Cage Framework 381k31

SAM MPLE 2. Diffferences an nd the CAG GE Distancce Framewoork1 Pankaj Ghhemawat Afteer analyzingg the casess in sectionn 1, the reallity of semiiglobalizatiion and the importtance of crross-countryy differencees should be b clear. Thhis secttion introduuces the CAGE C distaance framew work, whicch is used to idenntify and prrioritize thee differencess between countries c thhat companiies musst address when w develooping cross--border strattegies.2 Beggin by conssidering thee example summarized in exhibiit 2-1, whiich plotts Walmartt’s operatinng margin by countrry in 20044 against tthe distance betweeen each coountry’s cappital and Walmart’s W heeadquarters in Benntonville, Arrkansas. Thhe impact off geographicc distance iss obvious, bbut whaat other typees of differeence or distance can yo ou identify that t separatted the markets thaat were proffitable for Walmart W from m those thaat weren’t? Exhibiit 2-1 Wallmart Internnational’s Operating O M Margin by Co ountry (2004 estimates)

1

Thee CAGE disstance frameework disagggregates diistance or difference d innto fourr major categories: c Cultural, istraative, Geoggraphic, aand Ecoonomic. Diff fferences aloong these diimensions generally g haave a negatiive effeect on manny cross-boorder interractions, although in some casees, diffferences aloong a limited subset of o CAGE dimensions d can actuallly 3 encoourage rathher than disccourage succh interactio ons. Each of o these brooad typees of differeence or distaance is illusstrated by th he Walmart example. 

Culturaal distance:: Culture can be defin ned as the collection of beliefs, values, andd social norrms—the un nwritten, unnspoken rulles of the game—thhat shape the behaviior of inddividuals aand organizzations. Cuultural distaance encom mes differences d in religiouus beliefs, race/ethnicit r ty, languag ge, and sociial norms aand values. Societies even diffeer in their social attittudes towaard market power andd globalizaation in waays that haave importaant effects, both form mally via reegulation and a informaally, on hoow 4 businessses operate. Interesstingly, Waalmart’s foour profitabble marketss share lingguistic, religious and ethnic simiilarities or, at least, tiees through large l diaspoora.



isstrative disstance: Historical an nd political associatioons betweenn countriess—colonial links, freee trade agrreements, tthe tenor of o current relationshhips—profou undly affeect econom mic exchangge betweenn them—w which is th he same ass saying thhat differennces along these dimeensions mattter a greatt deal. So, of course, do nistrative attributes a specific s to a particullar countryy such as auutarchic pollicies or weeak institutiions and high levels of o corruptionn. In the Waalmart exam mple, note thhat two of tthe profitabble countriees, Canada and a Mexico o, partner wiith the Unitted States in i a regionnal free tradde agreemeent, the Norrth Americcan Free Trrade Agreem ment (NAFT TA). And a third profittable “counttry” as classsified by Walmart,, Puerto Rico, R is officially an unincorrporated terrritory of thee United Staates.



Geograaphic distannce: The geographic g dimensionn of distannce involvees more thaan just how w far two countries c arre from eaach other: other o attribbutes to bee considereed include contiguity, a countryy’s physicaal size, witthin-country y distancess to bordeers,

2

access to t the oceann, topographhy, and eveen time zonnes. Exhibit 21 makees it clear thhat the capital city off each of Walmart’s W foour profitabble “countrries” is geeographicallly closer to t Walmarrt’s headquaarters than the capitalss of any of the unprofitable ones; in additionn, Canada and a Mexicoo share a co ommon landd border wiith the Uniited States. 

Econom mic distancee: Consumeer wealth an nd income and a the cost of labor are a the most m obviouus (and reelated) determinants of econom mic distancee between countries. Others includde differencces in availlability (or lack) of reesources, in nputs, infraastructure aand complements, andd organizattional capaabilities. It seems a bbit harder for f Walmarrt to do welll in poorer countries— —although tthe numberr of data points p is very v limited. Note, however, h thhat econom mic distancee has not been b entireely or evenn primarilyy a liabilityy for Walm mart. The company saves morre money bby procurinng merchandise froom China— —exploitinng econom mic distancee, particularrly in s of labor co osts—than it i makes froom its entiire internatiional store network. We W will reeturn to suuch strategies in sectionn 5, which discusses d arrbitrage.

Whaat the Numbbers Tell Uss e Inteernational economists have adaapted Newtton’s law of universsal gravvitation too describe trade annd other internationaal econom mic inteeractions. Thus, T the simplest graavity modeel of internnational traade betw ween two coountries preedicts that trade t will bee directly reelated to theeir econnomic sizes (a unilateeral attribuute of each country) and a inverseely relaated to the physical p disttance betweeen them (aa bilateral orr country-paair attriibute). Auggmented graavity modeels add meaasures of other o types of diffferences as well w as unillateral attribbutes. Exhib bit 2-2 show ws the resuults of one o such annalysis that evaluated cultural, c ministrativee, geographhic, and economic effects e on trrade.

Exh hibit 2-2 Effeects of Simillarities Verssus Differences on Bilatteral Trade

3

D Dimensions of o D Distance/Proxximity

Determinaant

Channge in Trade

C Cultural

Common laanguage

+42% %

A istrativee

Common reegional tradingg bloc Colony/coloonizer links Common cuurrency Differencess in corruptionn

+47% % +188% +114% –11% %

G Geographic

Physical disstance: 1% inccrease Physical sizze: 1% increasse Landlockeddness Common laand border

–1.1% % –0.2% % –48% % +125%

E Economic

Economic size: s GDP (1% % increase) Income leveel: GDP per caapita (1% incrrease)

+0.8% % +0.7% %

Sourrce: Pankaj Ghemawat and a Rajiv Mallick, Ma “The Industry-Levvel Structure of Interrnational Tradde Networks: A Gravity-Based Approacch,” working paper, Harvaard Busiiness School Boston, B Februaary 2003. The estimates corrrect for unobbserved threshholds for participation in trrade and are all signiificant at the 1% 1 level but are, a in a numbber of cases, smaller s than those t reportedd in manyy other studiees, apparently due to the corrrection

Thee signs on most m of thee estimates in the table probablyy accord wiith youur intuitions (althoughh they cannot be reeconciled with w a fullly globbalized “flaat” world). What are probably more m surprrising are tthe maggnitudes off some of the t effects— —for exam mple, that countries wiith coloonial ties arre apt to trrade almostt three timees as muchh as countriies withhout them, or even moore if one allso ts for the roole of colonnial ties in generatting culturaal similaritiies! The peersistence of o such larrge effeects decadess and, in some instannces, more than a centtury after tthe origginal coloniial relationsships were dissolved reinforce r thhe conclusion thatt complete globalizatioon—as in the t disappeearance of the t effects of suchh considerattions—is exxtremely unnlikely anytiime soon. milarities verrsus differeences along many of th he same dim mensions allso Sim helpp explain fooreign direcct investmeent or comp panies’ foreeign presencce. Thuus, for U.S. companiess that operaate in just one o foreign country, thhat counntry is Canaada 60 perccent of the time t (and 10 0 percent off the time itt is the United Kinngdom).5 Grravity modeels have also o been adappted to explaain

4

crosss-border interactions as diverse as equitty trading, e-commerrce trannsactions, patent citatioons, immiggrant flows, air traffic,, phone callls, and even the inncidence of wars! The basic b conclu usion from this literatuure is thhat differennces betweeen countriess—and diffferences in differences— — mattter in signifficant, prediictable wayys. Idenntifying andd Prioritizinng Differencces Havving highligghted the peersistent imppact of crosss-country differences d or distances, the rest r of this section foccuses on ussing the CA AGE distannce fram mework to identify and a prioritiize the diffferences that t must be accoounted for in developiing global strategies. s Exhibit E 2-3 helps in thhis regaard by identtifying bilatteral and unnilateral facctors to conssider for eaach of thhe CAGE categories. c Exh hibit 2-3 Thee CAGE Fraamework at the Country Level

Cultural Distance Couuntry pairs (bilaateral)

Couuntries (uniilateral)

 Different languages  Different ethnicities; lack of connective ethnic or social networks  Different religions  Lack of truust  Different values, norrms, and dispositionns  Insularity  Traditionallism

nistrative Geographicc Distannce Distance

Econom mic Distance

 Lack of o colonial ties o shared  Lack of regional trading bloc o common  Lack of currency  Political hostility

 Physical distance  Lack of land border  Differences inn time zones  Differences inn climates / disease environmentss

 Rich/pooor differennces  Other differennces in cost or qquality of naturral resourcees, financiaal resourcees, human resourcees, infrastruucture, and infoormation or know wledge

 Nonmaarket/closed econom my (home bias vss. foreign bias) o  Lack of membeership in internaational organizzations  Weak institutions, corruption

 Landlockedneess  Econom mic size  Lack of internnal  Low peer capita navigability income  Geographic siize  Geographic remoteness  Weak transportationn or communicatioon links

5

Thee most disstinctive feeature of the t CAGE E frameworrk is that it encoomes the t bilateraal attributess of countrry pairs ass well as tthe unillateral attrributes of individuall countriess. Most of o the othher fram meworks that have beeen proposedd for thinking about thhe differencces acrooss countriees (or locatiions) focus on just uniilateral attributes; that is, theyy assume thhat countriess can be asssessed one by b one againnst a comm mon set of yardsticcks. Note that t this chharacterizatiion appliess not only to carddinal indicces such as the World W Econ nomic Foruum’s Globbal Com mpetitiveness Index or Transparency Intternational’ss Corruption Percceptions Inddex but also to ordinaal ranking schemes succh as Michaael Portter’s “diamoond” framework for diiagnosing th he (relative)) internationnal com mpetitivenesss of diffeerent counttries as home basess in speciffic induustries. Butt indexicalitty of this sort s is restriictive sincee it can’t deeal withh ideas suchh as “The U.S. U is closeer to Canadaa than it is to t Indonesiaa.” Morre generallly, indexiccality is incapable of capturring bilaterral diffferences of the t sort neccessary to ennvision cou untries as exxisting in (aand evenn occupyinng) space inn relation too each otheer, that is, as a nodes inn a 6 netw work insteadd of as an array along a common yardstick. y Havving drawn that distincction betweeen unilateraal and bilateeral influencces, it iss useful to add a that theey can be fitted f togeth her into the same overrall struucture. Speccifically, unnilateral meeasures of isolation i (oor integratioon) captturing counntry-specificc attributes that t generallly decreasee (or increasse) a country’s innvolvement in cross-bborder econ nomic activvities can be treaated as a common c coomponent of that co ountry’s disstances along variious dimenssions from all a other couuntries. Forr example, really r isolatted counntries (chharacterizedd by uniique, ingrrown cultuures, clossed ministrative policies, phhysical rem moteness, orr extremelyy high or loow incoomes) can be b thought of as beingg relatively distant from m everywheere elsee. That saidd, one needs to add biilateral indicators to suuch unilaterral conceptions too capture thhe idea thaat a compaany’s homee base affeccts whiich countriees are close and which ones o are farrther away. Thee other poinnt worthy off even moree emphasis is that diffe ferent types of distance matterr to differennt extents in i different industries. For instancce, sincce geographhic distancee affects thhe costs off transportaation, it is of partticular impoortance to companies c d dealing in heavy h or buulky produccts.

6

Culttural distannce, on thhe other hand, h shapes consum mers’ produuct prefferences andd should bee a crucial consideratio c on for a connsumer gooods or media m comppany—but is much leess importan nt for a cem ment or steeel busiiness. Exhibbit 2-4 provvides a sum mmary of th he characterristics that aare likeely to makee an industrry particularrly sensitive to a partiicular kind of distance. Exhibiit 2-4 The CAGE C Fram mework at th he Industry Level L

Culttural Disttance Culturral differences matterr the most when::  Products have highh lingguistic content (TV V programs)  Products matter to culttural or nationall idenntity (foods)  Product features varyy in of sizee (cars) or stanndards (eleectrical equuipment)  Products carry couuntry-specific quaality assoociations (winnes)

isstrative Distancee Governmentt involvementt is high in industries thhat are:  Producerss of staple goods (eleectricity)  Producerss of other “entitlemeents” (drugs)  Large empployers (farming)  Large supppliers to governmeent (mass transportaation)  National champions c (aerospacee)  Vital to naational security (telecomm munications)  Exploiterss of natural resources (oil, mining)  Subject too high sunk costs (infrrastructure)

Geographiic Distance Geography plaays a more importan nt role when:  Products hav ve a low value-to-weiight or bulk ratio (ceement)  Products aree fragile or perishablee (glass, fruit) vision  Local superv and operatio onal requirementss are high (servicees)

Economic Distance Econnomic differencces makke the biggest impaact when:  Naature of demandd vaaries with incom me (caars)  Ecconomics of staandardization oor scale are limited (ceement)  Laabor and other facctor cost diffferences are salient (garmentss)  Diistribution or buusiness systems aree different (innsurance)  Coompanies need to bee responsive andd aggile (home apppliances)

E Distance Framework F Appplications off the CAGE Thee CAGE fraamework, onnce it is takken down to o the industrry level, lennds itsellf to a veryy broad arraay of appliccations. Lett’s focus heere on four of the most imporrtant ones. Makking Differeences Visible

7

Onee applicatioon of the CAGE distance fram mework is to t make kkey diffferences vissible. Whilee this appliccation may seem too obvious o to be worrth belaboriing, most notable n inteernational business b debbacles can be tracced back too a failuree to apprecciate a key y type of cross-counttry diffference or distance. d Fuurthermore, in a very diverse d worrld, manageers cannnot simplyy fall back on personnal experien nce to enssure adequaate senssitivity to differences. Checklists C o the sort embedded of e inn exhibits 22-3 and 2-4 can help evenn experienced peoplee avoid errrors due to forggetfulness annd cognitivve overload in a compleex environm ment. Undderstanding the Liabilitty of Foreiggnness A second s appplication off the CAG GE framew work is to pinpoint tthe diffferences accross counntries thatt might handicap multinationnal com mpanies rellative to local comppetitors—thee so-calledd liability of foreeignness—oor more gennerally affect their relaative positioons. This ccan be a useful exeercise for both b multinaationals and d their locall competitors. When there aree substantiaal liabilities of foreignn ness, multinnationals oftten lookk to acquire or set upp t venttures with local l firms to overcom me thesse barriers. Asssessing Natuural Ownerss and Compparing Foreiign Competiitors Eveen if multinnationals caan be confiddent that th hey are goiing to prevvail overr local com mpetitors in a particularr market, th he CAGE frramework ccan be used u at a finner level of resolution to t shed ligh ht on the rellative position of multinationa m als from diffferent counntries. For example, e CA AGE analyssis can help explaain why Spaanish firmss do well in n many induustries acrooss Latiin Americca, but allso why success in n Mexico has provved com mparatively easier for U.S. firm ms. 7 Again, such anallysis is moost valuuable whenn conductedd at the inddustry level and is inddicative rathher thann decisive. Thus, T particcularly goodd or bad glo obal strategiies can mattter morre than “nattural ownersship” advanntages. mparing Maarkets and Discounting D Com by Distancce Thee CAGE fraamework caan also be used to com mpare markkets from tthe persspective of a particularr company. One method to conducct quantitatiive anallysis of thiss type is to discount (sppecifically, divide) raw w measures of 8

marrket size orr potential with meassures of disstance, broadly defineed. While such discounting d involves numerous n approximati a ions, making som me adjustmeents of markket potentiaal for distan nce is a betteer idea, givven how w much distance d m matters, thaan refraining from making aany adjuustments at all. Some companies c d formally do y use methods of this soort in deciding d to enter or exiit markets (as ( describeed in the firsst case in thhis secttion, on Groolsch). Connclusion Thee CAGE fraamework heelps identify fy the most important cross-counttry diffferences andd their impllications foor strategy. However, understandi u ng diffferences is not a suffiicient basis for setting g global strrategy. Think backk to the AD DDING valuue scorecardd from the previous seection and aask youurself how each e type off difference or distance affects the six levers ffor valuue addition and subtracction. Is it a challengee that must be ted for and addresssed? Or doees it offer an a opportunity to improove econom mic proffitability? The T next thhree sectionns help add dress these questions by introoducing thrree types off strategies for creatin ng and claim ming value in the presence of o cross-borrder differennces: adaptation, agggregation, aand arbiitrage.

1

Pankajj Ghemawatt And Jordan n I. Siegel, “Cases “ on Reedefining Global Strategy” , (Harv vard Business Review Preess, 2011):59 9-69 2

For a more extendeed treatment of this mateerial, see Pannkaj Ghemaw wat, “Disstance Still Matters: M The Hard H Reality of Global Ex xpansion,” Harvard Ha Business Reviiew, Septembeer 2001. This topic is also addressed at substantially s g greater lengthh in chappter 2 of Pankkaj Ghemawatt, Redefining Global Strateegy (Harvard Business B Schoool Press, 2007), andd chapter 3 off Pankaj Ghem mawat, World d 3.0: Globall Prosperity aand How w to Achieve It I (Harvard Buusiness Revieew Press, 2011). For a colllection of maaps that highlight distaance effects, see s www.ghem mawat.com. 3

For furtther discussionn of the ways in which CAG GE differencees can encouraage ratheer than discouurage cross-boorder activity, see the discussion of arbitrrage in sectionn 5 and the t referencess cited therein. 4

For an original disccussion of cultural distance and how itt affects foreiign direcct investmennt, see Jorddan Siegel, Amir Lich ht, and Shaalom Schwarrtz,

9

“Egaalitarianism, Cultural Disttance, and FDI: A New Approach,” working papper, Harvvard Business School, Bostoon, October 2008. 2 5

Susan E. Feinberg g, “The Expansion and Location Paatterns of U U.S. Mulltinationals,”” unpublished working paper, p Rutgerrs University y, 2005. 6

For a more m extendedd discussion of o indexicality y in a broadeer social sciennce context, see Andrrew Abbott, Chaos of Dissciplines (Chiicago: Univerrsity of Chicaago Press, 2001). 7

Subram manian Rangaan and Aldem mir Drummon nd, “Explaininng Outcomes in Com mpetition amoong Foreign Multinationaals in a Foccal Host Maarket,” Strateggic Mannagement Jourrnal 25, no. 3:: 285–293.

10