cultural dimension moderates therelationship between contractual governanceand relational governance in their complemen-tary effect on opportunism.
Hypothesis 6 (H6): The uncertainty avoidancecultural dimension moderates the relationshipbetween contractual governance and relationalgovernance in their complementary effect onopportunism.
A complete representation of our conceptualmodel is shown in Figure 2 below.
METHODOLOGY
Data collection, sample, and measurement
The target population for this study was largeU.S.-based firms engaged in domestic and/or inter-national outsourcing of business processes. Serviceprovider operations could be located in the U.S., inan international location, or both. Engaging organi-zations meeting these criteria began by compiling alisting of contacts obtained through multiple onlinedatabases. This search was restricted to only thosefirms found in the Russell 3000 Index as of October2009. Key contacts were identified by reviewingjob titles and, where available, job descriptions.
Culture and Governance in Outsourcing Relationships 1421
Governance
Governance Relational
Contractual
Individualism-Collectivism
H1
H2
H5
Uncertainty Avoidance
Culture
H3
H
Control Variables
H4
ProviderOpportunism
H6
Service
Figure 2. Conceptual model
Contact information was obtained for members ofmanagement at 2,356 companies. Each of theseorganizations was provided with an overview of theresearch project detailing what would be expectedof them. Interested firms were asked to provide theresearch team with the contact information for alloutsourcing relationships within a functional area.Following this procedure, an initial agreement toparticipate was obtained from 78 organizations (i.e.,customers). Telephone or email communicationswere established with 176 other firms in an effort todetermine their reason for not participating. Thesecorrespondences did not reflect an observable bias,with the majority citing company policy againstsharing sensitive information (i.e., names of theirservice providers) or a personal lack of sufficientknowledge with the firm’s outsourcing initiatives.
Web-based surveys were distributed to the 78customer firms who agreed to participate alongwith their identified service providers. As requested,some customer firms provided contact informationfor multiple service providers to reflect all of theiroutsourcing relationships within a particular func-tional area. As our analysis is dyadic in nature,we only included in the final sample relationshipsfor which we obtained a completed survey fromboth the customer and service provider organiza-tion. Completed surveys representing 134 outsourc-ing relationships were submitted from the customerfirms, while service provider surveys were submit-ted for only 105 of the relationships. After account-ing for the nonresponse from either the serviceprovider (mostly) or customer firm representative(a few instances) at the relationship level, both cus-tomer and service provider responses (i.e., matcheddyads) were received for 102 relationships. Whilemodest, the rate of response and sample size is
an anticipated by-product of the rigorous researchdesign (i.e., requesting the inclusion of all relation-ships within a functional area and requiring match-ing responses from service providers). Traditionalassessments of nonresponse bias were conductedby comparing early versus late respondents (Arm-strong and Overton, 1977). Moreover, a 𝜒2 testwas used to compare respondents to all firms inthe Russell 3000 Index in terms of industrial sectorrepresentation. These assessments did not offer rea-son for a significant concern of a response bias;although we acknowledge that we cannot fully ruleout this potential. The majority of the customerfirms in our sample came from four sectors (Con-sumer Discretionary, 22%; Technology, 22%; Pro-ducer Durables, 17%; and Health Care, 15%) andmost of the outsourced activity broke down into twofunctional areas (Logistics/Supply Chain, 49% andInformation Technology, 31%). For more details,see online Appendix S1, Tables S1 and S2.
In the sections below, the measurement basis foreach variable is described. The final items usedin the multiitem constructs are presented in theonline Appendix S1. All single-item constructs aresufficiently described within the text. Except for thebuyer-switching difficulty, previously in-house, andprovider expert constructs, all items are measuredfrom the provider’s survey.
Opportunism
Service provider opportunism is reflected by twofirst-order constructs representing salient forms ofopportunism noted in the outsourcing literature:shirking and poaching. The scales used to repre-sent these two constructs were newly developedfor this research project. While other measures ofopportunism exist, our objective was to assess these
two specific forms of opportunism highlighted inthe extant literature (Aron et al., 2005; Jap andAnderson, 2003; Wathne and Heide, 2000). To ourknowledge, no prior scales specifically measureshirking and poaching. The items were developedand refined through a multiround q-sorting exerciseinvolving academic peers and experienced sourcingprofessionals (c.f. Menor and Roth, 2007). The sort-ing exercise included shirking and poaching amonga battery of seven multiitem constructs. By the finalround, each item was correctly matched with itsintended construct at an acceptable rate. The psy-chometric results surpassed thresholds prescribed inthe literature (overall placement rate> 90%, meanproportion of substantive validity= 0.90, and meancoefficient of substantive validity= 0.81).
Governance mechanisms
The four-item scale used to represent contractualgovernance (CG) was developed for this study. Thescale was not included in the q-sorting exercise.Rather, it resulted from discussions with one of thesourcing experts who served as a “sorter” in theexercise. As previously explained, our CG constructspecifically represents the extent to which the ser-vice provider’s compensation and business opportu-nities are tied to their performance against contrac-tual service level agreements (SLAs). To our knowl-edge, previously validated measures of interfirmcontracting do not specifically assess these contrac-tual contingencies as conceptually specified herein.We sought to word the items in a vernacular familiarto the respondents. As such, the CG items all relateto the SLAs that exist within the contract betweenthe two organizations. SLAs are frequently noted askey elements in outsourcing contracts (Eastwood,2006; Landis et al., 2005; Robinson et al., 2008).
Relational governance (RG) represents thereliance upon committed and cooperative relationsto govern the commercial exchange. Relationalgovernance or similar constructs have been oper-ationalized in a multitude of ways in the literature.Drawing on multiple previously validated scales,we utilize a composite of four items to reflectRG (Benton and Maloni, 2005; Cannon andPerreault, 1999).
National culture
The culture scores are with respect to the locationof the service provider’s operations. Each culture
measure— individualism–collectivism (IDV)and uncertainty avoidance (UAI)—is basedon Hofstede’s cultural dimension scores (Hof-stede, 1980). Scores were obtained fromhttp://www.geerthofstede.nl in December 2011.While alternative cultural typologies exist in theliterature, Hofstede’s dimensions and scores arewidely utilized (Sousa and Bradley, 2008; Yeniyurtet al., 2009). As highlighted previously (Crosslandand Hambrick, 2011), using the Hofstede measuresallows for more direct comparison with other con-tributions in this literature. To determine the valuesfor IDV and UAI for each outsourcing relationship,the questionnaire presented the respondent witha listing of countries (along with space to type incountries) and asked him/her to indicate all coun-tries where the provider had operations servicing thefocal relationship. This information was combinedwith the culture scores for each country to arrive atan average relationship score for IDV and UAI.
Exchange hazards and other contextual controlvariables
Transaction cost theory posits that interorganiza-tional difficulties depend upon three characteristics:asset specificity, uncertainty, and frequency(Williamson, 1979). In this study, the focus ison recurring relationships and thus frequencyis not a distinguishing factor. Measures of assetspecificity and uncertainty, the most commonlyassessed exchange hazards in empirical studiesof TCT (David and Han, 2004), are included.It is important to control for these hazards incontractual relations as they have been linked todependency, vulnerability, and ultimately the risk ofopportunistic behavior (Wathne and Heide, 2000).Human or physical assets that are specialized toa certain relationship may be difficult to redeployor have limited value outside of the relationship(Holmström and Roberts, 1998). Their presencecreates a lock-in situation, leaving the owner vul-nerable to opportunistic action by their exchangepartner. To reflect these exchange conditions, thisstudy includes measures from both the buyer (i.e.,switching difficulty—SD) and provider perspectives(i.e., provider specificity—PS). The buyer’s lock-invulnerability is captured by a single-item measureof switching difficulty to provide a broader, andperhaps more practical, assessment of exchangehazards from their perspective. A two-item mea-sure of the extent to which the provider has made
Culture and Governance in Outsourcing Relationships 1423
investments in human and physical resources,which are unique to this relationship, is used toassess the provider’s lock-in vulnerability. Impor-tantly, switching difficulty and relationship-specificinvestments are also noted as key determinants ofinterfirm dependency and relative bargaining power(Buchanan, 1992; Heide and John, 1988). Theinclusion of both buyer and provider perspectivescapitalizes on the dyadic nature of our data, and isa strength of the analysis. Uncertainty representsthe rate and predictability with which the exchangeenvironment is changing. Environmental uncer-tainty exacerbates the difficulty in developing fullyspecified and complete contracts (Crocker andReynolds, 1993; Klein et al., 1978; Williamson,1979). This contractual incompleteness also leavesorganizations vulnerable to opportunistic behavior(Crocker and Reynolds, 1993). Technologicaluncertainty (TU) is the most commonly usedmeasure of uncertainty (David and Han, 2004).Our multiitem measure of TU is adapted from apreviously validated scale (Dröge, Claycomb, andGermain, 2003). As explicated in the Analysissection below, these exchange hazard variables areused as predictors for the use of the governancemechanisms as well as for provider opportunism.
Eleven additional contextual variables areincluded to control for their potential influence onthe use of the governance mechanisms, the risk ofopportunism, or both. Three dummy variables areincluded to reflect the functional type of outsourc-ing activity—IT outsourcing (IT), logistics out-sourcing (LOG), and finance/accounting outsourc-ing (FA) —and Other Business Processes was usedas the baseline. Next, data obtained from Yahoo!Finance (http://finance.yahoo.com/) was used todetermine customer firm size (FS) in terms of annualrevenue. Longevity of relationship (LR), measuredin years, represents how long the provider has beenrendering these services to the customer. Three vari-ables serve as instruments in our first-stage analysis(discussed in detail below). First, contract size(CS) represents the annual value of the outsourcingcontract (1=<$1 million; 2= $1.0–$24.9 million;3= $25.0–$49.9 million; 4= $50.0–$99.9 million;5=>$100.0 million). Second, a dummy variable isincluded to represent whether or not the outsourcedactivity was previously performed in house (PI).Third, a three-item scale is included to reflectthe extent to which the buying firm considers theprovider an expert (PE). The measurement for PEis based on a previously validated scale (Benton and
Maloni, 2005). Contract enforcement (CE) capturesthe strength of the legal institutions in each serviceprovider country and is based on the enforcingcontracts ranking in the World Bank’s Doing Busi-ness database (http://doingbusiness.org/). Finally,we include two variables that capture additionalaspects of the relationship between the buyer andprovider. The first, called Cross-cultural Relation-ship (CR), is a dummy variable coded 1 if the buyerand provider are located in different countries; 0otherwise. Inclusion of this variable allows us toisolate the potential influence due to the relation-ship being cross-cultural as opposed to intraculturalin nature. The second, Buyer Subsidiary (BS),reflects the average duration, in years, that thecustomer firm has had subsidiaries in the countrieswhere the service provider operates in support oftheir relationship (0= no subsidiary exists; 10= 10or more years). This serves as an indicator of thebuyers’ experience with the different cultures,which could impact the use and effectivenessof different governance mechanisms. As eachcustomer firm is a public company, the locationsof their subsidiaries were obtained from their10-K filings.
Measurement model evaluation
Confirmatory factor analysis (CFA) using LISREL8.8 was performed on all multiitem scales mea-sured from the provider survey: shirking, poaching,contractual governance, relational governance,provider specificity, and technological uncertainty.The sole multiitem scale measured from the buyer’ssurvey (i.e., provider expert) was not includedin this provider CFA. Analysis of this constructindependently demonstrated strong validity andreliability. The provider CFA exhibited sufficientoverall model fit (RMSEA= 0.077; 𝜒2/df= 1.59;CFI= 0.90; IFI= 0.90; RMR= 0.079). Strongconvergent validity is supported by all itemsloading significantly on their intended construct(all loadings> 0.45 with p-values< 0.001). Theaverage variance extracted (AVE) for each con-struct exceeds the square of its largest interfactorcorrelation; demonstrating significant discriminantvalidity. The Cronbach’s 𝛼 and composite relia-bilities of all constructs ranged from 0.63 to 0.87.These figures are within the routinely acceptedrange for scales with adequate reliability. Modelingopportunism as a meta-construct reflecting shirkingand poaching was empirically validated by the
a If provider operation is in same country as the customer headquarters (i.e., the U.S.), this was coded at the maximum scale value (10 ormore years).
strong second-order factor loadings (p< 0.001)and the Marsh and Hocevar target coefficient(T) equaling 0.994 (Marsh and Hocevar, 1985).Table 1 provides the descriptive statistics forall items (except those detailed in AppendixS1, Table S1) and the Cronbach’s 𝛼 for eachmultiitem scale.
The data for the dependent and independentvariables largely come from the same respondents.Hence, there exists a concern for a common meth-ods (CM) bias. The incorporation of secondary datato measure the culture variables helps diminish theCM concern. Moreover, all of our hypothesizedeffects involve interaction terms. Recent researchhas demonstrated that interaction effects cannot bedue to CM variance (Siemsen, Roth, and Oliveira,2010). However, one cannot definitively say thata CM bias does not exist. Therefore, we useHarman’s one-factor test to assess whether theresearch method has a large effect on the relation-ships observed (Podsakoff and Organ, 1986). Anexploratory factor analysis with all of the manifestitems resulted in the first factor representing amodest 24.6 percent of the total variance. Collec-tively, these data and model characteristics suggestthat the measurement approach is not substantivelybiasing the observed relationships. While we
recognize that some studies assess opportunismfrom the other firm’s perspective, our measurementof opportunism from the perspective of the oppor-tunistic party (i.e., the service provider in our case)follows prior studies in the interorganizationalmanagement literature (Brown, Grzeskowiak, andDev, 2009; John, 1984; Joshi and Arnold, 1997;Ping, 1993; Provan and Skinner, 1989). One mayargue that service providers would be inclined toanswer in a socially desirable manner when askedabout their firm’s tendency to act opportunistically.To address this, respondents were told that theresearch was voluntary and confidential. Follow-ing guidance from consumer behavior literature,the opportunism questions were asked using anindirect questioning technique, which has beenshown to significantly mitigate social desirabilitybias (Fisher, 1993). Finally, we used a statisticalprocedure (Randall and Fernandes, 1991) to controlfor the potential that social desirability bias mayexist and measured the extent to which respondentsfelt uneasy answering each of the opportunismquestions. A factor reflecting these “desirabil-ity” measures (opportunism desirability—OD)is included as a control variable in theanalysis.
Culture and Governance in Outsourcing Relationships 1425
ECONOMETRIC SPECIFICATION ANDANALYSIS
Our primary research objective is to determinethe potential moderating influence of culture onthe effect of contractual and relational governancemechanisms on service provider opportunism.Thus, moderated hierarchical regression is anappropriate analytic technique to partition thevariance cleanly (Cohen et al., 2003). Our econo-metric analysis also accounts for the likelihoodthat the use of contractual mechanisms and the useof relational mechanisms are jointly endogenousdecisions. Specifically, the use of contractual (orrelational) governance may be influenced by theuse of relational (or contractual) governance, theextent of exchange hazards, and other character-istics of the exchange. Contractual and relationalgovernance are also implemented, at least in part,to address opportunism, making them endogenousdeterminants of our dependent variable. Thesefactors could result in potentially biased estimatesfrom a single-stage analysis. To correct for thislikely endogeneity, we follow guidance in theliterature and employ a multistage hierarchicalregression approach (e.g., Poppo, Zhou, andZenger, 2008; Slotegraaf, Moorman, and Inman,2003). In the first stage, each of the two governancemechanisms is regressed on FS, LR, CE, CR,BS, TU, SD, PS and either RG or CG, depend-ing on the dependent variable. In the first-stagemodel with CG as the dependent variable, con-tract size (CS) and previously in house (PI) arethe instrumental variables. Prior work by Poppoand Zenger (2002) uses an instrument similar toCS (budget) for contractual governance. Otherresearch has demonstrated an association betweencontract size and the extensiveness of contractualcontrols (Anderson and Dekker, 2005). Whetherthe outsourced activity was previously performedin house serves as an indicator of the operationalknowledge possessed by the buying firm. Again,extant research has linked such knowledge to moreextensive contractual governance (Argyres andMayer, 2007; Mayer and Salomon, 2006). UsingStata’s estat overid procedure, it was determinedthat CS and PI do not correlate significantly withthe error term in the opportunism equation. Forthe first-stage model with RG as the dependentvariable, provider expert (PE) is the identifyinginstrumental variable. Prior research suggests thatwhen one party to an exchange perceives the other
to be an expert in their field, it facilitates knowledgeand information sharing (Ko, Kirsch, and King,2005), fosters the alignment of norms and values(Brown, Lusch, and Nicholson, 1995), and ispositively associated with committed relationships(Benton and Maloni, 2005). Moreover, our analysisrevealed an insignificant correlation between PEand opportunism (−0.10; p-value= 0.30). Giventhe association between the instrumental variablesand the governance constructs established in theextant literature, along with their observed insignif-icant correlation with opportunism, we deem theinstruments to be appropriate.
The first-stage regressions allow for the deter-mination of residuals for CG (i.e., CG_r) and RG(i.e., RG_r) that are absent the influence of the othergovernance mechanism and other characteristics ofthe exchange. These residuals are used in the sec-ond stage. The interaction terms in the second-stageregressions are also calculated using CG_r andRG_r rather than the original governance variables.In addition to the independent variables used inthe first-stage models (absent the instrumental vari-ables), the second stage also includes dichotomouscontrol variables indicating the functional type ofoutsourcing. Separate stage-two regressions are per-formed for each cultural dimension. The indepen-dent variables in the stage-two regression modelsare sequentially entered in seven blocks to clearlydemonstrate how the groups of variables contributeto the explanation of the variance in opportunism.The first block contains the control variables (OD,IT, LOG, FA, FS, LR, CE, CR, and BS) andexchange hazards (TU, SD, and PS). Subsequently,the variables representing the main effect for con-tractual governance (CG_r), the main effect for rela-tional governance (RG_r), and the interaction ofthese governance mechanisms (CG_r×RG_r) areincorporated in the second and third blocks, respec-tively. The fourth block introduces the individual-ism construct (along with the associated two-wayinteractions with the governance mechanisms) tothe model containing CG and RG and the fifthblock includes the three-way interaction with CGand RG. Similarly, the sixth and seventh blocksinclude uncertainty avoidance and the three-wayinteraction. Our data has multiple observations fromsome customer firms. Using Stata 13.0 we speci-fied the regression models to utilize robust standarderrors to account explicitly for the intragroup cor-relation among the multiple observations from thesame customer. Stata’s robust clustering procedure
adjusts the standard errors using the Huber-Whitemethod known as the “sandwich estimator” of vari-ance. Finally, the largest variance inflation factor(VIF) across all models was 8.1, below the sug-gested threshold of 10 (Cohen et al., 2003). Thus,excessive multicollinearity is not a concern.
RESULTS
The first-stage results (Table 2) for Model 1ademonstrate that relational governance (0.443;p-value< 0.01) has a significant positive effect oncontractual governance. In Model 1b, contractualgovernance has a significant positive effect on rela-tional governance (0.361; p-value< 0.01). Thesefirst-stage results support the argument that con-tractual and relational governance tend to be usedin a complementary manner. Further, the resultsconfirm the need to control for the endogenousinfluence that each governance mechanism has onthe other prior to evaluating their association withopportunism. Finally, the three instrumental vari-ables (CS, PI, and PE) are significant predictors ofthe potentially endogenous governance constructs;validating their strength as instruments.
The full second-stage models—individualism(see Table 3, Models 2d and 2e) and uncertaintyavoidance (Models 2f and 2g)—are used to evaluatethe research hypotheses. We proceed on the assump-tion that contractual and relational governance serveas complements in their effect on provider oppor-tunism but conduct a formal test to confirm this cri-terion (Howell and Dorfman, 1981). In the absenceof the interaction term CG×RG, both main effectsof CG and RG independently are negative andsignificant (see Model 2b). The interaction termCG×RG is not significant in the absence of theculture variables. This is not surprising consider-ing that our theoretical arguments suggest the con-textual importance of culture to this relationship.However, when the interactive effect of contractualand relational governance is added to the modelthat includes the IDV cultural dimension, it toois negative and significant (see Model 2e: −0.230,p< 0.05). These results are clearly not suggestive ofa substitutive relationship but do modestly supportprior research that identifies contractual and rela-tional governance as complementary in their effecton mitigating opportunism.
The hypotheses related to the moderating influ-ence of individualism–collectivism (H1, H2, and
H5) are assessed by the results of Model 2e. Theresults demonstrate that the negative effect ofcontractual governance on opportunism is signifi-cant in high individualism cultures (IDV×CG_r:−0.236; p< 0.05) and less so in more collectivistcultures. Conversely, the interaction between rela-tional governance and individualism (IDV×RG_r)exhibits the opposite effect (0.158; p< 0.05).Relational governance has a stronger mitigatingeffect on opportunism in collectivist cultures.These results support Hypotheses 1 and 2. Model2e also reflects a moderately significant three-wayinteraction between IDV and the joint effect of thetwo governance mechanisms (IDV×CG_r×RG_r:0.194; p−value< 0.10), supporting Hypothesis 5.
The hypotheses related to the moderating influ-ence of uncertainty avoidance (H3, H4, and H6)are assessed by the results of Models 2g. We findthat contractual governance is less effective in highuncertainty avoidance cultures (0.319; p< 0.01).The opposite effect is observed by the interac-tion of uncertainty avoidance and relational gov-ernance (−0.207; p< 0.10). These results supportboth Hypotheses 3 and 4. The three-way interactionbetween UAI and the two governance mechanismsis not statistically significant (UAI×CG_r×RG_r:−0.060; n.s.), thus, the results of Model 2g do notlend support for Hypothesis 6.
DISCUSSION AND IMPLICATIONS
In alignment with our theorizing, we find that con-tractual governance is not universal in its ability tomitigate opportunism: the effectiveness of contrac-tual safeguards is highly contingent upon the cul-tural context of the relationship. These results arepictorially depicted in Figure 3. The desired neg-ative effect of increased contractual governance onopportunism is stronger when the service provider’soperations are located in more individualistic andlow uncertainty avoidance cultures. In these soci-eties, the broader social framework is weak, indi-viduals value personal achievement and materialreward, and people are more comfortable withambiguity, uncertainty, and contingent compensa-tion systems (Doney et al., 1998; Redding et al.,1994). The key implication for outsourcing man-agers is that contractual governance cannot be reliedupon to the same extent across all cultural set-tings. Instead, managers must fully consider thesocietal norms and value systems in countries where
*p< 0.10; **p< 0.05; ***p< 0.01Standard errors are calculated using robust clustering on the customer variable.One-tail t-test used when a directionally specific effect was posited (i.e., instrumental variables and the effects of CG and RG on eachother). Two-tail tests used otherwise.The bold values represent statistically significant values.
their service providers operate. Several examplesfrom our data highlight this intriguing finding.For instance, a large U.S.-based department storesources logistics services from providers in Chinaand Vietnam (low individualism cultures) as wellas the U.S. (a high individualism culture). Acrossthese relationships, this firm tends to emphasizecontractual governance less than relational gover-nance. Our results suggest that this approach isaligned with the Asian providers, but misalignedwith U.S. providers. The opportunism assessmentsin these relationships reflect this expectation. Whenservice providers are located in more individualisticand low uncertainty avoidance cultures, formal con-tracts with strong financial incentive mechanismsshould hold a prominent position in the overall gov-ernance system. Conversely, in collectivist and highuncertainty avoidance cultures, managers need toconsider alternatives to contractual governance asthe desired effect of reducing opportunism will notbe realized, and in fact, opportunism is likely toincrease, as reflected in Figure 3. This particularfinding complements other recent work identifyinginstitutional contexts in which extensive reliance oncontracts may have adverse implications (Zhou and
Xu, 2012). One plausible explanation for this resultis that asking service providers in these cultural set-tings to agree to strong incentive-based contracts isso contrary to their societal norms that it is actuallyperceived as an injustice that needs to be covertlyrectified.
The interactions between relational governanceand the two cultural dimensions reveal that the abil-ity of relational governance to mitigate opportunismis strongest in collectivist and high uncertaintyavoidance societies. The graphs in Figure 4 suggestthat more extensive relational governance is notnecessarily counterproductive in individualistic andlow uncertainty avoidance cultures, but it is simplyineffective in these contexts. People and organiza-tions in collectivist and high uncertainty avoidancesocieties respond more favorably to the social andcooperative nature of relational governance, as wellas the certainty derived from the focus on solidarityand longer-term commitments under relationalgovernance. Again, the managerial implicationis that when the service provider has operationsin these societies, the overall governance systemshould emphasize relational mechanisms suchas socialization initiatives, information sharing,
Culture and Governance in Outsourcing Relationships 1429
(a) (b)
Figure 3. (a,b) Interaction graphs of contractual governance and national culture
cooperative problem solving, and open communi-cation. Doing so should have the most profoundimpact on reducing opportunism. Alternatively,when service providers operate in individualisticand low uncertainty avoidance cultures, these prac-tices will be only marginally effective at mitigatingthe threat of opportunism. Outsourcing managersshould seek alternative means of addressing thesebehavioral concerns.
Just as culture moderates the effectiveness ofeach mechanism individually, it is plausible thatculture moderates their joint effectiveness as well(i.e., a three-way interaction). Our results indicatethat the interactive effect of the two governancemechanisms is significantly moderated by theindividualism–collectivism cultural dimensionbut not significantly moderated by uncertaintyavoidance. To further our understanding of theinfluence of individualism–collectivism on thejoint effect of contractual and relational gover-nance, the three-way interaction is pictoriallypresented in Figure 5. In this graph, comple-mentarity between contractual and relationalgovernance would be indicated by “High RG,High CG” (line 1), which has lower levels ofopportunism than do any of the three othercombinations. These results depict an intricaterelationship between individualism–collectivismand the effectiveness of these two governancemechanisms. There are a few key takeaways fromthese findings. First, we generally observe that thecombination of contractual and relational gover-nance is an effective strategy; however, it is knownthat there is a cost to deploying each governancemechanism (Das and Teng, 1998). Therefore,outsourcing firms need to balance the potentiallysmall benefit of using both forms of governanceagainst their cost of deployment. In Figure 5, wefind that the incremental benefit of using the twogovernance mechanisms jointly, rather than relying
on a single well-aligned governance approach,may be modest. Bearing in mind the cost, the beststrategy for outsourcers may be using a contractualapproach with service providers in individualisticcultures and a relational approach with serviceproviders in collectivist cultures even though usingboth has a more pronounced effect. Next, it appearsthat contractual governance should not be usedas the sole governance mechanism in collectivistcultures. Not only is this approach ineffective, butit seems to be in such contrast to societal normsthat service providers perceive it as an injusticethey seek to right through more pronounced oppor-tunistic behavior. Thus, firms wishing to employcontractual governance should also incorporaterelational mechanisms. Similarly, we do not adviseoutsourcing firms to rely only on relational gov-ernance with service providers in individualisticcultures. Perhaps members of these societieshave a greater proclivity toward opportunism (onaverage), and the trusting nature of high relationalgovernance is dangerous without correspondingeconomic incentives to control provider behavior.It is important to note, however, that these insightsand our interpretations are with regard to the effectof governance on the risk of opportunism. Whilebeyond the scope of this study, we fully recognizethat strong relational governance and well-specifiedcontracts may benefit the exchange in other wayssuch as improved interfirm coordination.
CONCLUSION
Our study makes multiple theoretical contributionsto the outsourcing literature. First, the theoreticalframing of our model is grounded in Williamson’s(1991) “shift parameter” analysis. The resultsherein contribute much needed empirical evidencesupporting his assertion that the institutional
Figure 4. (a,b) Interaction graphs of relational governance and national culture
Figure 5. Three-way interaction of CG, RG, and theindividualism–collectivism cultural dimension
environment, as it relates to culture, impacts (i.e.,shifts) the comparative effectiveness of governancemechanisms. Previously, empirical evidence sup-porting this framework has been limited (Oxley,1999). Second, our findings complement prior workon the extent to which governance misalignmentsmatter (Argyres and Bigelow, 2007; Nickersonand Silverman, 2003; Sampson, 2004). We findthat firms relying on CG in low IDV or high UAIenvironments and firms relying on RG in highIDV or low UAI environments, may experiencedisappointing outcomes in terms of the mitigationof opportunism. Finally, the preponderance of priorstudies exploring the individual and joint effectsof CG and RG in international outsourcing, do sousing a sample of supplier relationships based ina single country such as China (Li et al., 2010;Liu et al., 2009) or India (Jayaraman et al., 2013).These prior contributions offer valuable insightsinto the effectiveness of governance mechanismswithin these specific countries. Yet, their empiricalsetting lacks sufficient institutional heterogeneity,restricting their ability to speak more broadlyabout the influence of specific societal factors (e.g.,culture). This leads Zhou and Xu (2012: 690) tonote that “a multicountry research setting is neededto better assess the effects of alternative governancemechanisms.” Directly to this point, we sought a
context under which domestic firms outsource to awide variety of countries, which inherently possessa wide array of cultures. This is theoretically andpractically interesting because it provides us withbroader insight about certain cultural constructs(i.e., IDV and UAI), while also controlling for othergermane factors.
There are limitations in our study that presentopportunities for additional research on the man-agement of international outsourcing relationships.Also, the nature of our results opens the door forfuture studies to explore alternative explanationsfor our findings. First, the cross-sectional natureof our data limits the extent to which assertionsof causality can be made. Additionally, althoughwe did not find any evidence of response bias, thevalidity of our inferences should be judged in lightof the modest response rate and sample size. Futureresearch can seek to further validate our findingsusing larger samples and longitudinal data allow-ing for stronger claims of causality. The currentstudy reveals the important role that national cultureplays in moderating the effectiveness of governancemechanisms, but alternative explanations are worthconsidering. While we control for several factors,it is also plausible that other national institutions(e.g., strength of contract law, interpretation of lawsand the legal environment, regulatory environment,protection of intellectual property, etc.) influencethe utility of interfirm management practices.Our model includes the buyer subsidiary variableto control for the buying firm’s prior experiencewith the provider’s culture; however, such culturalexposure can be gained through previous sourcingactivity in these or similar cultures as well. Inaccordance with TCT, our analysis controls forthe effects of technological uncertainty, buyerswitching difficulty, provider specificity, and otherfactors. However, the TCT literature recognizes
Culture and Governance in Outsourcing Relationships 1431
additional exchange hazards (e.g., site specificity,temporal specificity) that should be considered inexploring alternative explanations for the intricaterelationship between culture, governance, andopportunism. Similarly, while switching difficultyand provider specificity allow us in part to isolatethe potential influence of interfirm dependency onopportunism, closely related factors (e.g., supplymarket dynamics, operational disruption risk, etc.)may also contribute to the relative dependencyand vulnerability of both firms. These providealternative explanations for our findings that couldnot be ruled out without further research. In thecurrent study, all customer organizations are basedin the U.S. Future research could extend this lineof inquiry by studying relationships involvingnon-U.S. customer firms. Our theoretical logicis based on the culture in the provider’s countryshifting the effectiveness of CG and RG in miti-gating opportunism and not the buyer’s culture oreven cultural differences. Additionally, our modelcontrols for whether the outsourcing relationshipis cross-cultural. Thus, we would not anticipatea change in our findings with customer firms innon-U.S. countries. Yet, this is an empirical ques-tion worthy of future investigation. We believe it isalso important to study opportunistic behavior fromthe standpoint of the buying firm: we are not famil-iar with any study that simultaneously accountsfor double-sided opportunistic behavior. Finally,the focus herein was on the influence of culture onthe ability to control provider behavior. However,another managerial imperative is coordinating withthe provider (Dibbern, Winkler, and Heinzl, 2008).Future studies should evaluate the impact of cultureon the practices used by customer organizations tocoordinate with service providers. We encouragescholars to pursue these and other research ques-tions to continue to advance our understanding ofthe most appropriate means of managing complexglobal outsourcing arrangements.
REFERENCES
Anderson SW, Dekker HC. 2005. Management controlfor market transactions: the relation between transac-tion characteristics, incomplete contract design, andsubsequent performance. Management Science 51(12):1734–1752.
Angst CM, Agarwal R. 2009. Adoption of electronichealth records in the presence of privacy concerns: theelaboration likelihood model and individual persuasion.MIS Quarterly 33(2): 339–370.
Argyres N, Bigelow L. 2007. Does transactionmisalignment matter for firm survival at all stages ofthe industry life cycle? Management Science 53(8):1332–1344.
Argyres N, Mayer KJ. 2007. Contract design as a firmcapability: an integration of learning and transactioncost perspectives. Academy of Management Review32(4): 1060–1077.
Armstrong JS, Overton TS. 1977. Estimating nonresponsebias in mail surveys. Journal of Marketing Research14(3): 396–402.
Aron R, Clemons EK, Reddi S. 2005. Just right outsourc-ing: understanding and managing risk. Journal of Man-agement Information Systems 22(2): 37–55.
Barthélemy J, Quélin BV. 2006. Complexity of outsourc-ing contracts and ex post transaction costs: an empiricalinvestigation. Journal of Management Studies 43(8):1775–1797.
Beatty RP, Zajac EJ. 1994. Managerial incentives, moni-toring, and risk bearing: a study of executive compen-sation, ownership, and board structure in initial pub-lic offerings. Administrative Science Quarterly 39(2):313–335.
Benton WC, Maloni M. 2005. The influence of powerdriven buyer/seller relationships on supply chain sat-isfaction. Journal of Operations Management 23(1):1–22.
Bowman EH. 1963. Consistency and optimality in man-agerial decision making. Management Science 9(2):310–321.
Bradach JL, Eccles RG. 1989. Price, authority, and trust:from ideal types to plural forms. Annual Review ofSociology 15(1): 97–118.
Brett JM, Okumura T. 1998. Inter-and intracultural nego-tiation: US and Japanese negotiators. Academy of Man-agement Journal 41(5): 495–510.
Brown JR, Grzeskowiak S, Dev CS. 2009. Using influencestrategies to reduce marketing channel opportunism:the moderating effect of relational norms. MarketingLetters 20(2): 139–154.
Brown JR, Lusch RF, Nicholson CY. 1995. Power andrelationship commitment: their impact on marketingchannel member performance. Journal of Retailing71(4): 363–392.
Brown Johnson N, Droege S. 2004. Reflections on thegeneralization of agency theory: cross-cultural consid-erations. Human Resource Management Review 14(3):325–335.
Buchanan L. 1992. Vertical trade relationships: the role ofdependence and symmetry in attaining organizationalgoals. Journal of Marketing Research 29(1): 65–75.
Cannon JP, Doney PM, Mullen MR, Petersen KJ. 2010.Building long-term orientation in buyer–supplier rela-tionships: the moderating role of culture. Journal ofOperations Management 28(6): 506–521.
Cannon JP, Perreault WD Jr. 1999. Buyer-seller rela-tionships in business markets. Journal of MarketingResearch 36(4): 439–460.
Cohen J, Cohen P, West SG, Aiken LS. 2003. AppliedMultiple Regression/Correlation Analysis for theBehavioral Sciences. L. Erlbaum Associates: London,UK.
Couto V, Mani M, Lewin A, Peeters C. 2006. The Glob-alization of White-Collar Work. Booz Allen Hamilton:Chicago.
Crocker KJ, Reynolds KJ. 1993. The efficiency of incom-plete contracts: an empirical analysis of air forceengine procurement. RAND Journal of Economics24(1): 126–146.
Crossland C, Hambrick DC. 2007. How national systemsdiffer in their constraints on corporate executives: astudy of CEO effects in three countries. StrategicManagement Journal 28(8): 767–789.
Crossland C, Hambrick DC. 2011. Differences in manage-rial discretion across countries: how nation-level insti-tutions affect the degree to which CEOs matter. Strate-gic Management Journal 32(8): 797–819.
Das TK, Teng B-S. 1998. Between trust and control:developing confidence in partner cooperation alliances.Academy of Management Review 23(3): 491–512.
David RJ, Han S-K. 2004. A systematic assessment ofthe empirical support for transaction cost economics.Strategic Management Journal 25(1): 39–58.
Davis JH, Schoorman FD, Donaldson L. 1997. Towarda stewardship theory of management. Academy ofManagement Review 22(1): 20–47.
Dawson JF, Richter AW. 2006. Probing three-way inter-actions in moderated multiple regression: developmentand application of a slope difference test. Journal ofApplied Psychology 91(4): 917–926.
Dibbern J, Winkler J, Heinzl A. 2008. Explaining vari-ations in client extra costs between software projectsoffshored to India. MIS Quarterly 32(2): 333–366.
Doney PM, Cannon JP, Mullen MR. 1998. Understand-ing the influence of national culture on the develop-ment of trust. Academy of Management Review 23(4):601–620.
Dröge C, Claycomb C, Germain R. 2003. Does knowledgemediate the effect of context on performance? Someinitial evidence. Decision Sciences 34(3): 541–568.
Dyer J, Singh H. 1998. The relational view: cooperativestrategy and sources of interorganizational competitiveadvantage. Academy of Management Review 23(4):660–679.
Eastwood G. 2006. The BPO Market Outlook: ChangingCompetitive Dynamics, Key Players, and Best Prac-tices. Business Insights Ltd.: London.
Eisenhardt KM. 1985. Control: organizational andeconomic approaches. Management Science 31(2):134–149.
Eisenhardt KM. 1989. Agency theory: an assessment andreview. Academy of Management Review 14(1): 57–74.
Ellram LM, Tate WL, Billington C. 2008. Offshore out-sourcing of professional services: a transaction costeconomics perspective. Journal of Operations Manage-ment 26(2): 148–163.
Fisher RJ. 1993. Social desirability bias and the validityof indirect questioning. Journal of Consumer Research20(2): 303–315.
Gelfand M, Bhawuk D, Nishii L, Bechtold D. 2004. Indi-vidualism and collectivism. In Culture, Leadership,and Organizations: The GLOBE Study of 62 Soci-eties, House R, Hanges P, Javidan M, Dorfman P,
Gupta V (eds). Sage Publications: Thousand Oaks, CA;437–502.
Ghoshal S, Moran P. 1996. Bad for practice: a critique ofthe transaction cost theory. Academy of ManagementReview 21(1): 13–47.
Giannetti M, Yafeh Y. 2011. Do cultural differencesbetween contracting parties matter? Evidence fromsyndicated bank loans. Management Science 58(2):365–383.
Goo J, Kishore R, Rao HR, Nam K. 2009. The roleof service level agreements in relational managementof information technology outsourcing: an empiricalstudy. MIS Quarterly 33(1): 119–145.
Griffith DA, Myers MB, Harvey MG. 2006. An investi-gation of national culture’s influence on relationshipand knowledge resources in interorganizational rela-tionships between Japan and the United States. Journalof International Marketing 14(3): 1–32.
Gulati R. 1995. Does familiarity breed trust? The impli-cations of repeated ties for contractual choices inalliances. Academy of Management Journal 38(1):85–112.
Halford GS, Baker R, McCredden JE, Bain JD. 2005. Howmany variables can humans process? PsychologicalScience 16(1): 70–76.
Handley SM, Benton WC. 2012. The influence ofexchange hazards and power on opportunism inoutsourcing relationships. Journal of OperationsManagement 30(1–2): 55–68.
Heide JB, John G. 1988. The role of dependence balancingin safeguarding transaction-specific assets in conven-tional channels. Journal of Marketing 52(1): 20–35.
Heide JB, John G. 1992. Do norms matter in marketingrelationships? Journal of Marketing 56(2): 32–44.
Hewett K, Bearden WO. 2001. Dependence, trust, and rela-tional behavior on the part of foreign subsidiary market-ing operations: implications for managing global mar-keting operations. Journal of Marketing 65(4): 51–66.
Hofstede G. 1980. Culture’s Consequences, InternationalDifferences in Work-Related Values (Abridged edn).Sage Publications: Beverly Hills, CA.
Hofstede G. 1985. The interaction between national andorganizational value systems. Journal of ManagementStudies 22(4): 347–357.
Holmström B, Roberts J. 1998. The boundaries of thefirm revisited. Journal of Economic Perspectives 12(4):73–94.
Homburg C, Cannon JP, Krohmer H, Kiedaisch I. 2009.Governance of international business relationships: across-cultural study on alternative governance modes.Journal of International Marketing 17(3): 1–20.
Howell JP, Dorfman PW. 1981. Substitutes for leadership:test of a construct. Academy of Management Journal24(4): 714–728.
Jap SD, Anderson E. 2003. Safeguarding interorganiza-tional performance and continuity under Ex post oppor-tunism. Management Science 49: 1684–1701.
Jap SD, Ganesan S. 2000. Control mechanisms and therelationship life cycle: implications for safeguardingspecific investments and developing commitment. Jour-nal of Marketing Research 37(2): 227–245.
Culture and Governance in Outsourcing Relationships 1433
Jayaraman V, Narayanan S, Luo Y. 2013. Offshoring busi-ness process services and governance control mech-anisms: an examination of service providers fromIndia. Production and Operations Management 22(2):314–334.
John G. 1984. An empirical investigation of someantecedents of opportunism in a marketing channel.Journal of Marketing Research 21(3): 278–289.
Joshi AW, Arnold SJ. 1997. The impact of buyer depen-dence on buyer opportunism in buyer–supplier rela-tionships: the moderating role of relational norms. Psy-chology & Marketing 14(8): 823–845.
Kale SH, Barnes JW. 1992. Understanding the domainof cross-national buyer-seller interactions. Journal ofInternational Business Studies 23(1): 101–132.
Kalnins A, Mayer KJ. 2004. Relationships and hybrid con-tracts: an analysis of contract choice in informationtechnology. Journal of Law, Economics, and Organi-zation 20(1): 207–229.
Kessler JB, Leider S. 2012. Norms and contracting. Man-agement Science 58(1): 62–77.
Kirsch LJ, Sambamurthy V, Dong-Gil K, Purvis RL. 2002.Controlling information systems development projects:the view from the client. Management Science 48(4):484–498.
Klein B, Crawford RG, Alchian AA. 1978. Vertical inte-gration, appropriable rents, and the competitive con-tracting process. Journal of Law and Economics 21(2):297–326.
Ko D-G, Kirsch LJ, King WR. 2005. Antecedents ofknowledge transfer from consultants to clients in enter-prise system implementations. MIS Quarterly 29(1):59–85.
Lafontaine F, Shaw KL. 1999. The dynamics of franchisecontracting: evidence from panel data. Journal of Polit-ical Economy 107(5): 1041–1080.
Landis KM, Mishra S, Porrello K. 2005. Calling a Changein the Outsourcing Market: The Realities for theWorld’s Largest Organizations. Deloitte Consulting:New York.
Li Y, Xie E, Teo H-H, Peng MW. 2010. Formal con-trol and social control in domestic and internationalbuyer–supplier relationships. Journal of OperationsManagement 28(4): 333–344.
Liu Y, Luo Y, Liu T. 2009. Governing buyer–supplierrelationships through transactional and relational mech-anisms: evidence from China. Journal of OperationsManagement 27(4): 294–309.
Lumineau F, Henderson JE. 2012. The influence of rela-tional experience and contractual governance on thenegotiation strategy in buyer-supplier disputes. Journalof Operations Management 30(5): 382–395.
Lumineau F, Malhotra D. 2011. Shadow of the contract:how contract structure shapes interfirm dispute resolu-tion. Strategic Management Journal 32(5): 532–555.
Luo Y. 2001. Antecedents and consequences of per-sonal attachment in cross-cultural cooperative ventures.Administrative Science Quarterly 46(2): 177–201.
Luque SD, Javidan M. 2004. Uncertainty avoidance. InCulture, Leadership, and Organizations: The GLOBE
Study of 62 Societies, House R, Hanges P, JavidanM, Dorfman P, Gupta V (eds). Sage Publications:Thousand Oaks, CA; 602–644.
Macaulay S. 1963. Non-contractural relations in business:a preliminary study. American Sociological Review28(1): 55–67.
Macneil IR. 1985. Relational contract: what we do and donot know. Wisconsin Law Review 1985(1): 483–526.
Malhotra D, Murnighan JK. 2002. The effects of contractson interpersonal trust. Administrative Science Quar-terly 47(3): 534–559.
Marsh HW, Hocevar D. 1985. Application of confirmatoryfactor analysis to the study of self-concept: first-andhigher order factor models and their invariance acrossgroups. Psychological Bulletin 97(3): 562–582.
Masten SE. 1993. Transaction costs, mistakes, and perfor-mance: assessing the importance of governance. Man-agerial & Decision Economics 14(2): 119–129.
Masten SE, Crocker KJ. 1985. Efficient adaptation inlong-term contracts: take-or-pay provisions for naturalgas. American Economic Review 75(5): 1083–1093.
Mayer KJ, Salomon RM. 2006. Capabilities, contractualhazards, and governance: integrating resource-basedand transaction cost perspectives. Academy of Manage-ment Journal 49(5): 942–959.
Menor LJ, Roth AV. 2007. New service developmentcompetence in retail banking: construct developmentand measurement validation. Journal of OperationsManagement 25(4): 825–846.
Naor M, Linderman K, Schroeder R. 2010. The global-ization of operations in Eastern and Western countries:unpacking the relationship between national and orga-nizational culture and its impact on manufacturing per-formance. Journal of Operations Management 28(3):194–205.
Nickerson JA, Silverman BS. 2003. Why firms wantto organize efficiently and what keeps them fromdoing so: Inappropriate governance, performance, andadaptation in a deregulated industry. AdministrativeScience Quarterly 48(3): 433–465.
Oxley JE. 1999. Institutional environment and the mech-anisms of governance: the impact of intellectual prop-erty protection on the structure of inter-firm alliances.Journal of Economic Behavior & Organization 38(3):283–309.
Park SH, Ungson GR. 1997. The effect of national culture,organizational complementarity, and economic motiva-tion on joint venture dissolution. Academy of Manage-ment Journal 40(2): 279–307.
Ping RA. 1993. The effects of satisfaction and struc-tural constraints on retailer exiting, voice, loyalty,opportunism, and neglect. Journal of Retailing 69(3):320–352.
Podsakoff PM, Organ D. 1986. Self-reports in organi-zational research: problems and prospects. Journal ofManagement 12(4): 531–544.
Poppo L, Zenger T. 2002. Do formal contracts andrelational governance function as substitutes orcomplements? Strategic Management Journal 23(8):707–725.
Poppo L, Zhou KZ, Zenger TR. 2008. Examining theconditional limits of relational governance: specializedassets, performance ambiguity, and long-standing ties.Journal of Management Studies 45(7): 1195–1216.
Provan KG, Skinner SJ. 1989. Interorganizational depen-dence and control as predictors of opportunism indealer-supplier relations. Academy of ManagementJournal 32(1): 202–212.
Ralston DA, Holt DH, Terpstra RH, Yu K-C. 1997. Theimpact of national culture and economic ideology onmanagerial work values: a study of the United States,Russia, Japan, and China. Journal of InternationalBusiness Studies 28(1): 177–207.
Randall DM, Fernandes MF. 1991. The social desirabilityresponse bias in ethics research. Journal of BusinessEthics 10(11): 805–817.
Redding SG, Norman A, Schlander A. 1994. The nature ofindividual attachment to the organization: a review ofEast Asian variations. In Handbook of Industrial andOrganizational Psychology (2nd edn), Triandis HC,Dunnette M, Hough LM (eds). Consulting Psycholo-gists Press: Palo Alto, CA; 647–688.
Rigdon M. 2009. Trust and reciprocity in incentive con-tracting. Journal of Economic Behavior & Organiza-tion 70(1): 93–105.
Ring PS, Van de Ven AH. 1992. Structuring cooperativerelationships between organizations. Strategic Manage-ment Journal 13(7): 483–498.
Ring PS, Van de Ven AH. 1994. Developmental pro-cesses of cooperative interorganizational relationships.Academy of Management Review 19(1): 90–118.
Robinson P, Lowes P, Loughran C, Moller P, ShieldsG, Klein E. 2008. Why Settle for Less? OutsourcingReport. Deloitte Consulting: London.
Sampson RC. 2004. The cost of misaligned governancein R&D alliances. Journal of Law, Economics, andOrganization 20(2): 484–526.
Shane S. 1994. The effect of national culture on thechoice between licensing and direct foreign investment.Strategic Management Journal 15(8): 627–642.
Shenkar O, Ronen S. 1987. Structure and importance ofwork goals among managers in the People’s Republicof China. Academy of Management Journal 30(3):564–576.
Siegel JI, Licht AN, Schwartz SH. 2011. Egalitarianismand International Investment. Journal of Financial Eco-nomics 102(3): 621–642.
Siemsen E, Roth A, Oliveira P. 2010. Common methodbias in regression models with linear, quadratic, andinteraction effects. Organizational Research Methods13(3): 456–476.
Slotegraaf RJ, Moorman C, Inman JJ. 2003. The role offirm resources in returns to market deployment. Journalof Marketing Research 40(3): 295–309.
Sousa CMP, Bradley F. 2008. Cultural distance andpsychic distance: refinements in conceptualisationand measurement. Journal of Marketing Management24(5): 467–488.
Tosi HL, Greckhamer T. 2004. Culture and CEO compen-sation. Organization Science 15(6): 657–670.
Tversky A, Kahneman D. 1974. Judgment under uncer-tainty: heuristics and biases. Science 185(4157):1124–1131.
Ueno S, Sekaran U. 1992. The influence of culture on bud-get control practices in the USA and Japan: an empiricalstudy. Journal of International Business Studies 23(4):659–674.
Uzzi B. 1997. Social structure and competition in interfirmnetworks: the paradox of embeddedness. Administra-tive Science Quarterly 42(1): 35–67.
Wathne KH, Heide JB. 2000. Opportunism in interfirmrelationships: forms, outcomes, and solutions. Journalof Marketing 64(4): 36–51.
Williamson OE. 1975. Markets and Hierarchies: Analysisand Antitrust Implications. Free Press: New York.
Williamson OE. 1979. Transaction cost economics: thegovernance of contractual relations. Journal of Law andEconomics 22(2): 233–261.
Williamson OE. 1991. Comparative economic organi-zation: the analysis of discrete structural alternatives.Administrative Science Quarterly 36(2): 269–296.
Williamson OE. 2000. The new institutional economics:taking stock, looking ahead. Journal of Economic Lit-erature 38(3): 595–613.
Wuyts S, Geyskens I. 2005. The formation ofbuyer-supplier relationships: detailed contract draftingand close partner selection. Journal of Marketing69(4): 103–117.
Yeniyurt S, Townsend JD, Cavusgil ST, Ghauri PN. 2009.Mimetic and experiential effects in international mar-keting alliance formations of US pharmaceuticals firms:an event history analysis. Journal of International Busi-ness Studies 40(2): 301–320.
Zaheer S, Zaheer A. 2006. Trust across borders. Journal ofInternational Business Studies 37(1): 21–29.
Zhou KZ, Poppo L. 2010. Exchange hazards, relationalreliability, and contracts in China: the contingent role oflegal enforceability. Journal of International BusinessStudies 41(5): 861–881.
Zhou KZ, Poppo L, Yang Z. 2008. Relational ties orcustomized contracts? An examination of alternativegovernance choices in China. Journal of InternationalBusiness Studies 39(3): 526–534.
Zhou KZ, Xu D. 2012. How foreign firms curtail localsupplier opportunism in China? Detailed contracts,centralized control, and relational governance. Journalof International Business Studies 43: 677–692.
SUPPORTING INFORMATION
Additional supporting information may be foundin the online version of this article:
