‘The Hypermodern MNC—A Heterarchy? Gunnar Hedlund Commenting upon an early B76 (or, in an alterative notation, PQKn2-Kn3 as black), Aron Nimzowitsch advertised his move as one of “hypermodern daring.” The exaggeration contained in the expression served two purposes. It helped selling Nimzowitseh’s pathbreaking ‘books on chess strategy. It focused attention on the novelty of his ideas and thus inspired attempts to refat his conclusions. ‘The present paper is restricted to the second goal. The term “hyper: ‘modern MC is meant to convey the suspicion that some crucial a= pects of developments of and in multinational corporations (MNCS) fannot be grasped by notions in the merely “modem” schools of thought. Even more than in the case ofthe grandmaster’s rallying calls for the “hypermodern school,” "my system” ete, the departure from supposedly conventional views is bound to be exaggerated. However, some polarization of issues is desirable in order to arive at greater conceptual clarity. Inaddition it seems that concepts and theories older than present variations on the theme of “global strategy," and some- times not used by protagonists of the said theme, can usefully be ap- plied. ‘The other word in th title—heterarchy—is no less problematical. ft was used ina recent study by the Stanford Research Insite deserve a shift of perspective in a wide range of eienees. (See also Ogilvy, 1977.) ‘A key idea Is that of reality being organized non-hierarchicaly. A special case Is holographic coding where entize systems are represented and, a5 It were, “lenown” at each component of the system. (As in & hologram, each port contains information sufficient to reproduce the ‘whole original image, albeit somewhat blutred.) ‘The concept of heterarchy does not seem to have been used sue atall, in discussing MNCS. Nor has it inspited more than passing alu sions in organization studles in general. Sjostrand (1985) uses the con- ‘cept in contrast to hierarchy, but he does not give any definition, nor does the notion figure much in his discussion. The holographic para- ddigm is encountered more frequently (Mitroft, 193; EI Sewy, 1988). The iscussion in Faucheux and Laurent (1980) about integrating others Horan Recee Manspnent, Spring 95, Vo.25, Number LP. 935 (Th by Toba Wey Sos oe (CCE Oi. 010m 2750120‘oles and vantage point, and about “interaliing the envieonment in decisionmaking in a more indirect say touches on many’ ofthe issues brought up below concerning heterarchy. So do contributions on zt referential systems, such as Varela (1975). Also Laurent (1978) discusses the concept of hierarchy ‘As the previous discussion indicates, there ae no strict definitions to hold on fo. As always, consultation of the Oxford English Dictionary {ives food for thought. The only direct reference to “heterarchy” gives the meaning “the rule of an alien” This is exactly what heterarchy inthe present use ofthe term is not. Tis use, as well as SRI's and Opilvy’s, builds on putting “homo,” rather than “auto,” as the opposite of “hetero.” It's not easy to ative ata simple definition, for example by contradistinction in relation to “hierarchy.” The apparent supeciorty of this latter term as to darity of meaning derives mostly from the dulling effects of old habits, We have become so accustomed to the concept of Firarchy that we forget exactly what i is that we want to conceive of ‘with i, For example, the etymological meaning of “ruling through the sacred” or “rule of the episcopate” is rather alien to transaction cost analyses of markets and hierarchies.” Certainly it would not make uch sense to define heterarchy as nonhierarchy, meaning ruling through the profane. The abstruseness of novelty in “heterarchy” thus partly derives from the abstruseness of convention in “hierarchy.” Many authors hail hierarchy asthe dominant or even only stable form of organization of human as wellas other systems. Space limitations do not permit thorough discussion of this strain of thought, One contibution will, however, be beifly mentioned. Koestler (1978, p. 290) puts the rgoment for hierarchy very strongly: All complex structures and processes ofa relatively sa be character dispaly hierarchic organization, and thc Spples regardless whether we are consiering ina ‘ate systems, ving eganisms, socal organzaions, oF patterns of behavioue, Koestler himself mentions the suspicion that the hierarchic model's universal applicability may originate in the model being logically empty, or merely 2 reflection of the way in which a perceiver approaches an ‘object or situation. He rejects these possibilities but does not discuss them at length. The pervasiveness of hierarchical thinking models is treated by Ogilvy (1977) and Bouvier (1984), who support the view that hierarchy toa large extent isin the eye of the beholder. Some other comments should also be made in relation to Koestler’s arguments ‘A key idea with Koestler is the existence of parts which are self= regulating, relatively autonomous, and which exhibit properties not de- dducible from lower unite. At the rome time, they are parte of larger ‘wholes. Koestler calls these units “holon” Thisis not inconsistent with the hierarchy notion. However, he also discusses more complex net- 10.1 Humes Resource Management, Spring 1985works, where “vertical” and “horizontal” connections interusine. We are warned, however, not to forget the primacy ofthe vertical, hierarch cal, dimension (ibid. p. 298): Inia fhe sight of the lige ofthe entsined branches Ina forest made us forget tat tbe branches oxgnate in separate tees, The tees ae vereal statues, The ‘meeting poins of branches om neighbouring Wexs {orm hentrontl networks sever! levels Without the ttces there could be no entwining. and no network ‘Without the networks, each tee woul be isolated, and {hate would be no intepstion of functions. Arborasion ‘nd etiultion seem to be complementary pancples in the architecture of organisms, Ia symbolic universes of ‘scours afooition ie elected nthe verea eno lation (definton) of concepts, etzuation in their“ Sonat connotations in sasocalive network. ‘The last paragraph indicates a tendency to define whatever cannot be captured in hierarchical order as only a looser kind of “association.” ‘Another example is the discussion of “abstract” and “spolight” men ‘ary. The lattervery vivid, almost photographie images resembling to- tal recall of past situations—seems not to fit the hypothesis of hierarchic storing of information. Koestler (Ibid, p. 48 ff, 296-297) “solves” this problem by assuming that there is something he calls “emotional rele- vvance,” which leads to lack of schematization in hierarchies. He also regards spotlight memory as more “primitive” (bi, p. 53), and poss bly phylogenetically older than abstractive memory. ‘Koestler even regards the supposedly older principles of storing and managing information in the human brain as hacmaful. He suggest t0 initiate (id, p. 103} Not an amputation, but a process of harmanization Which assigns each level of te min, Irom rocerl i Piles to abatact thought, ls appropriate place nthe Rerarchy. This imple relforcag the new. brain's power of veto against that type of emotive behavior End that ype oniy=which canoot be reconcled wi reason, such a the lind” pasion of the group-ning, ‘The “process of harmonization” cannot be achieved by education: “It can be done only by ‘tempering’ with human nature itself to correct its tendemic schizopsychologieal disposition” (p. 104). He expects “the Iab- oratories to succeed in producing an immunizing substance conferring rental stability” (p. 108). To the present author, this line of reasoning seems like tying to ‘expurge and, if possible, ezadicate thought patterns which do not fit 2 hierarchies! model. The tree metaphor also hides an important aspect of ‘much social organization. Any given unit may be a member of several systems, which each may be conceived of asa hierarchy. Ina tree, every Hedlund: The Hypermedern MNC 111branch obviously primarily “belongs” to ane wee. Hossever, iit equally clear to what “arborizing structure” a US. citizen, born by Jewish pac. tents, working fora French company in Spain belongs? Koestler quotes Hyden (1961), who suggests thatthe same neuron may be a member of several functional “clubs,” ae suppor forthe distinction between arbo- izing and retiewating structures. However, it seems that this ould rather be taken as an example of non-hierarchy. (Below, a heterarchy 1 endowed withthe atiibutes of having many centers of different Kinds, This seems to fit the neuronal cubs better) ‘Also on the empirical level, some of Kestler’s examples of hierarchic organization may be questioned. Later research on memory, and on the lente functioning of the brain, does not appear to fit the hierarchic ‘model (McCulloch, 1965; Pribram, 1971). Organizations, in their actual functioning, are far less hierarchic than their organization charts would imply. Action systems do not always work a 2 hierarchy of stalegies transloemed into action programs and simple final acts (Alison, 1971; Mintzberg, 1978). The “holon property” of Koester is fully consistent with the het- cezarchy mode! outlined below. The supposed inevitability of hierarchy, however, seems tobe a Procrustean bed in describing life in real organi zations. Therefore, rather than continuing the conceptual discussion, 1 willy to sketch some developments in MNCs, ilstrating the need for 2 concept covering these developments. Thereafier, a tentative de~ mination of the concept of » hetearchicl MNC willbe provided. ¢* "A ploneer in reviewing the development of different kinds of MNCs was Howard V. Perlmutter (1965). His original scheme of an evolution of, ora least a distinction between, ethnocentie, polycentric, and geo ‘centric MNCs has hardly been improved upon. Therefor, this ia nat ral starting point to discuss tendencies of change in the nature of ‘mullinatonal business. ETHNOCENTRISM Almost all now existing firms have started on a national basis and only ‘graduelly developed international ties. Foreign business was inital ‘nly marginal, more so for companies from large nations than for those ‘with small “home markets." Intemationalization was often based on monopolistic advantages which could be exploited by internalizing, (ransactons within the firm. (See Hymer, 1976; and Dunning, 1977 for early and representative statements of theories of foreign direct in vestment based on “firn-specfic advantages.”) These advantages, in terms of, for example, proprietary technology were exploited in a slow, gradual process, by moving concentrically to markets further away fom the home country, and by investing in increasingly committing forms. From sales outlets in the neighboring county, the firm cautiously 12 J Huma Resource Menagemen, Spring 1986‘roved towards manufacturing plants on alien continents. (Vernon, 1966; Stopford and Wells, 1972; and Johanson & Vahlne, 1977 ae good examples of gradual leaching theories of foreign direct ivestment Ethnocentric companies are managed by home country people. and with time there is alot of rotation between HQ and subsidiaries. The control style will vary in accordance with practice inthe parent company and parent country. For example, Swedish firms transferred 2 celiance fon nonnative control (Etzioni, 1961) to this international operations US. companies used relatively more ofcalcultive and coercive control, with less autonomy forthe subsidiaries (Hedlund, B60, 1984; Hedlund snd Aman, 1983), The role of a foreign subsidiary in such 2 company is operational rather than stategic. Stcategis ate derived from the prospects of ex- fending the geographieal scope of firm-spectic advantages and formu- lated at the center. The subsidiaries implement, but there is also an entrepreneurial element to early stages of internaionalization, which is lost as the firm gots used to going to foreign lands. “The environment of the MNC-—as far as aspects of intemationallza tion are concerned—could be characterized as Type 1 (placid random) or ‘Type 2 (plac clustered) in the Emery and Trist (1968) classification. ‘That is, opportunities and problems are ether randomly distributed (ype I) oF clustered (Type 2), but competitive relations are not primary 25 in Type 3, nor is drastic environmental turbulence the maln issue (Type 4). The absence of strategy in a Type 1 situnton is particularly apparent inthe subsidiary. The best strategy is todo as well as possible fon a purely local and perhaps also short-term bass. Interdependencies between the center andthe subsidiaries in the en- terprse are primarily sequentisl (Thompson, 1967). Products, know- hhow, and money for investment are sent from the center to the periph- xy. There is vertical division of labor, so that activites up-stream the value-added chain are conducted atthe center and down-steam opera- ons at the periphery. (This, of course, does not hold for raw materials based MNCs.) The novelty and uncertainty of foreign operstions favor icrarchy rather than market or federation solutions (Willamson, 1975; Daems, 1980). That is, subsidiaries ere conteclod rather tightly, either through orders or shared outlooks; for example, by transferring peopie between units in the firm (Edstrom & Galbraith, 1977, POLYCENTRISM ‘As time goes by, foreign business may become dominant rather than ‘marginal, the subsidiaries get more activities and become more sell sufficient, management becomes more host-country oriented and con- sisting of host-country nationals. The MNC becomes an assemblage of semi-independent units. There is [ess rotation of personnel, and in a Hedhands The Hypenmoders MNC. 213,‘way, the polycentsic MNC isles trans-or international thom the ethie~ Centic version, Indeed, the teem multinational fits better forthe polyse tele fem than forthe other archetypes ‘The competitive strengths move from proprietary technology to 3 ‘essto distrbution channels, rand name, international experience, and finance. Econamies of scale and scope become important, New invest- ments are sought worldwide, almost as in a portfolio placement stra ‘egy. As long asthe firm stays in the original ne of business, the sizeof the host market will bean important criterion for the decision on where to invest. The ethnocentric stage of confinement to near and familar bodes looses in significance. (See Vernon, 1979; and Hedlund and Kverneland, 1984 for a discussion and some empirical support. Some of the results reported have to do with changes in the environment of MNCs, lowing “Instant poljentiom.” athe tha wih fem specc levelopments.) Subsidiaries are operationally independent and increasingly forced t0 take stratogic decisions with respect to their operations in theie market (Disturbed reactive” local environment forthe subsidiary). HQ consol ‘moves towards caleulative, based on financial results rather than on influencing the substance of decisions. The extreme is reached when :he parent company acts only as a holding company, buying and selling assets internationally, with no view fo anything but the financial out comes of its dispositions, Interdependence between subsidiaries and center is pooled, Financia resources and some specialist competence are kept at the center, ‘whereas product and technology flows ae les pronounced. Activities are duplicated internationally, 0 that manufacturing, for example, is ‘undertaken in most subsidiazies. “The tendency in terms of control mode is to move toward Jooser coupling betoreen units and from the hierarchy (inthis case also some- ‘what fo the etymological sence) of ethnocentrism to market ealutons. ‘Transfer pricing based on market prices rather than internal costs, fra ‘dom to choose extemal supplies, rewards and punishment in monetary terms, and elaborate bons payment systems accompany greater tuzi- lover fates of personnel and organizational units being sold olf and bought. Internationaization is more and more conducted though 2c- 4quisitions rather than greenfield ventures. The tendency to market so- Iutions could be interpreted in terms of increased routinization of inser national transactions, with consequent reduction of uncertainty. Also, the idiosyncracy of assets (technology, people, etc) is not as pro- nounced a in the initial tages. According to Williamson (1975), sis should lead to markets rather than hierarchies. 41 Hone Resource Management. Spring 1986GEOCENTRISM Perlmutte's (1965) original clasifiation defines the various “cen: teisms” primanly according to the attitides of management. Above such aspects have been linked to strategie situations, stages in the lite cycle of internationaization,” types of interdependence between parts of he firm, types of environment facing the company, etc. It Decomes even more necessary to discuss these other 2specs when de- sorbing a geocentric firm. One reason for this is that "geocentric strate- pies” may be accompanied by ethnocentric atitudes, Indeed, the sift from poly: to geacentri strategic fous is often perceived by host coun- tay management as a shift back to HO and home country attitudes ‘vriters on “global strategy” mostly mention interdependence be- toveen units in the firm asa distinguishing characteristic. (See, for eam- ple, Porter, 1980 and 1984.) The actions of a subsidiary in country A influence prospects for the subsidiary in county 8, perhaps because they face the same competitor, who has o divide his resources between the two markets. Thus, competition isnot confined within each national market, but system-wide, The MNC exploits systems advantages, sub- sidiaries, and country-specific advantages being considered the parts of the system. Thus, at the extreme, subsidiaries specialize and operate slobally in limited felds. The MNC inthis way intenlzes the eplotation Af (country) comparative advantages. This is very important fom the point (of view of theories of international trade and investment, Ricardo never fought that the same agent would produce both wine and cloth. Ax ssamptione that the MN is a reflection of firm-specific advantages 2 Ia Hymer must confront avery peculiar type of advantage, that is multina- tionality in self To say that the MNC. exists because it exploits the ‘advantages of being an MNC is tautological, so the convenient theoret- ‘al starting point in monopolistic advantages dissolves when applied to the geocentric fms. As Vernon (1979) himself has noted, the product ite cycle theory of international trade anid investment become less useful as the international spread of companies is extended. Global competition, where a firm faces the same rivals on most mar- ets, means that gradual interationalzaton strategies pose problen Hiedhund and Kvereland (1984) and Lundgren and Hedlund (1983) show how market entry by Swedish firms into apan and South-East ‘Asia respectively is faster and more committing than theories of gradual- jam woul lead one to expect. Firms do not follow a neat sequence from agent over sales subsidiary and some local manvfacturing t large-scale local production, Instead, the pattern Is one of jumping steps in the chain and building up positions very rapidly. Intemadonal stategy is increasingly driven by considerations of vals, and somesimes actual of potential cooperatos’, behavior, zather than by the exploitation of FSAs (Grmspecfc advantages) asin the ethnocenvic fir, or by the attrac Lveness of markets one By one as In the polycentric firm, Oligopolistic Healund: The Hypemodem SINC (15reaction, interme of initaling competitors’ moves (Knickerbocker, 1973) 35 well 25 awiding competitors and building up mal hostage postions, becomes carmmon, Perlmutter saw the use of third country nationals (TCNS) in manage: iment as a sign of geacentity. Other aspects concerning the manage ment process are reliance on global profitability goals and increased rotation of personel. Probably a shit back to les calculatve and mare rormative and coercive contrl is required inorder for global strategies to work. The subsidiaries have to implement strategies formulated ac: cording toa global logic, they have tobe able to act quicly in response to competitive conditions, they must be encouraged to look ata wider picture. Most writings on global strategy give the subsidaces a less Independent role than that implied ina polycentric MNC. A recent {zation of authority to HQ often follows, and is recommended to follow, a globalization of competition. (See, for example, Channon and Jalland, 1979; Hedlund and Aman, 1983.) Often, global divisions structured around products, technologies, or customer types are created Io coor rate activites in specific competitive niches. The basiness envizonment can be characterized as a global disturbed reactive one in the terme of Emery and Trist (1965) ‘Interdependence between parts ofthe firm moves from the polycen tric pooling of resources a the center to sequential and reciprocal. Prod: ucts, know-how, money, and people flow in increasingly complex pat ters, and not asin the ethnocentze frm from one core to the periphery. (See also Bartlett, 1964 and his discussion of the “integrated network model” of an MNC. His other concepts of “centralized hub and “de ‘centralized federation” can be compared with ethnocentrism and poly: centrism, respectively.) Particularly reciprocal intecdependence is ex- pected to lead to internalization ina hierarchy (f Thompson, 1967, 50 the trend towards markets in the polyeentic MNC is reversed. Also reversed isthe tendency to duplicate acavities in various subsidiaries. The discussion so far fs summarized in Table {. Obviously, it ignores many complexities tnd gives a very simple pleture ofthe range of poss- bites. For example, there certainly exst geocentric MNCs which build primarily upon sequental interdependences between center and pe- fiphery. A clear example would be mining companies in highly concen- trated industries. Nevertheless, one can better understand the character of most geocentrcity and globality of competition i everal strains f-— In this context—often forgotten theoretical heritage are applied: + Thompson's (1967) dassifstion of various types of interdepen- ence, end his hypotheses of mechanisms ofistegration related to those types. + Teansaction cost theorist” (Cosse, 1957; Willamson, 1975) notions. of altemative governance modes and the determinants of effective Solutions to the governance problem. 16-1 Ham Resoce Mangement, Spring 1986Sons woes ana ny efor foot “nsdeueus unos 90H ‘jen oot wawoiduy 4 sievog ——_Aoene ow ones pue ubuNMD9y anpsgne yo jor 2.0 ‘Apisgs o huoveiny wowvensue Kempisans cc pur ynvanbas yorwineds —__Aimpagne-saiues poyooy —_Luipisgns-sausoyenvanbog eon Sapog vopezeto “wowstap feuoreana an SNePLNON, ‘uompadent = ‘oui Sob aaisuenpe xyovoneunyous ‘uonadtiog _ adoos pur ajas "at er ypedeaany nodes ywewusod weunwopyeuesgns eutien ) waanaoAog aS “Taiger Ledland: The Hypernodem MNC 17+ Classification of mechanisms of socal integration such as Etzion’'s (98. + Typologies of orgeniantonal environments (Emery and Test, 195), and hypotheses about behavioral implications of those envieoa ments ‘Three entries in Table I have not been adequately foreshadowed inthe discussion above. Organization structure has been added, using the results of Stopford and Wells (1972), Franko (1976) and others linking strategy to the structure ofthe international organization. “Hierarchy” sone governance made in geocentric firms wil be explained inthe nest section. I believe that pure hierarchy wil be detrimental to many global strategies, and that there nevertheless isa strong possibility that this is ‘what wil happen in many firms. Emery and Test’s type 4 environ ment-—the turbulent one—-has been introduced asa Ukely development forboth the entre geocentrie MC and its subsidiaries. This wilco be discussed in the next section, STRAINS ON THE GEOCENTRIC MNC A radical view concerning geocentrism and global is that we are witnessing the disappearance of the Intemational dimension of bysi- ‘ness. For commercial and practical purposes, the nations do nat exist, and the relevant business arena becomes something Uke a big unified “home market.” Business action as well s¢ concepts to deccrbe Gms and the situatons they face will be similar to the ease of 2 company Working in one national market Hosrever, there are a nismber of dificlties facing the MNC, which ‘wants to act asf the world was one big market and competitive arena, 10 De adapted to in a scaledoup version of “ordinary,” national strategy. ‘In spite of proclaimed increased homogenization of demand (Vernon, 1979), theze are sil strong differences between ‘nations and regions. Protecionism is furthermore on the increase rather than the other way around. The loyalty of many employees i stil, primarily with their home county. (See Doz, 1979, and Doz and Prahalad, 1980.) + The need for cooperation, in joint ventures or in other forms, charac teristic of many branches of industry, makes undateral strategy mak- ing problematical. + Cultural differences in management style makes one at Fast ques- tion the viability of uniform, worKdorde control systems and other ‘management practices, + Beonomizing by sharing resources between diferent lines of busi 16 1 Human Resource Manegonet, Spring 1986ness, with diferent customers and competitors, mitigates against totally subduing the local country dimension in organization and strategy. Size itzelf may be severe problem in coordinating operations glob- ally in the same way as one would coordinate national business, The Complexity and variability of environmental drcumstances. com ‘pound the sizeof problems. Response times may be too long to keep Up with changes In markets, The cognitive limitations of integrating Information ae very real. Particularly atthe strategic level, advances {ninformation technology may not be sufficient, although no doubt be of value. ‘The supply of managers abe to cerry out ambitious global strategies already is 8 bottleneck today for most firms. If strategy making is recentralized to the HQ “brain,” it will become even more dificult to fil positions, since this requires more transfer of personnel With the development of specialization between subsidiaries, these will become 20 large and important that it will be detrimental to assign narrow strategic roles to them. For example, a research center in India serving the whole network of an MNC would probably, with time, develop ideas and products which do not Ait the prevail lng strategies ofthe group, but which could wellbe a bass fora new line of business. [t would be wasteful not to entertain a capacity to ttlize the creativity and entrepreneurship of people at all nodes of, the network. Besides, those people would probably resign if they id not get such opportunites, Finally, centrally guided global strategies for given products simed at beating given competitors, looking at the world as one market, nay lead to neglect of opportunities to exploit edting diferences between nations, Ifthe global thrust is combined with areemphasis ‘on HQ and home-country guidance, the company may return t0 etivocentism, only being able to exploit ideas originating at home. ‘Advantage seeking and advantage development will not be main, concerns, but only the exploitation of existing advantages. In the Jong ran, such frm may become sterile. The results in Davidson & Hlaspeslagh (1982) indicate that the global product division as an. organizational solution may indeed entall such risks. Most ofthe points iustrate the danger of seving geocentric just as the sealing up of the national corporation, thereby geting 1d of the international dimension of business and reestablishing central strategic direction from center, which i atthe apex of one, big global hierarchy. Even if this characterization of global strategy isa charcature and may ‘seem to be set up as a straw-man, I believe that both academic discus sion and—but less s0—practice in lsrge MNCs are affected by outlooks, Hedlund: Te Hypermedern MNC_/ 19philosophies, statagies, and management practices similar to the ones eseribed Perlmutter’ orginal conception of geocentricity was not as restricted asthe “mononational” version sketched above. He sketched situation ‘where subsidiaries were "parts ofa whole whose focus is on worldwide objectives as well as local objectives, each part making its unique contri- bution with its unique competence.” These lines do not clearly denote the attrbutes of geocentricity, but their connotative meaning is very ‘ich, believe one can usefully single out some of those connotations and specify a special case of geocentricty, an option which is sil not fully developed in actually but lowards which many companies proba bly wil, and should, move. This isthe hypermoderm MNC, and one of is distinguishing marks sits hetearchical nature. ‘THE HETERARCHICAL MNC The heterarchical MIC differs from the standard geocentric one both in terms of strategy and i terns of struc. Stately, the main dividing lines beeen exploting competiive advantages derived from {home country bazeon the one hand, and actively seeking advantages xiginating in the global spread of the frm on the other In its most extreme form, this would mean that one could not assign the company to any pardcular industry. Any opportunity which activates the poten {hl inherent in broad geographical coverage would be a candidate for inclusion in the company's zeperoie of products and services. Oby- ously, no MIC would ike to got this extreme. Specialization bens {ppl also to information search, and in many contests existing borers U entry ito a global indatry would be prohibitively reste. However, the concept of explting the advantage of mullinationalty as such also apples within a rather lied Bel of business information cn competition, tchrologcal wend, developments in tated eds, “spect of national environments, etc lead to opporaniies not easly ‘Rented by purely lca firms, or by polycentric MNCs, or even by ctimocentriclly tainted global MNCs. The diference between the lator fed the heterarchy is most pronounced when it comes tothe structure of the enterprise and the processes of managing it Indeed, X may be that the idea of strocare determing stategy (Se Hall and Saas, 1960) isa fundamental one for the heterarchicl MNC. Rather than dertiying properties of the industy in which it competes and then adapting its Structure othe demands thus established, the hypermodern MINC fst detnes its structural properties and then looks for strategic options fol Towing from these properties, in actual If, ofcourse, every candidate forheterachy wil have come from « story ina given set of industies, regions, et. The MNCs most Ht to face the indeterminacy of strategi- aay relatively open vistas ae probably those described by Bartlet 20.1 Humen Resource Management Spring 1985‘lott ntagatinyNations!Resposinenas Gis Consumar—Tecammricaons Fores fr Global Eoordnaton Invearton Packages Goods te Fore for Naona!RasponehnesiOiteretaion Figure 1 (Source: Bartlet, 1984). (1984) 26 “transnational.” Such firms are active in industries where itis Important both to achieve global integration and local differentation, for ‘example, adaptation to host government demands (see Figure 1). ‘Thus, strategie imperatives of dual focus force some MNCs to adopt structural solutions and managernent practice in consonance with these task demands. These adaptations then constitute an opportunity for sometimes much wider and diverse strategic options, or atleast more intensive ulizations of the global spread of the company. In order to achieve this, further development of structural tats are desirable. Their trchetypical expressions wil be ennumerated below. 1. Fut, the heterarchical MNC has many confers. One could speak of polyarchica rather than monarchical MNC, were It not forthe lack of integration implied inthe former term. The main dea is thatthe foundations of compedtive advantage no longer reside in any one country, but in many. New ideas and products may come up in ‘many different countries and later be exploited on a global sale. A [Beographically diffused. pattern of expertise is bult up, corre sponding to unique abilities in each node of the network- These ablities may be a reflection of dissimilarities between countries a5, in “demand theories” of international trade (Burenstam-Linder, 1961) or simply expressions of spatially distributed talents for tech- nological development within the firm. Atthe extreme, each “sub- sidiary” is at the same time a center for and perhaps 2 global coordinator of activities within one fed (such as fr one product), and a more peripheral agent for local distribution in another In diversified firms itis obviously easier to find examples of such Intemational specialization. For example, the Swedish company Hedlund: The Hypemodean WNC 1 21Alas Copco has the headquarters fois Ai Power division located in Belgium, whereas the other divisions are headquattered in Swe- den, Esselte (office equipment) has put the center ofits largest division in London, However, even within one product division there is scope for mult-centeredness. The dangers ofthe global product division un- dlluted by geographical considerations fave been discussed by Davidson and Haspeslagh (1982) Relations are restricted to those between one center and units in the periphery. Relations within, the supposed periphery are not exploited, and information over- load on the center and lack of motivation in subsidiaries create save problems. Hedlund (1960) documents the strategic alienation Df subsidiary managers in Swedish MNCs, and gives some sugges tions of how to involve the subsidiaries more in strategy formula- tion. In this case, the mother—daughter structure, rather than global product divisions, is the organizational background. This ‘seems to support the views of Bartlet (198, 1984) thatthe kmpor- tance of the formal orgenizaton structure is easily exaggerated. Simmonds (1985), reviewing the teature and discussing various ‘ways to “achieve the geocentric ideal,” concludes that other man agement systems, such as the planning, accounting, and reporting systems, ate important obstaes. 2. A key idea in the conception ofa eterarchical MN is that subi lary managers are also given a strategic role, nat andy jor User “own’ company, but for the MC ase whole. The notions of "headquarters," “center,” “home county,” and “corporate level” dissolve and are not synonymous. Corporate evel strategy has to be implemented and formulated in a geographically sattered network. 3, Heterarchy implies dfrent Kinds of centers. There is not only 3 set of globel divisions and subdivisions, or only a set of geographi- ‘al divisions further spit up in national and regional subunits. A heterarchy consists of mix of organizing principles. There may be an R&D center in Holland with global responsibilities for coocdi- hating product development, product division headquarters in Germany responsible for the main product, marketing center for ‘Asian Singapore, anda center for dealing with global purchases in Londan. The multidimensionality of orgenizing principles (fanc- lions, products, geography, customer type, ec) reflects the need to coordinate activities along each and all of those dimensions. In a hhetevarchy, tere i nat ane overeting dimension superordinale lo the ‘All this may seem an unduly complicated way of describing a matrix organization. Recognizing the probability of muddled think Ing and expression, I stl believe that there are important ditfer- 221 Hanan Resource Management, Spring 1986ences between what I call «heterarchy and a mate structure, for ‘eample ‘ay have a'aser" coratng te, Ferthermore, nots unin oo torent Senate decree SESE ait ere merece Rote eee mney ceyecctt Seba yaukeree rene Seaneneea ‘Admittedly, most writers on matrix organization would stress that the formal matrix is only & small part ofthe "matex way of life.” ‘They stress the need fr shift of focus overtime, Heil in apply Ing dual reporting relationships, and care in not overloading the arbiter role of top management. I. A further characteristic ofa heterarchy concerns the degree of cou: pling between organizational units, In many cases, there would be 8 deintegration of relationships. A subsidiary wil be given in ‘rensed freedom to purchase components extemally, and to sell to ‘customers outside the corporation, Joint ventures and other types of cooperation with other fizns will be more prevalent than in the tightly controlled globsl frm. 1 such freedom is not given, there willbe litle chance of really profiting from the opportunities pro vided by global reach. (Of course, negative effects on the rest of the operations of the MNC must be considered algo, An important subset of global op- Portunities consist exactly in interaicing flows of information, products, and money within the bounds of the fim, saving on Fransaction costs associated with market solions. Perhaps i i moze appropriate to speak of fleciliy inthe section of governance ‘ode, rather than reintegration. A hetcrarchical MINC will have no problem in entering joint ventures, externalizing production and hhandling internal transactions according to erms-length principles in one business ates, and insisting on unitary contel, intern- alization, and governance by management fat in another context. Full exploitation of global scanning and information processing capabiltes will lead toa range of opporturites, some best handled i hierarchie, others rather suited to marketlike governance. The Hedland: The Hypermodern MNC_ 1 23heteratchical MNC could be seen asa meta-instiuton, which con inwously creates new institutional arrangements, in the light of expertise concerning what works best for each specific purpose. ‘This, assuming that chojes ae rational, speeds up the process of Institutional evolution in comparison with a “Darwinian” process of selection. A “Lamarckian” development, where experience is accumulated, experiments fully exploited and memory over "en- erations” kept intact, guides the choice of govemance forms. ‘Thus, it may be correct to speak of deintgration for firms com- ing from 2 tradition of strict global contol from one center. For firms with a polycentric past, the tendency may be the opposite. Common to boths thatthe range of types of relationships between ‘units fn the company, as well a in clation to outside actors, will Increase, 5. Another atibute of heterarchy is that integration i ached primar- ily hough normative conta, and only secondarily through calcula- tive and coerdverbureavcratic regulations. "Corporate culture, “management ethos” (Garett, 1984), “management style,” "cul tural contol” Gaeger and Baliga 1985), ec. become critical. This is the only way fo assure coordination inthe diverse, extended, and. fluctuating environment and activities of a heterarchical MNC. Pure buretucratic control breaks down because of cognitive over load and motivational problems. Puce calculative conti, which may serve well inthe polycentric MNC, will ot establish the mu tual trust, the ability to “enrifice” the focal forthe global and the short tem forthe long term, and the shared code of commmunica- tion necessary for rapid action ina coordinated fashion. Thus, and paradoxically 60, a hetezarchy may contain more of hierarchy, in its ‘etymological sense, than does bureaucratic hierarchy itself. This i, the reason for “hierarchy” in Table 6, It was mentioned above that the hologram is @ special type of Iheterarchy, where information about the wes contained in ack par ‘Tis is a critical characteristic of the heterarchical MNC. Every member of the company will in the extreme case be aware of all Aspects ofthe firm's operations. Obviously, this is only a theoret- cal ideal. However, widely shared awarencss of ceisal goals and Strategies, and of critical interdependencies between units in the firm, is not an impossibility. Some remarkable corporate tum- arounds recently are no doubt putly attributable to initiatives in {his direction. In Sweden, the cases of SAS and ASEA are particu lady striking. The distribution of information in every part of a hologram is possible because of ase technology. One could say thatthe corpo Tate ethor is the analogue of the laser light. By sharing certain conceptions about the fim, and certain ways of acing in relation to 241 Human Resource Management. Spring 1985‘other members of the fem, it becomes possible to rapidly share Information, interpret the meaning of events in and outside the xganization in siilac way, and see opportunities for local action Inthe interest of the global good. The Iaser beam elec of corporate culture's the unifying element of a heterachical organization. Itis ‘rudial to support the formation of sucha culture, since the risks of anarchy ate otherwise very great Pessimism regarding the efficiency and integrity of non-hierar- chical and unified control is less warranted on empirical than on theoretical” grounds. Ogilvy (1977) discusses how ambitions to “organize societies and polities hierarchically are influenced by and influence modes of thought and even the stractre ofthe personal- ity. The fact thats phenomenon may be hard to grasp and expla in terms familar tothe grasper does not mean, however that the phenomenon does not exist. Ogtivy quotes McCulloch (1965) on hneterarchical patterns of preference in neural netorks. Geis in pte ined finding gansie tencies, scaly demonstate consetency of higher or der than had Been dreamed of tn our plosophy. An ‘ongerizm posseseed of the nervous systems new Fors—itsuisertly endowed to be unpredictable fom ny theory founded on scale of values has» het rch of values, and ls tus intetvly too ch to SUbmut ta summum bonum.(MeCullocs, 196, p. 3) Yet, although we cannot explain how we are able to wal, for example, we stilldo.Iis worth quoting McCullogh again, for some clues about the properties of heterarchies and possible analogies {a0 mote but also no less) with a discussion on organization and entra in human insitions. imphases added By the present ‘uthor.) ‘The deal of te (the bain) neurons and thee specie onnectons need not concern ws here In general You ‘may think ofa» computer any prt of hich Come $lghals rom many part of the Peay snd fom other pare of the brain spinal cord, Itt oly one cel rp En the pats fom input to ouput, uti can st he fiers fn allo ts inputs ond can ene the behavior of the Programmed acuvity the half centers, ana he rll gee» euatitute for depth by entra oredr eto. susie, given ts bral of este of he spl pane and the werd mpngent span to sce wheter he rl ore eng Sing Be Ing eating sleeping, ete. must do it wih ellsecond otiponent action snd conduction veloes of wally than 100 meters per second, and d it ara tine, Sy, in-athird of second That thas worked #0 sell ‘hrooghout evolution, without Kael evolving. pont Hedland: The Hypermedern MNC 1 25is siructre the atu solution ofthe organization Sy spproprie behav. We know much expertly of the Behavior of the components, bot stil have nd theory worthy ofthe name to explain 8 Secu actions. (tecCatoch, 1985, p. 397) 7. The foregoing discussion on the heterarchical nature ofthe ner- vous system leads to another, perhaps hairaising, analogy. The ‘metaphor underlying much thought on corporate strategy is one of the firm consisting ofa brain and 2 body. The strategy makers in the center ae the brain, and the implementors in the periphery are the body. Thinking and acting take place at different locations. Books ike The bra ofthe fr (Beer, 1972) testify tothe forcefulness ofthe metaphor. However, the dangers of separating thinking and acting too much in an organization have been well lustrated by the decline and often fall of formal long-range planning depart :ments in companies. One way of deserbing the heterarcical MNC. Js to say that thinking i nat only restriled to one exclusive center, ba {05 on inthe whole enterprise. Thus, at appropriate metaphor for discussing a heterarchieal frm would be & "fm ase brain” model rather than a "brain of the firm” model ‘A weakness of the metaphor is that it may lead one to see the firm as only a cognitive entry. However, the core of the dea is that ‘not only does thinking take place also in the periphery, but # goes logelher with and directly informs action. Coalitions with other companies and alo other types of actor are quent in the heterarchical MNC. Exploiting global reach will often ‘ean to serve as a catalyst bringing together elements with syner- istic potential, perhaps firms from diferent continents previously rot known to one another. It may be of interest to note that Emery land Trist (1968) saw as two pelmary ways of coping with turbat lence: (0) The eatin of common vals, binding people and organizations together and enabling them to expond quicdy to enviconmental change: This corespords to the emergence of corporate cure 354 binding element eis above. “ (ty Cooperation benweeneterogenous ements ater than competion bettcen homogenous elements (asin "ype 3) asthe primary et pation of top nodes, “The latter point inches things such as joint ventures and coopera~ on between firms and governments. A heterarchical MNC will share and pool its power with other actors in order to benefit mayi- rmolly From its global capabilites. This does not mean that twill do 420 in all elds of business. Again, itis the multitude of governance forms and degrees of internalization which characterizes a het- cerarchy, 26 1 Hunn Resource Management, Sprig 19869. Fially, and returning to the strategic ambitions rather than the structural properties of heterarchy, this type of MNC would be fit to attack the most difficult global problems of today. This may seem naive and even ridiculous to managers busy surviving pro- ducing and selling a narrow line of producis or service. However, assuming 2 type of company that Sees the expiotation of global. ity as such a ie main source of strength, it dots not soem that fr- fetched to consider raical prolem-orimiaton as guiding statepy formulation. (Rather than starting from existing physical or human resources, or from competitive positions in narzow fields of bust ress) HUMAN RESOURCE MANAGEMENT IN A HETERARCHICAL MNC No fll discussion of human resource management inthe context of 2 heterarchy wil be attempted here. Instead, afew important points will be brought up, without pretense of exhaustive treatment. 1 Concerning organiaton structure, many models wil be simulta- neously used in a heterarchical MNC. The fleility and multi rmensionality of the structure defy easy categoszation. Change of the formal organization will not give rise heterarchy. Subter changes in management processes are required. However, the for- ‘mal organization may stop a movement towards heterarchy. ‘One consequence of breaking down (up?) a large hierarchy is ‘that is a longer posse to promt people many by giving ther jos “higher up.” Movement between centers will be more common, and movement from periphery towards center in the sare writ will be less common, Also, the need to build up the "nervous systern” ofthe heterarchy is of importance here, 2s isthe need to use per~ sonal competence wherever it pays off best, 2, The core of a heterarchical enterprise will consist of pole with « long experience init. A frm invests considerably in the employee, and vice versa, The latter is part of the commanication system of the firm, and the history ofthe human system inthe company may belts most strategic resource, Thisis often said and may sound lke 8 platude. However, itis less 20 in an organization which builds ite strategy on advantage seeking and using its global coverage, rather than on advantage exploitation on the bass of known and Sable assets ‘This communication network is not easily imitable by other firms. Much les so can a small part of it be used by others, in limited sense, the employee is of value to competitors as a source of information, since he has lot of, aso ofa stategic nature. After Hedlund: The Hypemadem MNC / 27having interrogated and “emptied” the unfaithful soul, hossevee, ‘sof hitie use, Thus, from the point of view ofthe employee, the ‘osyncracy of his relation with the frm is very grat. This is also true the other way around, since i takes a long time to find and tain a replacement. However, the “hologram quality” makes the fire mare robust than in a hierarchy. Many employees will share the same information and be able to support or replace each other. This does not mean tha the firm can easly Sie the employee, since such behavior would undermine the mutual trust necessary to en courage investment inthe long-term future of the MN. Idiosyn- cratic assets should lead to internalization, according to Williamson (1925), so one can expect more encompassing and long-term contracts ‘with employees. Another possiblity is partdpation inthe owner- ship ofthe company. ne can exaggerate the need for permanence, however. There is considerable fox inthe activities of the heterarchical MN, and this requires Gexibility also concerning personnel Joint ventures and other forms of cooperation, sometimes on a project bass, by. ‘definition mean that new members continuously enter and leave the system. Perhaps one can speak ofa dual caver system. just as fone speaks of dua labor markets in some countries. There wil bea Timed but stil nzmerous core of almost ife-tme employees, and a ‘much larger number of people with more fleeting association with the frm. Tn the debate on the Japanese labor system. it is often pointed out thatthe core enjoying life-time employment and other rarvels ofthe Japanese employment system is rather small, What has surprised most analysts is thatthe duality of the labor market has not disappeared with modernization. Perhaps the solution ‘with an Integrated core surrounded by quastintegrated satellites (which themselves might constiate cores in other systems) is ‘good combination of stability and flexibility? ‘The core provides the memory and the information infrastruc: ture necessaty to grasp opportunites on a global scale. The looser links to the outside help against rgidfication of response by estab- lishing channels forthe communication of new ideas. In this con text, the balance between young and al members ofthe organization 1 probably enitical. (See the discussion by Lorenz, 1971, on the bal- dance between processes of acquiring, reining, and dismantling ‘cultural knowledge, and the importance of age in this respect.) Company demography needs to be planned more systematically than when the fim fa system of oles which canbe easily commis nicated and learnt. Not allowing steady recruitment of “new blood," or dismantling of knowledge by early retirement, ae taps inthis area. However, much more research i needed on company. demography before any strong statements can be made, 281 Haman Resource Managenent Spring 19863, In order for internalization of norms t take place, # at of rattan of personnel and international travel and postings ace necessary. The tendeney to man purportedly global firms with home country man- agers—and more so than in polycentric fms—will not work in a heterarchial MNC. Advances in information technology may help the formation ofthe nervous system ofthe frm, bu this will not be ‘enough for building strong internal cultures. ‘The problems an the practical level of international tansfers of people are well known. The solutions are less well known, apart from obvious hints such as paying well, giving spouses jobs, and being aware of reentry problems. Perhaps recruitment of cand dates for the core should be very selective, with 2 stong emphasis ‘on willingness ta travel and change function in the company. ‘Sending people abroad very early is probsbly a g004 ides, possibly ‘even before they have formed families, (Would the best be to have the new employee swear to chastity and keep unmarried, like in the very successfl international operations of some eclestiatical organizations such as the Jesuit Order?) 4. A rowch broader range of people inthe frm must develop capacity for Strategie thinking and action. This implies open communication of strategies and plane, decentralization of strategic asks, using task forces on stategic issues actively, and providing ealy opportuni tes for development of “top management capabilities” also for “subsidary” employees, (The words “subsidiary” and even "man- ager” sound a Bit funny in the context of a hetearchial MNC. ‘There is less obvious subordination, and the clear distinction be- ‘beeen managing and operational functions is es relevant than in a clear hierarchy. Heterarehy may mean the beginning ofthe deline of the professional manager as/a species within the organizational 00.) ‘Control systems which measure performance along many di- mensions (products, regions, short and long term, ot.) are neces- sary, This i also almost » plattude, but in actual practice many. ‘companies who claim they do this really do not. Even if the sys tems are there, they are not used for more than very limited pur poses. (Hedlund and Zander, 1985, report onthe economic contol Systems of Swedish MNCs. See also Business International, 1982.) 5. Reward and punishment systems are critical. Carters of bad news must ‘not be kale. Kobrin (1984) shows how MNCs neglect to use the expertise of host country managers for the assessment of political Fisk. The long term must not be sacrificed. Perhaps. bonus should. ‘oe given on the bass of profitability i the unit where the employee served five years ago? Particularly at very high levels, an effective career strategy is to turn “star” and “question mark” jobs into Hedlund The Hypermedern MNC_/ 23“cash cove jobs, and leave just before they stat looking the “dogs.” Top managers ae rather adept at taking credit or other people's work and avoiding erticam for thee own, and temporal Ecension of the review period may counter the tendency to misuse Sach talents ‘Similarly, the global aspects may be supported by rewarding people for lobal rather than local prot, or whatever the objec I The dicalty hes im matching rexponsibiity with authority Probably, a heterrchical MN has refrain from mechanical com pensation formulae to large extent. It snot possible to construct Perfect equations forthe datibuton of bonuses, for example, pat fealty when circumstances change often and drastically. Paying «employees pardy on the basis ofthe performance ofthe entice frm is one possibility. SAB-Nife, » mall Swedish MINC, as large ‘onus element in its system or paying subsidiary managers. Halt ‘ofthe bonus is based an the performance af the entre company (90 percent of ales are abroad), and half depends onthe results ofthe Individual subsidiary. Shareholding by employees may be a very poten instrument to stimulate action in the interest ofthe toa company, and to encour: age normative integration. Would not be beter fo have the ex- poyees in, aay, the Indian subsidiary own shares inthe parent company then the Indian goverment forcing the subsidy to join ventures with loa partes, some more sleping than other? [Not that the former wold stop the late, but inthe long, u tis ‘would constitate an important change in the identity of the MINC. ‘Glee mertlty mayb ried ar down in the organisation. Start tog up now and dosing down old acvites is helped by under standing ofthe reasons for change. Technlogia! developments axe turning many workers into technicians, and to technicians heeding 10 know a ot about costomers. Global compestion t changing the rales of the game forall employees. Some examples of action in Swedish firme in the direction indicated is given By + Volvo's gigantic program for improving substantially the tech seal know-how at all levels inthe company. : + SAS! focus on foreign competion in mobilizing for turn- around, and the very public nature of is corporate strategy + SKE Steets program to import ste! technology from Japan and teach its employees about competitive facts and, of course, technical matters) by sending workers on assignments with the Japanese licensor. (An informal race on productivity ensued, and the Swedes caught up with ther teachers in Japan.) 6, [tis hard to tell what the personality type best suited to eterarchy is. 30 1 Human Resource Management, Spring 1985Ogilvy (1977) argues that a sort of “polytheistic" personality, and acceptance of such Protean prospects, go together with more de- ‘centralized organizations and societies. Speculting on this, one could argue that people from polytheistic or atheistic cultures ‘would be most comfortable in such situations. Old Greeks, Vi ‘ings, Hindus, and Japanese would do well. Chestians (particu- Iney protestants), Moslems, Jews, communists, and people af- fected by "scientism” would do worse. The respresentatives of, ‘western culture included inthe former list are all dead, so many fiems would do well to look around a bit for new managers, ‘Such speculation aside, it seems clear that a heterarchical MNC would require many employees with the following qualities: + Aptitude for searching fr and combining elements in new way’. Probably good knowledge in several fields of ecience and tech: ‘nology is one precondition for ths. ‘Skil in communicating ion nl rapidly turing them into action Very good command of several languages and knowledge of and sympathy for several cultures. (Steiner, 1975, argues that bilie~ sguslsm is qualitatively different irom monolingualism, in that it gives a "stereo quality” to perception and interpretation, See also Maruyama, 1978.) Honesty and personal integrity. These old fashioned ideals are critical for heterarchy not to turn inte chaes. Willingness to take risks and to experiment. Advantage seeking is much more risky than advantage exploitaon. The orgenina~ Hon must support such learning from failures. The heterarchi- ‘al MC wold mean an attempt to innovate from the basis of 2 large firm, working across national boundaries at very early Stages in the innovation process. Asin all entvepreneutial sc tivity a igh failure rate is to be expected. Therelore, in prac- tice, every company needs also a part which makes money in mote stable and predictable ways, The theoretical alternative of a perfect extemal capital market can be ignored for the mo- ‘ment, because of gency cost considerations Jensen and Meck- ling, i976), frwould be very difficult for anonymous sharehold- fs; as well 25 for lenders, to assess ex ante, monitor constantly, and even evaluate ex post, the activities of s gent inely and entirely heterarchical MNC. This also means that the financial stength of a wel-run, fairy large traditional MNC makes it the nly realistic candidate for heterarchy on an inter- national level. Neither small ims on thelr own or together through market relationships nor governments, for various reasons but in both cases having to do with ageney cost prob- lems, are likely to succeed Hound: The Hypermoder MNC 131+ "Paths the company and is etiviies. Eathusiosm fer the company need not go to the etymological limits ofthe word but genuine appreciation of the company and ite eultucu ie ‘valuable, Perhaps this means thatthe widely admired sceptical thinking type of person is of Fess interest than the person able ‘o form strong attachmen's? 7. Management development activitis (in the more restricted sense) should be sen asa primary inerument to build «corporate culture or ‘ulate and disseminate strategies, and establish links inthe com ‘munication system of the firm. lis role for acquiring skills and for learning facts and methods is perhaps only subsidiary CONCLUSION The heterarchical MNC is a so far loosely defined concept Itcovers & particular brand of geocentric company, which differs significantly From 2 version that is likely to develop more rapidly in the immeciate future, ‘The importance in bringing up and further outlining the demands of and possbilies Inherent in heterarchy lies in the ask of the purele tlobal company regresting into a si2ed-up model ofthe lage rational firm, An ethnocentric backlash isa clear possibly, but mostly unneces sarily so. Therefore, firms should actively explore the dangers ofrecer tralizaion, ever fand when such moves are desirable, and find ways of ‘compensating for those dangers. Only the broad outlines of response ‘an be drawn without much experimentation and accumulation af expe- Fence, The MNC is » crucial arena for such institutional innovation, since i is uniquely powered to address some of the most urgent prob- Tems ofa global sale. ‘Where sould one look for signs of heterarchy?In terms of industries, probable fields are those characterized by the use of many different technologies, high but not maximum global homogeneity of demand, fast rate of technical and market change, ron-triviat scale economies (out not necessarily in manufechiring), and absence of strong local barriers to ‘entry. This means that information technology and biotechnology come to mind, which should make the reader (and weiter) saspicous, since this seems too obvious (and boring). However, also the automobile in- dusty, building and construction, and many services Rt many, but not all of the criteria in terme of geographical and comporate origins, heteratchical MSCS are mare likely 0 evolve from less than gigantic ims, and from contexts ‘with history of rather autonomous and entrepreneurial subsidiaries. This may give European firms an advantage over U.S. ones. Ina larger picture, MNCs from newly modernizing nations may stand an even better chance, Chandler and Daems (1980) show how instiutional iner: 32 J Human Rewure Management Spring 1986tia and established forms of corporate organization in Europe delayed the formation ofthe large, managerialy un firm as compared to in the USA. Olsen (1982) has discussed how the same mechanisms may make whole nations rise and fall. The heterachical prospect may seem 100 remote, or even silly, to peaple in auccessfel hierarchies likely to enjoy stil some time of harvesting the fruits of investments in 2 powerfal ‘organization for the maximum utilization ofenstng physical assets and know-how. It may seem less remote for people who have lite alterna- tive but to directly exploit the amazing global icity of capital, technol- ‘gy, and people to develop new products, markets, and competences. Gunnar Hedlund i fit with the Inttue of international Busines atthe Stockholm School of Economics in Stockholm, Sweten. REFERENCES Alison, GT. Esene of decision Expl the Cutan mise rs. Lite Brown, 197 Basle, CA. “Matnaonal Sucre Change: Brluon Vers Rergeis oats Obed (EL): The Magma son Sandy a {Sip atonal Cert, Comer Patkohing Company Lined, ioe Bale CA, “Organization and Cont of Global Ene: nunc, Chartres on Canletner Dat ved Bone Sool ose es 8 Te aioe frm: Aen Lane, 972 ‘Bouvier, PL Subject andthe concep of Nec: the dominant pr igm and the provaling work syste Prange fom th erm Cine of Sy or Count Sons Rest New Gore june Eh, enti 3 Ens on ded Tempio Nw ee Wey 1 tines internal Catpraton: Aes age sary perfomance ‘stchowiar fs Chas Be and Bash VBE w Cisne, A b,j nd Seem, He) Mangal Herrcs, Cartes, TMA tnd Londen Engle Harvard Grier Pee, 0. Channon: abd and Me. Mattos Pong, London: Benge, a Comte RH, “The Note ofthe Fin.” Econo N.S 97 ¢, 386-105, Seema “The dee ofthe modern instal eer a new perspective” in'h:B: Chander snd Baer (Edn), Monga Hiern’ Cage, MAand Lando Harvard Unversity Es, 10. Dunston, Woe and Pape, F.“Skaing» Clo Pootoct Organi Won”, Hard Bases Ree yup 2 a Smet one! an Manat Steg Manze: oe ie and Terman Eipnet ger 199 Don and Patalads © low MNEs Cope wth Hos! Goverment Tntrvesion” Hard Bess Rens, Marh-Apa 000 Dunning} TH. “rade, lator score aay an the mations ener, A search fran eco approch” In. Ohl, PO. Fessloom, Sea) Ws Ede) The hire losin of ene ety Lv don: rian 197. dso, fy and Galo JR "Tranter of Managua a Cordon and crt Sey intend Grprasone te See Ga ‘ery Tune 19, Bae, Hedlund: The Hypermodern MNC 33lSawy, 0. A. "From separation to hologeaphicenfolding.” Paper presente to TIN meeting, Boston May 1988 Smery, FE, and Tet EL, "The Cava Tete of Organizational Enviar tment” Hoan Relation, 196318, p. 21-32 Buion A. A Compuatve Als of Comper Organs. New Youk: Free rss, 1961 Faucheax, C. and Laurent, A. “Sigliance ofthe epistemological revolution “gra management scence.” in Prem fom the worse on he ptemaogy of manage, IASM, 150, Franko, LG. The Euopon Mulinatinate, Greenwich, CT: Greylock Press, 1976, Hal, Band Saas, Me "Soategy Follows Sractre.” Stele Management four- ‘ui, 198012 Hedlund, G. “The Roe of Foreign Sobsidates in Suatgic Decision Making in Svedth Malinao Conpations” Sie Mangement oral, 1905 9, ‘Hedland, 6. "Organization n-Berween-The Evolution ofthe Mother-Doughter Sastre a Managing Foreign Sur in Swed MN.” fore of 1984 cre Bees a lund, and Kverveland, A. being inthe experi of Sais ‘Bras Stodholm:Insbue of International Busines, Stockholm School of Managing Relainsips wih Facign Subir rea nt Cott Sdn be. Sook: Serges Meanie 198, Hedlund, ., and Zander, U. Fault of Gs nd Fell up of Peformencefor Tag Susie Seta Me Working Paper 8514 tat often tinal Busines, Stockholm Schoo! of Economic 5, Hyden, He "Control of the Mind.” In Farber SM, and Won, RH. 1, "ein, Control of the Mind, New Yor 18h, ymer The nmin Opin of Natal Fe: A Stay of Diet erie “iments MILT. Press, (Originaly publahed se decor dsacraton 1860), 1978. Ineger, A. Me and Baliga, B. R. “Conta Systems and Strategic Adaptation: ‘Wesson fom the Japanese Experience” Sater Managemen ura, 1985, «a Jensen, M,C. and Mecting, W. Hi “Theory ofthe fm: managerial behavio, agency costs and ownersip sre Joona of Frc Eawot, Ocaber reso ° ” Jobenson, [and Vahine, [6 “The Interationalzaton Proces ofthe Firas—A Modal of Knowledge Development snd Increasing Foreign Market Comait- ment” Journal o Management Studi 1977, 120). Kncberbockar, FT, Olgopoltie Rein and Malinao Energie, Boston Harvard Business School, 1973, Kobrin S.J Managing pola rik xsexment, Unversity of California Press, "ser Koestler, A anus—a summing wp. New York Random Howse, 1978 ucen A: “Manager swornacy” Acdemy of Manageme! Reins, 1978, “ant, Lots K. “Knowledge, bl and no InP, Wess (GH ay opened ste ahr and proc. tre Pubtshing Company, 197 Lindgreh Sand Hednd sou eg Spsasen, Stole: sate of international Business, Stockolm Scosl of Bonomi, 183. Maruyama, M. The epsccnalogis! evolution” Future, fone 1978, 20-242, MeCulloch, W. EmbodOnnt of mn Cambri, BA, 1963 SA. 1 Human Reswurce Management, Spring 1986Mintberg. HL “Patterns in Stntegy Formtion.”, Maeutgmont See, 1978 ‘Soa trol Ly Our Ot Pietro the Nord Dow' Work Angmar. Research Paper, ‘University of Souten Callers, 1383, vy. } Maltidmersiona ma Osiord University Press, 1977 Rn, Hesslbom, P- Ou, and Wiskman | (Ede) The ermal alc ow of sane actly London: Macmlan, 1977 (Olen, Me The sear delineation lew Haven and London: Yale Uni Pres, 1902 Pedmutter, H. V._“Uenterpse interationale—tois conceptions". Rewe ‘ronomique Scale, 1955, 3 Poder, Mc E: Compelive Srey, New York: Free Pres 1980 Paster, M. E. Campelion i geal industri comet fromacrk, Boston: ‘scvard Binet Schoo 1884 Pram, K. Languner ofthe bran, Englewood Cie, N} Prentice-Hall, 1971. ‘Simmonds, K. "Global Strategy” Achuving the Geocetrie Wea" Intrational Making Ree, Sing 1S Sjostrand, 5. Sembaluorgoation, Stockholm: Doxa, 1985. Sener, G. Ae Babel, Oxord Universy Pree, 1973 Stopford J. Mey and Well L. Ts Managing the Maltnation! Estep, New fork Baic Docks, 1972 Thompson]. D. Organon in Actin. Newe York: MeGraw-H, 1967 Varela Fy “A Coeds (or SeltRaference" Int. Gon. Sytem, 1975, 2, 5-2 Yemen, & “nieratonsl Invertment and intemationl Trade inthe Product (Gye, Quertrly Journal of Exon, May 196 Vemnon, R."The Breduct Cycle Hypoess tn 2 New Interasonal Environ Tent." Oued Bulletin of Eonar ond Static, 195, 41. Willamson,“O. E. Marks and Birr mls and AMinst Impl. "Neve York: Free Press, 1975 ty Hedlund: The Hypemodem MNC / 35