The Effect of National Culture on the Choice of Entry Mode Author(s): Bruce Kogut and Harbir Singh Reviewed

work(s): Source: Journal of International Business Studies, Vol. 19, No. 3 (Autumn, 1988), pp. 411-432 Published by: Palgrave Macmillan Journals Stable URL: http://www.jstor.org/stable/155133 . Accessed: 06/04/2012 23:49
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Palgrave Macmillan Journals is collaborating with JSTOR to digitize, preserve and extend access to Journal of International Business Studies.

http://www.jstor.org

THE EFFECTOF NATIONALCULTUREON THE CHOICEOF ENTRYMODE Bruce Kogut*

Stockholm School of Economics

Harbir Singh**
Universityof Pennsylvania

of Abstract.Characteristics nationalcultureshave frequentlybeen claimed to influence the selection of entry modes. This article investigatesthis claim by developinga theoreticalargumentfor why culture should influencethe choice of entry. Two hypotheses are derivedwhichrelatecultureto entrymodechoice,one focussing on the cultural distancebetween countries,the other on attitudes towards uncertaintyavoidance. Using a multinomiallogit model and controlling for other effects, the hypotheses are tested by analyzing data on 228 entries into the United States market by acquisition,whollyownedgreenfield,andjoint venture.Empirical supportfor the effect of nationalcultureon entry choice is found.

Foreign direct investment into the United States has grown dramaticallysince the early 1970s. Accompanying this increase has been a growth of academic work studying the phenomenon.1Whereas impressive informationconcerning foreign direct investmentin the United States in general is available, there has been surprisinglyfew statistical investigations concerning the choice of entry modes. The objective of this article is two-fold. First, original data regarding the choice of entry mode by foreign firms is described in terms of country and industrypatterns. Second, the factors that influence the choice between joint ventures,wholly owned greenfield(i.e., start-up)investments,and acquisitions are analyzed statistically. In particular, the statistical investigation seeks to

at of Businessat the *BruceKogutis a VisitingProfessor the Institute International and School. Stockholm Schoolof Economics is on leave fromthe Wharton *"*Harbir of at Schoolof Singhis an AssociateProfessor Management the Wharton and Business. His priorresearchis on corporate acquisitions on top management in Jounal and the of incentives,andhas been published theAcademy Management
Strategic ManagementJournal.
Of the manyhelpfulcontributions, authors the wouldlike to acknowledge Anderson, Bowman, Erin Ned JeanFrancois JakeJacoby,Tom MarkCasson,Richard Hennart, Gatignon, Caves,WujinChu, Hubert referees. authors The thank DileepHurry, Ommer Khaw,Eirene Pugel,SteveYoung,andthe anonymous was for research assistance. Stevens their for by Funding theproject provided theReginald Chen,andCriag H. JonesCenter theWharton a of Schoolunder grantfromAT&T. Received: 1987& February 1988. 1988;Accepted: February September 1986,March April1986;Revised: 411

412

FALL1988 OF BUSINESS STUDIES, JOURNAL INTERNATIONAL

to regarding theirpropensities explaina strikingdifferenceamongcountries propensities modes.Thesedifferences country in enter acquisition by versus other in whichrelatesaspectsof a towards acquisitions examined a framework are operations.2 regarding governance foreign the of nation's culture preferences to between culture Thisarticle test represents firststatistical of the relationship the of patterns entrymodeswhile of and entrychoice as an explanation country for variables.Becauseour measureof controlling firm- and industry-level is validatethe culture derived fromthe indicesof Hofstede[1980], the results of Moreover, usefulness hisconstructs, though wasnotourprimary this intention. cost for thefindings suggestthattransaction explanations modeof entrychoice context.3 stemming theinstitutional cultural from and mustbe qualified factors by
A PREFATORY NOTE ON TERMINOLOGY

It is important the outsetto defineterminology. This articlelooks at three at kindsof entrymodes:acquisitions, ventures, greenfield and investments. joint in referto the purchase stockin an already of existingcompany Acquisitions in an amountsufficientto confercontrol.All of the acquisitions our study withtheremaining shares dispersed across consistof a controlling equityshare many investors.A joint ventureis the poolingof assets in a commonand and separate organization two or more firmswho sharejoint ownership by investment a is controlover the use andfruitsof these assets.4A greenfield investment newfacilities.Suchaninvestment be whollyowned in can start-up of the or a jointventure.Forpurposes simplifying exposition, classifyall we whicharewhollyownedunder and investments greenfield thosewhich start-up involveshared under jointventure. ownership as greenfieldand acquisition Manystudies,as discussedlater, have treated alternative beingonly a question representing entrymodes,withjointventures This approach of the degreeof ownership. impliesthatentryandownership involvetwo sequential decisions,the first decidingwhetherto investin new should facilitiesor to acquire existingones, the secondone on how ownership is on Whereas be shared. suchan approach clearlydefensible boththeoretical as we andempirical jointventures a choicemadesimultaneously grounds, treat modesof entry. with otheralternative it as can Ourreasoning be terselysummarized follows. Conceptually, could of well be arguedthatjoint venturesare not merelya matter equitycontrol, for a characteristics strategic appropriate certain butrepresent set of governance orforthetransfer tacitorganizational of cost ortransaction motivations knowledge [Kogut1987]. Jointventuresarevehiclesby which to sharecomplementary be or which couldnot otherwise shared to coordinate but distinctknowledge of to a limitedset of activities influencethe competitive positioning the firm. on consider ventures the sequential managers joint Empirically, evidence whether It is of interest, therefore, to, or simultaneous with, otherentrychoicesis slim. in and thatGatignon Anderson [1987],whoseresultsaredescribed moredetail well modelof entrychoicediscriminates between later,findthattheirstatistical wholly owned and sharedcontrolchoice of entries,but not betweenwholly

CHOICEOF ENTRY MODE

413

ownedandthe degreeof sharedcontrol.This findingsuggeststhatmanagers perceivethe choiceas betweenwholly ownedandjointventure(andpossibly otherentrymodes),withdegreeof ownership beingexplained otherfactors, by the due suchas perhaps bargaining powerof the parties.5 Consequently, both and to the above conceptual empiricalreasons,we framethe joint venture in with choice as madesimultaneously consideration otherentryalternatives.
THEORETICAL FRAMEWORK

The theoryunderlying approach our reflectsin someways a return an older to in line of thought theworkon foreigndirectinvestment. Sincethe publication of StephenHymer'sthesis in 1960, the economictheoryof foreign direct has investment beendriven by country-level not suchas differences variables, in interestrates, but by industry-and firm-level variables[Hymer1960]. variablesreflectbarriers entryandpatterns oligopolistic to of Industry-level behavior. Firm-levelvariables related the conceptof transaction are to costs, of the assetsbetweenfirmsis impeded market whereby transfer specialized by the failures,thus necessitating expansionof the firm (in some cases across in to the To borders) order internalize transfer. theextentthatthesamevariables to influence whether enterby foreign directinvestment, or licensing, exporting, the choiceof the modeof entryis jointlyandsimultaneously determined.6 in Becauseouremphasis thisarticle uponcountry in is patterns theentrymode we propensities, do not seek to developa full theoryof entrychoice. Rather, we concentrateon only those factors likely to affect nationalpatterns. on Observations differencesamong countriesin their propensities joint to venture,acquire,or invest in greenfieldsites have been madeby Robinson and [1961],Brooke Remmers [1972],Franko and [1976],andStopford Haberich to [1978]in relation the lowerfrequency overseas of jointventureactivityby firmscompared thatby European to American firms. In his studyon foreign acquisitions, Wilson[1980]foundthatthere weresignificantly different patterns of acquisition BritishandJapanese amongAmerican, corporations. A numberof previousstudieslend theoretical empiricalsupport the and to betweena firm'scountryof originand the mode of entry.Two relationship isolatethe influenceof culture entrymodepatterns. on studies,in particular, The investigations researchers the University Uppsalarelated at of by foreign directinvestment to patterns the "psychicdistance"betweencountries. By psychicdistance,it is meantthe degreeto which a firm is uncertain the of characteristics a foreignmarket. of Psychicdistance,theyreasoned, wouldbe in influenced differences the cultureand language the home andtarget by of countries. Similarly, Puxty[1979]speculated therelationship on between cultural and differences ownership overseassubsidiaries. policiesregarding Neither of thesestudies, laid however, outsystematically cultural how differences influence statistical entrychoices, or provided evidence. large-sample in We seek to explaindifferences country in propensities the choice of entry modesfromthepointof departure differences cultures in that amongcountries influencethe perception managers of regarding costs anduncertaint of the

414

FALL1988 BUSINESS STUDIES, JOURNAL INTERNATIONAL OF

constant Assumingrevenues alternative modesof entryinto foreignmarkets. the will acrossalternatives, managers choosethe entrymodewhichminimizes management of perceived costs attached the modeof entryandsubsequent to in have been shownto the subsidiary. Becausedifferences nationalcultures and practicesand employee resultin differentorganizational administrative it that distant twocountries, are expectations, canbe expected themoreculturally characteristics average[Bendix on the more distantare theirorganizational differentially factors influence 1956;Lincoln, Hanada Olson1981].If cultural and or of the perceived realcosts anduncertainty the modeof entry,thereshould in of exist country patterns thepropensity firmsto engagein one typeof entry modeas opposedto others. an existingforeignmanagement, Due to the difficultyof integrating already in culturaldifferencesare likely to be especiallyimportant the case of an have studieson mostly domesticacquisitions acquisition. Indeed,empirical and by costs are substantial are influenced what shownthatpost-acquisition Jemisonand Sitkin[1986] call the organizational of the two firms. They fit fit between cultural defineorganizational as "thematch administrative practices, firms"[Jemison of and practices, personal and characteristics thetarget parent in and Sitkin 1986, p. 147]. Sales and Mirvis [1984] document detail the whenbothfirmsdifferstrongly an administrative conflicts following acquisition in theircorporate cultures. In contrast the integration a to costs of an acquisition, joint ventureserves who tasksto local partners frequently purposeof assigningmanagement the with the arebetter ableto manage locallaborforceandrelationships suppliers, 1971; Stopford Wells 1972]. Thus, a and buyers,and governments [Franko problemsensuingfrom cultural joint ventureresolves the foreignpartner's controlandownership. Unquestionably, factors,thoughat the cost of sharing But distance betweenthe partners. a joint ventureis affectedby the cultural to such conflict shouldnot obscurethe originalmotivation choose a joint an of appeared venturebecausethe initialalternative integrating acquisition Of than tasks management to a localpartner. course, moredisruptive delegating distance thepartners, of ajointventure be troubled onlyby thecultural not may assets. A wholly owned but also due to concernsover sharingproprietary and avoidsboth the costs of integration conflictover greenfieldinvestment assetsby imposingthe management of the investing sharing proprietary style full while preserving ownership.8 firmon the start-up Forthisreason,we expectthattheuse of acquisitions foreignfirmsentering by the United States shouldbe dissuaded,the more distantthe cultureof the of countryof origin.9The following analysistests the relationship cultural two in modechoiceunder different to hypotheses: factors country patterns entry of between country theinvesting the distance 1. Thegreater cultural the of firm andthe country entry,the morelikely a firmwill choose over an acquisition. a jointventure wholly ownedgreenfield or 2. The greaterthe cultureof the investingfirm is characterized by regarding organizational the uncertainty avoidance practices, more

OF MODE CHOICE ENTRY

415

likelythatfirmwill chooseajointventure whollyownedgreenfield or over an acquisition. distant Hypothesis1 is derivedfrom the premisethat firms from culturally will attachgreater costs to the management acquisitions countries of relative thanfirmsfrom to joint venturesor to wholly ownedgreenfield investments similarcountries.10 Thesecosts maybe perceptual or accurate culturally only difficultiesof managing foreignworkforce a a in appraisals the increased of culturallydistantcountry.Hypothesis2 is derivedfrom the premisethat confront firmswith greater acquisitions over the management of uncertainty firmsfromcountries foreign operations." characterized relatively Therefore, by in will high uncertainty avoidance theirorganizational practices tendtowards or investments. jointventures greenfield
LITERATURE REVIEW

Therehavebeen severalpreviousstudieswhichhave foundthatentrychoice is influenced the firm'suncertainty the characteristics the targetted by over of countries.In this section, the centralfindingsare reviewed.These studies differin termsof whichentrymodesarebeingcompared are, as a result, and The for complexto compare. implications the choiceof entrymodearesorted out moreclearlyin a subsequent sectionwhen discussingthe relationship of the explanatory variables. A common themein a number studies beentheidentification perceived of has of as in uncertainty a functionof a firm's experience a country.In developing their theoryof internationalization based on the Uppsalaschool's work on psychic distance, Johanson Vahlne and the [1977]attributed evolutionary process by whicha firmadvances fromexporting jointventuring whollyowned to and in subsidiaries the reduction perceived regarding foreignmarket to risk the as a firmgains in experience. They did not, however,explorethe implications in for countrypatterns entrymodebehavior frompsychicdistances between nor countries, stipulate howtheexperience thefirmmitigates of clearly perceived in uncertainty arisingfromdifferences cultures. The influenceof firmexperience entrychoicehasplayeda prominent on role in severalof the studiesemploying Harvard the Multinational Enterprise Data Base. In their pioneeringstudy on the ownershipstructure American of multinational and firms, Stopford Wells [1972] foundjointventures,relative to whollyownedactivities,wereless likelyto be chosen,the morecentral the product the core businessof the firmandmoreexperience firmhadin to the the relevantcountry.Similarly,they found that marketing advertising and and the intensity,as well as research development intensity,discouraged use of jointventures. Dubin[1975]turned aninvestigation thedeterminantsforeign to of of acquisitions firmsover the periodof 1948 to 1967. Using bivariate by American crosstabulations withoutstatistical testing, he foundthat the tendencyto acquire fell with the size of the firm, its foreignexperience, if the targetcountry and was an LDC. His findings,thus, suggestan increasing of acquisitions use the

416

JOURNAL OF INTERNATIONALBUSINESS STUDIES, FALL 1988

betweenthe homeandhost countries and lowerthe cultural physicalbarriers the andthe moreexperience firmhas in the foreignmarket. Data Enterprise Base a Davidson[1980]analyzed versionof the Multinational of the from1967 to 1975 andtraced establishment foreign whichwas updated of the from their inception.Through identification statistically subsidiaries 1) he correlations, foundthreepatterns: thatfirmswill morelikely significant have invested in invest where they or their competitors the same industry are similarcultures a preferred whichhavereputedly before;2) thatcountries in firm-levelexperience a countryof and target investment; 3) thatprevious likelihoodof no matterif licensingor joint venture-leads to an increasing for whollyownedinvestment laterentries. modeis influenced that suggest,therefore, thechoiceof entry Theabovestudies However,becausethe statistical and differences firmexperience. by cultural studiesby Dubin[1975] andDavidson[1980] did not test theserelationships could patterns for the variables, explanation country whilecontrolling other for The relationships. firstis the to be considered be derivedfromtwo spurious countries of involvement particular greater the between historically relationship and internationally the influenceof firm experienceon entry choice. The in among composition differences industrial between secondis therelationship expenditures in and of differences the intensity marketing research countries, across industries,and the influenceof the desire of firms to controlthe assets. or extensionof marketing- research-intensive international of the In the threestudiesthat investigated determinants entrymode while has for othervariables, experience not, however,been controlling statistically by entry acquisition in mode.Analyzing to proven be instrumentalchoiceof entry firms,Wilson and for versusgreenfield American, British,German Japanese the did influence decisionto that [1980]reported experience not significantly over establishment an acquisition.12 by investin foreigncountries a greenfield The decision to acquirewas found, instead, to co-vary positively with of and negativelywith the proportion recentlyestablished diversification was and to subsidiaries total establishments with whetherthe targetcountry an LDC. firmsto Caves and Mehra[1986] analyzed138 decisionsof non-American a versusacquisition through qualitative enterthe UnitedStatesby greenfield choice model with industry-and firm-level variablesas the independent for variables,while controlling joint ventures.Theirdatawas drawnfroma for announcements theyearsof 1974 to 1980. Theirresults listingof reported influenced in investments a country that the disconfirmed hypothesis previous into overacquisition theUnited a foreignfirm'sdecisionto enterby greenfield of States.Rather, theyfoundthatsize of theforeignfirm,diversity its product influenced and positively significantly range,andits degreeof multinationality durable industries goodswere the decisionto acquire.In addition, producing because,argue more likely to be characterized entrythrough acquisition by of requires goodsto localconditions CavesandMehra,the adaptation durable thanthrough investment. greenfield skills bettercaptured acquisition through

CHOICEOF ENTRY MODE

417

werefoundto be negatively Jointventures related thechoiceof acquisitions, to thussupporting premiseof this paperthatacquisitions joint ventures the and modesof entry.No control madefor country-level aresubstitute was variables. Ina recent and study,Gatignon Anderson 1267 decisions [1987]reanalyzed entry from the Harvard Multinational Databasefor the years 1960 to 1974, also choice model. As described using a quantal earlier,theiranalysisof entryas a three-way decision between whollyownedandvarious levelsof jointventure controlwas not able to discriminate betweenthe chosendegreeof joint well Theirbinomialtest of whollyownedversusjointventure, venture ownership. the and however,confirmed Stopford Wells [1972] bivariate results.Wholly owned subsidiaries(greenfieldand acquisition) were favored over partial the the ownership, greater R&Dandadvertising of intensity the foreignfirm. They also found supportthat the degreeof multinationality a negative had to effecton the likelihood jointventure.13 Theirdummy variables regions for tendedto show strongcountrypatterns. Based on the positiverelationships betweenR&D and marketing/advertising intensityto wholly owned entries, a transaction theoryof entrychoice is supported.14 cost they concludethat In summary, literature datehas foundthatuncertainty the foreign the to over marketinfluencesmanagers decisionson how to invest overseas,thatthere areclearbutunexplained in country patterns the selectionof entrymodes,and thatbothfirm- andindustry-level variables relatedto the choiceof entry are mode.The previous literature not, however,clearlyextrapolated the has from research cultural on traitsto implications countrypatterns the relative for in use of different entrymodes,norhasit testedthe relationship betweencultural factorsandentrymodechoicewhile controlling otherfactors. for This papertests explicitlythe influenceof countryculturalcharacteristics, toward attitudes including uncertainty, uponthe choice of the mode of entry intothe UnitedStates.Though variables currently country-level economic are discounted explaining firmsinvestoverseas,cultural as why differences among countries how this investment play a role, this articlecontends,in explaining is channeled.
DESCRIPTION OF ENTRY MODE PATTERNS

Becausedatacomparing entryactivitiesof foreignfirmsin the UnitedStates are not easily available,it is worthwhile reportthe patterns to foundin our samplebefore turningto statisticaltests of the above hypotheses.Whereas data aggregated on foreign are acquisitions routinely fromDepartment available of Commerce similar dataforjointventures generally are publications, lacking. In part,this imbalance be explained the significance acquisitions can by of as a modeof entryfor foreignfirmsintotheUnitedStates.Fortheyearsbetween 1976 and 1983, acquisitions were responsible over 50% of the foreign for in directinvestment the UnitedStates,risingas highas 79%of thetotalvalue in 1981.15 On the otherhand,dataon jointventuresas a modeof entryinto the United Statesis notaggregated published theDepartment Commerce. and by of While

418

JOURNAL OF INTERNATIONALBUSINESS STUDIES, FALL 1988

TABLE 1 Distribution of Modes of Entry by Industrial Sector

Joint Ventures 18 3 2 25 3 2 2 4 24 25 10 3 1 8 4 12 147 Acquisitions 35 26 15 35 3 20 8 7 14 21 3 10 4 17 30 25 274 Greenfield 2 5 3 17 2 6 4 0 13 13 10 0 2 3 0 5 85 Total N 55 34 20 77 8 28 14 11 51 59 23 13 7 28 34 42 506

Resource Paper Chemical Petroleum Rubber Primary Metal MetalFabrication Machinery Electrical Equipment Transportation Instrumentation Other Manufacturing Communication Wholesale Financial Services Other Services Total

of to it is thusimpossible have a value estimate jointventures,it is possible, to of the sourceslisted in the appendix, describethe frequency the basedon mode of entry across industriesand countries.This data is availablefor not greenfield,andjoint ventures,as well as otherinvestments acquisitions, in included this study. and of Table1 providesa breakdown joint ventures,acquisition, greenfield in for by industry theyears1981 to 1985. Thereis a cleardifference industry are morefrequent amongthemodesof entry.Jointventures relatively patterns andelectric nonelectric and inpharmaceuticals/chemicals machinery. Acquisitions in occur primarily naturalresources,financialservices, and miscellaneous are industries.Chemicaland electricalmachinery especially manufacturing At level of aggregation, for investments. a higher industries greenfield attractive morecommonthanothermodesof entryin tend acquisitions to be relatively sectorsof the economy. nonmanufacturing is The country pattern given in Table2. Again, thereare strongdifferences 46 are amongthemodesof entry.ForJapan, of its 114 entries jointventures.16 firmshave a high are Whereas acquisitions not common,Japanese Japanese and, of proportion the wholly owned greenfieldinvestments.Scandinavia United also Kingdom represents joint France, leantowards ventures. especially with the remainder the otherextreme;111 of its 141 entriesare acquisitions, and jointventures greenfield. evenlydividedbetween in in showcleardifferences countiy Thetrends oursample regarding propensities the selectionof the mode of entry. It is unclear,however,whetherthese for is whenthe relationship controlled firm- andindustryare patterns robust by pattern generated differences is levelfactors. couldwell be thatthecountiy It

CHOICE ENTRYMODE OF TABLE 2 Joint Ventures and Greenfield Entry by Country of Corporate Headquarters
Joint Ventures United Kingdom Japan Scandinavia Switzerland Germany France Italy Netherlands Belgium Malaysia S. Africa Canada Other Total 15 46 9 4 6 23 4 6 5 1 1 13 14 147 Acquisitions 111 35 5 20 10 6 3 24 10 1 0 28 20 274 Greenfield 15 33 4 3 8 4 1 7 2 0 2 3 5 85

419

Acquisitions,

Total N 141 114 18 27 24 33 8 37 17 2 1 45 25 506

in the sectoral characteristics foreigndirectinvestment of acrossthe countries of origin. The next section gives a formalstatisticaltest to determine the the factorsinfluencing choice of entry.
SELECTION OF VARIABLES

The hypothesesto be treatedposit that the choice of entryis significantly influenced the cultural characteristics the homecountry the investing of of by effects of the relationship firm- and of firm. Because of the confounding with countryidentification, is not possibleto test variables industry-level it for for country effectswithoutcontrolling otherinfluences.Consequently, the will investigate followingspecification: statistical the analysis firm Entry Choice = f(culturalcharacteristics; variables,industry variables) In previous studies,a number firmandindustry of variables havebeen tested and shown to be significantin explainingthe mode of entrychoice. These studiesindicateseveralproxies. Because, as discussedbelow, acquisitions form the baselinecase, we discuss the relationship these proxiesto the of variablein the contextof choosinga joint ventureor greenfield dependent relativeto acquisition. ourinterest in controlling specification As is for error, we merelysummarize conventional the the existingliterature of arguments on theexpected between control the variables thechoicevariables. and relationships
Firm-level Variables
Diversification [Diversified].

Dubin[1975], Wilson [1980], and CavesandMehra[1986] have foundthat firms following diversification strategiesare more likely to entera foreign countryby acquisition over greenfield.The presumable explanation this for

420

FALL1988 BUSINESS STUDIES, OF JOURNAL INTERNATIONAL

management on firmsare competing superior is pattern thatdiversified and/ are and, industries, therefore, notconcerned in efficiencies mature or production level. Analogously, or with a de novo transfer a product innovation brand of relativeto firmsshouldbe morelikely to engagein acquisitions diversified jointventures.
CountryExperience [Experience].

entrymode in the same country The effect of previousentryon subsequent studiesto be significant.17 multivariate has not been shownin large-sample to we theoretically, can expectthatthe propensity jointventure Nevertheless, shoulddeclineas a foreignfirmlearnsmoreaboutthe relativeto acquisition it It local environment. can also be expectedas a firm picksup experience, with relative jointventuring to its is morelikelyto increase use of acquisitions local partners.
MultinationalExperience [Multinational].

on to Contrary theirexpectation the sign of the coefficient,CavesandMehra in of (i.e., the number countries which a [1986] foundthatmultinationality with correlated the choiceof acquisition is firmhas subsidiaries) significantly of One overgreenfield.18 interpretation this findingis thata firmwith greater and is international experience ableto beartheriskof an acquisition to integrate of Along these lines, the greater nationality. subsidiaries diversemanagerial the the multinationality, greatera firm's ability to acquire;the lesser the the multinationality, morelikely a firmwill sharethe risksandmanagement should,thus,favorthe a through jointventure.Multinationality responsibility abilityto acquire.
Asset Size [U. S. Asset Size and Non-U. S. Asset Size].

its It standsto reasonthatthe largerthe investingfirm, the greater abilityto acquire.Despite the logic, the empiricalevidenceis mixed. Dubin [1975] morefrequently large than relatively firmstended acquire to foundthatsmaller firms, thoughhe did not controlfor other factors. In his cross-sectional Dubin'sfindings.However,thesestudiesdrew tests,Wilson [1980]confirmed of data of the largestcorporations the UnitedStates and other upon entry their CavesandMehra [1986]studydidnotrestrict attention countries. European Theirresultsshowedthatthe size of the to entriesof the largercorporations. over to related entryby acquisition firm entering is positivelyandsignificantly and morefinancial managerial require generally acquisitions Because greenfield. be firm'sassetsshould positively size resources jointventures, of theforeign than are to with correlated thetendency acquire. Conversely, acquisitions discouraged, size. targetfirm, or investment the largerthe assetsof the American partner,
Variables Industry-level IndustryVariables [R&D and Advertising].

differin theirindustrial is for Oneexplanation thecountry pattern thatcountries will modes be influenced thecharacteristics and structures thatchoiceof entry by confirming importance, their literature Becauseof a substantial of theindustry.

CHOICEOF ENTRY MODE

421

industryR&D expenditures sales and industrymedia and advertising to to expenditures sales were chosen as control variables in the statistical investigations.19 Data on both variablesare taken from the FederalTrade
Commission's Line of Business study for 1975.20

of the Conventionally, relationship thesevariablesto entrychoice is said to in to discourage ventures order preserve joint assets proprietary andto discourage unrelated The acquisitions. previous studieshaveassumed, empirical however, foreign entry was usually for the purposeof marketaccess or low cost manufacturing. Clearly,foreignentryintotheUnitedStatesmaybe motivated in orderto sourcetechnology purchase or brand labels. Themorediversemotivesof investing theAmerican in economymakeit more difficult signthestructural to variables. example, For firmsfromR&D-intensive industries mightjoint ventureif they possess the requisitetechnologies but lack the marketing depth.Or they may tend to acquireif they are investing for technology sourcing.Similarly, firmsfrommarketing-intensive industries mightengagein a jointventureif they possessthe brand labelbut lack other resources alongthevalue-added chain.Ortheymayacquire theyareinvesting if for market penetration lack labelrecognition. and and Stopford Wells [1972] foundthatAmerican firmspursuing advertising-intensive an tendto strategy full ownership theiroverseassubsidiaries. of Theirdatais drawn,however, froma timewhenAmerican firmswereinvesting overseas with clearstrategic For advantages. our study,it is equallylikelythatforeignfirmsareinvesting in the United States for technologyand brandlabel acquisition for the as of assets. No prediction made, therefore, exploitation theirproprietary is on the signs of the coefficientsforR&D andAdvertising.2'
Sectoral Dummies [Manufacturting Services]. and

Two sectoral dummies usedin orderto control otherexogenouseffects are for notcaptured theR&DandMarketing by variables. Thesedummies required are because thereareclearpatterns themodesof entryacross in services,extractive, andmanufacturing industries we wish to controlfor sectoraleffects not and variables.(See Table 1.) BecauseJapanese captured the structural by firms areactivein jointventures manufacturing, wouldbe a bias towards and there the in overstating Japanesecontribution the total numberof manufacturing entriesand in joint ventures.To avoid a bias, sectoraleffects are controlled for by using dummies whetherthe entryis in manufacturing in services. or
Variables Country-level

As notedearlier,previous studieshavepinpointed uncertainty a significant as influence decision.Whereas upontheinvestment has uncertainty beenmultiply one concerns ability theforeign the interpreted, interpretation of firmto manage the local operations its subsidiary. perceived of The abilityto managemaybe influenced two considerations, concerns absolute one by the cultural attitudes towardsuncertainty avoidance,the second concerningthe relativecultural distancebetweenthe countryof the investingfirm andthe country entry. of

422

JOURNAL OF INTERNATIONALBUSINESS STUDIES, FALL 1988

equations of in are Bothconsiderations proxied thespecification theregression distance. and avoidance cultural entitled uncertainty the through use of variables from are distance derived and avoidance cultural for Themeasures uncertainty cultures in foundthatdifferences national [1980].Hofstede theworkof Hofstede along four dimensions.These dimensionswere labeled vary substantially uncertaintyavoidance, individuality, tolerance of power distance, and for scalesfor countries each ordinal Hofstedecreated masculinity-femininity. factoranalysisof questionnaires basedon a standardized of thesedimensions employeesin more between1968 and 1972 to 88,000 national administered Bias for of than 40 overseas subsidiaries a majorAmericancorporation. controlled. As was in differences occupational positionsamongsubsidiaries it interval, by separated a four-year of the studyconsisted two questionnaires was possible to test for the reliabilityin scores over time; only questions in than.5 correlation scoreswereusedto derivethe scales. showinga greater of for The indicesof Hofstedecan be criticized a number reasons,especially of and of validity the dimensions themethod constructing the regarding internal Whereasthe criticismhas a soundbasis, Hofstede'sstudyhas the scales.22 of namely,the size of the sample,the codification some appealing attributes, in on index, and its emphasis attitudes the traitsalong a numerical cultural for conservative, if they Our workplace. use of the indicesare, furthermore, and they are less likely to be foundsignificant with the are poorconstructs, a prioripredicted sign. variablesto analysisused two cultural Based on these scales, the statistical test the two hypotheses. Distance[Cultural Distance]. Cultural of distantthe country the investing that We hypothesize the moreculturally firmfromtheUnitedStates,themorelikelythechoiceto setup ajointventure. basedon the deviation indexwas formed indices,a composite UsingHofstede's dimensions (i.e., powerdistance,uncertainty alongeach of the fourcultural of and avoidance,masculinity/femininity, individualism) each countryfrom in for werecorrected differences the The theUnitedStatesranking. deviations Algebraically, and averaged. of variances each dimension thenarithmetically we builtthe followingindex:
4

CDj -

i=l

{(Iij - Iu)2/J7}/4

and dimension jth country, whereIij standsfor the indexfor the ith cultural is the varianceof the index of the ith dimension,u indicatesthe United Vi fromtheUnitedStates. difference thejthcountry of States,andCDjis cultural Thoughthe scaling methodimposesweightsbased on index variance,any theoretically error measurement cannotbe expectedto be correlated resultant of and variables shouldreducethe significance the withthe otherindependent statistical relationships.

MODE CHOICE ENTRY OF TABLE 3 Summary of Predicted Signs

Variable Diversified Experience Multinational U.S. Asset Size Non-U.S. Asset Size R &D Advertising Manufacturing Services Distance Cultural Avoidance Uncertainty *NP- No Prediction JointVenture

423

Greenfield

+ NP* NP NP NP
+ +

+ NP NP NP NP + +

Uncertainty Avoidance [Uncertainty Avoidance]. Uncertainty avoidance should not be understood as referring to the individual's willingness to bear risk or as the risk profile of a firm regarding its product strategy. Rather, the elements making up the dimension are organizational and managerial in character. The construction is fortunate for our purposes, as we wish to isolate the influence of cultural attitudes towards uncertainty over organizational functions, such as employment relations. The more uncertainty avoiding a culture tends to be, the less attractive is the acquisition mode due to the organizational risks of integrating foreign management into the parent organization. The above discussion is summarized in Table 3. MODEL SPECIFICATION The decision to enter by acquisition, joint venture, or greenfield is modeled as a qualitative choice problem. A multinomial logit model is specified to estimate the effect of the explanatory factors on the probability that each of the three alternatives would be chosen. The multinomial logit allows the explanatory variables to affect differential odds of choosing one alternative relative to another. Thus, the coefficient vector is specific to the alternative, not to the firm making the choice [Judge et al. 1985, pp. 770-72]. Consequently, the specification of the probabilities is:
j=3

Pij = exp(xijBj) / E exp(x,Bj), where Pij is the probability that the ith firm will choose alternative j, xij is a vector of variables representing the variables characterizing the ith firm and thejth govearnancemode and Bj is the vector of coefficients to the independent variables. However, since the probabilities are constrained to sum to one, the system of equations are over-identified. The parameters can be estimated by setting the Bs of one of the alternatives to 0. In our model, it stands to reason

424

JOURNAL OF INTERNATIONALBUSINESS STUDIES, FALL 1988

the as to use acquisitions the baselinecase by which to compare estimated (oint ventureor greenfield). of parameters the otheralternatives is to: Underthis condition,the specification reduced
3

P11= exp(xqjBj)I1 E exp(x1/Bi), +

e=2

specifiedas with the baselinealternative

3

Pil = 1/1 + E exp(x1,Bj).

a by (Bs) The parameters are estimated maximizing log likelihoodfunction iteration procedure.23 usingthe Newton-Raphson Unfortunately,values for R&D and Advertising are only available for casefromthesample, the valueseliminate entire Since manufacturing. missing (1972, pp. 238-41).24We treat suggested Johnston by we follow a technique as withexplanatory the manufacturing, service,andextractive threeequations For sectors,R&Dand whicharenot identical. thenonmanufacturing variables termsare not are as Advertising recorded 0. If we assumethe disturbance variables estimation. dummy The we correlated, can runa singlemultinomial will pick up the sectoraldifferences. RESULTS coefficientsshouldbe The resultsare providedin Table4. The estimated utilityof choosinga jointventureor the as interpreted representing marginal A coefficient signifies relative anacquisition. positive to whollyownedgreenfield more likely the variable,the that the greaterthe value of the independent alternative (i.e., joint ventureor acquisition,as the case may be) will be are chosen;the converseis truefor a negativesign. T-teststatistics given in parentheses. showstrong distance for The estimated usingcultural parameters the equation Distanceis to increase The effectof Cultural for support the firsthypothesis. and a overanacquisition is significant of theprobability choosing jointventure at at the .001 level. Its effect is, however,only significant the .1 level for we two-tail test, thougharguably greenfield.(We are using a conservative could apply, followingCaves and Mehra,a one-tail test to the coefficients are Avoidance for forwhichwe havepredicted signs.)Theresults Uncertainty at with the coefficients correctly signedandsignificant .001 moreimpressive, and .05 for jointventureandgreenfield, respectively. The asset size variablesgenerallyare correctlysigned. The effect of U.S. is at AssetSizeon choosinga jointventure significant .001. Clearly,thelarger thanacquire. the morelikelyto jointventure the size of theAmerican partner, and is Theeffectof U.S. AssetSizeon choosing greenfield negative significant runsat the .1 and .05 and distance avoidance cultural in boththe uncertainty level, respectively.It is likely, however, that this result stems from the

OF CHOICE ENTRYMODE

425

.~~~~~~~~~~G) (D. ~~ ~
' 0> -0
D) C) l 0

~ ~ Cy)0 ~ ~ cv
I 0 r t

I11111 I I

oc'J

: C)

Cy)Lo

0)

0~~~~~~~~~~00
C

0
C) OC

I0

0
(

= ) 00 ~ L0() ~~~~~~0 0 00V) LO >^~~~~~~CCY5 o

0 (\ coc_ Ocx 0 Y COCO C:c) o

4.'
-

o o

) CZ
<

C\ r-

I Ot CY

C) LO

CY

C\

'It

(O

(O

O)

00 1

OD

xr0o

'

aCo

, o '

'

LO c

Lo

dl
111

_
*

C/3

N
Cj
Ca)

r-= CY) CY) 0) C00 r C\-)

o \' C)

- OLO 0)

LO
c\ Ici

-'

o
0j ..J
0) <

<
50
N

CE _'c

0
1OO0I "O
(D 0 tl

0)1
C
0) -C\'t 0)
c) C

0)O0
11 co

LU.0
4)

o00

cO C\
0
0)

CZ
(

~~~ ~ ~~~ ~~\

OD

() Cy CD 0 CD1\ _ O

5.0,
.

'l

to ) _

>
?

U0 1OiI

i..

o~~~~~~( C) 0
ooc o
X 'in

O D 03(Q

0)

I 0C

OC

Oo 00 ta

0(0 0t
m co co coo

0 CL ._,_V 0

*0
-

~~~~~N -o ~~~~0 )-cc l- co

c c co l co IlO

con co

CD
V

0
Q)

c:
>

75
o ,

n c.i

>

V~~~~~~~~~~~0 i Un F C
) O OX

x) a\) a) 0--U) D(10

coo

426

JOURNAL OF INTERNATIONALBUSINESS STUDIES, FALL 1988

and in of measurement asset size for greenfield termsof the investment for or acquisition jointventurein termsof the assetsize of the targetor partner. for The effect of the Non-U.S. Asset Size is insignificant the case of joint signed,thoughstill but distanceestimation, correctly venturein the cultural larger avoidanceestimation.Interestingly, for insignificant, the uncertainty at overacquisition the .01 level greenfield size of the foreignfirmencourages run; avoidance thisresult run in thecultural distance and.05 in theuncertainty confirms findingof Dubin. the are and Excperience Multinationality correctlysigned (with the exceptionof in Avoidance for to thecoefficient Multinationality jointventure the Uncertainty studies, Similar earlier to However,thet-testsarenotsignificant. estimation.). by effectsas measured priorentriesarenot shownto be therefore, experience estimates.Unlikesome other multipleregression robustunderlarge-sample [1987]), and studies(e.g., CavesandMehra[1986], andGatignon Anderson is experience not foundto be significant. ourmeasure multinational for To firmsmorelikely to enterby acquisition. We also do not find diversified that signed,showing diversified is Diversified positively the thecontrary, variable are The or firmstendto enterbyjointventure greenfield. results notsignificant on at andit wouldbe premature this time to speculate the causes. variables, are variables of someinterest.Of the dummy sectoral The industry in a indicating preference is only Manufacturing significant both equations, in sector. As over acquisition the manufacturing for greenfieldinvestment investments. shownlater,this effect is almostentirelydue to Japanese variablesis the positiveeffect of of The most interesting the industry-level ventureand greenfieldentry, thoughonly significantin the R&D on joint and estimation .05 for uncertainty distance formercase (at .1 for the cultural avoidance).Elsewhere,we have shown that joint venturesappearto be industries in [Kogutand encouraged growingandR&D-intensive particularly to findingsandsome transaction Singh1987]. This resultis counter previous A is cost arguments. possible interpretation that non-U.S. firms enterthe At to States tapintoAmerican technology jointventures. a minimum, by United and (though given the positivesign to R&Dfor bothjointventures greenfield in to appear be discouraged high for not significant the latter),acquisitions industries. R&D-intensive and investments. to related jointventures greenfield is Advertising negatively Thoughthe resultsare not significant,they are consistentwith Caves and of are for that Mehra's [1986]argument acquisitions favored thepurpose brand is to This labelor product adaptation. relationship expected be morepronounced work. whichwe will exploremorefully in further for mature industries,
CONTROLLING FOR JAPANESE ENTRIES

resultsaredrivenby outliers,namely,that that It couldbe argued the cultural in fromthe UnitedStatesandscoreshigh scoreshighlydistant culture Japan avoidance.At the same time, Japanesefirms tend toward on uncertainty

OF CHOICE ENTRYMODE

427

greenfield jointventureentries.Thus, the resultscouldbe interpreted and as a primarily effect. Japanese Fromone pointof view, Japanas an outlieris consistent with our argument andthis resultshouldbe expected holdfor entriesfromothercountries to that are culturally from the UnitedStatesbut whose firms have yet to different establisha strongforeigninvestment position. Nevertheless,the effects of cultural distance uncertainty and avoidance shouldbe expected holdfor the to samplein the absenceof Japanese entries.To show this, we reestimate the on of earlierequations a subsample the data, havingremovedthe Japanese cases. Theseresultsaregiven in Table5. Theeffectsof culture indeed weaker stillcorrectly are but signedandsignificant in two cases. Cultural Distanceis significant .05 forjointventures just at and to shy of .15 for greenfield.(Again,it is important note thatundera one-tail at Avoidanceeffect is negligiblein test, it is significant .1.) The Uncertainty the case of jointventures significant .05 for greenfield. but at The othereffects remainlargelythe same as before, except for changesin is significance. Interestingly, Multinationality positivelysigned,showingthat for are of On acquisitions discouraged the more multinational corporations. in in theother increases significance theruns,andis significant hand, Experience in threeof the runsat .1 usinga one-tailtest. The positiveeffectof R&Dfor remains at jointventures significant the .05 levelin bothruns.Themanufacturing dummycoefficientis highly insignificant. Clearly,then, the earliersectoral effect is drivenby the sectoral of firms. preference Japanese In summary,the statisticalestimations providestrongsupportthat cultural distanceand nationalattitudes towardsuncertainty avoidanceinfluencethe choiceof entry mode.It should underlined theserelationships robust be that are despitethe controlsaddedfor industry-and firm-level effects. The weaker resultsfor the subsample when the Japanese entriesare removedarepartlya resultof the reduced samplesize (the cases dropfrom228 to 173) andpartly a resultof theoutliereffectof Japan.25 is impressive, It that therefore, cultural to in effects appear be still persistent despitethe reduction samplesize and in the diminishment variance the cultural of variables.
CONCLUSIONS

The aboveresultsoffer the first large-sample test multipleregression of the view that entrymode selectionis influenced culturalfactors. prevailing by in The resultshavea secondary implication termsof validating usefulness the of Hofstede's measures cultural of a scalemeasuring dimensions. Unquestionably, the culturalcharacteristics the firm level would be preferable. at Yet, the collectionof suchdataappears formidable this time. It is, therefore, the at all more remarkable the strength the resultswere found, despiteusing that of of whichweredeveloped other measures national cultural attitudes for purposes. Theresults should interpreted care.Thevariable uncertainty be with of avoidance is definedin the contextof organizational managerial and it preferences; is not a measure cultural of attitudes towards in a larger risk sense. Furthermore,

428

JOURNAL OF INTERNATIONALBUSINESS STUDIES, FALL 1988

U '

o~~~~~~~~~ C\l o m
m

y -o

0~~~~~~c <1)~~~~~~~~" -co

@ a)

ai O ~~~~Cj o

CL

Ir

C)

C)

~~~~~~~~~~uO

co

u'

4) *o

>

-C .JL
C

~~~~~~~C\j
(I)\

C\j

rl

00 It

C\ CY)

~~~~~~~~.
100 Cf) N
j) C(1) oI}c0??1

C\l CY)cC)
Cl) CY (

oD Ct) I
O ndwd (

mo

zO

X~~~~~~~~~~~~
o oao CD oCD
CY o) oC

OL

t~~~~~~~~~~ QO _~~~~~~~
o.

u'

-r

oC

. _ . ...

@ C O. (\1dO. (\It

(C1 (C)

EQs -

0
c1c)

CD
O

o
(1C) Q\

v e
t tv

'I aw~~~~~~~r -r, OV SD

5~ ~~~~~~y

NS

CHOICE ENTRYMODE OF

429

historical time. Since the resultsmay only have validitywithin a particular betweenthe United has foreigndirectinvestment beenconcentrated historically whicharerelatively thereareconfounding StatesandEurope, culturally similar, As and distance experience. Japan otherAsiancountries and effectsof cultural in distance continueto increasetheiroverseasinvestments the West, cultural at may be increasingly offset by growingexperience the firm level. Though we have tried to controlfor such effects, it could well be that our proxy variables were insufficient. A final consideration which deservesfurtherexploration a more refined is analysisof entrydecisionsin the contextof oligpolisticgaming.Competitive dynamics,such as the rushto invest, arelikely to influencethe entrychoice. In addition,the relationship on amongthe variablesmay changedepending of of the functional are purpose the entry.Bothrefinements the subject current work. havea widerimplication outsideof country and Theresults patterns thechoice of entry.The above studysuggeststhatwhen economicchoice is compared across cultural characteristics likelyto haveprofound are countries, implications. theories internalization thefirmmaybe culturally of and Whereas robust,their in to the empirical application a comparative setting appears warrant consideration of cultural on differences thecostsandriskswhichmanagers attach different to modesof transacting. as Whether theseresultsareinterpreted contradicting internalization an theory of entrychoice is largelya question the definition transaction of of costs. To costs are broadlydefinedto includecommunication some, transaction and fromcultural control factors.In ourview, costs, evenif thesecostsarederived it is theoretically empirically and to betweentransaction interesting distinguish costs that are independent a firm's countryof origin and those that are of factors.The multinational determined cultural is by corporation the heir, to use PhilipCurtin's[1984] expression,of the historical cross-cultural broker in worldtrade.Butno matter superior current how the multinational corporation may be in replacingthe skills of tradersby the international extensionof the organizational of boundaries, management these firms are likely to be influenced the dominant by country culture.The resultsof this papersuggest thatfurther into of determinants managerial investigation thecultural decisionmakingis soundlywarranted.
APPENDIX DATA SOURCES

Data on joint ventures, acquisitions, and greenfield are not compiled and systematically theUnitedStatesgovernment must,therefore, gleaned by be froma number publicly-available of sources.Dataon acquisitions were taken fromtwo sources:the Department Commerce's of publication ForeignDirect
Investmentin the United States for the years 1981 to 1985 and Mergerstat

Review,W.T. Grimm Company, & Chicago,1984, for the years 1981, 1982 and1983. Acquisitions valuedless than$10 millionwereexcluded. addition In

430

JOURNAL OF INTERNATIONALBUSINESS STUDIES, FALL 1988

to theCommerce publication above,sources cited usedforjointventures were:

Mergers and Acquisitions and the Yearbook on Corporate Mergers, Joint

Ventures, Corporate and Policy. Forthestatistical data investigation, forjoint ventures weretakenfortheyears1981to 1985. Dataon greenfield investments
were found in Foreign Direct Investmentin the United States, again for the

years 1981 to 1985.

NOTES
'For a review, see Ajami and Ricks [1981]; Arpan, Flowers and Ricks [1981]; McClain [1983]; and Hood and Young [1980]. 2An exploratoryinvestigationof strategic motives for the choice of entry is provided in Kogut and Singh [1987].
3For an extensive argumentalong these lines, see Robbins [1987].

4A joint venture is both legally and conceptually differentfrom a minority equity participationinvestment, where a firm invests directly into a second company but does not share control with a thirdparty. 5We would like to thank Jean-Francois Hennart for this observation on the Anderson-Gatignon paper, which came to our attentionsubsequentto submittingthis article for review. of 6Caves(1982, chap. 3) argues similarlyin his discussion on the joint determination exportingand foreign direct investment. in [1973]. A brief Englishdescription 7Themain findingsare reported Hornell,Vahlne and Wiedersheim-Paul is given in Johansonand Vahlne [1977]. 8The numerous anecdotes on the motivations of Japanese firms to invest in greenfield sites in ruralareas are consistent with this argument. 91tis importantto note that our hypothesis is stated at the country level to representaverage tendencies. We cannot make statementsto the particularfirm without more detailed knowledge of the correspondence of the national to corporateculture. "Culturaldistance is, in most respects, similar to the "psychic distance" used by the Uppsala school. "1Acommon confusion is to treat uncertaintyavoidance as equivalent to risk attitudesin general. We, in agreementwith Hofstede [1980], use uncertaintyavoidance in the original sense of Cyert and March to is resolved as separatefrom the issue of whetheran organization referto the way uncertainty organizationally or firm chooses or avoids risky environmentsfor a given return. "2Wilson'sdata also came from the MultinationalData Base on 187 U.S. multinationalsfor the period up to 1967, plus from the activities of 202 foreign-based multinationalsthrough 1971.
13They called this variable an experience effect, but in orderto be consistentwith our descriptionof similar -variablesin other studies, we have relabeled it as a measure of multinationalityin accordancewith Caves and Mehra.

is 14Another interpretation that the results confirm that firms maintainin-house what Dunning [1977] calls "ownership advantages". Acquiredor Establishedby ForeignDirect Investorsin 1983," '5R. David Belli, "U.S. Business Enterprises Surveyof CurrentBusiness, Departmentof Commerce, 1984.
16Though high, it is lower than the 73% reportedby Tsurumi [1976] for Japaneseoverseas manufacturing subsidiariesin 1971. However, Tsurumiincluded subsidiariesin which Japanesefirms had less than 25% share. If we take out these subsidiaries (which are better considered as minority investments), then the percentageis 60%. Caves reportsTsurumi'sestimate as 82%, but we have been unable to locate the source of the figure [Caves 1982, pp. 89-90]. 17Caves and Mehra [1985] proxied experience by whether the firm had made a previous investmentin the chosen United States. Wilson [1980] used the proportionof subsidiaries established before an arbitrarily breakpoint. 18On the other hand, Steuber et al. [1973] found that the percentageof equity share in a United Kingdom subsidiaryby a foreign firm increasedwith the multinationalityof the parent.

CHOICEOF ENTRY MODE

431

19For summaryof studies on R&D and advertising,see Scherer[1980, chaps. 14 and 15]. For a summary a of researchon the relation of R&D and advertisingto foreign entry choice, see Caves [1982, chap. 1]. 20Theline of businessdata are drawnfrom confidentialgovernmentsurveysof businesses. Thoughsomewhat dated, the published summarystatistics for R&D have been found to be reasonablystable over time. For evidence, see Scherer [1982]. We have made a parallel assumptionfor Advertising expenditures. 21Thisargumentis consistent with Caves and Mehra [1986]. 22Hofstede[1980] points out, however, that the external validity is reasonable high when tested against other variables which should be correlatedwith culturaldifferences. 23Wewould like to thank HubertGatignonfor sharing his programand his advice with us. 24Weare indebtedto Tom Pugel for this suggestion. 25Attemptsto avoid the loss in sample size by using a country dummy for Japan with the full sample floundered due to the collinearity between the Japan dummy and cultural measures (.81 for Cultural Distance and .86 for UncertaintyAvoidance).

REFERENCES
Abbegglen, J.C. & G. Stalk. 1984. The role of foreign companies in Japanese acquisitions. The Journal of Business Strategy, 4:3-10. Ajami, R.A. & D.A. Ricks. 1981. Motives of non-American firms investing in the U.S. Journal of InternationalBusiness Studies, 13 (Winter):25-34. Arpan, J.S., E.B. Flowers & D.A. Ricks. 1981. Foreign direct investment in the United States. Journal of InternationalBusiness Studies, 12 (Spring/Summer):137-54. Bendix, Reinhard. 1956. Work and authority in industry: Ideologies of management in the course of industrialization.Berkeley:Universityof California. Brooke, M.Z. & H.L. Remmers, eds. 1972. The multinationalcompany in Europe: Some key problems. London: Longman. Caves, Richard. E. 1982. Multinationalenterprise and economic analysis. Cambridge,U.K.: Cambridge University Press. & S. Mehra. 1986. Entryof foreign multinationalsinto U.S. manufacturing industries. In M. Porter, ed., Competitionin global industries. Boston: HarvardBusiness School. Curtin, Philip. 1984. Cross-cultural trade in world history. London: CambridgeUniversity Press. Davidson, William H. 1980. The location of foreign direct investmentactivity: Countrycharacteristics and experience effects. Journal of InternationalBusiness Studies, 12 (Fall):9-22. Dubin, Michael. 1975. Foreign acquisitions and the spread of the multinationalfirm. D.B.A. thesis, GraduateSchool of Business Administration,HarvardUniversity. Dunning, John. 1977. Trade, location of economic activity and the MNE: A search for an eclectic theory. In B. Ohlin, ed., The internationalallocation of economic activity. London: Holmes and Meier. Franko, Lawrence G. 1976. The European multinationals. Stanford, CT: Greylock Publishers. Gatignon,H. & E. Anderson.1987. The multinational corporation's degreeof controlover foreignsubsidiaries: An empiricaltest of a transaction cost explanation.ReportNumber87-103, MarketingScience Institute, Cambridge,Massachusetts. Hofstede, Geert. 1980. Culture's consequences: Internationaldifferences in work-relatedvalues. Beverly Hills: Sage Publications. Hood, N. & S. Young. 1980. Recentpatterns foreigndirectinvestmentby Britishmultinational of enterprises in the United States. National Westminster Bank QuarterlyReview: 21-32. Hornell, E., J.E. Vahlne & F. Wiedersheim-Paul.1973. Exportoch utlandsetableringar. Uppsala:Almqvist & Wiksell. Hymer, Stephen. 1960. Theinternationaloperations of nationalfirms: A study of directforeign investment. Ph.D. thesis, Departmentof Economics, MassachusettsInstituteof Technology. Jemison, D.B. & S.B. Sitkin. 1986. Corporate acquisitions: process perspective.Academyof Management A Review,11:145-63.

432

JOURNAL OF INTERNATIONALBUSINESS STUDIES, FALL 1988

process of the firm-A model of knowledge Johanson, J. & J.E. Vahlne. 1977. The internationalization development and increasingforeign marketcommitments.Journal of InternationalBusiness Studies, 8 (Spring/Summer):22-32. Johnston, J. 1972. Econometricmethods. New York: McGraw-Hill. Judge, G., W. Griffiths, R. Hill, H. Luetkephl& T. Lee. 1985. The theory and practice of econometrics. New York: John Wiley. Kogut, B. 1987. Joint ventures: Theoretical and empirical perspectives. Strategic ManagementJournal, forthcoming. & H. Singh. 1987. Entering the United States by joint venture: Industrystructureand competitive & Strategiesin International Business. Lexington, rivalry.In F. Contractor P. Lorange,eds., Cooperative MA: Lexington Press. Lincoln, J.R., M. Hanada& J. Olson. 1981. Culturalorientationsand individualreactionsto organizations: A study of employees of Japanese-ownedfirms. AdministrativeScience Quarterly, 25:93-115. McClain, David. 1983. Foreign direct investment in the United States: Old currents, "new waves," and & the theoryof directinvestment.In C.P. Kindleberger D. Audretsch,eds., TheMultinational Corporation in the 1980s. Cambridge,MA: MIT Press. Pate, James L. 1969. Joint venture activity, 1960-1968. Economic Review, Cleveland: Federal Reserve Board. trade. In R.G. Hawkins Pugel, T.A. 1981. Technology, transferand the neoclassical theoryof international & A.J. Prasad, eds., Technology Transferand Economic Development. Greenwich, CT: JAI Press. Puxty, A.G. 1979. Some evidence concerning cultural differentials in ownership policies of overseas subsidiaries.ManagementInternationalReview, 19:39-50. Robbins, James. 1987. Organizationaleconomics: Notes on the use of transactioncost theory in the study of organizations.AdministrativeScience Quarterly, 32:68-86. Robinson, Richard D. 1961. Management attitudes toward joint and mixed ventures abroad. Western Btsiness Review. Sales, A.L. & P.H. Mirvis. 1984. When culturescollide: Issues in acquisition. In Managing organizational transitions. Homewood, IL: Irwin. Chicago: Rand McNally Scherer, F.M. 1980. Indutstrialmarket structure and economic performnance. College Publishing Company. 1982. Inter-industry technology flows and productivitygrowth. Review of Economics and Statistics, 64 (November): 627-34. Steuber, M.D. et al. 1973. The impact of foreign direct investment on the United Kingdom. London: Departmentof Trade and Industry. Stopford, J. & L. Wells. 1972. Managing the multinational enterprise: Organization of the firm and ownership of the subsidiaries. New York: Basic Books. Stopford, J.M. & K.O. Haberich. 1978. Ownershipand control of foreign operations. In M. Ghertman& J. Leontiades, eds., European Research in International Business, pp. 141-67. Amsterdam: North Holland. Wilson, Brent. 1980. The propensityof multinationalcompanies to expand throughacquisitions.Journal of InternationalBusiness Studies, 12 (Spring/Summer):59-65.