GLOBAL_COMPETITIVE_PRESSURES

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

1fves L. Doz Christopher A. Bartlett

C, K. Prahalad

Global Competitive Pressures and Host Country Demands

Managing Tensions in MNCs

After several years of proposals, persuasion, and subtle pressure from its customers, its em­ ployet·s, a11d \'aricn,s !oral and national govern­ ment bodies, the United States-based hospital supplies rnmpany, Baxter Traveno\, recently rnmmissioncd a new French plant to manufor­ turc its line of blood rnllcrtion and 1ransfusion products. •t'hc con1pany already produced similar products in four other plants in Europe, se\"cral of which had the potential capacity to supply the French market The company was conscious of the fart that, from a purely eco- 11omic poi111 of view, a single European supply source would probably have allowed the great­ est ma11ufacturing cfficie1H:y.

In the fall of 1~179, the Nigerian government 11otified Fre11cl1 a11tomohilc manufacturer Peu­ J.\'<'Ol that its n.'<-cntly expanded $100 million assembly plant would have to import all needed components and supplies through the under­ utilized port facilities in l.agos. Under this ship­ ping plan, the parts and components had to be trucked anms five hundred miles ofbacl road to the plant. Not only was such an operation more time-consuming cmd less reliable, it was 11111rc expensi\·e than the previously employed daily rargo flights from Franre direct to the

SPRIN(; / l!lHI / \'OJ .. XXlll / NO. '.\

plant. Ne\'crtheless, the company complied with the request. The two situations desnibed above arc illustra­ tive of the scores of management derisions faced daily by multinational corporations, where the dri\'c for international competitive advantage through global rationalization of ac­ tivities is tempered by the needs and wishes of host nations and the diverse demands or their 1narkcts. As global competition imensifies and the variety of national demands increases, more decisions are affected by those opposing demands for natimrnl responsiveness and for global rationalization. Management is continu­ ally faced with the problem of managing the tensions that result. Analysis of these tensions has been discussed elsewherc. 1 The most dif­ firult challenge management faces is adminis­ trative: the structuring of the company's inter­ nal decision-making process to allow the organ­ ization to sense, interpret, and respond to ten­ sions, and the resolution of the often contra­ dirtory demands for global competitiveness and national responsiveness. After detaibl study of the administrative pro­ cesses of a srnre of large multinational com­ panies (l\.·fNCs), we have identified several ways

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

l"l·,, /. /) ; !I'.:< lw, h11,i1w1, pol it 1· ,111d i111,•111,l11on,1\

111.m,1g,·1m·111 ,11 I \SI-' :\I). 1'1 iol'!o joining ]\SF:\]'.

ht· 1.111gl11 al 1l1t· lla1,.11,l llu1im·"

',1110,,\ .111!\ .11 tlw d'F1111·ig1u·m1·111 su:)('1 •

irn1 rk, .\ll.1i11·1 .. \ g1,1d11:1tt· ol Ltult• ,k, I Ltllh'I

l-1mk, ( u111111l'llialn i11 h,IIHt' ,11111 ,l d1J<l01 ol hu,11n·1, .ul111i111111.11i,,11 (lla11a11I l'ni1,•1,itr). 1'11,

!,·,1,11 ll,,1 ,,11u!11tt1 1111•.11,h ,,11 tlw ,,11.1wgi1•1.

,111u 1111(·,. 111;111,1gt'llll'III fll(!!('\lt'~. ,llld ad111ini1llk

111,· I\ 11,·1111.,J 11111hi11,1tio11,d ,rn po1,11io11,. ! Ii, moil

11·1 t·111 1t·\\',!ll h l!H 1111·, 011 how 111,rn.1g,·1, i11 m11\1i

11.1111,11;il ,omp.rnit·, l.11,· d1t· ,011!1i,1, !ll',lh'd

l>I 111011, :1111,t• goilt'llllll!'!ll ll\ll'!\t'llli<111a11<I 111<>1<'

i11t1·11\\' i11h·111;1ti,111.il <••lll[Wli1io11. Iii, 1·,ulit·1

11·,t·.11,h. 011 tl1t· 111,111.1g1·11w111 ot 111uhi11atio11al, i11

llll' ,,,l1·,1,1111111111i1.11io11 ,·quipt11('lll ;wd 1-il,oi,.11

fi"\11·1 11 ,,,·111 i11dm11 in. \\!Ill 1lu· .\.I. ~:(',ti 11,·1

\!.1dl'll11 ul \11,11d i11 l!fii. l'1u!t·1·

/ /.,1,!i,Ji/i,1 .\ /l"'l/,/t i, ,Ill ,l"i,1.1111 p1olt-"rn ol

IJ11,111,·" ,1drn1111,11.11i,,n ,11 tlw (;1,ul11.11t' ',,1111111 ol

l\11,i11n, .\d111i11i,11.111n11. I l.111.11d \ '11i1t·1,i!1. l'w

111,1,1,·1, ,111d d,a 1111.11,· !l<-g1,T, in l,111i11n, ,11!11rn1i,

t1.111011 l1.,111 ll.11,,11d \'11iH·1,i1, ll!l/1 .11ul l~l/\ll

(ii

,ii 1lwll.111,11d B11,i1u·"

l'll!{llll'!'llllg II! 1\l,i1',lgt•1

t11.111.1g,11w111 \\'lli,1·, lk 1,,1, ;1 1i1iti11g t1·,ea1,li

lt-ll011 .11 ll.11-..11d .111d .1 ptotn,01 .11 11w [11,!i,111

l11,1i111h- fll \l.111,1g,·11u·n! 111 .\li111t·d.1h,1d, l11d1.1 111,

lt·.,,Ji1111;. Tl'"',ll<ii .. 11ul !Ollllllti11g i11t1·1,·,1, ,II<' ill

tll<' ,11,·,, ,,t ,11,11q;i, 111.111.1g,·111,·m ,111<11"1111,,l 111

l.111:t·. ,1u11pl1·\ 01g.111i,,11in11,

lor lop manag,·mclll to 1lcal with the variety and ;1111higuitv of national <kmauds while maintaining a glohal husi11css stiategy am! rnr­ porate din:rtion. ·1 ·hc al'lirlc outlines why these dual (kmamls rt·sult in administrative prob­ lems for most ~INC:s, arid present four ma11a­ l.(l'l'ial sol11tio11s to 1ht· prnlilt·m.

Conflicting DemlJnds

:\ rn111p:111y wishing to ht• sunnsful in tlw i11- t·n·asi11gly n1111pt·titi\'l' gl<1hal 11iarkct 11eeds 111

takt· full ad\'antagc of it~ glnha! n.'somTcs and multinational po,;ition. By rn11n·n1rating its worlc\widt· 111;11mfanming n·sm1rrcs, for cx­ ampll', a11 ~1:-,,/C may ohtain important global rn.'>I sa\·i11gs. By prntt·rting- its prnclwt and pro­

n·ss tt'chnoloMit's, it ca11 rc1ai11 its rnmpctiti\'e ach'antagt'S i11 rn~t or prodm t diararteristirs. ldcall•:, a dt·;ir g-lobal st1 ◄ 1lt'){}' e11.sun·.s maxi­ mum kn·ragt' ol a rnr lj>;111~··s skills and rapa· l,ilitic~. · 1 ht· lirm nt"eds an nrgani,ational strur­ tun· that pnrnit~ the dc\t'lopmcnt of a rlt•ar, ,011sis1,·n1 g-!!>hal din·ttio11 an(l primitics am\ farilitatl'S the rnonli11atim1 ;rnd integration of till' n1111pa11y\ rcso111Tes alld t'apahilitics worldwide.

I 'ht· i(leal of at ll'ar, t 011sistt·111, well-intcgTated gloh;il stratl'gy i~ limited h~· powerfol forct"s

that push ~!:'\Cs in the dircrtion of a mon• a111hig11011s, less ll'cll-intcgra!cd strategy that rl'sponds to national ditlt-w1H<'S. Siun· nation­ al l't'~ponsi''t'IH'S\ ohl'n I an ht· ad1icH·d only at the t·xpt'l\Sl' ol global rnnsi~tt'llt ~· and darity, lt'll'ii!>ll\ arc I rmted. ·1 he need ti)r a gloli,1ll~· nl11-

~ist1·111 stratq,:-y a11d full~· i1llt·g1atcd operation dqw1Hb 011 l'a!'h irn\11st1 y's l harancrislil~. 'J'hl' idt·al, rH·n·s~<HT in thl' :rntomohill' induslr\' a11d dl'sirahk i11 till' drng industry, may be rn1111tt•r­ pt ml11ctiw in tht' loud pron·s,i11g im\ustl)", 2

l'hl' product. 111a1111fo1 t11ri11g-, ;\lid distribution d1aractt'ris1 in of food prnn·s\illg tend to limit the g-ai11s to ht· 11@lt- hi· i11tt'grated g-lohal

'l'hosl' irnlustries whnl' global rnonli11atim1 ran pro\'idt· a rompt·titin· atk1111age still fatt'

thl' dcmaml~ ol 1·a1io11s host rn1111try interest gt,Hlps. \'a1io11al g(>\'1'111111,·1ll~ 1101 0111~· rcgu­ latl' dll ill( 1easi11g 1111111lw1 of dl'('isions---sud1

as tht· ~trw run· ,111d h't lllS ol thl' i11itial im·t'.'il­ lllt'lll awl the l(Jt'dtion nf a llt'W pla111-but

(."ali/m11i11 ,\fa1111g1'1111'11f U1•1•i1'11'

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

i11rrea~ingl~· intt'IW'lll' directly in the rnm­ pany\ ongoing drrision procc~ses. Employt'es are demanding positions on ~uh~icliary rnm­ pany hoards, and ho,1 go,·c1 nmcnts arc req11ir­ i11!-{ ~tth~ta11tial 11wrwrsl1ip and rnanag-en1ent p:11 titipation by local 11atio11als.

·1 he purposes of 11ational interest g-roups ap­ pL'ar to he trending towards greater variety. 'l lw go\·enmwnt of :\Ial.1ysia is att<·mpting to redre.'is the social pnililems r:1used by c:hinese and Indian dominalH't' of its ernnorny hy ask­ ing MNCs to ofler at least ,I() percent of their equity to ethnit ~falays. lkmands may reflect the muhiplirity of' g-ownunent omstituendcs in the host rnutllrv, am! often result from th1: interplay ht·t1,·ct•n' national it1teres1s (govern­ ment ,tgcnri<'s, lahor unions, industry asso­ ciations).

An ~INC tr~·ing to adapt to host rnuntry de­ iuand~ is frcqul'ntly confronted with a moving targe1. ·1·1ierc is a grnwinl-{ di\'t'r~ity or so111u·s, mearn, and ohjt·t ti, t·~ of national demaml~. ~ational dl'mands rarely appL'ar in a dearly art it ulatl'd, !ir1,1I foi 111. lt11her, thl'ir form and d1,ira( ll'ristics tend to t'\'Olre orer time, aml ohen hernrne lull~· re,ealed onl~· through nego­ tiations \\'ith the ~f~(:. ·1·heir ambiguity means that 111a11y important national demands GIil he guag-l'd 011ly 1hrough daily i11terartio11 with the app1opriate gme11mie1ll oflirials, union ll'ade1s, aud indu~t, y reprc~e111a1i,·es. 'fhe alter­ nati, e, the anal.11ical approach typically under­ takt'n hy a .,taff g1011p at rnrporate headtpiar­ teVi. i~ 11orrnallv !(,~~helpful.Top management, limed tn rclv 011 the recommendations of na­ tional suhsidi.11 y managers, i~ fared repeatedly with plans for eronrnnically questionable at.:­ tio11s, the app1ornl of which is supposedly im­ porta11t in retaining good relationships with key rninisu y oflicials. T~·pirally, there is liule top 111a11agem('11t < a11 do to evaluate the arrn­ racy of surh proposals. ;\11 inddent that illustrates tht' expcr;em·e of th(' rnmpanies west11clied inrnl\'ecl a suhsidiarr manager in a Srandinavian rnuntrr who dt·­ rided to manufacture rnmponents for a rertain rlas~ of elet'troni(' prorlun~ lol';1lJ~, rather than import them from he,1dtjllilllers. This meant hig-h<'rrnsts and lower profits for hoth the suh­ ~idia1 y and th<' headquarters. I !is justilication

Sl'IUN(; / l!IHI / VOL. XXIII / ~O. :l

was that government officials had strongl" hinted that, should the rompany manufacture components-acquiri11g what, in the govern­ ment ofticial's dew, w;is a "sell~containcd capa­ bility"-large pwernment contracts would be forthcoming.

The need to respond to host country demands often conflicts with a dear, consistent strategy. Host country demands put global coordination and in(cgrati011 of the company's resources and rapahi!ities in jeopardy. The ongoing, direct interaction with host country forces that is needed to assess their demands rnmpromisl's a coordinated, analytic approach lo strategy eren more.

Managerial Implications The conflict between host (ountry demands ancl rompctiti\·e fort:es tun1s stratrgi( decision making in the '.\[NC into an adrncacy process between two rnmpeting perspe(tires. The com­ pany finds itsl'lf trying to maintain a dear, cohesire gfoh<1l strategy while keeping poliries flexible enough to satisf}' diverse, ofkn ambig­ llllltS national demands. This imokes negotia­ tions with host gorernments and among mana­ gers within the rnrnpany.

Some managers, attuned to local needs and sensiti,·e to the power of host gorernments and national interest group3, fornr, on almost any issue, more subsidiary aut,momy and greater frcc<lorn in n·spomli11g to 11ational clemamls. Other m,magcrs, more concerned with world­ wide competitire strategi<'s, stri,·e to increase rnordination and imegration across g:eo­ graphir boundaries. In trying to gain power and in defending their own areas of responsi­ bility, these maiiag:ers unearth different facts, analyze tl1cm differe11tly, an<! propose differ­ ent strategic decisions. By confronting rnnflict­ ing ricws construdively the organiwtion cm learn. 3

A most difficult task for top ma1rngcmcnt is <K­ rnmodating these varying analyses while ensur­ ing that the resulting derisions add up to a rnnsistent strategic approach. Because of the inherent rnnflict between national and glob,il \'iews, managemenl cannot limit commitments to th,isc projects agreed upon br both sub­ sidiary managers and product di\'ision head-

65

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

quartcr exernti\'CS. Such a process mu Id lt•ad tn p,11,1lysi~, ,i11cc only a ft·w, insignifkant pn!jerts may he mutually sati~factory.

To pH'\t'IW' hoth national a11d glohal views, innca~i11g: agreement should not he soug:ht. Top 111a11aMemcnt wants both pc1spt'<:tiws to lie ,111)11g:, and mu,t de\'elop a way !<1 ~clecti\'ely farn1 1hc \'icw~ of one or the other .,icle for any pa1 tirular dee i~ion. Soll\(' projen~ ne('d to he appro\·t·d e1·cn thoug:h they do not draw rnn­ M'l\~u, l1(llt1 tll(' rnriom subunits i11 theorg,111i- 1a1ion. Top manag:cment thus seeks to nt'<lll' asyrumt'll y i11 dee ision making.

I 11 ,um111a1), when both a ronsisll'nt gloh;1I ,11all'g1 i, tl{'cdecl aml hmt rnunll)' d1•111ancl~ a1t.· ,!Ion~ at1d lar-rcarhin){, the ~llateg:ie dn i,i01H11aking: piorcss within the ~l~C tlt'ech to pronde lor ~e\'eral d1,ir,ictni~tir~: I ha1 t citifl;_( 1i11g: 1 iews be imbed<le1! in !ll i-:ani,a­ tional ,uhu11i1,, that those \'iews he clfcni\'elv 1 qin.'.,l'lllcd at thl' e orporalt' k•1el hr mana){el:s lrom !he ~11h1111i1, ;ind that top ma11a){t'111t·nt t:111 ,cl<'tt ci11t· Ill the otl1er r011l1itti11i-: 1·ie11 atrnrding to the spccilit~ of the dt'ri~ion at hand. :-.tuhi-climemionalitv and tlexihilitv arc the k<·~· H·quin·111en1s nf 1\,e desired st1.;tc){i1 de< i~1on pi nn'"

-, lit· llt'cd to maintain h,1lam1.· <l'ld llexihility hc1wt-c11 lrn al and global reptt''it·rnatin·.~ in 1(',H hmg: ,1,atcg:it dcri~iom sug:g:ests that tradi­ tic>1i;1l, t1riieli111en~ic111,1l r,n1dm1- or arca-ha,(·11 orga11i,atio11, ;ire likelr to fot·e rnnside1al,lt' ddfit 11hy.~ ,\ wo1 ldwidc prod1tttdi\'i,im1 ~11 ll( -

11111· i, likely to le,ul 1uan:1ge1 s to 1011rent1 ate on mai11tai11i11i.: !he tedmolo){ital superiority ol their p1odwt while hcittg: relati\'cly imt'nsililt· to 11ational demamh for product fkxihility. \\'ht·n \\'e~tini-:hou~c g:ave worldwide rcspomi­ bility lo it, prod11tt di\'i~ion mana){el~, those whme prodlH t~ bad a rnmpctitive edge in 1hc wc,1 Id 111a1 k1·1 (jUit kly affepted the rhallt·11w• of dneloping a11 international orientation. \\'hilt' tedrnolog:ical ~11pe1i01ity may harl' ){hen it a 1 ompt'titin· edge i11 foreign markl'ts, the tom­ pa11y t olltinually I an into dilfirnltie\ lit_•cmst• ol ih unrt·,pomivem·~~ to !oral, national clc­ mamh. In nm lear power 1.·ng:i11ecri11g, fot ex­ ample, ilw a111hitio11s pla11s set for growlh in E11ro1>e· in tlw late l~JHO., were thwa1ted hr go\'­ (•rn1111·111 to11te1ns that its worldwide p1rnlll(t

orientation would not allow Westinghouse enough tlcxihili1y its dealings with hmt g:o\'crn­ ments. ~

A g-ec1graphirally based slnKturc whirh ){i\'CS

national ~ul1sidiary ma11agers suhstantial autonomy results in different, Inn equally tax­ ing prnhkrn.~. 11lthou1-:h llexibility to local de­ mands is normally a< hicved, opcr,uions may liecmnc frnrtio1Mted and poorly LC)()rdinatcd, with i1<'adqua1 tt-rs management unable lo mak(' the terhrnral judgments nc(·dcd to main­ lain a11 eniricnt i.:lohal rompctiti\"c position. A senior rnrporalt' ma11ag:t·r of one rnmpany studied arknowlcdi-:cd that local suh,idiary au101111my a1Hl llexil1ility had s11cngtlw11c1l thc nmipany, ln1t < cu1111lain(·el that, a~ 1·xi~ti11i-: pro­ durts and .• 1arkeh had matur<'d, the rnmpany had little in the h'< lmoloi.:y pipcli11c. "We need a handle on somt'thini-:, hut lat-k a clear ap- 11n)a(·l1. ~fa11~· husi11t•sscs rc1111irt• num• n•ntral effort\: tedmoloµ,y has to ht· acquired or de­ \"('l(lll('li r<'ntrally, a11d tra(lc-ofls l1ctwee11 risky i11n·st1m·n1s 11t·<·d 1,, ht· 111aele n·ntrally."

1·11t' t1;11liti!i11al, u11itli1nt·n~icmal prmlm.t or ).:('1>gr,11>hie f>ri.:a11ilatio11al strnnun•.~ we ~tmlicd were unable to rope with the dual cn­ viro11mt·t1tal 1lem,111el~. S<llllt' rn11ipanies trie<I 11, s1111pk-mt•111 tlll'ir ~1111ttt1t\·~ witl1111lcs ,u1d 1m1ct'cl11n•s to r<'g"t1lat<• tht• rl'latiomhips hl'­ IWl'l'II r1a1i<>Jl.ll ~11h~idiaril's a11d Jll"rnillfl divi­ ~ions, hut tomplex and highly ,ariahle environ­ llH'llt,11 demands t H'ilt<'d it temion betwcc11 11a­ tici11,1I a1HI ){l1>llill \)t'I \jlt't ti1·t·~ that rnuld no\ he rt'solwd till ough a I ii-:id, hun\HHTatir franw• ,,ork. Eithn the prnr<'dme~ wcrt' not adhered to ;111d pown relationships d,·,eloped around them, or they led to .,trale){it iiillexibility­ rnnllirts heing: rt'ml\'l'd "hy the hook" rather than on tht' llll'l i1, of the pmitiom of \'ario11s ma11ag:e1 s 011 the issue~ at hand.

rlw rnmpanies ll't' s111died that had den·lopt'cl lll'xihlt· det is ion pi rn·ew·s, adjmtt'd lo hoth liHal 11a1i1111al ck111a11ds aml global rn111pt•titi,·c pn·ssun·s, had not wlied on traditional organi- 1ational .. 11 ue tut<·, or rii-:id rules and prnre­ dun·s. l 'nlik<' tbmt' in simplt' geoi-:raphic m prmlun organi,atiom, lop manai.:ers in these firms 1erng11ilt'd th(' Ht't'd for their org:aniza­ tion tn rdlcrt tht' dual envi1011111ental ck­ mamls. llnlike rnmpanies that simply ~upplc-

(:11/i/omia ,\la1111t;1'llll'III H,,1•il'111

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

mcnted their struclures with rules aml proce­ clmes, the.,e rnmpanies rerngniu:d thal the complex and highlr \'ariable decis:on pror~ss was intrinsirally nmllktual and could not be leg:islated through a ln1reaucratic framework, 'l'hesc nm11ianies sl1ared two charatJeristics: they had informal structures, and their top ma11agement cxercised great care in maintain­ ing the <1uality of the <ledsion process he tween m,1nag~rs with tct I itorial responsibilities and 1nanagers with pn,1luct re.~po11sihilities. 'l'hey needed more asy1:1111ctry in the decision pro­ cess 1ba11 rnuld he prmided by formal struc­ tures aml ad111i11istrati\'e prot·edures, which tend to dcline permanent, shared responsibil­ ity. These rnmpanit·s n.-cop;ni,.cd that to solve rnnflins they needed to learn to ma11age ten­ sions, not just to formulate procedurt's. In these rompanies, top management's role in maintaining the dedsicm process was twofold. First, they tried lo prevent ex<·nuin.'.~ se11siti\'e to national demands or international rnmpeti­ tion from dorninati11g the process. The viability of both perspecti\'es had to be maintained. (\Vhe11 asked how he dcricled with which mana­ gers to side, the d1ief cxernti\'e orfirer from a major rnmpany who practircd the balancing act sun:essfullr answered, "With the weaker.") Second, hecal1se of the nmstc\tll d,111p;er of horse trading am! improper rnnllict resolu­ tion, top m,maµ;t·ment had to monitor the qual­ ity of the derision making processes. Harold ( ;cneeu's famed large meetings at ITT pro­ \'i<kd this capability: Cenecn put managers of widely different 11rienlatim1s i11 the sa111c roOlll and asked them prnhing, well-prepared qucs­ tio11s to ensure all rclevants !'arts were carefully ,·onsiclcrecl. 6

Not all the senior exernti\'es of the companies we studied went to the same length as Ccneen to monitor processes through whid1 stralq\"ir decisio11s were made, yet ;ill reroinizcd the sub­ tle, politicd nature of thl' process of halanring 11atirnial dema11ds a11d internatirn1al competi­ tive pressures.

Managing Flexible Asymmetry

\\'ithin 1l1e rn111pa11it'S studic<I, a wide \'ariety ol solutio11s had been de\'eloped lo deal llexiblr with the nrnflirls helween national demands

Sl'RIN(; I !!IHI / VOi.. XXIII / NO. :1

and international cnmpctitive pressures. All were trying to provide the rnpaoility to explore in detail and from c:ontlidng perspectives the facts pertinent to a decision, and to reach 1hc decision on the merits of facts rather than on some dominance of worldwide product execu­ tives or of subsidiary managers. Nol only did top management of thest· rnmpanies hare to develop lhe various nrgani1ational groups to represent the different perspectives, they had to devise structural arrangements giving top manageml'nt the ability to rnnlrol the groups' an:ess to and inlluenre in the decision-making: processes. BcGtUse the formal organizational structure tended to he too inflexible to provide a responsive decision process, all the companies relied heavily on informal structure.s. We classified four mode~ of administrati\'e solutions to managing tt•nsions. Some com­ panies' decision-making proresses incorpor­ ated characteristics of several modes; in some, all four rntep;ories O\'erlapped. The four ap­ proaches arc prcs~ntcd in order of complex­ ity, The first two aim at building into the or­ ganization the rnpability for top management to rapture the tensions in the critical emiron­ ments and manage them. The third is a decision-making process that imolres the de­ velopment of coalitions and their skillful man­ agement from one derision to the next. In this approach, top management creates and man­ ages temporary bodies surh as task forces, com­ mittees, and sperial teams, using them to intlu­ enre the outcome of the issues to which they are assig-11cd. In the fourth .~olution, management employs a \'ariety of de\'kes, formal and infor­ mal, to create a stable organizational derision conlext in which asymmetriral responsibilities for renain dasses f)f derisions become institu­ tionalized.

S1tbsta11ti1•1' /)l'Cisio11 i\la11age111l'lll. In this, the simplest of the four management modes desnibed, top management simply ensured that the managemeut groups representing eac:h of the critical defision perspecti\'eS was strong enough to de\'dop and ad\'orate its par­ ticular viewpoint. In one company, whose operations were geographically organized, strong corporate produrt and functional groups pro\'ided top management with well-

(i7

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

supponcd lase." for global strategics. Thl'se globally oric111ed managers tc11dcd to highlight the l'OSI ofheing nationally rcspousire.

Obviously die national am! the global perspl'c­ tiw·s l1<·q11er1tly \1'<'r<' in n1111liu ahout ma_jm derisiom, and SlKh del·isions were m1>n·d up­ \qrd for arbitration. A subsidiary m;mager in om· n1mpany· rnmmented: In tlw10111p.1111 th1·1(· i,,1 wmlcm1· !(•tll'lq:att• "upward" j,iq 111 be ,u11· 110! to ht' (,1t1gh1 lllll on ;1 limb. Fo1 ma111

i\\ut·,. d,·1 i,i,,11, 11.nt• 1<1 lw ( l,•an•d hi' 1h1• l1lwl, ol 1lu p1odmti1m ;rnd nwk1·ti11g 1kp.1tt1m·111,. If 111· di,agHT w1d1 du·m IH' go 11! 1!w p1t·~id1·11L ·1 he1t· a11· f11·qm·111 1i,ir- l,1· nw1o1lwr, ol a t('<HII ol ,,um.,lll'r, who iq,out di11'11h 11J 111•.· p1,·wk111. ·1 ht•11· ton1rnllt•1~ kt·rp hun 101\·

,1.111111' i11lo111wd 011 cl1•11·lopmt•11t in t'<H h ~uh,icli.111

lkspit<' tht' obrious a<h·a11tagt' of simplititr. ~ut h .t lll()(I(· mar dt'arlr rt'~ult in at lt·as! th rel' maior p1ohlt·111s.

Till' lir~t dcrircs from top 111,mag-t'1t1t·nt's 1eliaw1· rn1 two opposing- ad\'oratt· groups fo1 the info1 rnatio11 011 whid1 it mmt m;1kt· its fi11.1l de( i\i >ll. !11 \llf h a ~ituation, the infonnation fl.'d 11, lop 111a11aµ;c1m·nt may lw rarcfullr sl'll't ted It> l:irnr a p,111irnlar point of \'i('W. ·1·,,p m;mal-(l'IIH'llt\ dee i-;ions 1uarhc made 011 i11- 1·0111pit·tt·, tonllitting, or inanurate data, and ('hoirl.'~ may he amo111-: a limi1cd, somt'timc.~ t'Xltt·1m·, st•1 of alternati\'('S. The best cll'frrnl' a1-:ains1 thi\ i\ a clctaill'd, \1·orki11g- knowledg-t· ot the C!Ht1pa11,·'s operatiom ;me\ irnl11s11y· rnndi­ tions so that hia~t·s i11 information r;m be de­ Ice tC"d and ;iltcrnati\'t's t·xpandt'd. \'et 110 top 111anagt·1m·111 g-roup cu1 he so well-infornwd 1ha1 i1 ( ,111 cktnt biases in all d,11,1. partinila1 ly that relatin!-{ lo host rnurllrit·s

Tbt' Sl't ond problem is top 111,111agt·111t·nt on·r­ !oad. Top m;maµ;cnH·nt quickly bernrm·s O\'t'J"­

pm,·ncd with dnisiom, s11µ;1-:cstio11s, ,rnd a11al~·st·s th,111·;111 only he rt'~ol\'l'd at their lcn·L I hi\ slow~ down the dct ision prmcss and

da111pt·11s the gennation of proposals lrn111 within the mgani1atim1 clue to thl' time and toffo1 ts m·rcs~,11 r to oh1,1in top m,u1agc111e11t's apprn\'al, C:e111rali1ation may fmthl'r ht· de­ fcatl'd hy· frustrated managers who rnakl' honlcrlim· dl'C"isiom on their 01m, kari11g- that s11hrnis~io11 to top manag-('111e1n would n·~ult i11 no derisions ,11 ,di.

·1·1ie third ma_jorprohlem e1Kou11tered h~·somt· rnmpanit·.~ is in implementation. After top

111,magl'ml'nt arhinatt'S on issues of rnnflkt that ha\'t' been d1·ra1t·d for resolution, s11rress­ f11l implcm<'ntatim1 oftt·11 1k1)t'11ds 1111 tlH' 11n­

rn111promisi111-: 1ommi1111e11t of managers who han· just li11ishcd lig-hti11g for a different sol­ ution. F\'ell in a \1·1•11-di~ciplined management team sudi a sit11ati,,11 ran lead to problems.

'J'lu· s11l1s1;111ti\-l' 1ll'cisi(ltl rna11a!{C11Jc11t mmlt• ofll·rs one major ad\'alltil!-{t': it is simple. It re­ quires little ol' the ~.uhdt· halanl'ing of groups tbal olhl'r solutions c;1ll for. Top managemcm retains di1t·ct n·111ral rnntrol orer the key deci­ sio11s. This dl·t isio11 pron•-~s is well-suited to rnmpa11ies whne rnnlfo ts lwtWl'l'll rn11111ry and global dcm:_,rn\s an· k11·

!'he rnmp<III\' in our sarnplt' that hesl illus­ lratcd lht· suh~ antiH· dt"risio11 111a11ag-eme111 apprn<trh wa~ a ph.mn;u t·tttit al rnmpa11y that had g1 cl\\"11 lnHn a lint' rht·11tirals busi11css with large-sc,tk, rc11tr.1!i1crl 111am1foC"t11ri11g- am! rc­ st·ard1 and a kw 1.trgt· rns\omt·is. In this rom­ pan~·. li11,11Kt·, 1u·1,onnel, and !!'gal staffs wt·re l'<'lllrali1.cd at 1lw rnip<ll'att· lcr<'I. The ){l'O­

g-r.iphi1 ally 01 gani/l'd i11t1·1 national di\'ision had built up at its IH·adqu;tr!t'l'S \taff g-roup.~ in ma1111fot tming and markl'ling th,11 were two to four time\ larg-t'I' t\i;rn t'([lli\alrnt g-roups in ~imil.tr t (llnp:rnies ~tudit'd. · 1·rtt'St' di\'ision-le\'el g-nlllps had both 11w n·so111n·s and thl' m,111- d:tt(' ro dt'n·lop global 1)('rspcr1ircs of the rnm­ pan(s husinr.,s to ( m111tnhalamc tht' nation,11 \'il.'wpoi11ts pro\'idcd hv g-rngraphir linl' maria­ ).(t'rS, · 1 ·1il'\' rl'ril'll't'd and rnmmemed on tht' major 11adrn1al pbn\ and propos;ds. lkrisinns 011 whirh Ilic grng-r,1pliir and h11Ktion,il m;rn.1- gcrs hl'ld difft·H·111 dt'ws ll't'l'l' t'lt·vatt"d to the I111t·111;1ti(lllitl l·.xt'ntlin· ('.ommilh't', whirh met l0ad1 Wt'l'k at ilt'adqt1<ll'lt'r~. Om· lint' 1qanagt•1 t·xpla1m·d solllt' of lht· l111~ttatiom with the .',\'.',lt'Jll:

I 1it"11· i, 110 q1u·,111u1 111.11 ihi, i, ,11(·11 ,1·m1,1li,1·d um1- p,w1 ,11ul g1·11i11g 1\t-(i,ion, 1r,11hnl ,an lw a ti11w- 1011,\1111iug ;uni 0111·11 t1mt1.1ti11g ploH'"· \\'1· ha1t• ju\\ h1·t·11 1h1011gl1 111'1! rw1uti11· 10111111ith'(' 11·1i1·11·, (JI\ tlu· 11un1i1J11 iii lioll" Oil\ ,,tin l,1111· \hu11l1I lw ,11111t111t·d to

h.1111\11· :1 1u·11· p1,1dw l inl1lld11< tio11. I hP 111;11 k1·1ing ',\;ill di,,1g11·1·d with 0111 ,1pp10.uli and 1.1111·d tlH· \.,.,lit' Im 11·1 i,·11. \\"l'\t· h;1d 10 li.1, k ol! mu 01 igi11,tl p1opo,al

:\notlwr rnmpa11y· lollc>wing surh an approach was an umlin·t ~ilicd European clenrical cq11ipmc111 firm that bad a hmm· market am!

(,'11/iforni(I ,\lr111agm1m/ itl'l•int•

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

1:xported, 111ostly to developing mun tries. As it engaged in prod11ctio11 in more countries and adapted its product line to \oral needs, the com- 1>a11y faced diflirnltics maintaini11g its central­ ized 111anagemc·nt.

Sufo·trmtil'f /)l'fisio11 Arbitmli,m. A simple answer to some of the problems of suhstanti\'C decision managenwnt is to delegate the important halanring responsibility to trustt'd arbitrators. ·ro ensure that these arbitrators ha\'e the power to make dcrisiom on top 111a11ageme1lls' behalf, both gnigrapliiral units aml proclun di\'isions must dcpeml on them. A Swedish telecommun­ ication cqui1ime11t mant1farturcr adopted this approach. LM Ericsson had about forty manu­ fonmin,l{ or assembly suhsidiarit·s abroad, and three major produfl dirisiollS for switrhing equipment, transmission equipment, and tde­ pl11111c sets 1111<\ arress,iries. ·1·1ie divisions were ha~cd in Sweden, with all deH'lopmcm and most nHnponcnt manufat'turc there. ~lost for­ t:ign suhsidi;1ries were aniw onlv in assembly_ i11stallatio11, am! mai1llenanre wo;-k, · Both the produn dil'isions and the subsidiaries wen· entirt'ly ~cp;1rale suhunits with no direc' fi,n11al c·c>ntal'I. Relatio11shiJlS hctwel'n 1l1e two took plare primarily through \'arious corporate .~ervin·s. For i11sta11re, the Oflire fo1 O\'erseas Produrtion pnn·ided support for asmnblr and manufarl u1 in14" operatio11s for all prodtKt lines prnduc<'d hy the sub.,idiarics. ~lost prominent among the roordinatill).{ staff managers were area marketing managers, ead1 of wholll was respm1sihk for relationships with a group of subsidiaries. ·1·11cir 1es1ic,nsil,ilities ranged be­ yond marketing to many general management issues. Their balancing of dil'ision executi\'es prcocrnpied with Ol'crall rompetitireness and national subsidiary managers concerned with 11ational rcspo11sin:11c~s was dt'scrihed by one area 111·.1rketing 111;1nager: I help tht' or;il ma1i,1ging din·r101 s. The1· ( ;in rail ou me 11·hc111·1·1·1 tl1t"1· get into a 111;J!1t•r .1how 11hi1 h th(·\· an- not too \'.lie ot themwll('\ .. I gt"t into a 11·hok ,pen1um ol ljllt''>!lo11s. flrnn Lite clt-lin·rit'-; lo the 01('1;ill 1t1at(·gir rn• 01di11;11i()lllOIOll('nmti1w11t!

1·1l 1lw 1li1i~i(>l1, i·1,u lt']>l('\t'nt l]w Jello\\' i111lw !ielcl, ancl 1 ice n·1.,;1. You h.tH' to lt-;ul tht'm to rnmpromi,t· IO make dee i,io11s rnrne 0111 i11 a 11·a1· Wt' 1hi11k hes! tor 1he whole 11u1111a111·.a11d ;utl'pt,dik- fi)I hotl1 ofthe1nl People knoll' 1h;1t i! they di1;1grc·(• with lll(',lhe1· nta)· hal't• to p1t•.1ent a s!nmg ;11g11UH'lll. In a ~t'mc it i, pmn•1· without lo1mal

Sl'RI;,/(; / 1!181 / \'OJ.. XXIII / ;>/0, :l

authorit}', OI rnurse if I am wrong too often my poll'er decreases. m I mm·c onlr 11'11en I am sure of myself. Sometimes you ha1·c to gin! mnrc support to the loG1I, weak rnrnpanr than to the hig. toughdi\'lsion, hut the kcr to s11ccess i., to remain nemral, tn hd1a1·ea~a hroker. The situation h halarKed enough for that.

It's ;1 game where there (<m be 110 winr1e1 lmt \\'here ~ome mm petition het\1'<:'CII m:mager~ i~ built into the sy.\tcm. A m,magcr nut in th: field c;mm11 pull ,111 the strings himself, 11orra11 a dilfoon h.iH· !ornl"agcms" within a mhsidiar)'. 7

Compared to the dedsion m,magcmcnt ap­ proac:h, decision arbitration adds a screening level with top management intervening in ded­ sions only when corporate staffs fail to balance decisions effectively. Yet, like the decision man­ agement approad1, the decision arbitration mode tends to neate divisive forces within the organization. Key decisions arc often reached hy top management and for<:cd on rnanagers who had argued against them. Reliance on de­ pendencies between subunits to gire power to arbitrators reinforrcs the di\'i:;i\'c forces. At LM Ericsson, se\'eral medianism, offset this divi­ s_;reness: many managers (Swedish engineers I rorn the Stockholm telecommunication engi­ neering school) shared a rnmmon cultural identity; job rotatiou between units was com­ mon; and long years of indoctrination usually were a 1>rerequisite for a numagcr with respon­ sibility for a suhsidiat)'. National demands for local management in the subsidiaries chal­ lenged this approad1, and th:: informal unity of the company was threatened. As in the previous apprnad1, decision overload threatens the derision arbitration process. l.~·f Eri<:sson rnuld operate effectively because it was a single industrr rnmpany with few divi­ sions. C:iven the long-term relationships with postal and telephone authorities in the host rnulllries and the dear-cut generations of equipment, m;~jor strategic decisions were few aml for between,

Tn11/101wJ Coalition Ma11agr111n1I. In companies facing rnmplex tedmological or competitive enriro11ments and those where tension be­ tween national and global demands was in­ tense, top management had to retreat from direct imolvernent in the nmnerous critical de­ cisions and manage the process by which such decisions were made.

Top management may build asymmetry in de-

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

< is ion prcrt'~S<'~ hy inllucm·ing 1hc rnmposilion 11! th<· g1 (m11~ 1ha1 nmkc dc<·isi1ms. 'l'hcy di1cn their .11t<·111io11 to the inputs of the derision jll 1Jrcss, clet iding wbid1 nianagets slHn1lu par• ti< ipatt· i11 makir1g.whirh derision.<;, awl in what role. For ~ome ch'ci~ions. geoµ;1aphil, produrt, and furn 1io11,d management groups were all gin·n a ~trong rnict'; for other.<;, diffen_•nn·s in a·-.pomih1li1~· :md power were mon· dea1 ly a, .. igm·d.

n\(' need ro eng;1ge indi\'idual mauagcrs i11 < c1 lain dn i-.ion prore.'>Ses and di~engag<' thl'm lrom qthcl"'i, and to prcserre flexibility ill thi<; rhoir<', led top 111a11ageme111 to 1elr c·xtemiH·l~· on tcmpoi,11~· <;l1w1ures aml ad hor group<;, P1 ojt·t I l<'itlll~. joint la<;k lorn_•<;, rnmmittee<;, and wot king- partie~ helped !Op m,mav.emt'lll dl'~ign tht· rnalit1011 ofvit·\\')loi1m it felt appro• p1 iate for a pa1 ti< ula1 deri<;ion. The~l' 1,.•111p1ir­

;11 ~· '>11 lH tun•<; olte11 p10\'ided stalf group~ with a powt•1 lul me;m., ofinfhtl'tl((' in .~tratt•gi1 de! i­ ~ioi1 JHote~,t·s, and W<'r<' forums i11 whid1 111,magn, n1uld rt·soh-e dilferemcs aml reach joi11t de1 i~iom to wliitli 1ommit111ent~ 1ould be m;tde. ( :omiderahll' tboug-ht wa~ put into tht· tompo-.itic111 ,,1"1\wse g10l1ps to emutc they n·­ He1 tnl top rnan.igt·rnt•nt's pcr~pecti\'l' on the de< i~ion. i.l'adc1~hip of a group, ~rnp<· of i1, a'>\ignmt·111. ,md 11a1un· ol tlw output (a11aly~i,, lt'tOlllllH'tHlatiun,. dcri,iom) were <a1clulh clt•[("llllill("Cl.

\bior tomide1a1iom in ;Migning manage1s to tcmpot ar~· d<·t ision rnalitions Wer<': ., l,eg:itirnar~: ·1·owhatcxtcntdoesa111anage1

add to or rnmp1ornis<• the legitima<) of tht' < otH lu,icm of the g1m1p?

• Dilll'tt·11ti,1tio11 ol i11fo1 mation: \\'bit b lat b

and poi111, ot ,iew are Lo bt' represented iu lht· group 01ml whit h arc to be ig11med?

• C:nmmitment: Docs the 1-:roup haw the < lout to make rl'\'ersing- or igno1 ing its findings clil!i1 uh for top rn,111agct11t•11t? Is that 1hcolijl'c­ tiH· ,ought: ('-;irnilar que,tions ma) be 1ai~ed lor ,1dH111i1 111a11ai.;-c1~ who di~ag1ee with the gtoup\ < ow lu~ion<;.)

l II Olll' ( 0111p,111y ~clli11g- rnll<;U111er produt t~. a ltcq11t·11t 1e~pollM' of the divi~ion prcsidl'nt to '>ales dow11111rns or othn dillirnlt!rs in a ~11h­ si<lia1y was tn appoint a marketing task fot<e to

iO

analyze th<' situation, to make rcrnmmcnda­ tiom to tht' country manager with n,pil's to him, am\ to report on progress. Hy appointing ;1 l1cmlquar1ns proclurt 1narketi11g manager to head up 1he task forrc am! imlirating that ac­ tion would he agreed upon, the president cffrc­ tivcly < 01111tc1 halanred the power advantage of th<' line rnanag<'r in d1argc of the subsidiary. Saicl ()Ill' (livision 11iarkcti11g 111anagcr: I hi, ,11lp1 t>,u h h,1, h(·t·n I l"I r ~ul te,~lul IOI 11s berau~e the

(OJllllCI\ lll,lll,1gt'l\lH)\l"IC\f)l'lltlw1ap,1h1hlll'SOfth!'prn• dut l Ill.ti kt·tilll,l' 111,rn.1ge11w11t l,l'rm1p, that ha1e ll(·r11 (\e- 11·l111>t'd m1·1 clwp,1,1t1·11·ir,11,.llu111m1elumlanwn1ally, 11t· all 11-.,lill' th.u IH' limf 10 11ml-. wge1he1 to \oln· thr p1ohll'm, \\'ith till' d11i\iun pH·,iclt•flt t·~1n·t1inl{ reml11- 1i,,11, it\ ,11n;i1mg 11011 111u1 h 1>1ng1,-,, W(" mah· together.

l\rist1il ~lye1s neatccl a pl1annan~u1iral nmnril within ii<; international oq(ani1a1io11 to allow

g-t'ogt aplii: line 111;111ag-ns and lmsiness devcl­ opn" 'll ,ia l 1-•a11al{er.<; to debate 11ew produd prioi 1ti<·s 1., .ore dt·t iding- on the long-term p1odt1< t dt•\"t'loprnenl .~tratcgr to be suhmiued 111 it~ ror ,(JI alt· re~ean·h 1mli1·~· rnmmitlee. Pro­ pmals cornirw out of the pharnlat·eutit-al t·otm­ ril imorpor;:t1•d the appropriate hal<UH't' hc- 1wee11 the clcrnaiH!s for global competitiveness and lm.11 n·,ponshl'tll',.,.,, and typit"ally had the '>llppoi t and n1mmitmelll of the m,magemt·nt

g-1011p imolred iu 1lwirdt'\'clop111ent.

For t<'mporary rnalitions to work, the orga11i1.a- 1io11al .~l1 w ture a·id ~ph·ms tatmot he Ml over­ whelmingly fm m<"d on 011e management pcr­ spcrtirc as to rnak{' irnpmsihlr the introdurtion of a diffl'rcnt point of view. The perrci\'cd kgitiruary and power of the individuals brought in mu.'>l be adcquatl' to prnvide the rcquirt'd pt'1spet ti\'e. lk\'clopi11g rnmpany­ wid<' r('~jlt'l'I and an-cptauce of the various p<'rspet tiv<'s that 11ct'd to be cng-aged in the decision p1 oress is a prerequi~ite to the use of this app1 oarh. Surh a ta~k may take months and t·,·en years to an omplish.

Swis~ dt·n1i1·,il giant Brown l\ovcri found it diflirnh to tcly on surh ~uuctures to develop rnonli11atec\ produrt line strategies among its \'anous national afliliatcs. Sinrc the rompany's <>rg-a1ii,ati(111al ~trul·turt' a111\ administrati\'e systems were built almmt exdusively around a g-cographic foLUS, the intersuhsidiaq' produd and business rnmmittt't'S were hardl}1 more than mt'ful forums for cxl·hanging- ideas. All

C'fllifornifl 1\lfl1wgi'111t'11f R1•1•i11w

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

husiness results were measured country b)' country, all <'aree1· path.~ were uatioual, profit <'('liters were national divisions, and business plans and hudg('ts were consolidated on a mt­ tional basis. There was liulc support of views 1101 aligned with a geographic focus, and little moti\'ation for husillcss ur product teams to reach results dirergent from a geographic forus.

Th(' ro11sta111 forming and reforming ol dc<·ision-maki11g groups awl manag:emem of roalitio11 formation put g:re;;t cletn;!11ds on top 1na11age111e11t, rcq11iri11g u;,11tinual l(Kus on group processes, the ability to neate and man­ agl' ohl'll sensitir<' intcrpnsonal si1u;111011s, and alll'lllio11 to the stallls awl pow<'r that follow any d1a11gc in .,trurture and decisio11 proresses. ()11(' 111a11av;er n)lllllll'nted: ~;lllll('tilll('' pn1cl11, { t'X('! 111i1t·, ig111ue tlw ;ll('il ma1i.1gt·1~ i11 ~onw (HI rl'!>IH>11(k!Ht' cu· (l(u umt·rl!~. ()1 pl'c1pl1• srht·d­ uh· llll't'ting~ ,o .i, to make ,mt• a m1•mh1.•r of tll<' group will 1w1 li1• ,1hll· 1,i ,Hlt·1ul. I 1rn1,1a111l1 ha1·1• to t;ik (arc o! who i, .11i1·1uliug 11hit h nuTting,. ol 1,how•c,whirh drn ll· nw111,. aml ;111 ~Ul h n1i11,1·11w. I ahl-ay, m·t·d tu h1 ing h,1t k 1lwli,1!:t1Ke

If top 111a11agcme11t (·1111 build a11 orp;anililtion in whirh strong indi\'iduals representi11v; \'ari­ c>11s i111porta11t 111a11agc111ent 1>crspectin·s ran ht· readily mored into am! out of key decisions, it has control of a powerful am! lll'xihle deci- sion prorl'ss.

D1·(i,1io11 Conltxl ,\1m1agl'l11m/. Like the tempor­ arr rnalitio11, decision context ma11agemem de­ ~nihe~ an cll\'iromm·111 in whirh !Op manage­ lllt'llt is rn11cn1lt'd with the prorcss of decisioll making as rnurh as the s11hsta11re. Rather than requiri11g top management to ronstrw:t and monir.or the patterns of inlluc11rc in srnrcs of difti.·rent derisi<>ll gnn1ps, tliis approach at­ tt'mpts to ncatt' an ovnall organizational envi­ ronment th;1t rnotiratt·s manag:ers of \'arious perspcctin·s to initiate int1.•r<Ktion 011 majo1 prohlt·ms in which dilf<.'n.'11n·s in pcrspecti\'e are niti(al. · !'his mork is hased ml a belief that rn,ma~e111ent can set a derision rnntext in which II' magers (an themsclres form the ncrcssarr alli,rnn's to deal with issues where multi1>k e11,·in111111c111s 11ced rnnsidcration.

Dcrisio11 romexl manag-emelll allows many m11ltidime11sio11al issues to l,,e negotiated and resoh'('(\ al i111ermcdiatt' rnanav;cmrtll lc\'els,

Sl'Ri7':<: / l!lHI / \'OL. XXIII / '.':O. :1

without inrnl\'ing lop management. InM's "concurrencn system" is based on this pre;nise. Each subunit, defined by geography or pro­ duct, has,,,~ right to object to the plans of any other 1 , r T\cgatirdy affected in any way. When an objccl'.on is made, the two units en­ gage i11 negoi · ·,rion; when resolution is impossi­ ble at middle rcls, the difference is fonnally registered as a 11011-rnncurrl·nce" and brought to the attention of corporate staffs and top management. l's11ally these require further cffon, this time by an ad hoc, intenmit task force. Should an agreunent fail to emerge, lop uianaKement can rule din:ctl~·. hut the overall dcrisio11 rnntext discourav;es surh e~calation.R

Beyond using formal mcd1anisms, such as the strurturc of the planning ;111c\ budgeting sys­ tems or the patterns of measurement and ernl­ uation, management also tries to influence rc­ latio11ships tlirough informal structures. 9

Examples include top management's derclop• mcnt of a strong internal \'alue s~·stcm, a strong rnrporate identitr, and a set of beliefs which cle!incs norms for interpersonal interacti011s. Career patterns, the .mcrcssion of assignments in a manager's career, and a disassociatior. of business risks from personal risks can be l'.scd to foster a coopcrati\'c ralue system whid: dis­ courages narrowly defined, short-term, ielf­ interested bcha\'ior. The objective of this ap­ proach is to de\'clopa gro11pof managers with a willingm:ss and the ability to represent a partir­ ular perspective, an understanding of other needs, and an apprni,1tion that 1J\'erall corpo­ rate ohjectires may require different perspec­ tires to pre\'ail.

One of the companies we studied that ap­ peared to be makinv; major efforts to manage its decision context was Eli Lill\', 1he L'nited States-based phannan:utical c~mpany. The ability of this company to develop an emiron­ ment in which managers fostered rnoperatirc rather thall adversarial relationship5 on rnn­ troversial issues deri\'cd larv.ely from long­ established internal ra\uc systems. There was a stroo1~ norm that manav;ers he open, honest, and mutually trusiing. Many of its administra­ ti\'c systems rcllected and rcinfon:cd the com­ pany's informal, rnoperative e11\'ironmr1n. 'I'he evaluation system, for example, required

i\

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

that managers he c,·aluated not only by their line superiors, bm also h}' nmnagers in other parts of the rnmpany with whom they had regu­ lar wm-1..ing rclatinnships. Career de,·clnpment was carcfull}' monitored to ensure that key m;magers were exposed to corporate \':tines at headqual'lers, and that the}' gained an expos­ ure to a ,·,nicty of pcrspecth·es and indi\'iduals. Hy trnnslenin,; a manager from a line to a staff position, from a subsidiary to headquarters, and from domestic to international responsi­ hili1ics, it was hoped th.u the incli\'idual would g;1in ;111 .ippredation of the muhidimensional­ it)' of most problems and would de\'clop pcr­ s011;1I rela1io11s:1ips with those with whrnn he or she might later ha\'c to ncgotialc, This< nrnur­ <tgl'{I rnopcration and mutual acrnmocla1im1 Oil

many issues that might othern·ise have 1csuhed in rnnflirt or required arhitratin11 at a higher h·\·el.

\\'hen rritiral derisions are frc,1uc111 and inwr­ clt·11t·mk11dl's rnmpll'X, dcdsirn1 rn111ex1 llliUI· a1,;emt•11! may Ix· neeck·cl. It is flexihlt• in rnu- 11li11g: cir u11tc111pli11g ~uhunits and integrati11g1,r st•parn1i111,; different pcrspt•l:li\'cs. It is inrum­ hent for middle managers 10 deddc when and where multiple pcrspc<:ti\·csare needed, and to t1kc initiali\'l' amm:lingly, lnformation and nwasun•mclil systems arc set to enmre 1ha1 t•xn·ssi\'e asynunetry ist:allcd to the anenlion of

High

top management. A national subsidiary man­ aging dircrtor of a company using such an approach commented: I dt·,clop a rnmpany hmlgt'I. ,\ prndutl ,·kc-prrsidrnl ftom 1hc l'.S. "'ill romt• IU me to lalk ufmo\jor dr,·cl11p- 1111•111s lw widlt's tu Sl't' i11rnrpo1atrd into our plam. llc nmy .il,o go IO onrof my di1·ision 111:um1-1crs; howc,·cr, any imt·· •mr111 m d1;1111-1e in 1hr hudgc1, if ii in\'Oh-cs a large am mt, tll'ed\ tu Ill" pan of an mcrJI\ Mrntegir plan disrnsst•d with nil'. If I deride to ill\'rst, it ought to he a , iahle hmitll'~\ from mr point of' 1·il'11·, n1hrrwisl' the op­ t•1atin11 "·ill 1101 Ill" hmdt•d. If hr thinks tliffC1cntly, the prntlurl rirc-ptr~idcm "'ill rrnm· hc1c and uy to mm-ince mt·. Lllt'r I ~lmll di~tuss my bmlgct with thr hcml of illlt't 11;11io11al 111K•1atiom. I Ir makt•s tht• bud gel al\nrnlinn. I gr11c1allr follow hut rnin· my 1csern11ions. I must S.l)' tht•\' U\' IO rnm·imt· me. I do nut retail somrthing l'\'Cf h.i,:ing 'IK·t·n impmt·d on 1111•,

To a large exlent, smTCssful derision context management is a product of history, rooted in the rompany's leader's values. The IHM system owes murh to Thomas Watson Sr.'s \'alucs. Eli 1.illy's \'alue system ;1lso developed over many years. ·1·hc late f\1r. I.illy oli.t•n cmplmsi1.ed the key role 11f the company's \'aluc system: \'ah11•s a1t• c111iu· ~imph- tht• mlt' 11flm1h men and insti- 1111ic111,. Hy wmhini111-1 11111 thought~ ;md helping one mmtlicr, "'t' ,111· ahle In mt•1w· the p:1r1~ lofthl' 111rpora­ tiu11J i11tu i• r,.1icmal, workalilt· 11mna1wmcnt sysl('III,

Successful derision context management hi11ges upcm two complemcnt,uJ top mam1ge­ me11t tasks: the 11urtmi11g of a rnrporatc \'.tine

Lo11._ ________ _,___ ______ ..,,__ ______ ~

Low

i2

Frequency ol dec1s1ons mqumng t1alancin!J between national demands and compM!llve pressure

High

C11lifi11·11ia Mmwgr111r11t llrt•irw

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

sysLC'm which creates a legitimate, orga11ization­ al stnu:ture and provides a sense of common identity for the organization's members; and the legitimization of the duality of perspe(tivcs by imhcdding them into the stnu:ture and rnm­ ple111c11ti11g this structure with numerous, merlapping administrative systems whid1 (ol­ lerti\·cly create a rnoperati\'c cmironmcnt.

Conclusion

·1·hc four ma1mge111e11t approarhe.l just des­ cribed arc neither mutually cxdusi\'c nor total­ Ir intcn:hangc:1hk'. They rnrrespoml to differ­ ent levels of ('()lllplexity ;ul(( tension in the ded­ sions being f:Ked by the company.

The rnmplexity of nitiral dec:i.lions varies sub­ stantially hctwt•en a company with a single busi­ ness in a few foreign markets, and a diversified <:ompany operating in a larg"e number or 11a­ tio11al environments. NOl onlr is there likely to he an inrreascd ncecl l<lr rnordinated manage­ ment approach in th1} latter situation, hm top management will be less able to maintain a de­ tailed understanding of the suh.~tarKc of the 1111n1erous n1111plex derisions (entral to thr• company's surresshrl operations.

1 n husincsse.\ where tension between global in­ tegration and national responsiveness is high, the need fordetisions that reflect both perspec­ tives may force ma11age111t'tlt 10 cle\'clop pro­ cesses that provic\e su(h outrnmes rather thall try to personally integrate the disparate \'iews on each issue.

A flm· dwmirals rnmpany probably face.~ fcwct nimplex a11d ron!lirting decisions tha11 an ethi­ cal drng linn. :\ telephone swit:hing" equip­ ment business probably invoh·es fewer c:om­ plex strategic issues than a global romputcr husiness. The rnmplexity and potential rnnflin in clet·isions depends, in part, on the nature of the business am\, to a great extent, on the strat­ egy a tirm adopts to manage that business. One drui-:- rnmpany may ht' willing to yield to host t·ou11t1-r pressure to huild d1e111iwl plants in ma:1y t011ntries arou11CI tile world; a11otl1cr may choose to retain a single global wurce for dli­ rie11t-y's sake and not rmnpetc in national mar­ kets where g11ven1n1e11ts rc11uire !oral produc­ tic111. ·1·11e sem11d rm11pa11y's strategy clearly re­ duces the tension between global and national

Sl'RINC / 1981 / \'01.. XXIII / NO. :t

perspectives and can be implemented with mmh simpler .idministrativc systems.

The adoption of the more complex, more sub­ tle means of managing tension between nation­ al responsiveness and global integration docs not require companies to disrnntinue other Tr,odcs of management. Top management of companies we studied that employed the deci­ sion rnntext approach still engaged in substan­ ti\'C decision making: on maj~r or unusual issues. The broadening 1)f their repertoire of dedsion making simply allowed them to avoid some of the problems and limitations of the simpler approad1cs.

The solutions outlined here were presented in order of increasing complexity; it is likely that managers of large, complex multinational cor• porations will need to employ the temp0rary rnalition and decision context approaches (see Figure I). As a multinational company at­ tempts to develop and impleml'nt a more com­ plex strategy, with diflicult trade-offs between national responsi\'enc:;s and worldwide inte­ gration in each of' its lwsinesses and extensive interdependencies atrnss businesses, it has to adopt a more complex management mode. The limitations on the :;tratcgy aie not analyti­ cal, but rather administrati\'c: the complexity of the strategy is rnnstraincd by what is ad­ ministrati\'cly feasible.

REFERENCES i. See \°\cs I.. llw. ,111d ( . K. l'rahatad, "llow ~f;\C:,;

(:ope 1•·ith I !mt Cm·crnme111 lk111aml,;," !lr11Hml Uu1ilff" lfr!'i1'!I', \'ol :"1H, :-,;o. 2 (~lard1-.-\p1ii l!lHO). pp. H!l-WO; also \\cs I .. Dn,. "Str,llq~k ~fonagrnH'IH in ~luhinatiou;il Co!llpanics." Simm ,\lm1111;rmml Rn•i,,,,,, \'nl. 2 I ;\o. :! (\\'in1c1 1!180), pp. '.!i··l!i.

~. See Christopher :\. 11:11 tic!!. "~luhinational Stnic­ llll'al Ernluticm" {unpuhli\lwd doctoral di1~c1 tatirn1, 11.lr• 1,11d Bmim·s~ Sthonl, ]!Ii()); Yrcs I.. 1)(1/, .\fultim1/i(ll111/ S//'/llfgir ,\lm111g1·11u•11/: Eo11wmlf 1111(/ /'0/1/i,al /111/1r111/n·r1 (fol thrnming), arnl l 'liid1 \\'ri, luu;um .. l/1ukrli!l!f ,\fa11- 11g1•111,,11f m .\lulti1111/io1wl Firm, (:-,;l'\\' \'ork: 1'1aegcr. 1!178).

:t Sn· Chri~ :\q.rytis a!HI llunalcl :\. Schon. Orw111iu1- 1i,mr1/ l.mmi111;: ,\ 1'!1111ry ,if" :\r1im1 !'r11/wrlirr (RC'ading. ~I:\: :\ddi,(m \\'t''ill'\', 19iK).

I. c:h1istoplwr :\. llanktt, "~!11hinati1mal Strurll1r.il Ch;111gr: Ernlution n·rsus Rcorg,rni1a1io11." in J.ar~ Ot- tcrlll'dt (ed.), of llrmlqmnlm•.\'11/J1idim_,· Ud11tiomhi/H i11 Cm/i!m1/i,,m (l.';irnhorough, Fiigbml: (;owcr. 1!l81)

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

1•1,\111/t11m/1111m/(.'.,tf11,w/1t,11, (l.1111b01,>11gl1.

11!1'11)

", \'H·, I 1l1J1. (,,,.,,,,.111u·11I c,.,.iwl m11/ .\fu//m11/i1111,i/

',/111/,1;•1.\/1111111;11111•11/./',..11·1 \_\1/fl/ll(l/ll/"f"r/l'!lltrlllilllll111(1m1

/:11111/m1,11/(\,·11 York: ]'1,1rg.11. l!li!l).pp. 111)-l l!l

h (;,111·111 I' lh.1,. "'',11.itq.(I ,mcl ',t11utu11· ,H I J 1 ··

il.\',l ,\ll111:11w1,. l!li'I)

i. ]<)I ,1 dt"t,11ll·,I dl0 ,tll]>ti1H1 ,,I I \I F1i,wu1 ,1ml

1\1111111 I\Ol!"\L ,1·1· [lo/, (;1111111m,11/ Cm1/1,./. ]),ll,l 011

1l11•ph,u111.1tl'lltl1.1\;i11d:H',thhp1odt11t,1omp,lllil',llll'II·

1io111d 111 il1l" .111i,ll" 1,111 ht· l1J111ul in l\,111!t"t1, " 1 1uh1.

11.1t1<>11.1l',11u·1u1.1lE111hu11111 ··

,'I 'it't".\IH,th.1111 K.111. ··1'L11rniugi111lwlll\1( 1upo1,1•

7,\

tHm,·• papn JHt'st•nwd d! tlit• ·1 [\[',.()RSA "ip1•,ial

l1ttt'tt·,tC011k1l"ll1t•o11"i11,111•gi, l'l.u111ing. '.\t•w(l1lt•,111,,

l!lil

!I. \\'c li.i1t· di,111"l"d 1lww ,1dmi11i,11,1ti1t· ,1,t(•111,,

,llld 1li1·i1 p111pmi1t· l!"' 111 ddini11g ,mrl i•1plt·11w111i11~

,11.11,•g1,·, 1·,t1·11,i11·h 1·hP1d1t·H·. ~tT C. K \'1,il1,1bd.

··~11.1wg11 Ch1m1·, i11 1li1("1,1tit-d ~f;\C\./11m'/lld/J1111w11

1/11•1,w, \'ul :)1. So. I (111h•.\11gu,1 l\!il)). pp. Hi;

Y1c, I .. ll1Jt.,l/•1//ma/1,1111i/ <;1.,1/i';;u .l/111m;;,1111·1il: l·.",0111111m

,w,/ J>,,/1/111,/ (!011luomi11g) (.h.1ptt•1~ !l-1'.!. \\1, I. llw ( h.. 1'1.1h,il,1,I. '"ll1·.i1!,p1:111,:1 Co1111ul

,111d ~11,tll"gn Ch,111g1· in ~1ull111,1tio11,d ("1Hp1J1,1tio11,," in

Oi1t•i!)1'1h.op 1i1

(,'a(1fm11111 ,,!1111a1;1'//ll'III lfr1•iiw