Acquisitions Versus Greenfield Investments - International Strategy and Management of Entry Modes

7/30/2019 Acquisitions Versus Greenfield Investments - International Strategy and Management of Entry Modes

1/18

Strategic Management JournalStrat. Mgmt. J., 23: 211-227 (2002)

DOI: 10.l002/smj.218

ACQUISITIONS VERSUS GREENFIELDINVESTMENTS: INTERNATIONAL STRATEGY ANDMANAGEMENT OF ENTRY MODESANNE-WIL HARZING*Faculty of Economics andCommerce, University of Melbourne, Parkville Campus,t\/letbourne, Victoria, Australia

This paper adds an im portant explanatory variable to the well-established list of factors shown toinfluence the choice between foreign acquisitions and greenfield investments: the internationalstrategy followed by the multinational company (MNC) in question. The MNC's internationalstrategy is subsequently linked to the management of the two different entry modes by showingthat differences in strategy are reflected in different headquarters subsidiary relationships foracquisitions and greenfields. Some aspects of this relationship are also shown to change overtime, a process that is mediated by the MNCs strategy. Copyright 2002 John Wiley & Sons,Ltd.

INTRODUCTIONThe choice of entry mode into foreign mar-kets has received a lot of attention from inter-national business researchers in recent decades.An expansion into foreign markets requires adecision on two related but distinct issues. First,a company has to choose between non-equityentry modes such as exporting through agentsand licensing, and equity-based entry modes,in which the local enterprise is either partiallyor wholly owned. Many studies have investi-gated factors that might influence the choice fordifferent entry modes, often focusing on threealternatives: licensing, joint ventures and whollyowned subsidiaries and usually underpinned byeither transaction cost theory or the Owner-shipLocationInternationalization framework(see, e.g.. Caves, 1982; Anderson and Gatignon,1986; Kogut and Singh, 1988; Gomez-Casseres,

Key words: acquisition: greenfield: international strategy:entry modeCorresp ondence to: Anne-Wil Harzing, Department of Manage-ment, Faculty of Economics and Commerce, University of Mel-bourne, Parkville Campus, Melbourne, Victoria 3010, Australia.

1989; Hill, Hwang and Kim, 1990; Agarwaland Ramaswami, 1992; Kim and Hwang, 1992;Erramilli andRao, 1993; Kwon and Konopa, 1993;Bell, 1996; Benito, 1996; Erramilli, 1996; Aroraand Eosfuri, 2000; Makino and Neupert, 2000; Panand Tse, 2000).Second, if an equity mode of entry into a foreignmarket is chosen, the issue of whether to acquirean existing local firm (acquisition) or to set up acompletely new plant (greenfield investment) hasto be decided. A substantial number of studies haveinvestigated factors that migh t influence this cho ice(see, e.g., W ilson, 1980; Eorsgren, 1984; Cave s andMehra, 1986; Kogut and Singh, 1988; Andersson,Arvidsson and Svensson, 1992; Hennart and Park,1993; Anderson and Svensson, 1994; Cho and Pad-manabhan, 1995; Padmanabhan and Cho, 1995;Hennart, Larimo, and Chen, 1995; Larimo, 1996,1998; Barkema and Vermeulen, 1998; Brouthersand Brouthers, 2000) and usually based their stud-ies on a transaction cost framework or derived alist of factors from the literature. As a result ofthese efforts, there is substantial agreement on thefactors that have an impact on the choice betweengreenfields and acquisitions. This study, however,will introduce a key explanatory variable that has

Copyright 2002 John Wiley & Sons, Ltd. Received 19 September 2000Einal revision received 30 July 2001


2/18

212 A.-W. Harzingnot been studied before in the context of the choicebetween greenfields and acquisitions: the M N C sinternational corporate strategy. A recent study(Davis, Desai, and Francis, 2000) found that busi-ness unit strategy was not a significant determinantof entry mode choice, but recommended studyingcorporate-level strategies such as global integra-tion. In our study, international corporate strategyis defined as the way in which the organizationpositions itself with regard to the global businessenvironment and creates and sustains competitiveadvantage across national boundaries. The first aimof this paper will be to explore how internationalcorporate strategy infiuences the entry mode choiceof the MNC, focusing on the distinction betweenglobal and multidomestic strategies.

As discussed above, a substantial number ofstudies have been published on both types ofentry mode decisions. Recently a number of stud-ies (Gannon, 1993; Sarkar and Cavusgil, 1996;Andersen, 1997, Harland and Wheeler, 2000) havereviewed the progress in the field and the theoret-ical and conceptual frameworks that are used toexplain the entry mode decision. Somewhat sur-prisingly, however, none of the earlier studies orthe review studies has posed the question: whathappens after the choice of entry mode has beenmade. Are greenfields and acquisitions managedin the same way or do headquarterssubsidiaryrelationships differ between the two entry modes?And does the way in these two distinct entry m odesare managed remain similar or does it change overtime? To our best knowledge there is not a singleprevious study that has investigated these ques-tions. The fact that most previous entry modestudies have used secondary data has made thistype of analysis very difficult. Our study combinessecondary and primary data on greenfields andacquisitions and can provide both sides of the pic-ture, using MNC strategy as the link between thechoice for a particular entry mode and the subse-quent managementoi the acquisition or greenfield.The second aim of this paper will be to analyzewhether greenfields and acquisitions are managedin different ways; i.e., whether the two types ofentry modes are characterized by different head-quarterssubsidiary relationships.

The remainder of this paper is structured as fol-lows. In the following section, we will introduce anew variableinternational strategythat mighthave an impact on the choice between acquisitionand greenfield entry mode. The introduction of

the strategy variable will then be followed bya set of hypotheses about the impact of strate-gic choices on the subsequent management ofacquisitions versus greenfields. We will also dis-cuss how we might expect the management ofthese two different entry modes to change overtime, taking the international strategy of the MNCinto account. A subsequent section describes ourresearch methodology, more fully explaining ourdata collection, measures, and statistical methods.We then report our results, offer a discussion oftheir interpretation, and finish the paper with aconclusion.

LITERATURE REVIEW ANDHYPOTHESESInternational strategy and the choice betweengreenfields and acquisitionsPrevious studies have identified and tested a sub-stantial list of variables that might have an impacton the choice between greenfields and acquisi-tions. In Appen dix 1 we have summarized theresults for six of the most commonly distinguishedvariables: R&D intensity, the degree of diversi-fication, the level of foreign experience, culturaldistance, the size of the foreign direct invest-ment in comparison to the size of the investingcompany, and the time of entry. These six vari-ables will be included as control variables in ourstudy. A variable that has received very little atten-tion in previous entry mode studies, but one thatmight have a big impact on a firm's choice ofentry mode is the firm's international strategy.Typologies of international strategy have receiveda lot of attention in the international managementliterature (for an overview see Harzing, 2000).Nearly all studies distinguish two different typesof international strategies: "global" and "multido-mestic/multinational", while many include a thirdhybrid strategy often called "transnational" andsome include an "international" strategy. In thisarticle we choose to focus on global and multido-mestic strategies only, because these two strate-gies are the most commonly accepted and clearlydefined. Global strategies are characterized by ahigh level of globalization of competition withnational product markets being interconnected anda focus on capturing economies of scope and scale.The dominant strategic requirement is efficiency.

Copyright 2002 John Wiley & Sons. Ltd. Sirat. Mgml. J.. 23: 211 -227 (2002)


3/18

Acquisitions vs. Greenfield Investments 213and as a result these companies integrate and ratio-nalize their production to produce standardizedproducts in a very cost-efficient manner. Sub-sidiaries typically fulfill a role as "pipeline" forheadquarters and they are not supposed to respondactively to local market demands. Multidomesticcompanies experience a lower level of global com-petition and compete predominantly on a domes-tic level, while adapting products and policies tovarious local markets. The company can be char-acterized as a decentralized network. Subsidiariesare relatively autonomous and are allowed to bevery responsive to the local market (Bartlett andGhoshal, 1992; Harzing, 2000).

The impact of the M N C s international strategyon the choice between greenfields and acquisitionscan be argued from two theoretical perspectives:the different firm-specific advantages associatedwith the two strategies and the different levelsof internal (parent) versus external isomorphismthat are portrayed by subsidiaries in companiesfollowing the two strategies.

Global and multidomestic strategies are associ-ated with different types of firm-specific advan-tages. Rugman and Verbeke (1992) link the own-ershiplocationintemalization theory of inter-national production to different types of interna-tional strategy. Their analysis takes as a startingpoint that foreign direct investment has been cho-sen as the most efficient mode of entry, henceinternationalization advantages are assumed to bepresent. They distinguish two types of ownershipadvantages, which they call firm-specific advan-tages (FSAs). The first are location-bound FSAswhose benefits depend on their being used in oneparticular location (or a set of locations). Theycannot easily be transferred and cannot be usedin other locations without significant adaptation.Nonlocation-bound FSAs do not depend on theirbeing used in one specific location. They canbe used on a global scale, because transferringthem to other locations can be done at low costand without substantial adaptation. With regard tolocation advantages they distinguish two sources:home and host country. Linking these two con-cepts to the two intemational strategies we havedistinguished, global companies tend to focus onthe exploitation of non-location bound home-basedFSAs , such as for instance a proprietary technol-ogy. They do exploit location advantages in hostcountries, but this is usually limited to the exploita-tion of low cost locations which allows globalCopyright 2002 John Wiley & Sons, Ltd.

companies to pursue their strategy based on costefficiency (Bartlett and Ghoshal, 1989). Buildingup a low-cost production site is easier when thesite can be set up from scratch, so that it can incor-porate the latest production technologies and canbe built to match the company's exact productionrequirements rather than having to accept exist-ingpossibly too large or inefficientoperationsin an acquired subsidiary. The core capabilities ofmultidomestic companies lie in the exploitation oflocation bound FSAs using host country specificadvantages. These companies have to deal withmarkets that require tailoring products and poli-cies to local circumstances. In order to be ableto do so companies need to be well aware oflocal circumstances and well-integrated into thelocal market. This will be easier to achieve byacquiring an existing company with a knowledge-able work-force and good connections in the localmarket, than by setting up a new subsidiary fromscratch.

Recently, several studies have recommendedtaking an institutional perspective to look at entrymodes (Rosenzweig and Singh, 1991; Haveman,1993; Davis et al., 2000). Institutional theoristsfocus on the impact of extemal institutions onorganizations to try to provide an explanationfor the high degree of homogeneity in organi-zational forms, behaviors and practices for dif-ferent firms (DiMaggio and Powell, 1983). Theprocess of homogenization is coined isomorphismand the extemal institutions can include the state,professions, interest groups, but also other orga-nizations within the firm's industry. In an inter-national setting this means that subsidiaries areconfronted with an extemal environment that couldinclude parent and host government, local inter-est groups and other organizations which maybe subsidiaries of other MNCs. However, sub-sidiaries of MNCs are also subject to institu-tional pressures from within the organization tobecome isomorphic to the parent organization'snorms (Kostova and Zaheer, 1999). Recent pub-lications have therefore suggested that subsidiariesof MNCs have to conform to both internal (theparent organization) and external (the host coun-try institutional environment) sources of isomor-phism (Rosen zweig and Singh, 1 991; Kostava andZahee r, 1999; Dav is et al., 2000). Although alllarger companies with different business units willexperience this tension between intemal and exter-nal isomorphism, it is particularly important for

Sirat. Mgmt. J., 23: 211 -227 (2002)


4/18

214 A.-W. HarzingMNCs. Porter (1986) and Bartlett (1986) havedeveloped these ideas in the intemational businessstrategy literature as the tension between forcesfor global integration and national responsiveness(Rosenzweig and Singh, 1991). The resultant inte-gration/responsiveness framework (Prahaiad andDoz, 1987) has become one of the comerstones ofthe intemational business strategy literature. Usingthis framework, companies following global strate-gies strive for a high level of integration and a lowlevel of local responsiveness, while the reverse istrue for multidomestic strategies. In terms of iso-morphism, global strategies will focus on intemalisomorphism, while multidomestic strategies willfocus on extemal isomorphism. Linking this withthe choice of entry mode, we argue that partic-ular modes of entry facilitate either internal orexternal isomorphism. Establishing foreign sub-sidiaries as a mirror image of headquarters or atleast making sure that key stmctures, policies andprocedures are similar is much easier to realizefor greenfields than it is for acquisitions. With theformer, headquarters can mold structures and poli-cies to their specific preferences, while the lattercome with established stmctures and policies thatmight be much more difficult to change. Align-ment with host country conditions (external iso-morphism) is much easier for acquisitions than itis for greenfields, since the former are local firmswith an established local network. Other thingsbeing equal, companies following a global strategywould, therefore, prefer greenfields, while com-panies following multidomestic strategies wouldprefer acquisitions.

In sum, both the FSA perspective and the iso-morphism perspective would lead us to expectcompanies following a global strategy to pre-fer greenfields over acquisitions, while compa-nies following a multidomestic strategy would pre-fer acquisitions over greenfields. Hence, in termsof the actual distribution of entry modes in ourempirical study we can put forward the followinghypothesis:

Hypo thesis 1: Relative to compa nies followinga multidome stic strategy, compa nies following aglobal strategy will have a higher proportion ofgreenfield subsidiaries, w hile compa nies follow-ing a multidome stic strategy will have a higherproportion of acquisitions relative to com paniesfollowing a global strategy.

Headquarterssubsidiary relationships ingreenfields and acquisitionsWe have argued above that the intemationalstrategy of the MNC will have an impact on thechoice of entry mode. Subsequently, we wouldalso expect this choice to have an infiuenceon the way subsidiaries are managed after theyhave been set up/acquired. We would henceexpect headquarterssubsidiary relationships todiffer between greenfields and acquisitions. Theheadquarterssubsidiary relationship can be seenas a classic control problem, whose attributes aresimilar to principalagent relationships (Nohriaand Ghoshal, 1994). Headquarters, the principal,cannot make all decisions because it does notpossess all the necessary knowledge or resources,but it cannot leave all decisions to subsidiariesbecause the interests of subsidiaries might bedifferent from that of headquarters or the MNCas a whole. Therefore, the key aspect of theheadquarterssubsidiary relationship is the wayin which headquarters ensures that subsidiariesare working towards common organizational goals.The different types of control mechanisms arethe tool that headquarters has to achieve thisalignment. Hence, the level of control exercisedby headquarters by means of the different typesof control mechanisms is the first element ofthe headquarterssubsidiary relationship that wewill investigate. As we will see below, thereis a range of control mechanisms available thatgoes beyond the level of autonomy granted tosubsidiaries. The second element that we willlook at is the level of expatriate presence insubsidiaries. Expatriates can perform many roles inthe headquarterssubsidiary relationship, amongthem control and knowledge transfer. The finalelement that we will study is the level of localresponsivenessin terms of local production,local R&D and adaptation of products andmarketing to local conditionsthat headquartersallows to the subsidiary. We will now discusshow we would expect these three elements ofthe headquarterssubsidiary relationship to differbetween greenfields and acquisitions.

With regard to the FSA perspective, we arguedthat companies following a global strategy prefergreenfield subsidiaries to exploit their non -locationbound home-based FSAs. Transfer of these FSAsis difficult when subsidiaries are allowed to operateindependently, so we would expect headquarters to

Copyright 2002 John Wiley & Sons, Ltd. Strat. Mgmt. J., 23: 211 -227 (2002)


5/18

exercise a rather high level of control over thesegreenfield subsidiaries. A high level of expatriatepresence might complement this high level of con-trol, while expatriates can also serve as the embod-iment of the FSAs to be transferred. Therefore, arelatively high level of expatriate presence is tobe expected in greenfields. Since the main role ofthese greenfields is to serve as a "pipeline" forHQ-based FSAs, it is unlikely that headquarterswould grant them a lot of opportunity to be locallyresponsive. Companies following a multidomesticstrategy were predicted to prefer acquisitions toexploit location-bound FSAs using host countryspecific advantages. In order to be able to do so,these acquisitions will need a certain level of inde-pendence, so they are unlikely to be strongly con-trolled by headquarters. Continued employment oflocal managers rather than their replacement withexpatriates will be preferred in order to tap intolocal knowledge, so the level of expatriate pres-ence is likely to be low in acquisitions. Finally,since FSAs are host country-based, it is likely thatheadquarters will expect these acquisitions to con-tinue any local production/R&D that existed beforethe take-over and/or will allow them to adapt prod-ucts/marketing to local circumstances.

With respect to the isomorphic perspective, weargued that companies following a global strat-egy would choose greenfields to facilitate inter-nal isomorphism. Exercising a high level of con-trol, either through the control mechanisms weidentified or through sending out expatriates, canfacilitate this intemal isomorphism. So the iso-morphic perspective joins the FSA perspective inpredicting a relatively high level of control anda relatively high level of expatriates for greenfieldsubsidiaries. An attempt to realize internal isomor-phism will not be helped by allowing subsidiariesa high level of local respon siveness, so just like theFSA perspective, the isomorphic perspective pre-dicts a relatively low level of local responsivenessfor greenfield. Companies following a multido-mestic strategy were argued to prefer acquisitionsthat would facilitate achieving the extemal isomor-phism that is important for this strategy. Alignmentwith host country cond itions is more difficult, how -ever, if the acquisition is strongly controlled oris not allowed to be locally responsive. It is eas-ier if the management of the acquisition is leftto local managers rather than to expatriates. Inconcordance with the FSA perspective, we expecta relatively low level of control and expatriate

Acquisitions vs. Greenfield Investments 215presence in acquisitions and a relatively high levelof local responsiveness. Hence:

Hypo thesis 2: Headq uarters' control over theirgreenfield subsidiaries will be higher than theircontrol over acquisition subsidiaries.Hypo thesis 3: Headq uarters will assign m oreexpatriates to top positions in their green-field subsidiaries than in their acquisitionsubsidiaries.Hypo thesis 4: Headq uarters will permit theiracquisition subsidiaries a higher level oflocal responsiveness than their greenfieldsubsidiaries.

Development of tlie headquarterssubsidiaryrelationship over timeAs we have argued above, companies followinga multidomestic strategy will prefer acquisitions,while companies following a global strategy willprefer greenfields. However, companies might be"forced" to accept a non-preferred entry mode,i.e., a greenfield for multidomestic companies andan acquisition for global companies. A reasonfor a company following a multidomestic strategyto decide to set up a greenfield could be thatno suitable take-over candidates are available inthe country in question. A reason for a companyfollowing a global strategy to acquire an existingcompany might for instance be a (temporary) lackof managerial resources or govemm ent regulationsin particular countries preventing new entries.

However, if companies are forced to enter amarket via a non-preferred entry mode, we wouldexpect that over time they would try to changethe headquarterssubsidiary relationship of thissubsidiary to make it resemble the headquar-terssubsidiary relationship of subsidiaries withtheir preferred entry mode. In this way, head-quarters might be able to overcome part of thedisadvantage of entering a market through a non-preferred entry mode. For example, if a companyfollowing a global strategy was forced to acquirean existing company in a particular country ratherthan set up a greenfield, it might try to increase itslevel of control over this subsidiary over time andlimit its local responsiveness by cutting back thelevel of local production. Hence:

Copyright 2002 John Wiley & Sons, Ltd. Strat. Mgmt. J., 23: 211-227 (2002)


6/18

21 6 A.-W. HarzingHypo thesis 5: Over time the characteristicsof the headquarters subsidiary relationshipin subsidiaries will converge towards thecharacteristics ofthe "preferred" entry mode.Hypo thesis 5a: In multidome stic comp anies thecharacteristics ofthe headqua rters subsidiaryrelationship of greenfields will come to resemblethose of acquisitions.Hypothesis 5b: In multidomestic companies the-characteristics of the headquarters subsidiaryrelationship of acquisitions will not change.Hypothesis 5c: In global companies the charac-teristics of the headquarters subsidiary rela-tionship of greenfields will not change.Hypo thesis 5d: In global comp anies the chara c-teristics of the headquarters subsidiary rela-tionship of acquisitions will come to resemblethose of greenfields.

METHODOLOGYData collectionData for this study were collected by means of anintemational mail survey between October 1995

and March 1996. This mail survey was conductedas part of a study that focused on control mecha-nisms in multinational companies. Questionnaireswere mailed to CEOs and Human Resource Man-agers at the headquarters of 122 multinationals andto the managing directors of 1650 subsidiaries ofthese multinationals in 22 different countries. Thisarticle only uses the data collected at subsidiarylevel. The overall response rate at subsidiary levelwas 20% , varying from 7. 1% in Hong K ong to42.1 % in Denmark. Table I summarizes the num-ber of respondents by industry, country of head-quarters and subsidiary country. The total numberof 287 subsidiary responses represents 104 differ-ent headquarters (85% of our population) and thenumber of responses per headquarters varies fromone to eleven.

MeasuresTo ascertain the entry mode acquisition vs.greenfield investmentrespondents were askedwhether the subsidiary had been acquiredby another owner after its foundation. Ouroverall sample included 97 acquisitions and 190greenfields. Respondents were also asked to statethe subsidiary's year of foundation and wh ere

Table 1. Num ber of respondents by industry, subsidiary country and HQ countryIndustry

Electronics, electrical equipmentComputers, office equipmentMotor vehicles and partsPetroleum (products)Food & beveragesPharmaceuticalPaper (products)Chemical (products)VariousCountry of locationof headquartersFinlandFranceGermanyJapanNetherlandsSwedenSwitzerlandUKUS A

Number ofrespondents412630203446255510


SubsidiarycountryArgentinaAustriaBelgiumBrazilDenmarkFinlandFranceGermanyHong KongIrelandItalyJapanMexicoNetherlandsNorwaySingaporeSpainSwedenSwitzerlandUKUS AVenezuela



7/18

Acquisitions vs. Greenfield Investments 217applicableacquisition by another owner. Inorder to test the hypotheses with regard tosubsidiary development, we calculated the timefor which subsidiaries had been under ownershipof headquarters at the time of the study. Forgreenfields this was simply the year of datacollection minus the year in which they wereestablished. For acquisitions this equated to theyear of data collection minus the year in whichtheir current owner had acquired them.

Four statements were constructed that measuredwhether competition was predominantly globalor local and whether the corporate strategy wasfocused on achieving economies of scale or onachieving local differentiation.' It was expectedthat global companies would be characterizedby global competition and a strategy to achieveeconomies of scale (Bartlett and Ghoshal's costefficiency), while multidomestic companies wouldpredominantly compete on a domestic level andstrive for national responsiveness. These questionswere subjected to cluster analysis and a two-cluster solution resulted in clusters that could eas-ily be identified as global and multidomestic (seeTable 2). Subsidiary managers, however, might notbe fully informed of the strategy applied by theMNC as a whole and their perception of this strat-egy might be influenced by their own characteris-tics. Therefore, for each headquarters we verifiedwhether subsidiaries were classified in the samecluster. This turned out to be the case in gen-eral, with a very limited number of exceptions thatinvolved for instance one out of five subsidiariesbeing in a different cluster. These divergent caseswere recoded to the cluster that contained themajority of subsidiaries. In total, 101 subsidiarieswere classified as multidomestic and 186 as global.

After a review of over 15 publications on con-trol mechanismscovering classic authors suchas March and Simon (1958), Thompson (1967),' The que stions used for the strategy construc t and the variousaspects of the headquarterssubsidiary relationship can befound in Appendix 2.

Child (1973, 1984), Ouchi (1977, 1979, 1980), andMintzberg (1979, 1983), as well as more recentauthors such as Martinez and Jarillo (1989), Hen-nart (1991) and Merchant (1996)four main cate-gories of control mechanisms were identified: per-sonal centralized control, bureaucratic formalizedcontrol, output control and control by socializationand networks, each being composed of 2 - 4 indi-vidual control mechanisms. To measure these dif-ferent control mechanisms empirically, we adaptedand supplemented the questions that were used byMartinez and Jarillo (1991). The items formed areliable scale (Cronbach's alpha = 0.74) and wereaveraged to form a composite score for the levelof control.

Three questions were used to assess the pres-ence of expatriates in a given subsidiary. Thesequestions asked respectively for the nationality ofthe managing director, the number of top five jobsheld by expatriates, and the total number of expa-triates working in the subsidiary. Since the numberof expatriates might be influenced by the size ofthe subsidiary, we calculated the number of expa-triates as a percentage of the total workforce of thesubsidiary. Because the items had different scalesthey were standardized before they were averagedto form a composite score for the level of expatri-ate presence.

Local responsiveness was measured with fouritems asking for the percentage of local R&D andlocal production incorporated in products sold bythe subsidiary and the percentage of products andmarketing that was substantially modified for thelocal market. The items formed a reliable scale(Cro nba ch's alpha = 0.73) and were averaged toform a composite score for the level of localresponsiveness.

Statisticat methodsTo explore the infiuence of the variables describedabove on the likelihood of an acquisition as foreignentry, we conducted a binomial logistic regressionanalysis. The entry mode is captured by a dummy

Table 2. ClusterCluster namesMultidomesticGlobalr-value

analysis of strategy variables (scale 1 - 5 )Global competition

3.194.08- 7 . 1 3 4 , 0 . 0 0 0

Domestic competition3.722.2714.807, 0.000

Differentiation3.813.30

4.387, 0.000

Economies of scale3.163.81

- 5 . 0 7 3 , 0.000Copyright 2002 John Wiley & Sons, Ltd.


8/18

218 A.-W. Harzingvariable that takes the value of one if the entryis made by acquisition and zero if the entry ismade by greenfield. In the binomial logistic modelthe probability of an acquisition is explained bythe variables R&D intensity, diversification, for-eign experience, cultural distance between invest-ing and target country, relative size of the invest-ment timing of the investment and intemationalstrategy. The regression coefficients estimate theimpact of the independent variable on the prob-ability that the entry mode is an acquisition. Apositive sign for the coefficient means that a vari-able increases the probability of an acquisition, anegative sign indicates the reverse. The model canbe expressed as:

where Y is the dependent variable (the selec-tion/occurrence of an acquisition in this case), Z isa linear combination of the independent variables

7. = P Q -\- fi\X\ -\- /32X2 - I - - I - /SpXn,where /So is the intercept, P\ ... (in the regres-sion coefficients and X\...Xn the independentvariables.The models were estimated with SPSS 10.0using the maximum-likelihood method. The nullhypothesis that all /3's, except j8o are zero can betested with the model x^- When the model x^ issignificant, this null hypothesis can be rejected.A test that a specific coefficient is zero can bebased on the Wald statistic. Significance levels ofseparate coefficients based on the Wald statisticare indicated in the models in Table 4. The partialcorrelation of each predictor variable with thedependent variable is indicated by R. R can rangein value, from 1 to + 1 . A positive value indicatesthat as the variable increases in value so does thelikelihood of the event occurring. If R is negative,the opposite is true.

To investigate the hypo thesized differencesin the headquarterssubsidiary relationship ofgreenfields and acquisitions, MannWhitneytests were used. Non-parametric tests were used,because the distribution of many of the dependentvariables included in the analysis was non-normal.Correlation analysis was used to investigate thehypotheses relating to subsidiary development.

R E S U L T S A N D D I S C U S S I O NInternational strategy and the choice betweengreenfields and acquisitionsTable 3 shows the means, standard deviations, andcorrelation coefficients for all study variables. Theresults of the binomial logistic regression are pre-sented in Table 4. Since some cases had missingvalues for one or more variables the total num-ber of observations is 277. The model has a highexplanatory power, with a high and highly signifi-cant x^- Another way to assess the performance ofthe maximum likelihood models is to measure thepercentage of correct observations and compare itto the classification rate that would be obtainedby chance (the baseline rate, which is equal toa^ -I- (1 a)^, where a is the proportion of acqui-sitions [34.3%] in the sample). As Table 4 shows,our model predicts the likelihood of an acquisitionbetter than a random model would, with a clas-sification improvement of 45.24%, which is wellabove the minimum improvement of 25% as sug-gested by Hair et al. (1995). The specificity (itscapacity to correctly predict greenfields) of themodel is very good to excellent (86.26%), while itssensitivity (its capacity to correctly predict acqui-sitions) is good (67.37%). Pseudo R-square mea-sures confirm that the model has a very goodexplanatory power.

Six of the variables in the model are signifi-cant: R&D intensity, level of foreign experience,cultural distance, relative size, year of investment,and strategy. Acquisitions are less likely for R&Dintensive firms investing in culturally distant coun-tries and more likely for firms with a high levelof foreign experience that follow a multidomesticstrategy, when the relative size of the investmentis large and when the investment is recent. Ascan be verified in Appendix 1, four of these vari-ablesR&D intensity, cultural distance, relativesize and year of investmenthave also receivedunambiguous support in previous studies, so wecan have confidence in both the explanatory powerof these variables and in the comparability ofour sample to previous entry mode studies. Ourresults with regard to intemational strategy sup-port our Hy poth esis 1: acquisitions are mo re likelyfor multidomestic companies and greenfields aremore likely for global companies. Below, we willlook at the implications of this in terms of theheadquarterssubsidiary relationship in green-fields and acquisitions.


9/18

Acquisitions vs. Greenfield Investments 219

XJCS

>

00o

ouT3cCO

1C

Qt/2

Od

o oe n d d

d dI d

CN enen -H p -Hd d d d dI I enCN p t^ O VO pd d dI I I

end2n

^ 'w ^

lO VO ~~ W") 00p CN iri CN 00d "*' d CNp vq Ov 3 end vd en iri o

oI

CNd00d

pdI

O N

-0.

o

d1

o1

CNd

d d d

^ en Ov CNd d d

00 p CN pd d dI I

od

od1gd

dsd

d1CNOd

CNd100od

od

sdod1od

od

odI I

o CN CNd d dO Ov Ou-i r-; _: d

in o oen en od ' ^ d

o .2

uS ou Sa. In

^ v= c 3O" c .i b ^ "O CN en 1^5 00

SJ

IIo

.2

dV

o g-V S

2-2

Copyright 2002 John Wiley & Sons. Ltd. Strat. Mgmt. J.. 23: 211 -227 (2002)


10/18

22 0 A.-W. HarzingTable 4. Logistic regression modelsVariable

InterceptR&D intensityDiversificationForeign experienceCultural distanceRelative sizeYear of investmentStrategy (multidomestic)

Coefficient(S.E)- 243 .633(35.9787)- 10 .9841(4.4552)- 0 .0281(0.0429)0.8432(0.3276)- 0 . 3 4 5 7(0.1325)0.2089(0.0894)0.1224(0.0180)0.7466

(0.3392)

Significance/R-value0.0000*"0.0068"- 0 . 1 0 7 00.51270.00000.0050"0.11390 .0045"- 0 . 1 1 6 20 .0097"0.09860.0000"*0.35140.0138*0.0894

Model x^N% correctBaseline rateImprovementSpecificitySensitivityCox & Snell R'Nagelkerke R^

114.9900.0000***27 7

79.78%54.93%45.24%86.26%67.37%0.3400.470

* " / ) < 0.001; " p < 0.01 , p < 0 . 0 5 , ^ p < 0 . 1 , all one-tailedPositive signs indicate a higher likelihood of acquisitions, negative signs a higher likelihood of greenfields.

Headquarterssubsidiary relationships ingreenfields and acquisitionsThe results of the MannWhitney tests are pre-sented in Table 5. The direction of the difference isas predicted for all three variables and is (highly)significant in all three comparisons. The data lendsupport to Hypotheses 2 to 4. Acquisitions expe-rience a lower level of control from headquarters.A separate analysis for all ten individual controlmechanisms showed that this difference in overalllevels of control was mainly due to a higher levelof autonomy for acquisitions (Z-value = 2.342,p 0.01, one-tailed) and a lower level of sharedvalues (Z-va lue = 3.156, /? = 0.001, one-tailed),standardization (Z = 1.864, p = 0.031, one-tailed)and informal com mun ication (Z -value = 1.601,p = 0.054, one-tailed). Acquisitions also show alower level of expatriate presence. This is causedby two factors. First, a higher percentage of acqui-sitions have no expatriates at all among theirworkforce. Second, for those acquisitions that doemploy expatriates, the various measures showlower expatriate presence than for greenfields.Finally, acquisitions show a significantly higherlevel of local responsiveness than greenfields. Itis remarkable that the smallest difference is foundwith regard to the most limited form of localresponsiveness: adaptation of marketing.

Overall, the results found in this section supportthe hypotheses that we put forward. Greenfields

Table 5. Differences in various characteristics of theheadquarterssubsidiary relationship between green-fields and acquisitionsVariable

Level of controlLevel of expatriatepresenceLevel of localresponsiveness

Z-score

- 1 . 8 0 1- 3 . 4 4 0- 3 . 4 4 7

ExpecteddirectionYesYesYes

Significance

0.0360*0.0005*"0.0005***

*" p < 0.001, " p < 0.01 , * p < 0.05, ' p < 0.1, all one-tailed

are more strongly controlled by headquarters thanacquisitions and have a higher level of expatriatepresence. For the subset of subsidiaries that haveexpatriates among their workforce, functions ofexpatriation that imply a dependent role on head-quarters are more important for greenfields thanfor acquisitions although the differences are onlymarginally significant.^ Greenfields are more likelyto be used in a "pipeline" role for headquarters.This is supported by the significantly higher level^ These functions are knowledge transfer (transfer of specifictechnological or managerial knowledge from headquarters tothe subsidiary, Z = - 1 . 2 5 1 , p = 0.105, one-tailed), positionfilling (sending an expatriate because headquarters feels localpersonnel is not qualified, Z = 1.591, p = 0.056, one-tailed),and training for a position at head quarters (Z = 1.719, p =0.043).

Copyright 2002 John Wiley & Sons, Ltd. Sirai. Mgmt. / , 23 : 211-227 (2002)


11/18

Acquisitions vs. Greenfield Investments 221of purchases from headquarters by greenfieldswhen compared to acquisitions (Z = - 3 , 5 4 3 , p =0,000). Differences in local responsiveness againsupport this picture, with local responsivenessbeing significantly higher for acquisitions. Theoverall picture that emerges from our results isthat of acquisitions as subsidiaries which havestronger external links with the local environmentthan internal links with headquarters, while thereverse is true for greenfields. We will now inves-tigate whether these differences persist over time.

Development of the headquarterssubsidiaryretationship over timeA correlation analysis between the time a sub-sidiary had been under headquarters ownershipand the various aspects of the headquarterssub-sidiary relationship discussed above showed sev-eral significant results. For greenfields there was anegative relationship between length of ownershipand expatriate presence ( -0 ,2 16 , /? = 0,003 , 2-tailed). This is to be expected since expatriates areoften used to set up operations, transfer knowledgeand train local managers and expatriate presencewould be expected to be much lower after this ini-tial period. For acquisitions there was a positiverelationship between the length of ownership andthe level of control (0,209, p = 0,042, 2-tailed)and a negative relationship with the level of localresponsiveness (:-0,224, p = 0,029, 2-tailed), Thiswould seem to indicate that acquisitions over timeare becoming more integrated in the corporate net-work and more similar to greenfields.

However our hypotheses proposed that theremight be a difference in the development of theheadqua rterssub sidiary relationship for multido-mestic and global companies. More specifically,we proposed that over time the management of a

subsidiary would converge to the management ofthe preferred entry mode option for that type ofstrategy. Hence in global companies the manage-ment of acquisitions would come to resemble themanagement of greenfields, while in multidomesticcompanies the management of greenfields wouldcome to resemble the management of acquisitions.We did not expect any changes in the managementof the subsidiaries that were established accordingto the preferred entry mode for both strategies.

Table 6 presents the results of this analysis. Wesee that, over time, the management of green-fields in multidomestic firms comes to resemblethe management of acquisitions only in respect ofexpatriate presence, while the level of control andlocal responsiveness does not change. Hypothe-sis 5a is only partially supported. As expected themanagement of acquisitions in multidomestic com-panies does not change much. The only variablethat shows a marginal change is the level of con-trol, A further analysis, however, showed that thishigher level of control was mainly due to a higherlevel of subsidiary managers going on corporatetraining programs (0,327, p = 0,005, 1-tailed) an da higher level of shared values (0,233, p = 0,05,1-tailed), An increase of these informal type ofcontrol mechanisms can be seen as a rather indi-rect consequence of the integration of acquisitionsin the MNC as a whole and is not likely to bethe result of a conscious effort of headquartersto limit the acquisition's "autonomy". The moredirect control mechanisms such as centralization,direct surveillance, procedures and standardizationdo not show any change. Hypothesis 5b is thussupported.

In global companies, no change is found for themanagement of greenfields in the level of controlby headquarters or the level of local responsive-ness. The level of expatriate presence, however.

Table 6, Develop ment of internal subsidiary characteristics over timeMultidomestic Global

Greenfield Level of control NCLevel of expatriate presenceLevel of local responsivenes s NCNCNegative (expected), p = 0,077* Negative (reverse), p = 0 , 004"

Acquisition Level of controlLevel of expatriate presence Positive (reverse), p = 0,099+NCLevel of local responsivenes s NC

NCPositive (expected), p = 0,084*NCNegative (expected), p = 0,019*

NC = no significant change,"* p < 0,001, * p < 0,01, > < 0,05, t p < 0.1 , all one-tailed.Copyright 2002 John Wiley & Sons, Ltd.


12/18

22 2 A.-W. Harzingis lower for subsidiaries that have been underownership from headquarters longer. This is notwhat we predicted for greenfield subsidiaries inglobal companies, since we did not expect theheadquarters subsidiary relationship to change inthis situation. However, as mentioned before adecline in expatriate presence over time is logical,especially in greenfield subsidiaries where expatri-ates are often used to set up subsidiaries. Over-all, we find a high level of support for Hypothe-sis 5c, For acquisitions in global companies boththe level of control by headquarters and the levelof local responsiveness change in the expecteddirection; i,e,, come to resemble the way green-fields are managed. There is no change in thelevel of expatriate presence. In contrast to acqui-sitions in multidomestic companies, however, thetype of control that does change in acquisitions ofglobal companies is not the more indirect infor-mal type of control, but the direct formal typeof control in the form of standardization (0,272,p = 0,037, 1-tailed) and procedures (0,377, p =0,013, 1-tailed), An increase in these types of con-trol mechanisms can clearly be seen as an attemptof headquarters of global companies to get theiracquisitions more in line with the company as awhole. Hypothesis 5d is partially supported.

Overall, we find a high level of support for the"no change" hypotheses and a more limited levelof support for the "change to the way of manage-ment of the preferred mode of entry" hypotheses.This could be partly due to small sample sizes(below 50) for the "less preferred mode of entry".However, the results do provide very interestingindications that MNCs might indeed try to changethe management of their subsidiaries over time.There is of course one important limitation to ourstudy in this respect. Although we have been ableto show that the time the subsidiary has been underheadquarters ownership is related to developmentof the headquarterssubsidiary relationship, ourresearch design does not allow us to verify whethe rindividual subsidiaries do experience this change.In order to do so, a longitudinal design would benecessary.

CONCLUSIONSThis paper showed that one variable that hasnot been considered in previous entry modestudiesthe strategy followed by the MNCs in

questionhas a significant explanatory powerin the choice of entry mode, with acquisitionsbeing more likely for multidomestic companiesand greenfields being more likely for globalcompanies, A limitation of our studywhich itshares with other entry mode studiesis thatwe have not been able to test whether ourmodel has normative merits. There is a modestindication, however, that choosing the preferredentry mode for the type of strategy has apositive performance effect. Subsidiaries with thepreferred entry modes had higher profits (p =0,059, 1-tailed) in comparison to other subsidiariesthan subsidiaries with the non-preferred entrymode. However, since performance was measuredin a rather indirect way (a 1 - 5 scale in comparisonto other subsidiaries [1 = m uch lower, 5 = muchhigher]) we cannot accept these results withoutfurther verification.

The analysis of the differences in the manage-ment of acquisitions and greenfields provided avery coherent pattern that was clearly linked tothe different strategies. Compared to greenfields,acquisitions were allowed to operate more inde-pendently with lower levels of control exercisedtowards them. This was also refiected in the lowerlevel of expatriate presence in acquisitions in gen-eral and the lower importance of functions of expa-triation refiecting a depen dence on headqua rters foracquisitions that did have expatriates among theirworkforce. Consistent with this picture, acqui-sitions were displaying a higher level of localresponsiveness in the form of local production andR&D and the modification of products and market-ing for local markets. Although empirical supportwas more limited, there were some indications thatover time the management of subsidiaries con-verged to the way subsidiaries with the "preferred"mode of entry for firms following a particularstrategy were managed, so that in multidomesticcompanies the management of greenfields becomesmore similar to that of acquisitions, while inglobal companies the management of acquisitionsbecomes more similar to that of greenfields.

Although our study was limited to the manufac-turing sector only, it has been the first to providedata for entry modes on both sides of the picture.It looked at factors infiuencing the choice betweentwo distinct entry modes and at the differences inthe management of both entry modes, including thedevelopment over the headquarterssubsidiaryrelationship over time. It showed that a comparison

Strat. Mgmt. J., 23: 211-227 (2002)


13/18

Acquisitions vs. Greenfield Investments 22 3of the differences in the characteristics of the head-quarterssubsidiary relationship between green-fields and acquisitions offers a useful addition tothe conventional entry mode studies and enhancesour understanding of the daily operations of differ-ent types of subsidiaries. We strongly recommendfuture researchers in this field to look beyond theinitial choice of entry mode to include a furtherexploration of the operational challenges of man-aging greenfields and acquisitions,

ACKNOWLEDGEMENTSThe author would like to express her gratitudeto the two anonymous SMJ reviewers for theirdetailed, constructive, and insightful comments,

REFERENCESAgarwal S, Ramaswami SN, 1992, The choice of foreignmarket entry mode: impact of ownership, location andinternationalization factors. Journal of InternationalBusiness Studies 23(1): 1-27,Andersen O, 1997, Internationalisation and market entrymode: a review of theories and conceptual framework.Management International Review 27(2) : 27-42,Anderson E, Gatignon H, 1986, Modes of foreign entry:a transaction cost analysis and propositions. Journalof International Business Studies (Fall): 1-26,Andersson T, Arvidsson N, Svensson R, 1992, Recon-sidering the choice between takeover and greenfieldoperations. Working paper. The Industrial Institute forEconomic and Social Research: Stockholm,Andersson T, Svensson R, 1994, Entry modes for directinvestment determined by the composition of firm-specific skill, Scatidinavian Journal of Economics96(4): 551 -560 ,Arora A, Fosfuri A, 2000, Who lly o wned subsidiaryversus technology licensing in the worldwide chemicalindustry. Journal of International Business Studies31(4): 535- 554 ,Barkema HG, Vermeulen F, 1998, International expan-sion through start-up or through acquisition: a learning

perspective. Academy of Management Journal 41(1):7 - 2 6 ,Bartlett CA, 1986, Building and managing the transna-tional: the new organizational challenge. In Compe-tition in Global Industries, Porter ME (ed,). HarvardBusiness School Press: Boston, MA; 3 6 7 -4 0 1 ,BartleU CA, Ghoshal S, 1989, Managing Across Borders:The T ransnational Solution. Harvard Business SchoolPress: Boston, MA,Bartlett CA, Ghosh al S, 1992, Transnational Matiage-ment: Text, Cases and Readings in Cross-Border Man-agement. Irwin: Homewood, IL,Bell J, 1996, Joint or single venturing? An eclecticapproach to foreign entry mode choice. Dissertation,Tilburg University,

Benito G, 1996, Ownership structures of No rwegianforeign subsidiaries in manufacturing. InternationalTrade Journal 10(2): 157-198,Brouthers KD, Brouthers LE, 2000, Acquisition of green-field start-up? Institutional, cultural and transactioncost influences. Strategic Management Journal 21(1):89-97,Caves R, 1982, Multinational Enterprise and EconomicAnalysis. Cambridge University Press: New York,Caves RE, Mehra SK, 1986, Entry of foreign multina-tionals into U,S, manufacturing industries. In Compe-tition in Global Industries, Porter ME (ed,). HarvardBusiness School Press: Boston, MA; 4 4 9 - 4 8 1 ,Child J, 1973, Strategies of control and organiza-tion behavior. Administrative Science Quarterly 18. (March): 1-17,Child J, 1984, Organization: A Guide to Problems andPractice. Harper and Row: London,Cho KR, Padmanabhan P, 1995, Acquisition versus newventure: the choice of foreign establishment mode by

Japanese firms. Journal of International Management1(3): 2 5 5 - 2 8 6 ,Davis PS, Desai AB, Francis JD, 2000, Mode of interna-tional entry: an isomorphism perspective. Journal ofInternational Business Studies 31(2) : 239-258,DiMaggio P, Powell W, 1983, The iron cage revisited:institutional and collective rationality in organizationalf\c\ds. American Sociological Review 48: 147-160,Eramilli MK, 1996, Nationality and subsidiary owner-ship patterns in multinational corporations. Journal ofInternational Busine.'ss Studies 27(2) : 225-248,Eramilli KM, Rao CP, 1993, Service firms' internationalentry mode choice: a modified transaction-cost analy-sis approach. Journal of Marketing 57(3): 19-38,Forsgren M, 1984, Foreign d irect investment in aninterorganizational perspectivea model for foreignacquisition behaviour, CIF W orking P aper no , 1,Uppsala University,Gannon M, 1993, Towards a composite theory of foreignmarket entry mode choice: the role of marketingstrategy variables. Journal of Strategic Management1(1): 4 1 - 5 4 ,Gomez-Casseres B, 1989, Ownership structures of for-eign subsidiaries: theory and evidence, Joumalof Eco-nomic Behavior and Organization 11: 1-25,Hair J, Anderson RE, Tatham RL, Black WC, 1995, Mul-tivariate Data Analysis with Readings. Prentice-Hall:Englewood Cliffs, NJ,Harland S, Wh eeler C, 2000, Reflections on entrymode choice. Paper presented at the 26th meetingof the European International Business Academy,Maastricht,Harzing AWK , 2000, An em pirical test and extensionof the Bartlett and Ghoshal typology of multinationalcompanies. Journal of International Business Studies31(1): 101-120,Haveman H, 1993, Follow the leader: mimetic isomor-phism and entry into new market. Administrative Sci-ence Quarterly 38: 593-627,Hennart J-F, 1991, Control in multinational firms: therole of price and hierarchy. Management InternationalReview, 31 Special Issue: 71-96,

Copyright 2002 John Wiley & Sons, Ltd. Smt. Mgm t. J.. 23: 211 -227 (2002)


14/18

224 A.-W. HarzingHennart J-F, L arimo J, Chen S-F, 1995, Does n ationalorigin affect the propensity of foreign investors toenter the United States through acquisition? In Pro-ceedings of the 21st EIBA conference, vol. 2, Urbino;4 0 1 - 4 2 2 ,Hennart J-F, Park Y-R. 1993. Greenfield vs. acquisition:the strategy of Japanese investors in the United States.

Management Science 39(9): 1054-1070,Hill CWL, Hwang P, Kim CW. 1990, An eclectic theoryof the choice of international entry mode. StrategicManagement Journal 11(2): 117-128,Kim WC, Hwang P, 1992. Global strategy and multina-tionals' entry mode choice. Journal of InternationalBusiness Studies (First Quarter): 2 9 - 5 3 .Kogut B, Singh H. 1988. The effect of national cultureon the choice of entry mode. Journal of InternationalBusiness Studies (Fall): 411 -4 32 .Kostova T, Zahe er S. 1999. Organizational legitimacyunder conditions of complexity: the case of the multi-national enterprise. Academy of Management Review24: 6 4 - 8 1 .Kwon Y-C, Konopa LJ. 1993, Impact of host countrymarket characteristics on the choice of foreign mar-ket entry mode. International Marketing Review 10 :6 0 - 7 6 ,Larimo J, 1993. Foreign direct investment behavior andperformance: an analysis of Finnish direct manufactur-ing investments in OECD countries. Acta Wasaensiano . 32, University of Vaasa,Larimo J, 1996, Mode of entry in foreign direct invest-ments: behavior of Finnish firms in OECD mar-kets. Proceedings of the 22nd EIBA Conference,vol, I, Stockholm School of Economics:409-434 .Larimo J, 1998, Form of investment behavior by Nordicfirms in world markets. Paper presented at the AIB1998 Annual Meeting, Vienna, Austria.Mak ino S, Neupert KE, 2000, National culture, trans-action costs, and the choice between joint ventureand wholly owned subsidiary. Journal of InternationalBusiness Studies 31(4): 705-714 ,March JG, Simon HA, 1958, Organizations. John Wiley:New York.Martinez JI, Jarillo JC. 1989, The evolution of researchon coordination mechanisms in multinational corpora-tions. Journal of International Business Studies 20(3):4 8 9 - 5 1 4 ,Martinez JI, Jarillo JC. 1991. Coordination demands ofinternational strategies. Journal of International Busi-ness Studies (Third Quarter): 429-444,

Merchant KA. 1996. Management Control Systems: Textand Cases. Prentice-Hall: Englewood Cliffs, NJ,Mintzberg H, 1979. The Structuring of Organizations.Prentice-Hall: Englewood Cliffs, NJ.Mintzberg H, 1983, Structure in Fives: Designing Effec-tive Organizations. Prentice-Hall: Englewood Cliffs,NJ.Nohria N, Ghoshal S. 1994. Differentiated fit and sharedvalues: alternatives for managing headquarters-sub-sidiary relationship. Strategic Management Journal15(6): 4 9 1 - 5 0 2 ,Ouchi W G. 1977. The relationship between organiza-tional structure and organizational control. Adminis-trative Science Quarterly 22(March): 95-112.Ouchi WG, 1979, A conceptual framework for the designof organizational control mechanisms. ManagementScience 25(9) : 833-848 ,Ouchi WG. 1980. Markets, bureaucracies and clans. Ad -ministrative Science Quarterly 25(March): 129- 141 .Padmanabhan P, Cho KR. 1995. Methodological issues

in international business studies: the case of foreignestablishment mode decisions by multinational firms.International Business Review 4(1): 5 5 - 7 3 .Pan Y, Tse D. 200 0. The hierarchical model of marketentry modes. Journal of International Business Studies31(4): 535-554 .Porter ME, 1986. Changing patterns of international com-petition. California Management Review 27(Winter):9-40.Prahalad CK, Doz YL. 1987. The Multinational Mission.Free Press: New York,Rosenz weig PM, Singh JV, 1991, Organizational envi-ronments and the multinational enterprise. Academyof Management Review 16(2): 3 4 0 - 3 6 1 .Rugman A, Verbeke A, 1992. A note on the transnationalsolution and the transaction cost theory of multina-tional strategic management. Journal of InternationalBusiness Studies 23(4): 7 6 1 - 7 7 1 .Sarkar M, Cavusgil T, 1996. Trends in international busi-ness thought and literature: a review of internationalentry mode research: integration and synthesis. TheInternational Executive XXXVIIl: 825-847 ,Thompson JD, 1967, Organizations in Action. McGraw-Hill: New York,Wilson B, 1980, The propensity of multinational compa-nies to expand through acquisitions. Journal of Inter-national Business Studies 12 : 59 -65 .

Zejan MC. 1990. New ventures or acquisitions: the choiceof Swedish multinational enterprises. Journal of Indus-trial Economics 38(3) : 349-355,


15/18

Acquisitions vs. Greenfield Investrrtents 225APPENDIX 1Factors influencing the choice between greenfields and acquisitionsCharacteristic Effect* Sup port found by(only studies with statistically significant results are included)R&D intensity

Degree of product diversification +

Foreign experience +

Cultural distance

Relative size of investment +

Time of investment +

Andersson et al., 1992; Andersson and Svensson, 1994;Brouthers and Brouthers, 2000; Cho and Padmanabhan,1995; Hennart and Park 1993; Hennart et al., 1995;Kogut and Singh, 1988; Larimo, 1996; Padmanabhan andCho, 1995Andersson et al., 1992; Brouthers and Brouthers, 2000;Caves and Mehra, 1986; Larimo, 1996; Larimo, 1998;Wilson, 1980; Zejan, 1990Barkema and Vermeulen, 1998; Larimo, 1993Andersson et at., 1992; Andersson and Svensson, 1994;

Caves and Mehra, 1986, Forsgren, 1984; Larimo, 1993Barkema and Vermeulen, 1998; Brouthers and Brouthers,2000; Larimo 1996; Larimo, 1998; Wilson, 1980Barkema and Vermeulen, 1998; Hennan et al., 1995; Kogutand Singh, 1988; Larimo, 1996; Larimo, 1998;Padmanabhan and Cho, 1995Brouthers and Brouthers, 2000; Caves and Mehra, 1986;Hennart and Park, 1993; Kogut and Singh, 1988;Padmanabhan and Cho, 1995Andersson et al., 1992; Andersson and Svensson, 1994;Barkema and Vermeulen, 1998; Larimo, 1996; Larimo,1998; Wilson, 1980; Zejan, 1990

*+Increases probability of acquisition.

In our study we used the following operationaliza-tions of the six control variables: R&D intensity: MNC R&D expenses divided byits total level of sales, Level of diversification: the number of different4-digit SIC codes in which the MNC operates, Foreign experience: the number of years thathad passed since the company established itsfirst foreign subsidiary, A logarithmic form of

this variable was used, since we expected yearsto have a decreasing impact on the firm's overallforeign experience, Cultural distance: Kogut and Singh's (1988)composite index was used to summarize thedifference between home and host country oneach of Hofstede's dimensions, Relative size of investment: number of employ-ee s of the subsidiary in question divided by thenumber of employees at headquarters. Since theresulting variable was badly skewed, the naturallogarithm of this variable was used as the final

measurement of relative size.

Timing of investment: year of foundation ofthe subsidiary andif applicablethe year ofacquisition.

APPENDIX 2Construct: corporate strategy(Likert scale 1-5, scale anchors: strongly dis-agreestrongly agree) Our company's strategy is focused on achieving

economies of scale by concentrating its impor-tant activities at a limited number of locations.

Our company's competitive position is definedin world-wide terms. Different national productmarkets are closely linked and interconnected.Competition takes place on a global basis. Our company's competitive strategy is to leteach subsidiary compete on a domestic levelas national product markets are judged too dif-ferent to make competition on a global levelpossible.

Copyright 2002 John Wiley & Sons, Ltd.


16/18

226 A.-W. Harzing Our company not only recognizes national dif-

ferences in taste and values, but actually triesto respond to these national differences by con-sciously adapting products and policies to thelocal market.

Construct: control mechanisms (Cronbach's alpha:0,74)(Likert scale 1 - 7 )Personal centralized control Autonomy (reversed scored): In some multi-

national firms, (strategic) decision-making islargely centralized at headquarters, in otherfirms subsidiaries have a large amount of auton-omy. In general, what is this subsidiary's auton-omy to decide its own strategies and policies?(scale anchors: very little autonomy very highautonomy) Direct supervision: In some multinational firmsheadquarters' mana gers strive for a close per-sonal surveillance on the behaviour of their sub-sidiaries. Other firms do not use this kind ofdirect personal supervision. Please indicate thedegree of personal surveillance tha t headq uar-ters' managers execute towards this subsidiary,(scale anchors: very little surveillance veryhigh surveillance)

Bureaucratic formalized control Standardization: In some multinational firms,all subsidiaries are supposed to operate inmore or less the same way. In other firms,such standardized policies are not required. Ingeneral, what is the degree of standardizationthat headqu arters requires from this subsidiary?(scale anchors: very low standardization veryhigh standardization) Formalization: Some multinational firms havewritten rules and procedures for everything andemployee s are expected to follow these proce-dures accurately. Other firms do not have suchstrict rules and procedures, or if they have,there is some leniency tow ards following them.Please indicate the kind of rules and proceduresthat headquarters exerts towards this subsidiary,(scale anchors: very loose/no procedures verystrict procedures)Output control Output evaluation: Some multinational firms ex-

ert a high degree of output control, by means

of a continuous evaluation of the results of sub-sidiaries. Other firms exert very little o utputcontrol beyond the requirement of occasionalfinancial reports. Please indicate the degree ofoutput control that headquarters exerts towardsthis subsidiary, (scale anchors: very little outputcontrol very high output control)

Planning: Some multinational firms have a verydetailed planning, goal setting and budgetingsystem, that includes clear-cut (often quantita-tive), objectives to be achieved at both strate-gic and operational level. Other firms have lessdeveloped systems. Please indicate the type ofplanning/goal setting/budgeting that headquar-ters uses towards this subsidiary, (scale anchors:very simple/no planning very detailed plan-ning)

Control by socialization and networks Shared values: Some multinational firms attach

a lot of value to a strong "corporate culture"and try to ensure that all subsidiaries share themain va lues of the firm. Others do not ma kethese efforts (or, having made it, have had nosuccess). To which extent do the executives in thissubsidiary share the compa ny's main v alues?(scale anchors: no shared values at all fullyshared values)

Informal communication: Some multinationalfirms have a very high degree of informal com-munication among executives of the differentsubsidiaries a nd headquarters. Other firms d onot foster that kind of informal comm unica-tion and rely exclusively o n formal comm unica-tion channels. Please indicate the level of infor-mal communication between this subsidiary andheadqua rters/other subsidiaries of the group,(scale anchors: no informal comm unication atal l daily informal communication)

Formal networks: Some multinational firms makeextensive use of comm ittees/task forces/projectgroups, both temporary and permanent, madeup by executives from different subsidiaries andheadqua rters. To what extent have this sub-sidiary 's executives participated in these groupsin the past couple of years? (scale anchors: noparticipation at all very highparticipation)

Strat. Mgmt. J., 23: 211 -227 (2002)


17/18

International management training: Some multi-national firms make extensive use of interna-tional (as opposed to purely national) mana ge-ment training program s. In these program s exec-utives from different subsidiaries and headqu ar-ters follow courses that deal mostly with thetransfer of compa ny-specific kn owledge. Wh athas been the participation of this subsidiary'sexecutives in these kinds of training program sin the past couple o f years? (scale anchors: noparticipation at all very high participation)

Construct: expatriate presence How many of the top five jobs in this subsidiary

are held by expatriates (employees on temporaryassignment from either headquarters or othersubsidiaries)? Tick boxes 05 .

Wha t is the nationality of the mana ging direc-tor of this subsidiary? Tick boxes: nationalityof parent/headqu arters country, nationality ofsubsidiary country, other (third country)nationality.

Please indicate the number of expatriates cur-rently working in this subsidiary.

Acquisitions vs. Greenfield Investments 227Construct: local responsiveness (Cronbach's alpha:0,73)(Six-point scale 0%, 1-25%, 2 6 - 5 0 % , 5 1 - 7 5 % ,7 6 - 99% , 1 0 0 % ) Please give your best estimate of the % of R&D

incorporated into products sold by this sub-sidiary that is actually performed by this sub-sidiary.

Please give your best estimate of the % of com-pany products sold by this subsidiary that havebeen manufactured (to any degree) by this sub-sidiary.

Please give your best estimate of the % of com-pany products sold by this subsidiary that havebeen created or substantially modified for thismarket.

Please give your best estimate of the % of mar-keting for compa ny products sold by this sub-sidiary that is consciously adapted to local cir-cumstances.

Copyright 2002 John Wiley & Sons, Ltd. Strat. Mgmt. J., 23: 211 -227 (2002)


18/18

Documents

Acquisitions Versus Greenfield Investments - International Strategy and Management of Entry Modes