MNE institutional advantage: How subunits shape, transpose and … · 2017. 9. 9. · ary...

MNE institutional advantage: How subunitsshape, transpose and evade host countryinstitutions

Patrick Regnér1 andJesper Edman2

1Department of Marketing and Strategy,Stockholm School of Economics, Stockholm,Sweden; 2Graduate School of InternationalCorporate Strategy, Hitotsubashi University,Tokyo, Japan

Correspondence:P Regnér, Department of Marketing andStrategy, Stockholm School of Economics,Box 6501, Stockholm S-113 83, Sweden.Tel: +46 (0)87369516;Fax: +46 (0) 31 99 27;email: [email protected]

Received: 4 May 2011Revised: 8 October 2013Accepted: 22 October 2013Online publication date: 19 December 2013

AbstractScholars increasingly emphasize the impact of institutions on multinationalenterprises (MNEs), but the opposite relationship has attracted less research –that is, MNE agency in relation to institutions. Based on a comparative case studyof six MNEs from the United States and Sweden, this paper remedies this. Itexplores and explicates MNE subunits’ strategic responses to host country institu-tional constraints and opportunities in five different regions. A new-institutionalapproach is adopted, which allows for an investigation of MNE subunit agency inrelation to normative and cognitive institutions, as well as regulative ones. Thisfine-grained analysis reveals not only what kinds of responses MNE subunitsinvoke, but why and how they are able to respond. We identify four strategicresponses by which subunits shape, transpose and evade institutions in the pursuitof competitive advantage: Innovation, Arbitrage, Circumvention and Adaptation.These responses are driven by three key enablers: multinationality, foreignness andinstitutional ambiguity – that serve to enhance and heighten three mechanisms:reflexivity, role expectations and resources. By linking the enablers and the mechan-isms to specific types of strategic responses in a framework and typology, thepaper not only contributes to emerging research on the interplay between MNEs,institutions and strategy, but to strategy practice.Journal of International Business Studies (2014) 45, 275–302. doi:10.1057/jibs.2013.66

Keywords: institutional context; competitive advantage; case-theoretic approaches;multinational corporations (MNCs) and enterprises (MNEs); neoinstitutional theory;strategy and business strategy

INTRODUCTIONApplying both economic (North, 1990) and organizational (Scott,2008) institutional theories, international management research haswitnessed an increasing interest in the impact of institutionson multinational enterprise (MNE) activities and behaviors(Cantwell, Dunning, & Lundan, 2010; Henisz & Swaminathan,2008; Jackson & Deeg, 2008). Recognizing that MNEs are uniquein simultaneously operating across multiple institutional environ-ments, researchers have investigated how country-level institu-tions and institutional distance affect entry mode behavior (Davis,Desai, & Francis, 2000; Lu, 2002; Meyer, Estrin, Bhaumik, & Peng,2009), performance (Chacar, Newburry, & Vissa, 2010), learning(Ghoshal, 1988; Henisz & Delios, 2002), legitimacy (Chan & Makino,

Journal of International Business Studies (2014) 45, 275–302© 2014 Academy of International Business All rights reserved 0047-2506www.jibs.net

2007; Kostova & Zaheer, 1999) and the transfer ofstrategic practices (Kostova & Roth, 2002).While these works have opened up an exciting

new area of inquiry, scholars have also lamentedthe absence of agency in institutional analyses ofthe MNE (Kostova, Roth, & Dacin, 2008; Phillips &Tracey, 2009; Phillips, Tracey, & Karra, 2009;Saka-Helmhout & Geppert, 2011). Even as theyhighlight the important role played by institutionalcontexts, extant studies say relatively little abouthow MNEs and their subunits respond strategicallyto host country institutions. Nor do they identifythe drivers that underlie such response strategies.Although several studies investigate the impactof institutional factors on entry mode strategies(Brouthers, 2002; Xu & Shenkar, 2002), these studiesgenerally maintain a focus on static structure (e.g.,joint venture vs wholly owned subsidiaries) and anarrow emphasis on how institutional constraintscan be reduced and legitimacy increased. Moreover,responses to institutions have predominantly beenconsidered on the level of the MNE and corporatemanagement, as opposed to subunit and subsidiarymanager level (cf. Xu & Shenkar, 2002). As a result,there is a dearth of studies that investigate howsubsidiaries and subunits respond to local institu-tional contexts.The limited number of studies that do address

MNE subsidiary agency in relation to institutionshave focused on the adaptation of home-countrypractices to local institutional settings (Gooderham,Nordhaug, & Ringdal, 1998; Kostova & Roth, 2002)or on processes of translation underlying globaliza-tion and trans-national institution building (Djelic& Quack, 2003). There has, however, been little ifany explicit emphasis on how such agency may linkto the advantages of the MNE in host countries. Thisis despite recent research suggesting not only thatMNEs may have advantages in responding to insti-tutions compared with domestic firms (Kostovaet al., 2008), but that experimental processes inMNE subsidiaries are where the main drivers of suchresponses lie (Cantwell et al., 2010). It would thus beof great interest to understand the unique traits thatprovide MNEs with an institutional advantage overdomestic firms.Against this background and various calls for

research on MNE responses in relation to institutions(Cantwell et al., 2010; Kostova et al., 2008, 2009;Phillips et al., 2009), this paper identifies specificMNE strategic responses to institutions, as well as theirunderlying mechanisms and enablers. The paper spe-cifically addresses the following questions: Do MNE

subunits employ strategic responses to institutions? Ifso, what do these look like, that is, what are thedifferent types of strategic responses? Why are MNEsable to provide these responses, that is, what are theenablers? How do MNEs respond to institutions, thatis, what are the primary mechanisms? Finally, arethese responses, enablers and mechanisms unique toMNEs, composing a particular advantage over domes-tic actors?We address these questions by exploring how the

subunits of six different MNEs from the United Statesand Sweden responded to local institutions in fivedifferent regions. Our study contributes to interna-tional management by identifying not only in whatway MNE subunits respond to institutional settings,but also delineating the underlying enablers andmechanisms of such responses. We find that whileMNEs are exposed to highly complex pressure acrossregulative, normative and cognitive institutions, theyalso respond forcefully. We identify four distinctMNE strategic responses, implemented using threespecific mechanisms. These mechanisms are enabledboth by the MNE’s unique social position across andwithin countries, as well as the degree of ambiguityin the local organizational field. By identifying theseenablers and the mechanisms they engender, ourwork breaks new ground in identifying how MNEsgarner competitive advantage vis-à-vis local actors, byshaping, transposing and evading institutions.In addition to contributing to international man-

agement, our study also offers insights for organiza-tion theorists by extending research on agency inrelation to institutions in a multinational context. Indoing so we delineate conditions for enhanced strate-gic responses to institutions and also illustrate possibleboundary conditions. We contribute to strategic man-agement by illustrating the importance of managinginstitutional and social factors for achieving competi-tive advantage, and by delineating what the under-lying mechanisms of such management processesmay look like. By building on recent developments inneoinstitutional theory and integrating these withinsights on MNE strategic responses, we seek tostrengthen the linkage between international man-agement studies and institutional and strategic man-agement theories (cf. Cantwell & Brannen, 2011).

THEORYAND RESEARCH OBJECTIVE

Institutions and MNEsCentral to the recent focus on institutions amonginternational management scholars is the realizationthat context has a direct influence on the strategies

MNE institutional advantage Patrick Regnér and Jesper Edman276

Journal of International Business Studies

of MNEs (Jackson & Deeg, 2008; Mudambi &Navarra, 2002; Peng, Wang, & Jiang, 2008). Whileregulatory, normative and cultural-cognitive institu-tions arguably shape the operating conditions facedby any firm (Scott, 2008), scholars have suggestedMNEs may be particularly exposed to these effectsbecause they span multiple contexts and nationalboundaries (Westney, 1993; Morgan, Kristensen, &Whitley, 2001). Consequently, multiple studieshave explored how local institutional conditions, aswell as the institutional distance between home andhost country, affect the competitive strategies ofMNEs (Boubakri, Mansi, & Saffar, 2013; Chung &Beamish, 2005; Henisz & Delios, 2001; Kostova,1997; Kostova & Roth, 2002; Lu, 2002; Meyer &Nguyen, 2005; Xu & Shenkar, 2002).While these studies make important contributions

to our understanding of how MNEs relate to localinstitutions, areas for further inquiry remain. First,many extant studies have operationalized institutionson the country level, focusing in particular on thepolitical, legal and societal aspects of institutions(Brockman, Rui, & Zou, 2013; Meyer, 2001; Penget al., 2008; Wan, 2005; Jackson & Deeg, 2008). Thismacro level of analysis, however, fails to take intoaccount the industry-specific normative and cogni-tive institutional challenges that MNEs often encoun-ter when entering host countries (Orr & Scott, 2008).Examining institutions on the level of the countrythus risks excluding important insights that can begained by analyzing the specific context withinwhich the MNE subunit conducts business (Phillipset al., 2009; Saka-Helmhout & Geppert, 2011).Second, drawing on the classic tenets of institu-

tional economics (North, 1990; 2005) and new-institutionalism in organization theory (DiMaggio &Powell, 1983; Meyer & Rowan, 1977), a majority ofstudies view institutions as constraints that increaseoperating costs for the MNE (Eden & Miller, 2004;Rosenzweig & Nohria, 1994). Consequently, scholarshave predominantly focused on how firms reduceinstitutionally driven transaction costs through theirchoice of country markets, ownership strategies andentry modes (Brouthers & Brouthers, 2000; Chan &Makino, 2007; Meyer et al., 2009; Xu & Shenkar,2002). Viewing institutions as entrenched and lar-gely immobile facets of the environment, extantresearch has particularly emphasized how firmseither adapt to or offset pressures for conformitywith firm-specific advantages (Zaheer, 1995). Therehas been far less emphasis on howMNEs proactivelyengage with and strategize around their institutionalenvironment.

In one of the few exceptions to the dominant viewof institutions as inertial macro-level constraints,Saka-Helmhout and Geppert (2011) explore varyingforms of active agency employed by local subsidiarymanagers introducing new products and practicesinto host country markets. This study is importantboth because it identifies an MNE advantage inresponding to institutions, and because it places thelocus of this response at the subunit level, therebyadding to the growing recognition that local subunitsplay important roles in MNE activities in general(Rugman & Verbeke, 2001; Andersson, Forsgren, &Holm, 2002, 2007; Almeida & Phene, 2004) and in thedevelopment of firm-specific advantages in particular(Birkinshaw & Morrison, 1996; Ghoshal & Nohria,1989; Gupta & Govindarajan, 1994; Roth &Morrison,1992). A more precise focus on the subunit level isparticularly vital, as theMNEs’ response to extraordin-ary institutions (Kostova & Zaheer, 1999) often lies atthe industry or business-specific level (Kostova et al.,2008; Phillips et al., 2009). Building on these initialefforts, this paper seeks to understand the dynamicsbetween MNEs and host country institutions byexploring responses on the subunit level.

A New-Institutional Approach to MNE SubunitAgencyTo explore agency on the subunit level, we explicitlyadopt a new-institutional framework (Greenwoodet al., 2008; Scott, 2008). Starting with DiMaggio(1988), new-institutionalists have devoted consider-able attention to various forms of agency, rangingfrom the muscular and purposive actions of institu-tional entrepreneurs (Barley & Tolbert, 1997;Greenwood & Suddaby, 2006; Maguire, Hardy, &Lawrence, 2004) to the reproduction of instituti-onalized routines by deeply embedded actors(Lounsbury & Crumley, 2007). Most recently, theinstitutional work literature has identified how thepurposiveness of agency varies, ranging from passiveand routinized iterations on the one hand to highlycalculative projective agency on the other (Battilana& D’Aunno, 2009). A primary purpose of this paperis to identify different types of strategic responses – thatis, the more calculative and projective forms ofagency (cf. Saka-Helmhout & Geppert, 2011) – thatMNEs employ vis-à-vis institutional environments.Following Oliver we define strategic responses as

“the strategic behaviors that organizations employin direct response to the institutional processes thataffect them” (Oliver, 1991: 145). This definitionexplicitly views conformity and acquiescence aspossible forms of strategic agency; it hence departs



from definitions that primarily emphasize actionsthat either deviate from and challenge prevailinginstitutions (e.g., DiMaggio, 1988; Maguire, Hardy,& Lawrence, 2004), or specifically seek to maintainthem (Lawrence, 1999; Lawrence & Suddaby, 2006).We use Oliver’s broader definition because for mostforeign entrants, conformity with local institutionsis not an example of taken-for-granted behavior andhabit, but rather an explicit form of strategic agency(Kostova & Zaheer, 1999; Luo, Shenkar, & Nyaw,2002; Saka-Helmhout & Geppert, 2011).The new-institutional approach to agency is parti-

cularly useful for our purposes for several reasons.First, new-institutionalism explicitly recognizes thatinstitutions are multifaceted and operate along notonly the regulative, but also normative and cognitive“pillars” (Scott, 2008); such non-regulatory institu-tions often constitute the most challenging aspects offoreign operations (Orr & Scott, 2008). Second, whilea view of agency linked to new-institutionalismallows for a more fine-grained analysis, the threepillars also provide a framework for generalizingfindings, both across our empirical cases and vis-à-vistheory at large. Finally, as our emphasis is on subunitagency in response to local institutions we are lessinterested in how differences in home and hostcountry institutional environments provide strategicadvantage (as is often the focus in comparativeinstitutional analysis, e.g., Jackson & Deeg, 2008),and more concerned with identifying common stra-tegies that can be employed by all MNE subunits,regardless of home country, when interacting withhost country institutions.

Enablers and Mechanisms of Agency: Why andHow Organizations are Able to Respond toInstitutional ConditionsAs scholars recognize the organizational capacity fortaking purposive action vis-à-vis institutions, theirfocus has increasingly shifted towards understandingwhy such responses arise, that is, what their under-lying enablers are. To date, research has largelyfocused on two enabling factors of strategic responsesto institutions: the organization’s social position,and the field-level conditions (Battilana, Leca, &Boxenbaum, 2009). Boundary-spanning social positions,that is, those that cross multiple institutional fields,can enable agency by exposing actors to alternativelogics, practices and norms, as well as status-specificexpectations from local actors (Rao, Monin, &Durand, 2003; Sauder, 2008). By contrast, marginaland fringe social positions within a field can enableagency by limiting the organization’s embeddedness

in local networks, thereby reducing pressures tomaintain established norms and practices (D’Aunno,Sutton, & Price, 1991; Lawrence & Suddaby, 2006;Leblebici, Salancik, Copay, & King, 1991; Scott,2008: 102).In turn, the broader field conditions of the institu-

tional environment can enable agency by providingactors with greater room for maneuverability andaction. Specifically, institutional fields characterizedby significant ambiguity and uncertainty (Lawrence,Suddaby, & Leca, 2009; Seo & Creed, 2002) can pro-vide greater opportunities for actors to employ poli-tical and social skills to engender change (Fligstein,1997). Similarly, actors may be more prone to iden-tify and leverage institutional contradictions (Creed,DeJordy, & Lok, 2010) in fields that operate onmultiple, and competing, logics (Greenwood, Diaz,Li, & Lorente, 2010). Conversely, more innovativeforms of agency may also arise in highly stable fields,since these are easier to interpret and understand(Beckert, 1999).Social positions and field conditions hence enable

strategic responses because they provide orga-nizations with opportunities to leverage particularmechanisms of agency. These include (but are notlimited to) the ability to see beyond current norma-tive and cognitive constraints (Bourdieu, 1977,1990; Emirbayer & Mische, 1998; Giddens, 1976;Sewell, 1992), heterogeneous pressure for confor-mity (Leblebici et al., 1991; Zuckerman, 1999), andaccess to particular capabilities, assets, relationshipsand networks (Greenwood & Suddaby, 2006). Thesemore specific mechanisms provide insight into howorganizations are able to engage in strategic res-ponses to institutions.In identifying enabling factors and their under-

lying mechanisms, the vast majority of extantresearch has focused on purely domestic actors.Notably, such enabling factors and mechanismshave clear parallels to key traits that set MNE apartfrom domestic actors. To begin with, MNEs are, bytheir very definition, boundary-spanning organiza-tions, with significant exposure to alternative prac-tices, norms and behaviors (Ghoshal & Bartlett,1990; Meyer, Mudambi, & Narula, 2011; Molina,2012; Rosenzweig & Singh, 1991; Westney, 1993).Second, the subunits of MNEs often inhabit mar-ginal or weakly embedded positions in host countryinstitutions and networks (cf. Mezias, 2002). Whilethis outsider position may result in increased pres-sures for conformity (Kostova & Zaheer, 1999;Rosenzweig & Singh, 1991), it may also mitigateexpectations of isomorphism (Kostova et al., 2008;



Shi & Hoskisson, 2012), allowing MNEs to challengelocal norms and behaviors (Edman, 2009). Finally,by virtue of their search for new market opportu-nities, MNEs often enter institutional environmentscharacterized by significant heterogeneity, uncer-tainty and ambiguity (Meyer, 2001; Meyer, Estrin,Bhaumik, & Peng, 2009; Clark & Geppert, 2011;Molina, 2012; Coeurderoy & Gordon, 2008; Jackson& Deeg, 2008).Viewed in light of new-institutional theory, these

studies suggest the MNEs’ unique social positions, aswell as their exposure to ambiguous field conditions,may strengthen their ability to undertake strategicresponses to institutions that are unavailable todomestic players, thereby resulting in a particularinstitutional advantage (Kostova et al., 2008, 2009;Phillips et al., 2009). This potential for agency vis-à-vis institutional host country environments andadvantage over domestic competitors may play animportant part in the overall corporate strategy ofthe MNE. To date, however, few studies have soughtto explore how MNEs actively respond to hostcountry institutions. While there is general recogni-tion that the MNE’s boundary-spanning may beadvantageous (cf. Grant, 1987), there is little insightinto exactly how this specific trait contributes to theMNEs ability to respond to host country environ-ments. This paper contributes to filling this researchgap. Specifically, our aim is to explore what kinds ofresponses MNE subunits invoke vis-à-vis host coun-try institutional environments, why they are able torespond (i.e., their enabling underpinnings), andhow they undertake the responses (i.e., the specificmechanisms).

RESEARCH DESIGN AND METHODSGiven both the explorative nature of our study,and our intention to extend existing theory, wefollow recent calls for qualitative methods andpluralist approaches in international businessresearch (Birkinshaw, Brannen, & Tung, 2011;Welch, Piekkari, Plakoyiannaki, & Paavilainen-Mantymaki, 2011) by employing a comparativeretrospective multiple case study design (Eisenhardt,1989; Langley, 1999). This approach was suitablebecause it allowed us to explore site-specific contexts,as well as mechanisms (Stake, 1994; Yin, 1994); bothof these were crucial, given our focus on culturaland institutional effects which vary in industriesacross different countries (Marschan-Piekkari &Welch, 2004).We chose subunits as the level of analysis because

our interest lies primarily in how responses are

created in specific institutional settings; this is oftenthe work of subunits dealing with local business-specific regulations, norms and customs (Regnér,2003; Saka-Helmhout & Geppert, 2011), as opposedto work at global headquarters (Birkinshaw, 1997;Orr & Scott, 2008). In particular, it has recently beensuggested that this level of analysis is most relevantwhen trying to determine MNE idiosyncrasies inrelation to institutions (Kostova et al., 2008) andthat MNE subunits may be particularly relevantwhen managing host country institutions (Cantwellet al., 2010). Our unit of analysis is thus the MNEsubunits’ diverse responses to institutional pres-sures. Note that we made observations across avariety of subunit and subsidiary types and that wecollectively label them “subunits”. Below we outlinethe case-selection criteria used to ensure sufficientvariety in responses.

Case SelectionAlthough we ground our study in an institutionalframework, the explorative nature of the workmeant that we had few a priori criteria to guide ourinitial case selection, other than that the casesshould embody situations in which firms activelyresponded to host country institutions. Given thetacit and taken-for-granted nature of institutions,Schneiberg and Clemens (2006) suggest that institu-tional interactions are most visible during times ofupheaval and change, when prevailing behaviorsand norms are challenged.We therefore purposely chose to focus on firms

introducing new products, services and practicesbecause such actions often provoke institutional con-flict and exceptions, thereby triggering purposiveresponses on the part of MNEs (Orr & Scott, 2008);this approach made the strategic responses to institu-tions more observable. Given the ambiguous andmultifaceted nature of institutional settings, we delib-erately searched for cases where MNEs responded toall three of Scott’s (2008) institutional pillars. Ourcase selection follows previous work on institutionsand MNE behavior (Kostova & Roth, 2002; Orr &Scott, 2008; Xu & Shenkar, 2002), and recognizes thecomplex and multilevel nature of institutions.We explicitly focused on cases wherein MNEs had

introduced novel and norm-deviant practices. Basedon interviews within 12 different multinationalsconducted for two earlier research projects, we wroteup case narratives outlining major challenges andopportunities faced by MNEs introducing novelproducts and practices into new environmental con-texts. From these we identified six cases wherein the



MNEs’ introduction of new products and practicesresulted in interactions vis-à-vis local regulatory,normative and/or cognitive institutions. This typeof focused theoretical sample of multiple casesallows for a replication logic in which cases aretreated as experiments from which inferences aremade, producing stronger theoretical generaliza-tions (de Vaus, 2001; Yin, 1994). The six cases permitvariation in external conditions by representingproduct and service introductions into various mar-kets – including the United States, Japan, the EU andSoutheast Asia – and by representing various indus-tries, products and services, ranging from pharma-ceuticals to automotive components to financialservices. Table 1 provides a summary of the six firmsand their host markets, products/services and insti-tutional conditions. As responding to institutionssometimes involve rather sensitive strategies, oneselection requirement was that interviewees werewilling to share delicate information and that richarchival data was available. To ensure this require-ment, one trade-off was that our cases differed intheir timing, including some from the mid- to late1990s (AGA, Autoliv, Citibank and Pharmacia) and

some from the late 1990s and early 2000s (Ericsson,The Hartford).

Data Collection and ValidityThe case studies were primarily composed of first-hand interviews, augmented by secondary sourcesincluding company documents, industry associationstudies and statistics, newspaper clippings andannual reports. We conducted background inter-views to identify key informants (usually subsidiarypresidents, heads of local product units or practiceandmanagers at headquarters with responsibility forcertain regions and projects), who subsequentlyintroduced us to further interview subjects. Weinterviewed an average of 13 informants per com-pany and the interviews lasted 1–2.5 h. To reducepotential bias and ensure diverse views of institu-tional conditions and responses, we conductedinterviews with not only subunit managers but alsoheadquarters staff and external experts (includinganalysts, industry associations, customers and com-petitors). In total, we conducted 110 interviews atvarious stages between 1998 and 2008, of which athird primarily concerned the respondents’ views of

Table 1 Firms, geographical markets, products/services and institutional conditions

Company Industry Objective Host market(s) Salient institutional conditions

AGA(Sweden)

Industrial Gas Introduction of Westernsales, distribution andmarketing formats forindustrial gas

Eastern Europe Lack of or extreme uncertainty regardingproduct regulations, sales and distributionpractices and norms in emerging hostmarkets

Autoliv(Sweden/US)

Automotive supplier Introduction of new airbagproduct

Continental Europe Product seen as unnecessary andexpensive by car manufacturers in hostcountries since mandatory seatbeltregulations were already in place

Citibank(US)

Banking and Finance Introduction of new lendingproduct

Japan New lending format ran counter to normsand assumptions dominant in theJapanese banking industry; opposed bylocal competitors

Ericsson(Sweden)

Telecommunications Introduction of mobiletelephony systems

Southeast Asia Product seen as less appropriate since itinterfered with existing radio frequencies(often of military use) and fixed telephony;unclear standards and lack of regulation inhost markets

The Hartford(US)

Life Insurance andConsumer Finance

Introduction of newdistribution strategy

Japan Existing norms and regulations impededfirm’s ability to introduce its distributionformat

Pharmacia(Sweden/US)

Pharmaceuticals Introduction of unique andnovel smoking-cessationproduct

Anglo-Saxonmarkets (Canada,UK & US)

Product ran counter to normative andcognitive foundations in host markets;seen as illegitimate and questionable. Lackof regulations and standards



institutional conditions and responses, while theremainder of the interviews concerned this togetherwith wider MNE strategic considerations. Our archi-val search employed public media sources like news-papers and magazines, as well as internal companydocuments, industry analyses and annual reports, asshown in Table 2.All interviews were recorded and transcribed,

except in a few instances when this was not possible.Interviews conducted in non-English languages weretranslated and double-checked by native speakers.Interview data was coded and organized inductivelyinto subject categories. Our approach was qualitativeand interpretative: we did not count words andsentences but rather sought to draw conclusionsabout meanings, mechanisms and actions from theavailable data (Langley, 1999). We began by usingdetailed coding, essentially summarizing individualbits of data and information; this yielded severalhundred categories per firm. During multiple codingiterations we extracted and evaluated diverse units ofcoded data and checked consistencies in the emer-ging codes; as the analysis progressed, categories were

combined and refined to form higher order themesand constructs. As data collection was split betweenthe two authors, codes and data classifications werecritically examined and challenged by the author notinvolved in a particular data collection.Because the research was exploratory in nature, we

iterated between data collection, coding and theoriz-ing throughout the project, revisiting old data andgathering new information which was subsequentlyadded to existing findings (Dubois & Gadde, 2002).Emerging concepts were evaluated by activelysearching for contradictory information through arefutation strategy (Spiggle, 1994), as well as byassembling data in temporal order using a process-tracing approach (Gerring, 2004). Building on thedata collection and sorting, we developed individualcase analyses for each firm, ranging in length from20 to 40 pages, which were validated by informants.We subsequently used cross-case comparative analy-sis, contrasting the firms across multiple categories(Gerring, 2004; Miles & Huberman, 1994). Thebetween-case approach increased external validityby both sharpening findings from the within-case

Table 2 Overview of data sources

Company Annualreports

Other secondary sources(company histories, company

internal documents, industry reports,press releases, etc.)

Number of informantsintervieweda

Informantsinterviewedby type

AGA 17 11 5 Corporateexecutives: 2Business unit/project managers: 2

Autoliv 18 20 15 CorporateExecutives: 4Business unit/project managers: 5

Citibank Japan 3 11 19 Corporate executives: 3Business unit/project managers: 11External experts: 5

LM Ericsson 20 14 22 Corporate executives: 3Business unit/project managers: 7External experts: 8

The Hartford 1 35 7 Corporate executives: 3External experts: 4

Pharmacia 30 8 16 Corporate executives: 5Business unit/project managers: 6Business developmentmanagers: 2External experts: 3

aSome informants were interviewed twice.



studies, and generating new concepts from thesystematic combination of various case aspects(Adcock & Collier, 2001; Collier & Mahon, 1993).

Data Coding and AnalysisOur findings emerged in a three-stage process. In thefirst stage we constructed firm-specific narratives toidentify the extent and type of institutional condi-tions faced by each MNE subunit. Separating theinstitutional conditions according to Scott’s pillars –that is, regulative, normative and cognitive – weidentified 16 distinct institutional conditions (6regulatory, 6 normative and 4 cognitive) whereMNEs used strategic responses. As the first columnin Table 3 shows, these various institutional condi-tions involved considerable uncertainty and com-plexity for the MNE.While all firms face some degreeof complexity and heterogeneity when challenginginstitutional settings, the findings suggest complex-ity may be particularly severe for MNEs.In the second stage of the analysis, we sought to

identify the strategic responses MNEs employed indealing with each of the 16 institutional conditions,as well as the responses’ enabling factors and theirunderlying mechanisms. Figure 1 outlines the pro-gression of data coding, categorization and analyti-cal dimensions, and thus includes the raw datacoding in the first column. The figure provides abridge between our method, data, findings andemergent constructs. As Figure 1 indicates, our dataanalysis in this stage was guided by our three over-arching research questions: What were the MNEresponses to the institutional conditions? Why werethey able to undertake these responses? How didthey implement these responses?We began by sorting data from the narratives into

descriptive categories based on these three overarch-ing research questions. Because the three researchquestions address different levels of abstraction (ran-ging from the broader typology of strategic responses,down to the more specific underlying mechanisms),we posed all three questions to each data point. Table 3provides concrete evidence of this analytical process.For example, as the table indicates, a quote from TheHartford’s experience in dealing with regulative insti-tutions provided insight not only into what theMNE’s strategic response was (“Worked to changeexisting institutions”) but also why the firm was ableto do this (thanks to its “International status”) andmore specifically how (by “Leveraged local audienceexpectations”). While these categories naturally boresome resemblance to each other, they each sought toanswer one of the specific questions of what, why and

how the MNEs responded to various institutionalconditions.We subsequently sorted the categories into three

overarching themes, depending on which of thethree research questions they addressed, as shownin column three of Figure 1. Once the second-orderthemes reached theoretical saturation we comparedthe categories across cases to identify similarities anddifferences. This analysis led to the conceptualiza-tion of our three overarching dimensions: strategicresponses to institutional complexity, the multipleenablers of these strategic responses, and the particu-lar mechanisms underlying the strategic responses.The answers to our what, why and how questionsthus compose our principal constructs, as shown inthe fourth column of Figure 1.In the final stage of the analysis, we continued the

cross-case analysis to identify how particularresponses, enablers and mechanisms were interre-lated. This analytical process involved iteratingbetween the codified and sorted data, the narrativecases and pre-existing studies of agency. Based onthis iteration, we developed propositions for howthe various enablers and mechanisms are interre-lated and result in various strategic responses bysubunits in institutional environments.

FINDINGS: A TYPOLOGY OF MNE STRATEGICRESPONSES TO INSTITUTIONS

As indicated in Figure 1, our analysis identified fourdistinct strategic responses types: Innovation responsessought to create and/or change host country institu-tions. Arbitrage responses sought to exploit differencesbetween host and home (or third-country) settings.Circumvention responses involved sidestepping andeffectively dodging the demands of local institutions.Finally, theAdaptation responses sought to conform tothe institutions of a host country.In addition, we also identified three unique enablers

of these responses. First, we found that responses wereenabled by the MNEs’ multinationality; through itsboundary-spanning multinational social positions,the MNE is exposed to alternative regulations, normsand understandings. Second, we found that responseswere also enabled by the MNE’s foreignness inhost country institutions. Foreignness effectivelyamounted to a weakly embedded outsider social posi-tion in host country institutions and networks. Thisforeignness trait is also unique for the MNE anddifferent in kind, like the case of multinationality.Finally, host country field conditions also enabled thestrategic responses in the form of the MNE’s extensiveexposure to institutional uncertainty and ambiguity.



Table 3 Representative quotes and first-order categorization by research question

Institutional conditions Representative quotes from case narratives Representative first-order categorization of narrative data in response to research questions

What did thesubunit do?

Why was the subunit ableto do this?

How did the subunitdo this?

AGA regulative and normative (A,B):Ambiguity due to collapse ofindustrial gas regulations andnormative assumptions in virtuallyall aspects

“…we repeated similar things everywhere …we moved from the first wave of efforts in theEast [to the second] and then we will moveinto a third…explicit knowledge transfer iscrucial…and now we are travelling to Chinaand do the same thing…” (AGA manager)

Introduced institutionsestablished elsewhere

Had experience from otherinstitutional locations;Uncertainty in local institutionalenvironment

Transferred institutionalknowledge; seeing beyondexisting institutions

AGA cognitive (C): Littleunderstanding of Western industrialgas sales and distribution formats

“…we solved it step by step and…we did it ina rather smooth way…when in Rome youmust do as the Romans…and adapt” (AGAmanager)

Adapted to localinstitutions

Low acceptance as foreign firm Followed local audienceexpectations

Autoliv regulative (D): Lack of airbagregulation; uncertainty as towhether it would develop

“We did not use lobbying or work withlegislators…It is the [US] legislation…thatpropelled it…” “The US legislation wassignificant and distinct event influencing this[airbag introduction]”


Had experience from otherinstitutional locations


Autoliv normative (E): Airbagsviewed as too large, unnecessaryand inappropriate

“…we produced airbags for Volvo’s US car…[then we] developed…airbag[s] for Europe[cars]” (Autoliv manager)

Adapted to fit locallyinstitutionalized norms



Autoliv cognitive (F):Difficult to understand any need forairbags

“…European consumers started to questionwhy they were not protected in the same way[w. airbags]”


Had experience from otherinstitutional locations;Uncertainty in local institutionalenvironment

Transferred institutionalknowledge

Citibank regulative (G): Existingregulations unclear; syndicationpossibly illegal

“It takes effort to explain to the market, toinvestors and borrowers…but [the products]will be adopted when they make sense”(Citibank manager)

Worked to change existinginstitutions


Seeing beyond existinginstitutions

Citibank normative (H): Loansyndication unknown, potentiallyinappropriate

“As a foreign bank you don’t hesitate topropose something or to do something. TheJapanese banks will get into situations maybethey shouldn’t because of relationships…wewon’t” (Citibank manager)

Sought to be outside hostcountry institutions

Was outsider, not part of localinstitutions; uncertainty in thelocal institutional environment

Leveraged local audienceexpectations

MNEinstitu

tionalad

vantag

ePatrick

Regnérand

JesperEdm

an283

JournalofInternationalBusinessStudies

Table 3: (Continued )

Institutional conditions Representative quotes from case narratives Representative first-order categorization of narrative data in response to research questions

Citibank cognitive (I): mechanisms ofloan syndication not understood

“People have different expectations of foreignand Japanese firms…we don’t really competewith them. We [are] outside the system butstill respectable…do things [they] couldn’t”(Citibank manager)

Sought to be outside hostcountry institutions

Was outsider, not part of localinstitutions

Leveraged local audienceexpectations

Ericsson regulative institutions (J):Lack of radio/mobile telephonyregulations; radio spectrumsprohibited

“[we]…believed in walking the talk…workedin different directions, influencing standards,influenced authorities, influencedcorporations…making sure the standardssuited us in some way” (Ericsson manager)



Used international status toinfluence local actors; Seeingbeyond existing institutions

Ericsson normative (K): Negativeattitudes towards telephony basedon radio

“…we had knowledge few had on the worldmarket…we had people travelling aroundAsia talking to…powerful people in themilitary…convincing them...” (Ericssonmanager)


Had international status; Hadexperience from other institutionallocations

Used international status toinfluence local actors;recognizing alternativeinstitutions

Hartford regulative (L): Existingregulations did not supportdistribution format

“We worked with the FSA…they knew wewere number one in the US, so that helped usa lot…we got in the door” (Hartfordmanager)


Had international status Leveraged local audienceexpectations

Hartford normative (M): Distributionformat seen as inappropriate

“We happen to have an insurance license, butwe’re in the business of doing somethingelse…our main competitors are bank depositsand investment trusts” (The Hartfordmanager)

Ignored local institutionalnorms

Ambiguity about industrymembership

Leveraged local uncertaintyabout firm’s position

Pharmacia regulative (N): Noregulatory framework in place tosupport nicotine chewing gum

“…what we have done, based on ourknowledge in medication and registrationfunctions, is rather remarkable…we wrote theFDA guidelines” (Pharmacia manager)

Created new institutions inthe host country


Transferred institutionalexperience and knowledge

Pharmacia cognitive and normative(O,P): Nicotine as medicationdifficult to understand and viewedas inappropriate

“…we had a very strong vision for how toboost Nicorette globally and the wholesmoking cessation market…it is a form ofvisionary fore-checking…” (Pharmaciamanager)

Created new institutions inthe host country


Transfer of institutionalexperience and knowledge;recognizing alternativeinstitutions

MNEinstitu

tionalad

vantag

ePatrick

Regnérand

JesperEdm

an284


In contrast to multinationality and foreignness, theMNE’s greater exposure to ambiguous institutionalenvironments is a difference in degree rather thankind, but the elevated exposure to institutionalambiguity appears to offer the MNEs greater oppor-tunity for engaging in strategic responses. In sum-mary, we found that the responses were enabled byboth the MNEs’ unique social positions – that is,their multinationality and foreignness – and by

host country field conditions – that is, the level ofinstitutional ambiguity.Focusing more specifically on how these enablers

underpinned the strategic responses, we found thatmultinationality, foreignness and host country insti-tutional ambiguity enhanced three mechanisms thatunderlie strategic responses. First, we found thatMNEs leveraged reflexivity, that is, the ability to seebeyond local norms, cognitions and regulatory

AGA

Autoliv

Citibank

Ericsson

Hartford

Pharmacia

A. Regulative

B. Normative

C. Cognitive

D. Regulative

E. Normative

F. Cognitive

G. Regulative

H. Normative

I. Cognitive

N. Regulative

O. Normative

P. Cognitive

L. Regulative

M. Normative

J. Regulative

K. Normative

What?

Why?

How?

What?

Why?

How?

What?

Why?

How?

Strategic Responses

Innovation

Arbitrage

Circumvention

Adaptation

Mechanisms

Reflexivity

Role-expectations

Resources

Enablers

Multinationality

Foreignness

Institutionalambiguity

Sorting of narrative data into first-order categories

Sorting first-order categories into second-order themes

Conceptualizing dimensions of second-

order themes

Case narrativesraw data

1. Recognizing alternative institutions (K,O,P)

2. Seeing beyond existing institutions (A,B,D,E,G,J)

3. Sought to be outside host country institutions (H,I)

4. Worked to change existing institutions (G,K,I,J)

5. Adapted to local institutions (C)

6. Created new institutions in the host country (N,O,P)

7. Ignored local institutional norms (M)

8. Leveraged local uncertainty about position vis-à-vis local institution (M)

9. Had experience from other institutional locations (A,B,D,E,F,G,J,N,O,P)

10.Low acceptance as foreign firm (C)

11.Introduced institutions established elsewhere (A,B,D,F)

12.Leveraged local audience expectations (H,I,L)

13.Ambiguity about industry membership (M)

14.Was outsider, not part of local institutions (H,I)

15.Transferred institutional experience and knowledge (A,B,D,E,F)

16.Uncertainty in local institutional environment (A,B,F,H)

17.Used international status to influence local actors (J,K)

18.Had international status (K,L)

19.Follow local audience expectations (C)

20.Adapt to fit institutionalized norms (E)

Strategic Responses

3. Sought to be outside institutions 4. Worked to change existing

institutions 5. Adapted to local institutions 6. Created new country

institutions 7. Ignored local institutional

norms11. Introduced institutions

established elsewhere20. Adapted to fit local institutions

Mechanisms

1. Recognizing alternative institutions

2. Seeing beyond existing institutions

8. Leveraged local uncertainty about positions vis-à-vis local institutions

12. Leveraged local audience expectations

15. Transferred institutional experience and knowledge

17. Used international status to influence local actors

19. Followed local audience expectations

Enablers9. Had experience from other

institutional locations10. Low acceptance as foreign firm13. Ambiguity about industry

membership14. Was outsider, not part of local

institutions16. Faced uncertainty in local

environment18. Had international status

Figure 1 Progression of data coding, categorizations and analytical themes and dimensions.



conditions of their surrounding host country envi-ronments. Reflexivity allowed MNE subunits torecognize institutional differences and contradictionsto a far greater extent than their local competitors;this in turn made it possible for the MNE subunits toidentify opportunities for agency in relation to insti-tutions, as well as possible deviations from prevailingnorms and practices.Second, we found that the enablers resulted in

MNE-specific role expectations, that is, different audi-ence assumptions and beliefs about the MNE subu-nits. These role expectations resulted in differentregulative, normative and cognitive conditions, effec-tively allowing the MNEs to engage with local institu-tions in a different manner, as compared with localdomestic actors. Finally, multinationality, foreignnessand exposure to ambiguous institutional environ-ments also provided MNEs room to transfer resourcesfrom MNE operations in other organizational fields,including institutional experiences, knowledge, socialskills and capital. In sum, the MNEs built on threemechanisms in their strategic responses: (1) increasedreflexivity (i.e., the ability to see beyond currentnorms and regulations); (2) differentiated role expec-tations (i.e., particular assumptions and beliefs aboutwhat the organizations could legitimately do); and (3)transfer of institutional resources (i.e., assets and/orpractices from diverse institutional settings).Combining these insights, our analysis suggests

MNE subunits crafted different strategic responsesby drawing on diverse mechanisms that wereenabled by their multinationality and foreignness,as well as their exposure to institutional ambiguityin the host country. Table 4 provides an overviewof the strategic response typology and the linkagebetween the strategic responses, the enabling factorsincluding social positions and host country field con-ditions, and the underlying mechanisms employed ineach response. By linking the enablers and themechanisms to the strategic responses, we highlighthowMNEs have particular institutional advantages inshaping, transposing and evading host country insti-tutions. Below we discuss these findings in greaterdetail.

Institutional InnovationIn institutional innovation, MNEs respond to institu-tional complexity by purposely seeking to work with andcreate new institutions, and/or change prevailing institu-tions. They are primarily able to do this by relying onincreased reflexivity, role expectations and resources thataccrue from their boundary-spanning multinationalitypositions. In this strategic response, the institutions

of the host country are relatively fixed and clearlydemarcated; MNEs hence do not actively leverageambiguities in local field conditions when engagingin institutional innovation, but instead build onstable extant institutions or the lack of them. In ourstudy, the institutional innovation approach is mostclearly exemplified by Ericsson and Pharmacia’sapproaches to the institutions of their host countrymarkets; in addition, The Hartford and Citibank usedinstitutional innovation when dealing with the reg-ulatory pillar of host country institutions.

Ericsson’s introduction of mobile telephony intoSoutheast AsiaWhen Ericsson entered Southeast Asian telecom mar-kets in the mid- to late 1990s, after having establishedstrong positions in Europe, it faced several institutionalchallenges and complexities. As in the case of Europein the late 1980s, the regulations of most countries“provided no room for mobile telecommunications”,beyond that of the military; hence, it was technicallyillegal to introduce handheld portable phones. Interms of societal norms, clients in host countries –including both governments and individual consu-mers – were uncertain as to the value and necessity ofmobile telecommunications, much as they had beenin Europe initially. To overcome these challenges,Ericsson worked with different constituents to changenorms and regulations to support its mobile telephonybusiness. As two managers observed:

[W]e were very early…we travelled around Asia and triedwith officials…people with power…traditionally the mili-tary…asked for a small part in radio for private use…we werevery early in the markets…we circled our competitors…wewere very proactive.

It was individuals that worked broadly on all frontiers; fromcomponents to laws and regulations…all of us we traveledthe world, to standardization organizations, etc.

Ericsson hence overcame challenging regulativeconditions by actively helping local governmentsdevelop the rules, technical requirements and legalstandards necessary for mobile telephony. Ericsson’sability to do this stemmed from its significant expo-sure to institutions in multiple country contexts. Thecompany had gained valuable experience workingwith regulatory bodies and authorities when settingup mobile telephony systems in the Middle East,Continental Europe and Scandinavia. By absorbinginsights from these events and building up a stock ofinstitutional knowledge and capabilities, Ericssonbecame an expert at establishing host country mobile



Table 4 MNE strategic responses: Definitions, enabling factors and primary mechanisms

Strategic responses Innovation Arbitrage Circumvention Adaptation

Purposely changing or creatinglocal institutions

Leveraging differences in host vshome and third-countryinstitutions

Leveraging ambiguities andoutsider social position in hostcountry environments

Conforming to local institutionalpressures

Enabling factors Social position Boundary-spanning throughMultinationality

Boundary-spanning throughMultinationality

Weak embeddedness due toForeignness

Weak embeddedness due toForeignness

Field conditions Low ambiguity;transparent and coherentinstitutions

High ambiguity;existence of institutionalcontradictions

High ambiguity;existence of institutionalcontradictions

Low ambiguity;transparent and coherentinstitutions

Mechanisms Reflexivity High:Ability to see institutionaldifferences key to identifyinginnovation opportunities

High:Ability to see institutionaldifferences and contradictionskey to identifying arbitrageopportunities

High:Ability to see institutionalcontradictions key to identifyingcircumvention opportunities

Low:Ability to see institutionaldifferences and contradictionsless important in adaptation

Roleexpectations

High:Local expectations on the MNE tointroduce new practices permitslocal institution-building

Low:Local actors’ assumptions andbeliefs about the MNE lessimportant in arbitrage

High:Local actors’ assumptions ofoutsidership provide greaterroom to defy prevailingpractices

High:Local actors’ assumptions andexpectations of adaptationprecludes active attempts tochange or defy institutions

Resources High:Transfer of alternative institutionalexperiences and knowledgesupports institution-building

High:Transfer of alternativeinstitutional experiences andknowledge makes arbitragepossible

Low:Transfer of institutionalexperiences and knowledge lessimportant in circumvention

Low:Transfer of institutionalexperiences and knowledge lessimportant in adaptation

MNEinstitu

tionalad

vantag

ePatrick

Regnérand

JesperEdm

an287


telephony regulations. As reported by one Ericssonmanager:

We placed people out there, those with home [country]knowledge, moved them to new locations…it was verychallenging in some countries.

[T]he Ministries answered they did not have 450 MHz avail-able…our people became acquainted with them and workedwith them…located opportunities in their frequency plans.

Pharmacia’s introduction of Nicorette into the UnitedStatesIn introducing the anti-smoking product Nicoretteinto the United States, Pharmacia similarly faced alack of regulations. In particular, there was consider-able confusion as to which regulatory body hadjurisdiction over the product, since authorities wereuncertain whether a nicotine chewing gum shouldbe considered a food or a drug. In addition to theregulatory hurdles, the product itself was consideredhighly illegitimate by virtually all actors. Nicotinewas viewed as a poison and an anathema to thepharmaceutical industry; this was reflected in thenorms of medical practices, the cognitive mind-setsof doctors, and the regulations of the Food and DrugAdministration (FDA). Doctors and pharmacists haddifficulty understanding how nicotine could be con-sidered a medication, as well as how and if it couldbe used to reduce smoking. This made it almostimpossible for the company to find a businesspartner since no one quite understood the product.As one central figure in the original venture noted:

We had a long-standing relationship with…[pharmaceuticalcorporation] Warner Lambert, who [also] owned the world’slargest manufacturer of chewing gum…they lost interest…they felt that selling a chewing gum containing a poisonmight have an adverse effect on their ordinary market.(Fernö, 1994: 1224)

Like Ericsson, Pharmacia sought to overcome theseinstitutional constraints by working intensively toinfluence regulations and norms and gain supportfor their smoking-cessation product Nicorette. Thisincluded work both among individual doctors andother medical practitioners, and on the level of theFDA. Notably, this was an explicit tactic, as one ofthe architects behind the entry noted:

[W]e had a very strong vision for how to boost Nicoretteglobally and the whole smoking cessation market…it is aform of visionary fore-checking.

[B]uilding a market…and providing product informationthrough…sales, conferences and advertising in industrypress. Physicians and pharmacists were the target…thegoal…[was] to create a market.

Pharmacia not only gained the FDA’s confidencein delineating the US regulation regarding smokingcessation, it managed to advance even furthercompared with other markets by registeringthe product as “Nicotine polacrilex” (from poly-methacritic acid) rather than “Nicotine chewinggum”, which was considered less legitimate in theindustry. In informing the FDA and helping itrewrite regulations, Pharmacia relied on experi-ence from both its home-country and other host-country institutional environments. A Pharmaciamanager reported:

[W]hat we have done, based on our knowledge inmedicationand registration functions, is rather remarkable…we wrotethe FDA guidelines.

In line with Ericsson, Pharmacia’s efforts thusstemmed from its boundary-spanning multinationalposition, particularly its efforts at working withregulators in its home country of Sweden. TheEricsson and Pharmacia cases illustrate not only towhat extent they influenced and changed institu-tions, but how they initiated and created them fromscratch.

The Hartford’s introduction of variable annuities salesthrough banks in JapanIn the case of The Hartford’s introduction of variableannuities into Japan, the primary institutional chal-lenge stemmed from what the company perceived asoverly restrictive regulations on sales channels. Inparticular, The Hartford’s strategy relied on selling itsinvestment-oriented insurance products throughindependent financial advisors, including banksand securities companies; the extant regulatory fra-mework, however, prohibited sales of insuranceproducts through financial institutions. Like Erics-son and Pharmacia, The Hartford was able to over-come this institutional hurdle by actively workingwith local regulators, convincing them to reclassifyvariable annuities as a mutual fund-type product,which could be distributed through banks and secu-rities companies. The primary reason the firm wasable to do this was its experience from the USinstitutional environment, and the regulators’ recog-nition of The Hartford as a leading firm in its homemarket. As one manager noted:

We worked with regulators…we convinced them that theproduct should be categorized as a mutual fund…and I think,they recognized that we were number one in the US…theyrecognized our expertise and knowledge of regulations andthe market.



Citibank’s introduction of loan syndication into JapanWhen Citibank sought to introduce loan syndicationonto the local Japanese financial services market, oneof its biggest challenges consisted of convincing localcustomers of the value of the new product. Loansyndication, a lending format in which multiplebanks jointly lend to individual borrowers, ran coun-ter to many of the fundamental characteristics ofJapan’s pre-existing bilateral lending tradition (Scher,1998). Consequently, customers were uncertain aboutthe value and merits of the new practice. Citibankovercame this uncertainty by educating customersabout loan syndication’s financial benefits. The abilityto do so came from the company’s experience ofovercoming similar normative constraints in othercountries, as one manager noted:

Being the first company to try something new is alwaysdifficult, but we took our knowledge of being first anywhere,in any market, and we explained to customers, pointed outthe benefits. So, this is a company that has been doing loansyndication for 20 years, so introducing it was not a problem.

Mechanisms and enabling factors of institutionalinnovation: Reflexivity, role expectations and resour-ces through multinationality and low institutionalambiguityIn each of the above cases, the focal firms faced veryclear regulatory and normative constraints from hostcountry institutions. To overcome these constraints,the firms purposively sought to innovate by introdu-cing new regulations, norms and values that wouldengender greater acceptance for their products.Notably, innovation and change in institutions canoccur through spillovers from foreign entrants andincremental learning; by defining innovation as apurposive attempt at change, we exclude such spil-lover effects.The cases highlight how the active implementa-

tion of new institutions was primarily enabled bythe combination of the MNE subunit’s boundary-spanning social positions as a multinationals, andthe low ambiguity of the host country institutionalenvironment (see Table 4; column 1, rows 1 and 2).Specifically, the subsidiary’s boundary-spanningsocial position engendered three underlying mech-anisms that were key to institutional innovation:reflexivity over alternative institutions, transfer ofinstitutional resources and specific role-expectationson the part of local audiences.Reflexivity was central to institutional innovation,

as the MNEs’ boundary-spanning activities in nume-rous institutional settings not only made them see

beyond immediate constraints, but also made themrecognize the potential for changing extant insti-tutions (see Table 4; column 1, row 3). For example,Ericsson could see beyond the normative and regu-lative resistance to mobile telephone systems andlocate opportunities in the new markets. Pharma-cia’s advantage was that it was able to see beyond theclassification of nicotine as a poison and an illegiti-mate drug and instead envision nicotine as amedicaltreatment. Likewise, The Hartford and Citibank sawthe Japanese regulations from the perspective oftheir presence in many other markets and thusmanaged to influence regulators.The transfer of institutional resources was in turn

crucial to institutional innovation because it pro-vided MNEs with a template on which to base theirinstitutional innovation (see Table 4; column 1,row 5). For example, both Ericsson and Pharmaciachanged local regulatory institutions by drawing ontheir institutional experience and practices fromhome and other host countries. The Hartford simi-larly introduced ideas from its home market, as didCitibank in drawing on its experience of “being firstanywhere” to educate and convince local constitu-ents. As these examples suggest, these resources wereavailable to the firms because of their boundary-spanning social positions and multinationality.Only through transfer of institutional resources frommultiple institutional locations and markets, werethese firms able to develop the insights and knowl-edge to introduce particular regulations, norms andcognitive models into the host countries. Naturally,capital resources relating to institutional change alsoplayed a role. Pharmacia, for example, faced severalyears of lobbying and discussion before it was able toconvince local regulators and doctors of the value ofNicorette; Ericsson similarly invested significanttime and energy in convincing local regulators notonly of the merit of its technology, but also of howthe market should be structured.Finally, role expectations were also critical for

institutional innovation because they providedMNEs with the legitimacy and status needed tochallenge prevailing regulations, norms and mentalmodels (see Table 4, column 1, row 4). For example,Ericsson gained access to key regulators and was ableto introduce its mobile system ideas largely thanks toits reputation as a world leading telecommunica-tions firm. Similarly, The Hartford’s reputation asbeing the number one seller of variable annuities inthe United States helped them initiate regulatorydiscussions with Japanese officials. The expectationsof local audiences on the MNEs’ thus played a



central role in their ability to introduce new institu-tions. In the case of our MNEs, these unique role-expectations were a direct result of their boundary-spanning positions, and the expectations andassumptions placed on them as large multinationalentities.Notably, the innovation response was also enabled

by the low ambiguity of institutional conditions ineach of the subunits’ host country markets. Specifi-cally, the lack of institutional ambiguity meant thatreflexivity around discrepancies and contradictions inthe institutional settings was not an importantmechanism underpinning institutional innovation.Instead, low ambiguity made it easier for the MNEsubunits to engage in institutional innovation byrelying on reflexivity, resources and role-expecta-tions directly related to their multinationality.In the case of Pharmacia, for example the unam-

biguous lack of FDA regulation of nicotine productsmade it easy for the firm to directly apply its knowl-edge from other markets. Pharmacia’s success hencedid not depend on any unique or idiosyncratic partof US regulations, but rather on resources and role-expectations accruing from its boundary-spanningposition and that they could build new institutionswithout having to steer and balance between diver-ging institutional perspectives. Similarly, the strate-gic responses of Ericsson, Citibank and The Hartforddrew on their reflexivity over and knowledge aboutalternative institutions based on their multination-ality, as opposed to reflexivity around institutionalcontradictions.

Institutional ArbitrageIn the institutional arbitrage response strategy,MNEs leveraged differences in institutions of homeand host country contexts. In contrast to theinnovation response, institutional arbitrage wasenabled by both the MNE’s boundary-spanningposition and ambiguities in local field conditions.Our cases suggest reflexivity and resources are impor-tant mechanisms underlying institutional arbit-rage, while role expectations are less important.We found particular evidence of arbitrage res-ponses in the cases of Autoliv’s introduction ofairbags in Europe with regulative and cognitiveinstitutions, as well as AGA’s interaction withregulative and normative institutions surroundingindustrial gas in Eastern Europe.

Autoliv’s introduction of airbags into EuropeOne of the salient challenges faced by Autoliv in itsattempts to introduce airbags to the Europeanmarket

was the lack of recognition and acceptance of the newproduct. Although car safety was a strong normamong European car consumers, this had alsoresulted in an earlier focus on seatbelts; consequently,car manufacturers had little understanding for intro-ducing yet another safety device. For Autoliv, thisproblem was a repeat of the situation it had faced inthe US market in the late 1980s and early 1990s a fewyears earlier: US car manufacturers had broadlyopposed the introduction of airbags, calling them“red herrings” and “a complete waste of time”; thisopposition was overcome by relying on the introduc-tion of new safety legislation. This experience fromthe US market benefitted the firm when dealing withsimilar opposition from European car manufacturers.In addition, the company was also able to takeadvantage of growing ambiguities in European end-consumers’ understandings of safety. As continentaldrivers became increasingly aware of the use of air-bags in the United States and in US-made cars, theyactively began to question why European automo-biles did not have similar features. Autoliv was able toleverage these increasing contradictions to its advan-tage. As one senior manager noted:

The established [European car-] manufacturers…did notbelieve in it…tried to combat it…European consumersstarted to question why they were not protected in the sameway…[but] we produced airbags for Volvo’s US cars…theAmerican type.

Building on their experience from the stronglynegative US car manufacturers and their realizationthat European consumers still valued safety, Autolivmanaged to introduce airbags, as one managerechoed:

[W]e were successful with the European airbag concept…[with] a “face bag”…an add-on safety system…. We createdthe market!

AGA’s introduction of new industrial gas distributionformats into Eastern EuropeAGA’s entry into Eastern Europe in the years afterthe fall of communism was characterized by con-siderable uncertainty and turbulence. While oldinstitutions had been discredited, new ones hadyet to take their place: this was true in the regula-tive, normative and cognitive domains. For AGA,the uncertainty arose in particular from the under-developed nature of the local distribution system;while there was little ambiguity in terms of itsunderlying product (industrial gas), there was con-siderable confusion as to the norms and practices



of how industrial gas should be stored, transportedand distributed and what the dominant regulatoryframework was. In AGA’s case virtually all dimen-sions were complex. As two managers noted:

How would the state-owned gas monopolies be divided up?…Which parts were offered to foreigners? How would thegeneral industry structure, our customers, be re-formed andprivatized?

The other issue was…the continuous problems with autho-rities…with bottled gases…they were constantly concerned.

In response to these uncertainties, AGA chose tobase its distribution practices on the regulatory andnormative institutions governing gas-distributionpractices in Western Europe. While these regula-tions were not the de facto standards of its EasternEuropean markets, there was nothing to preventthe firm from following them in its daily opera-tions (unlike, e.g., in the cases of Ericsson, Pharma-cia, The Hartford and Citibank, discussed above).In doing so, the firm not only overcame the uncer-tainties of the local markets, it also managed togain a competitive edge, thanks to the more devel-oped nature of the Western distribution practices.AGA’s arbitrage response thus built on both itsboundary-spanning social position as a multina-tional, and the opportunities that arose due to theambiguities and uncertainty of local field condi-tions. As one manager noted:

Besides the headquarters’ staff support…we had a “godparenting” sponsorship system…where each new subsidiarywas supported by an established subsidiary [in the West].

Mechanisms and enabling factors of institutionalarbitrage: Reflexivity and resources through multi-nationality and high institutional ambiguityThe cases suggest that AGA and Autoliv’s institu-tional arbitrage responses were enabled by both theirsocial positions as boundary-spanning multina-tionals, and the particular field conditions includinghigh level of ambiguity in the host country insti-tutional environment (see Table 4; column 2, rows 1and 2). As the above examples demonstrate, aninstitutional arbitrage strategic response calls forexploiting differences in institutional environments.Specifically, the boundary-spanning multinationalposition gave the firms access to important resourcesin the form of alternative institutional templates (seeTable 4; column 2, row 5); it was thanks to thesetemplates that AGA was able to implement its gasdistribution format, and that Autoliv was able topropose the adoption of airbags. For example, AGA’s

gas distribution practices were institutionalized andtaken for granted in Western Europe both legallyand cognitively; at the same time, however, theyconstituted an institutional novelty to newly openedmarkets in Eastern Europe. In the case of Autoliv,airbags had become an established norm in theUnited States, but an emergent – and contested – ideain the European market. By leveraging institutionaland building on the transfer of institutional resou-rces, both of these actors were able to introduce theirnew products.To engage in arbitrage responses, however, MNEs

must be able to recognize the institutional ambigu-ities and voids in host countries. As a result, reflex-ivity constitutes an important mechanism alongsideresources, for MNEs engaging in institutional arbit-rage (see Table 4; column 2, row 3). In the specificcase of AGA and Autoliv, the two companies had thereflexive capability to effectively identify and lever-age the contradictions and ambiguities of theirspecific host country institutional environment.This reflexivity mechanism was enabled by the com-panies’ boundary-spanning multinationality, butalso by the contradictions and ambiguities of thehost country institutional environment. Hence,AGA’s and Autoliv’s ability to introduce new ideasthat challenged the existing institution was alsodependent on host country field conditions. In bothcases, host country institutional conditions exhib-ited considerable uncertainty and ambiguity, withparticular institutional voids: AGA was hence able tointroduce its new gas distribution format into EasternEurope precisely because hitherto taken-for-grantedpractices and regulations had lost their legitimacy.For Autoliv, the combination of institutional experi-ence from the US airbag market and the pre-existingexperiences in car safety, in conjunction with itsdiverse constituents’ inconsistent understandings ofairbags (car manufacturers opposing and customerssupporting them), opened up a unique opportunityto introduce airbags in Europe.In contrast to the case of institutional innovation,

where firms simply introduced new innovations ontop of pre-existing institutions or lack thereof, insti-tutional arbitrage relies on introducing practices inresponse to particular ambiguities and institutionalvoids in local field conditions. Importantly, the roleof local institutional ambiguities constitutes themain difference between innovation and arbitragestrategic responses. In the case of innovation, firmsdraw on their transfer of institutional resources frommultiple institutional environments to change extantinstitutions; institutional conditions are hence the



object of their acting and do not provide for responseopportunities as such. In the case of arbitrage, how-ever, the response opportunity arises from alreadypre-existing institutional ambiguities; in this case,local institutional conditions are hence the subject –that is, the source – of their agency.

Institutional CircumventionWhile institutional arbitrage connotes a responsewhere an MNE actively leverages differences in insti-tutional pressures, our third strategic response –institutional circumvention – constitutes scenarioswhere MNEs avoid institutional pressures. Whileextant studies in international management havehighlighted how firms avoid institutional pressuresby relying on firm-specific capabilities and assets, ourcase is somewhat different. The circumvention strat-egy constitutes a response where firms leverage both theambiguities of host country environments, and their socialposition as a foreign outsider. These factors enable thecircumvention strategy because they result in bothparticular role-expectations among domestic actors,and because they provide the MNEs with reflexivityand the ability to identify ambiguities in the hostcountry institutional environment. Among our cases,Citibank and The Hartford’s responses to host coun-try normative and cognitive institutions are clearexamples of a circumvention strategy.

Citibank’s introduction of loan syndication into JapanOne of the constraints Citibank faced when introdu-cing loan syndication onto the Japanese market wasthat the new practice ran counter to many of thedeeply institutionalized norms and practices ofJapan’s pre-existing lending format, known as themain-bank system. For example, while loan syndica-tion was predicated on joint lending by multiplebanks, fees for arranging banks and the trading ofloans on a secondary market, Japan’s main-banksystem was built on the notion of bilateral lendingwith low interest and no loan trading. Conse-quently, the new practice faced significant norma-tive and cognitive constraints among customerssteeped in the ways of the main-bank system; more-over, an arcane usury law raised the possibility thatthe new practice might even technically be illegal.While these institutional constraints prohibited

many of Citibank’s domestic competitors fromintroducing the practice, they did not prevent Citi-bank from doing so. A major reason for this wasambiguity in the local institutional environment.Specifically, while Japanese banks were expected tomaintain pre-existing practices, Citibank and other

foreign entities were seen as uncertain elements andhence given greater leeway to challenge local norms,practices and products. As one manager noted:

[Y]ou could be outside the system, but still viable…you couldsolve political problems in a way that Japanese bankscouldn’t…if you think of the domestic [Japanese] banks, ifone of them upset the apple cart, that would have been ahuge deal.

The difference in customer perceptions and expec-tations on the banks is also evident in the words of aJapanese bank manager:

[W]hen the foreign financial institutions did loan syndica-tions in Japan, customers did not expect these banks, theforeign banks, to take the same share of lending on theirbooks as the Japanese banks had. In the case of the foreignbanks, the role of arranger and lender was seen as separate.

Notably, the introduction of loan syndication wasnot the first time that Citibank had taken advantageof its position as an outsider firm; in fact, the com-pany had come to recognize that it played a particularrole or position in the local market, enabling it tooften introduce new product innovations. In sum,Citibank’s ability to evade and circumvent the nor-mative opposition to loan syndication was groundedin both institutional ambiguities in host countryfield conditions, and its outsider position as a foreignfirm.

The Hartford’s introduction of variable annuities intoJapanLike Citibank, The Hartford also faced considerableopposition to its introduction of variable annuities.While this product already existed on the Japanesemarket, it was largely characterized by stable annualreturns; by contrast, The Hartford’s variable annu-ities were investment-oriented and thus similar tomutual funds. The company also sought to sell themvia banks and securities companies. These noveltiesresulted in not only regulative constraints discussedunder the heading of institutional innovation, butalso normative opposition, particularly within theJapanese life insurance. One president of a Japaneselife insurance company even went as far as to writeto the US life insurance association, complainingthat The Hartford’s variable annuities were “invest-ment products and…not fit for a safety-orientedindustry like life insurance”.Like Citibank, The Hartford nonetheless deviated

from these expectations and norms, actively market-ing its investment products through banks andsecurities companies. The reason for this was the



significant uncertainty and ambiguity that surroundedboth its identity and stature as a life insurancecompany. As a foreign entity that had enteredJapan relatively recently, The Hartford was not anactive member of the Japanese Life InsuranceAssociation; nor had it adopted the strategy of alarge dedicated sales force preferred by many localactors. This set the firm apart from many of itsdomestic competitors. Moreover, with its empha-sis on products that were very similar to mutualfunds, the company often seemed less of an insur-ance firm and more of a securities company. As onemanager noted:

Do you see the word insurance in our name? No, it’s notthere…that’s because we are not an insurance company…weare an investment company which happens to have a lifeinsurance license, but we are not an insurance company,we’re an investment company.

The outsider position of The Hartford, and thedistance between its product distribution strategiesand those of the main-stay Japanese firms, is further-more evidenced in a response from a manager at aJapanese insurance company, who noted:

Hartford’s products are too risky…that’s their problem, youcan’t sell…insurance like that.

Another manager of a Japanese insurer explained:

That style of things doesn’t work here in Japan; it’s not howthis market functions.

Like Citibank, The Hartford managed to circum-vent pressures for institutional conformity by rely-ing on ambiguities and uncertainty in the insti-tutional setting. Notably, the firm’s ability to lever-age this heterogeneity was not based on its globalinstitutional experience, but rather on the localsubsidiary’s foreignness and outsider position in theJapanese institutional environment. Recall that Japa-nese life insurance companies had significant experi-ence of variable annuities and were also aware ofalternative sales formats used in the United Statesand other areas; yet they were deeply embedded inthe prevailing norms and logics of the Japaneseinsurance market, hence they were largely unwillingor unable to act on their knowledge. On account ofits foreignness and uncertain status, The Hartforddid not face the same isomorphic pressure to con-form to traditional practices of the Japanese lifeinsurance industry; as a result, the firm wasable to break from local norms and practices tointroduce the new practice.

Mechanisms and enabling factors of institutional cir-cumvention: Reflexivity and role expectations throughforeignness and reflexivity due to high institutionalambiguityThe circumvention response relied on leveragingboth high institutional ambiguities in local fieldconditions, and the outsider social position thatcomes from foreignness (see Table 4; column 3, rows1 and 2). These factors enabled twomechanisms thatwere crucial for institutional circumvention: reflex-ivity and role-expectations.As the preceding cases show, reflexivity was crucial

because it allowed the MNEs to recognize institu-tional voids and opportunities available in the hostcountry (see Table 4; column 3, row 3). In the case ofCitibank’s introduction of loan syndication, mostcustomers remained firmly embedded in the beha-viors and mind-sets of the pre-existing bilateral lend-ing system, even after the bank had helped establishnew regulations. Citibank’s reflexivity enabled it tosee beyond these constraining norms and mentalmodels, to envision alternative institutions and lend-ing formats. The Hartford’s introduction of variableannuities similarly involved identifying blind spotsfor new distribution norms, beyond the taken-for-granted norms of Japanese life insurers. Notably,Citibank and The Hartford’s reflexivity were notprimarily a result of the firms’ boundary-spanningsocial positions as multinationals. In fact, many oftheir domestic Japanese competitors were active in theinternational market; hence they too were aware ofalternative regulations, norms and mental models.What enhanced Citibank and The Hartford’s reflexiv-ity was instead their position as outsiders in the localJapanese market. By being less embedded in localnorms and mental models, the firms were able to lookbeyond prevailing institutional constraints. Reflexiv-ity was therefore enabled both by local ambiguitiesand their foreign and outsider position.Role expectations were also a particularly impor-

tant mechanism underlying circumvention res-ponses (see Table 4; column 3, row 4). When Japa-nese banks sought to introduce loan syndication, forexample, they faced considerable opposition fromdomestic clients, who viewed the new practice asillegitimate and falling outside the expectationsapplied to traditional Japanese banks. As a foreignbank, Citibank did not face these role expectations;it was therefore able to break from institutionalizedpractices in ways that Japanese banks could not (e.g.,by selling loans to third-parties). Similarly, TheHartford’s lack of embeddedness in the local Japa-nese life insurance industry made it easier for the



firm to act on the idea of selling insurance throughbanks; this idea was, by contrast, an anathema toJapanese insurers, who complained that “life insur-ance is a safety-oriented product” and should not besold through investment companies.Notably, transfer of institutional resources did

not constitute a key mechanism in circumventionresponses. Citibank and The Hartford’s strategywas to circumnavigate extant host country normsand cognitions, not change them. As a result, theydid not actively leverage institutional experience orknowledge available from their boundary-span-ning multinationality. This stands in stark contrastwith the institutional innovation responses theyemployed when dealing with the host countryregulative institutions, as discussed earlier. In thiscase, resource transfer of institutional experienceand knowledge was crucial, given that the under-lying approach was aimed at actively altering pre-existing laws and regulations.

Institutional AdaptationOur final strategic response corresponds to the well-established adaptation strategy in previous interna-tional business literature (Bartlett & Ghoshal, 1989).In this response, MNEs actively conform to hostcountry institutions. The primary mechanism thatdrives this response is the role-expectations the MNEfaces from foreign actors and they are in turnenabled by its social position as a foreign outsider.In our cases, AGA’s responses to host country cogni-tive institutions are a clear example of this responsetype and Autoliv’s response to normative institu-tions is also of this type.

AGA’s introduction of new industrial gas distributionformats into Eastern EuropeWhile AGA used an institutional arbitrage responseto introduce distribution formats taken fromWesternEurope, it also balanced this approach with explicitattempts to adapt its behaviors and strategies toprevailing local understandings and interpretations.In particular, the company recognized that not all ofthe practices used in Western Europe were applicableto its Eastern European markets; moreover, several ofthe Western European markets’ institutions servedto constrain the firms’ opportunities in the localmarket and put it at a disadvantage to other compe-titors. In response, AGA balanced its reliance onarbitrage with practical adaptations to local normsand practices. For example, in each market

MNE institutional advantage: How subunits shape, transpose and … · 2017. 9. 9. · ary...

Documents

Transcript of MNE institutional advantage: How subunits shape, transpose and … · 2017. 9. 9. · ary...