International business ch13_view

Hill: International Business Competing in the Global Marketplace, Fourth Edition

Part 5 The Strategy and Structure of International Business

13. The Organization of International Business

© The McGraw−Hill Companies, 2002

Chap

ter 1

3

TheOrganizationofInternationalBusiness





Unilever is one of the world’s oldest multina-tional corporations with extensive product offerings in thefood, detergent, and personal care businesses. It gener-ates annual revenues in excess of $50 billion and sellsmore than 1,000 branded products in virtually every coun-try. Detergents, which account for about 25 percent ofcorporate revenues, include well-known names such asOmo, which is sold in over 50 countries. Personal careproducts, which account for about 15 percent of sales, in-clude Calvin Klein Cosmetics, Pepsodent toothpastebrands, Faberge hair care products, and Vaseline skin lo-tions. Food products account for the remaining 60 percentof sales and include strong offerings in margarine (whereUnilever’s market share in most countries exceeds 70 per-cent), tea, ice cream, frozen foods, and bakery products.

Historically, Unilever was organized on a decentralizedbasis. Subsidiary companies in each major national marketwere responsible for the production, marketing, sales, anddistribution of products in that market. In Europe the com-pany had 17 subsidiaries in the early 1990s, each focusedon a different national market. Each was a profit center andeach was held accountable for its own performance. Thisdecentralization was viewed as a source of strength. Thestructure allowed local managers to match product offer-ings and marketing strategy to local tastes and prefer-ences and to alter sales and distribution strategies to fitthe prevailing retail systems. To drive the localization,Unilever recruited local managers to run local organiza-tions; the U.S. subsidiary (Lever Brothers) was run byAmericans, the Indian subsidiary by Indians, and so on.

To knit together the decentralized organization, Unileverworked to build a common organizational culture amongits managers. For years, the company recruited peoplewith similar backgrounds, values, and interests. Thestated preference was for individuals with high levels of“sociability” who embrace the company’s values, whichemphasize cooperation and consensus building amongmanagers. It is said that the company has been so suc-cessful at this that Unilever executives recognize one an-other at airports even when they have never met before.Unilever’s senior management believes this corps of like-

minded people is the reason its employees work so welltogether, despite their national diversity.

Unilever has also worked hard to periodically bring thesemanagers together. Annual conferences on company strat-egy and executive education sessions at Unilever’s man-agement training center outside of London help establishconnections between managers. The idea is to build an in-formal network of equals who know one another well andusually continue to meet and exchange experiences.Unilever also moves its young managers frequently, acrossborders, products, and division. This policy starts Unileverrelationships early as well as increases know-how.

By the mid-1990s, the decentralized structure was in-creasingly out of step with a rapidly changing competitiveenvironment. Unilever’s global competitors, which includethe Swiss firm Nestlé and Procter & Gamble from theUnited States, had been more successful than Unilever onseveral fronts—building global brands, reducing cost struc-ture by consolidating manufacturing operations at a fewchoice locations, and executing simultaneous productlaunches in several national markets. Unilever’s decentral-ized structure worked against efforts to build global or re-gional brands. It also meant lots of duplication, particularlyin manufacturing, a lack of scale economies, and a highcost structure. Unilever also found that it was falling be-hind rivals in the race to bring new products to market. InEurope, for example, while Nestlé and Procter & Gamblemoved toward pan-European product launches, it couldtake Unilever four to five years to “persuade” its 17 Euro-pean operations to adopt a new product.

Unilever began to change all this in the mid-1990s. In1996, it introduced a new structure based on regionalbusiness groups. Within each business group are a num-ber of divisions, each focusing on a specific category ofproducts. Thus, within the European Business Group is adivision focusing on detergents, another on ice cream andfrozen foods, and so on. These groups and divisions havebeen given the responsibility for coordinating the activi-ties of national subsidiaries within their region to drivedown operating costs and speed up the process of devel-oping and introducing new products.

Org

an

iza

tion

al C

ha

ng

e a

t Un

ilev

er Introduction

Organizational ArchitectureOrganizational Structure

Vertical Differentiation:Centralization andDecentralizationHorizontal Differentiation: TheDesign of StructureIntegrating Mechanisms

Control Systems and IncentivesTypes of Control SystemsIncentive SystemsControl Systems, Incentives,and Strategy in theInternational Business

ProcessesOrganizational Culture

How Is Organizational CultureCreated and Maintained?Organizational Culture andPerformance in theInternational Business

Synthesis: Strategy andArchitecture

Multidomestic FirmsInternational FirmsGlobal FirmsTransnational FirmsEnvironment, Strategy,Architecture, andPerformance

Organizational ChangeOrganizational InertiaImplementing OrganizationalChange

Chapter SummaryCritical Discussion QuestionsClosing Case: OrganizationalChange at Royal Dutch/Shell





“Lever Europe” was established to consolidate the company’s detergent opera-tions. The 17 European companies now report directly to Lever Europe. Using its new-found organizational clout, Lever Europe consolidates the production of detergents inEurope in a few key locations to reduce costs and speed up new product introduction.Implicit in this new approach is a bargain: the 17 companies relinquished autonomy intheir traditional markets in exchange for opportunities to help develop and execute aunified pan-European strategy. The number of European plants manufacturing soaphas been cut from 10 to 2, and some new products will be manufactured at only onesite. Product sizing and packaging are harmonized to cut purchasing costs and to ac-commodate unified pan-European advertising. By taking these steps, Unilever esti-mates it has saved as much as $400 million a year in its European detergent operations.

Lever Europe is also attempting to speed development of new products and to syn-chronize the launch of new products throughout Europe. Nonetheless, history still im-poses constraints. While Procter & Gamble’s leading laundry detergent carries thesame brand name across Europe, Unilever sells its product under a variety of names.The company has no plans to change this. Having spent 100 years building these brandnames, it believes it would be foolish to scrap them in the interest of pan-Europeanstandardization.

Source: Guy de Jonquieres, “Unilever Adopts a Clean Sheet Approach,” Financial Times, October 21,1991, p. 13; C. A. Bartlett and S. Ghoshal, Managing across Borders (Boston: Harvard Business SchoolPress, 1989) H. Connon, “Unilever’s Got the Nineties Licked,” The Guardian, May 24, 1998, p. 5;“Unilever: A Networked Organization,” Harvard Business Review, November–December 1996, p. 138;and C. Christensen, and J. Zobel, “Unilever’s Butter Beater: Innovation for Global Diversity,” HarvardBusiness School Case # 9-698-017, March 1998.

This chapter identifies the organizational architecture that international businessesuse to manage and direct their global operations. By organizational architecture wemean the totality of a firm’s organization, including formal organization structure, con-trol systems and incentives, processes, organizational culture, and people. The core ar-gument outlined in this chapter is that superior enterprise profitability requires threeconditions to be fulfilled. First, the different elements of a firm’s organizational archi-tecture must be internally consistent. For example, the control and incentive systemsused in the firm must be consistent with the structure of the enterprise. Second, theorganizational architecture must match or fit the strategy of the firm—strategy and ar-chitecture must be consistent. For example, if a firm is pursuing a global strategy but ithas the wrong kind of architecture in place, it is unlikely that it will be able to executethat strategy effectively and poor performance may result. Third, the strategy and ar-chitecture of the firm must not only be consistent with each other, but they also mustbe consistent with competitive conditions prevailing in the firm’s markets—strategy,architecture, and competitive environment must all be consistent. For example, a firm pur-suing a multidomestic strategy might have the right kind of organizational architec-ture in place. However, if it competes in markets where cost pressures are intense anddemands for local responsiveness are low, it will still have inferior performance becausea global strategy is more appropriate in such an environment.

The opening case on Unilever touches on some of the important issues here. His-torically Unilever has competed in markets where local responsiveness has been veryimportant. The production and marketing of food, detergent, and personal care prod-ucts have traditionally been tailored to the tastes and preferences of consumers in dif-ferent nations. Unilever satisfied this environmental demand for local responsivenessby pursuing a multidomestic strategy. Its organizational architecture reflected thisstrategy. Unilever operated with a decentralized structure that delegated responsibil-ity for production, marketing, sales, and distribution decisions to autonomous nationaloperating companies. This allowed local managers to configure product offerings, andmarketing and sales activities, to the conditions prevailing in a particular nation. For

434 Part 5 The Strategy and Structure of International Business

Introduction





a long time, this fit between strategy and architecture served Unilever well, helping itto become a dominant consumer products enterprise.

However, by the early 1990s the competitive environment was changing. Tradebarriers between countries were falling, particularly in the European Union followingthe creation of a single market in 1992. This made it possible to manufacture certainitems such as detergents and margarine at favorable central locations in order to real-ize the benefits associated with location and experience curve economies. Also, newproducts in areas such as frozen foods and margarine were gaining regional or evenglobal acceptance. Unfortunately for Unilever, some of its global competitors movedmore rapidly to exploit this change in the competitive environment. Unilever founditself disadvantaged by a high cost structure (caused by the duplication of manufac-turing operations) and an inability to introduce new products in several national mar-kets at once. In other words, the competitive environment changed, but Unilever didnot change with it. By the mid-1990s, Unilever had recognized its problems andchanged both its strategy and its organizational architecture so that it better matchedthe new competitive realities. Unilever began to adopt a transnational strategic ori-entation, seeking to balance local responsiveness in marketing and sales with the cen-tralization of manufacturing and product development activities to realize scaleeconomies and execute pan-regional product launches. To implement this strategy,Unilever introduced a new organizational architecture based on regional businessgroups, each of which contained product divisions. These divisions were given the re-sponsibility for centralizing manufacturing and product development activities, whichimplied a reduction in the autonomy traditionally granted to operating subsidiaries. Toreestablish a fit between strategy, architecture, and environment, Unilever had to em-brace the difficult process of strategic and organizational change.

To explore the issues illustrated by cases such as Unilever’s, we open the currentchapter by discussing in more detail the concepts of organizational architecture and fit.Next we turn to a more detailed exploration of various components of architecture—structure, control systems and incentives, organization culture, and processes—andexplain how these components must be internally consistent. (We discuss the “peo-ple” component of architecture in Chapter 18, when we discuss human resource strat-egy in the multinational firm.) After reviewing the various components ofarchitecture, we look at the ways in which architecture can be matched to strategy andthe competitive environment to achieve high performance. The chapter closes with adiscussion of organizational change, for as the Unilever case illustrates, periodicallyfirms have to change their organization so that it matches new strategic and compet-itive realities.

As noted in the introduction, the term organizational architecture refers to the totalityof a firm’s organization, including formal organizational structure, control systems andincentives, organizational culture, processes, and people.1 Figure 13.1 illustrates thesedifferent elements. By organizational structure, we mean three things: First, the formaldivision of the organization into subunits such as product divisions, national opera-tions, and functions (most organizational charts display this aspect of structure); sec-ond, the location of decision-making responsibilities within that structure (e.g.,centralized or decentralized); and third, the establishment of integrating mechanismsto coordinate the activities of subunits including cross functional teams and or pan-regional committees.

Control systems are the metrics used to measure the performance of subunits andmake judgments about how well managers are running those subunits. For example, his-torically Unilever measured the performance of national operating subsidiary compa-nies according to profitability—profitability was the metric. Incentives are the devices

Chapter 13 The Organization of International Business 435

Organizational Architecture





used to reward appropriate managerial behavior. Incentives are very closely tied to per-formance metrics. For example, the incentives of a manager in charge of a national op-erating subsidiary might be linked to the performance of that company. Specifically, shemight receive a bonus if her subsidiary exceeds its performance targets.

Processes are the manner in which decisions are made and work is performedwithin the organization. Examples are the processes for formulating strategy, for de-ciding how to allocate resources within a firm, or for evaluating the performance ofmanagers and giving feedback. Processes are conceptually distinct from the location ofdecision-making responsibilities within an organization, although both involve deci-sions. While the CEO might have ultimate responsibility for deciding what the strat-egy of the firm should be (i.e., the decision-making responsibility is centralized), theprocess he or she uses to make that decision might include the solicitation of ideas andcriticism from lower-level managers.

Organizational culture is the norms and value systems that are shared among theemployees of an organization. Just as societies have cultures (see Chapter 3 for details),so do organizations. Organizations are societies of individuals who come together toperform collective tasks. They have their own distinctive patterns of culture and sub-culture.2 As we shall see, organizational culture can have a profound impact on how afirm performs. Finally, by people we mean not just the employees of the organization,but also the strategy used to recruit, compensate, and retain those individuals and thetype of people that they are in terms of their skills, values, and orientation (discussedin depth in Chapter 18).

As illustrated by the arrows in Figure 13.1, the various components of an organiza-tion’s architecture are not independent of each other: Each component shapes, and isshaped by, other components of architecture. An obvious example is the strategy re-garding people. This can be used proactively to hire individuals whose internal valuesare consistent with those that the firm wishes to emphasize in its organization culture.Thus, the people component of architecture can be used to reinforce (or not) the pre-vailing culture of the organization. This seems to have been the practice at Unilever,where an effort was made to hire individuals who were sociable and placed a high valueon consensus and cooperation, values that the enterprise wished to emphasize in itsown culture.

If a firm to going to maximize its profitability, it must pay close attention to achiev-ing internal consistency between the various components of its architecture. Let us lookat how structure and control systems might be inconsistent with each other. Figure13.2 shows an organizational chart for how Unilever’s European operations might be

Figure 13.1

Organization Architecture


ControlsAnd

IncentivesProcesses

Structure

People

Culture





structured (this chart is hypothetical). Note that there are several country subsidiaries,one for France, one for Germany, one for Spain, and so on, each reporting to the Eu-ropean Business Group. There are also several pan-European product divisions, one fordetergents, one for frozen food, one for margarine, and so on, again each reporting tothe European Business Group. Within this structure, responsibility for marketing,sales, and distribution decisions might be given to the country subsidiaries, while re-sponsibility for product manufacturing might be given to the product divisions. As forcontrol systems, imagine that profitability is the metric used to evaluate the perfor-mance of the country subsidiaries.

One problem with this set of arrangements is that the profitability of the countrysubsidiaries depends on manufacturing costs and new product development, and yetthe managers running the various country subsidiaries are not responsible for those im-portant functions—responsibility resides in the product divisions! Thus, if the man-agers of the product divisions do not do their job properly, production costs may riseand the profitability of the country subsidiaries might fall. In other words, the man-agers of the country subsidiaries are being evaluated according to a metric over whichthey do not have total control. If the performance of a subsidiary declines, they mayargue that this is not their fault; it was due to the inability of the managers in the pan-European product divisions to drive down manufacturing costs. Thus, there is a po-tential conflict between structure and the control systems used; they are potentiallyinconsistent.

Some inconsistency is a fact of life in organizations. Perfection in the design of or-ganization architecture is very difficult to achieve. Nevertheless, the inconsistency be-tween different components of an organization’s architecture can be minimized throughintelligent design. In the example just given, if the performance of each product divi-sion were assessed on the basis of manufacturing costs, it would give the managers ofthe product division the incentive to optimize manufacturing efficiency. The problemmight be further alleviated if the heads of both the country subsidiaries and the Euro-pean product divisions were rewarded according to the profitability of the entire Euro-pean Business Group (for example, by having their bonus pay linked to the profitabilityof the entire group). This would give the heads of the divisions a further reason to re-duce manufacturing costs, and it would create an incentive for the heads of each sub-sidiary and division to share any best practices developed in their operation withcolleagues across Europe to the betterment of the entire European Business Group.

Internal consistency is a necessity but not a sufficient condition for high perfor-mance. Consistency between architecture and the strategy of the organization is also

Figure 13.2

Fictional OrganizationalStructure at Unilever


France

MargarineFrozenFoodDetergents

Germany

Spain

European BusinessGroup





required; architecture must fit strategy. When Unilever began to emphasize cost re-duction as a major strategic goal, the firm had to change its architecture to match thisnew strategic reality. It had to move away from a structure based primarily on stand-alone operating subsidiaries in each country and toward one that looks more like thestructure depicted in Figure 13.2. Unilever had to create some entity, in this case theproduct divisions, that could reduce the duplication of manufacturing operationsacross country subsidiaries and consolidate manufacturing at a few choice locations.Such change is easier said than done. It is relatively easy for senior managers to an-nounce a radical change in strategy, but it is much harder to actually put that changeinto action. Doing so requires a change in architecture. Strategy is implemented througharchitecture, and changing architecture is much more difficult than announcing a change instrategy. We shall discuss why it is hard to change architecture later in this chapter.As we shall see, a prime reason is that organizations tend to be relative inert; they areby nature difficult to change.

Even with internal consistency and a fit between strategy and architecture, highperformance is not guaranteed. The firm must also ensure that the fusion between itsstrategy and architecture is consistent with the competitive demands of the market, ormarkets, in which the firm competes. In the 1980s Unilever had a good fit between itsstrategy and architecture—it was pursuing a multidomestic strategy. A decentralizedarchitecture composed of self-contained country subsidiaries was well suited to imple-menting this strategy. However, by the 1990s the strategy no longer made much sensedue to a change in the competitive environment. Trade barriers between nations hadfallen and more efficient global competitors were emerging. Unilever’s strategy nolonger fit the environment in which it competed, so it had to change both its strategyand architecture to match the new reality. This type of organizational challenge is notunusual; markets rarely stand still, and firms often have to adjust their strategy and ar-chitecture to match new competitive realities.

Organizational structure means three things: (1) the formal division of the organiza-tion into subunits, which we shall refer to as horizontal differentiation; (2) the loca-tion of decision-making responsibilities within that structure, which we shall refer toas vertical differentiation; and (3) the establishment of integrating mechanisms. Webegin by discussing vertical differentiation, then horizontal differentiation, and thenintegrating mechanisms.

|Vertical Differentiation: Centralization

and Decentralization

A firm’s vertical differentiation determines where in its hierarchy the decision-makingpower is concentrated.3 Are production and marketing decisions centralized in the of-fices of upper-level managers, or are they decentralized to lower-level managers?Where does the responsibility for R&D decisions lie? Are strategic and financial con-trol responsibilities pushed down to operating units, or are they concentrated in thehands of top management? And so on. There are arguments for centralization andother arguments for decentralization.

Arguments for Centralization

There are four main arguments for centralization. First, centralization can facilitate co-ordination. For example, consider a firm that has a component manufacturing opera-tion in Taiwan and an assembly operation in Mexico. The activities of these twooperations may need to be coordinated to ensure a smooth flow of products from thecomponent operation to the assembly operation. This might be achieved by central-


Organizational Structure





izing production scheduling at the firm’s head office. Second, centralization can helpensure that decisions are consistent with organizational objectives. When decisionsare decentralized to lower-level managers, those managers may make decisions at vari-ance with top management’s goals. Centralization of important decisions minimizesthe chance of this occurring.

Third, by concentrating power and authority in one individual or a managementteam, centralization can give top-level managers the means to bring about needed ma-jor organizational changes. Fourth, centralization can avoid the duplication of activi-ties that occurs when similar activities are carried on by various subunits within theorganization. For example, many international firms centralize their R&D functionsat one or two locations to ensure that R&D work is not duplicated. Production activ-ities may be centralized at key locations for the same reason.

Arguments for Decentralization

There are five main arguments for decentralization. First, top management can be-come overburdened when decision-making authority is centralized, and this can resultin poor decisions. Decentralization gives top management time to focus on critical is-sues by delegating more routine issues to lower-level managers. Second, motivationalresearch favors decentralization. Behavioral scientists have long argued that people arewilling to give more to their jobs when they have a greater degree of individual free-dom and control over their work. Third, decentralization permits greater flexibility—more rapid response to environmental changes—because decisions do not have to be“referred up the hierarchy” unless they are exceptional in nature. Fourth, decentral-ization can result in better decisions. In a decentralized structure, decisions are madecloser to the spot by individuals who (presumably) have better information than man-agers several levels up in a hierarchy. Fifth, decentralization can increase control. De-centralization can be used to establish relatively autonomous, self-contained subunitswithin an organization. Subunit managers can then be held accountable for subunitperformance. The more responsibility subunit managers have for decisions that impactsubunit performance, the fewer alibis they have for poor performance.

Strategy and Centralization in an International Business

The choice between centralization and decentralization is not absolute. Frequently itmakes sense to centralize some decisions and to decentralize others, depending on thetype of decision and the firm’s strategy. Decisions regarding overall firm strategy, ma-jor financial expenditures, financial objectives, and the like are typically centralizedat the firm’s headquarters. However, operating decisions, such as those relating to pro-duction, marketing, R&D, and human resource management, may or may not be cen-tralized depending on the firm’s international strategy.

Consider firms pursuing a global strategy. They must decide how to disperse the var-ious value creation activities around the globe so location and experience economiescan be realized. The head office must make the decisions about where to locate R&D,production, marketing, and so on. In addition, the globally dispersed web of value cre-ation activities that facilitates a global strategy must be coordinated. All of this cre-ates pressures for centralizing some operating decisions.

In contrast, the emphasis on local responsiveness in multidomestic firms createsstrong pressures for decentralizing operating decisions to foreign subsidiaries. In the clas-sic multidomestic firm, foreign subsidiaries have autonomy in most production and mar-keting decisions. International firms tend to maintain centralized control over their corecompetency and to decentralize other decisions to foreign subsidiaries. This typicallycentralizes control over R&D and/or marketing in the home country and decentralizesoperating decisions to the foreign subsidiaries. For example, Microsoft Corporation,which fits the international mode, centralizes its product development activities (whereits core competencies lie) at its Redmond, Washington, headquarters and decentralizesmarketing activity to various foreign subsidiaries. Thus, while products are developed at






home, managers in the various foreign subsidiaries have significant latitude for formu-lating strategies to market those products in their particular settings.4

The situation in transnational firms is more complex. The need to realize locationand experience curve economies requires some degree of centralized control overglobal production centers (as it does in global firms). However, the need for local re-sponsiveness dictates the decentralization of many operating decisions, particularly formarketing, to foreign subsidiaries. Thus, in transnational firms, some operating deci-sions are relatively centralized, while others are relatively decentralized. In addition,global learning based on the multidirectional transfer of skills between subsidiaries,and between subsidiaries and the corporate center, is a central feature of a firm pursu-ing a transnational strategy. The concept of global learning is predicated on the no-tion that foreign subsidiaries within a multinational firm have significant freedom todevelop their own skills and competencies. Only then can these be leveraged to ben-efit other parts of the organization. A substantial degree of decentralization is requiredif subsidiaries are going to have the freedom to do this. For this reason too, the pursuitof a transnational strategy requires a high degree of decentralization.5

|Horizontal Differentiation: The Design of Structure

Horizontal differentiation is concerned with how the firm decides to divide itself intosubunits.6 The decision is normally made on the basis of function, type of business, orgeographical area. In many firms, just one of these predominates, but more complexsolutions are adopted in others. This is particularly likely in the case of internationalfirms, where the conflicting demands to organize the company around different prod-ucts (to realize location and experience curve economies) and different national mar-kets (to remain locally responsive) must be reconciled. One solution to this dilemmais to adopt a matrix structure that divides the organization on the basis of both prod-ucts and national markets (as Unilever apparently did in Europe). In this section welook at different ways firms divide themselves into subunits.

The Structure of Domestic Firms

Most firms begin with no formal structure and are run by a single entrepreneur or asmall team of individuals. As they grow, the demands of management become too greatfor one individual or a small team to handle. At this point the organization is split intofunctions reflecting the firm’s value creation activities (e.g., production, marketing,R&D, sales). These functions are typically coordinated and controlled by top man-agement (see Figure 13.3). Decision making in this functional structure tends to becentralized.

Further horizontal differentiation may be required if the firm significantly diversi-fies its product offering, which takes the firm into different business areas. For exam-ple, Dutch multinational Philips NV began as a lighting company, but diversificationtook the company into consumer electronics (e.g., visual and audio equipment), in-

Figure 13.3

A Typical FunctionalStructure


Buying Units Plants Branch Sales Units Accounting Units

Purchasing Manufacturing Marketing Finance

TopManagement





dustrial electronics (integrated circuits and other electronic components), and med-ical systems (CT scanners and ultrasound systems). In such circumstances, a func-tional structure can be too clumsy. Problems of coordination and control arise whendifferent business areas are managed within the framework of a functional structure.7

For one thing, it becomes difficult to identify the profitability of each distinct businessarea. For another, it is difficult to run a functional department, such as production ormarketing, if it is supervising the value creation activities of several business areas.

To solve the problems of coordination and control, at this stage most firms switchto a product divisional structure (see Figure 13.4). With a product divisional structure,each division is responsible for a distinct product line (business area). Thus, Philipscreated divisions for lighting, consumer electronics, industrial electronics, and med-ical systems. Each product division is set up as a self-contained, largely autonomousentity with its own functions. The responsibility for operating decisions is typically de-centralized to product divisions, which are then held accountable for their perfor-mance. Headquarters is responsible for the overall strategic development of the firmand for the financial control of the various divisions.

The International Division

When firms initially expand abroad, they often group all their international activitiesinto an international division. This has tended to be the case for firms organized onthe basis of functions and for firms organized on the basis of product divisions. Re-gardless of the firm’s domestic structure, its international division tends to be organizedon geography. Figure 13.5 illustrates this for a firm whose domestic organization isbased on product divisions.

Many manufacturing firms expanded internationally by exporting the productmanufactured at home to foreign subsidiaries to sell. Thus, in the firm illustrated inFigure 13.5, the subsidiaries in Countries 1 and 2 would sell the products manufacturedby Divisions A, B, and C. In time, however, it might prove viable to manufacture theproduct in each country, and so production facilities would be added on a country-by-country basis. For firms with a functional structure at home, this might mean repli-cating the functional structure in every country in which the firm does business. Forfirms with a divisional structure, this might mean replicating the divisional structurein every country in which the firm does business.

This structure has been widely used; according to a Harvard study, 60 percent of allfirms that have expanded internationally have initially adopted it. Nonetheless, it givesrise to problems.8 The dual structure it creates contains inherent potential for conflictand coordination problems between domestic and foreign operations. One problemwith the structure is that the heads of foreign subsidiaries are not given as much voicein the organization as the heads of domestic functions (in the case of functional firms)

Figure 13.4

A Typical ProductDivisional Structure


Buying Units Plants Branch Sales Units Accounting Units

DepartmentPurchasing

DepartmentMarketing

DepartmentManufacturing

DepartmentFinance

Division Product Line B Division Product Line CDivision Product Line A

Headquarters





or divisions (in the case of divisional firms). Rather, the head of the international divi-sion is presumed to be able to represent the interests of all countries to headquarters.This effectively relegates each country’s manager to the second tier of the firm’s hierar-chy, which is inconsistent with a strategy of trying to expand internationally and builda true multinational organization.

Another problem is the implied lack of coordination between domestic operationsand foreign operations, which are isolated from each other in separate parts of thestructural hierarchy. This can inhibit the worldwide introduction of new products, thetransfer of core competencies between domestic and foreign operations, and the con-solidation of global production at key locations so as to realize location and experiencecurve economies. These problems are illustrated in the Management Focus that looksat the experience of Abbott Laboratories with an international divisional structure.

As a result of such problems, most firms that continue to expand internationallyabandon this structure and adopt one of the worldwide structures we discuss next. Thetwo initial choices are a worldwide product divisional structure, which tends to beadopted by diversified firms that have domestic product divisions, and a worldwidearea structure, which tends to be adopted by undiversified firms whose domestic struc-tures are based on functions. These two alternative paths of development are illus-trated in Figure 13.6. The model in the figure is referred to as the internationalstructural stages model and was developed by John Stopford and Louis Wells.9

Worldwide Area Structure

A worldwide area structure tends to be favored by firms with a low degree of diversifi-cation and a domestic structure based on function (see Figure 13.7). Under this struc-ture, the world is divided into geographic areas. An area may be a country (if themarket is large enough) or a group of countries. Each area tends to be a self-contained,largely autonomous entity with its own set of value creation activities (e.g., its ownproduction, marketing, R&D, human resources, and finance functions). Operations


Figure 13.5

One Company’s International Divisional Structure

Country 1

General Manager(Product A, B,and/or C)

Country 2

General Manager(Product A, B,and/or C)

Functional units

Functional units

Headquarters

Domestic Division

General ManagerProduct Line A

Domestic Division

General ManagerProduct Line B

Domestic Division

General ManagerProduct Line C

International Division

General ManagerArea Line





443

443M

an

ag

em

en

t Fo

cu

s

The International Division at Abbott Laboratories

www.mhhe.com/hill

for the U.S. market and then modifying those productsfor foreign customers is a slow and expensive process.Instead, across all four of the company’s businesses,Abbott is trying to build global products that can belaunched simultaneously around the world. Thischange is pulling Abbott toward adoptingglobal product divisions for all four of its busi-nesses. Some argue that only global productdivisions would give Abbott the tight controlover product development and productlaunch strategy that is deemed necessary.

On the other hand, bigger organizationswith greater purchasing leverage, such as largehospital groups and health maintenance organiza-tions, are coordinating their buying across a range ofproduct lines in both the United States and elsewhere.These powerful customers prefer to have a single con-tact point at Abbott. Abbott develops stronger relationswith key customers by having a single marketing orga-nization in each country in which the company doesbusinesses. This organization sells the products fromeach of Abbott’s four product divisions.

Executives at Abbott’s international division sup-port maintaining the geographic organization, whilethe heads of the product divisions favor a shift towardfour global product divisions. Top management seemsto have decided there is no perfect solution to thecompany’s organizational problems, and that imper-fect as the current structure is, it works too well tocontemplate a major change.

Sources: R. Walters, “Two’s Company,” Financial Times, July 7,1995, p. 12; Abbott Laboratories 2000 Annual Report; and M.Santoli, “Patient Reviving,” Barron’s, February 28, 2000,pp. 24–26.

With sales of about $14 billion in 2000, AbbottLaboratories is one of the world’s largest healthcare companies. The company split itself intothree divisions—pharmaceuticals, hospital pro-ducts, and nutritional products—in the 1960s,a structure that still exists. Each division oper-ates as a profit center, and each is relatively au-tonomous and self-contained, with its ownR&D, manufacturing, and marketing functions.By the late 1960s Abbott’s foreign sales weregrowing rapidly; the company added an inter-

national division to handle the firm’s non-U.S. operationson geographic rather than product lines.

Alongside these four divisions, however, a new busi-ness has grown up that is organized differently. Abbott’sdiagnostics business was established in the 1970s andbecame a world leader with global sales of $3 billion in2000. Unlike the other divisions, the diagnostics busi-ness is organized on a global basis, operating in foreigncountries through its own staff, rather than through theinternational division. Thus, Abbott handles global salesin two different ways—through an international divisionand through a global product division (the diagnosticsbusiness organization). The company is debating thebest way of organizing international operations.

This debate is being informed by two changes oc-curring in Abbott’s environment, changes that arepulling the company in different directions. One changeis a shift toward global product development in thehealth care industry. To quickly recapture the costs ofdeveloping new products, which for pharmaceuticalscan sometimes top $500 million, companies are tryingto introduce new products as rapidly as possible world-wide. Abbott has found that developing products first

authority and strategic decisions relating to each of these activities are typically de-centralized to each area, with headquarters retaining authority for the overall strate-gic direction of the firm and financial control.

This structure facilitates local responsiveness. Because decision-making respon-sibilities are decentralized, each area can customize product offerings, marketingstrategy, and business strategy to the local conditions. However, this structure en-courages fragmentation of the organization into highly autonomous entities. Thiscan make it difficult to transfer core competencies and skills between areas and torealize location and experience curve economies. In other words, the structure isconsistent with a multidomestic strategy but with little else. Firms structured onthis basis may encounter significant problems if local responsiveness is less criticalthan reducing costs or transferring core competencies for establishing a competi-tive advantage.





Worldwide Product Divisional Structure

A worldwide product division structure tends to be adopted by firms that are reason-ably diversified and, accordingly, originally had domestic structures based on productdivisions. As with the domestic product divisional structure, each division is a self-contained, largely autonomous entity with full responsibility for its own value creationactivities. The headquarters retains responsibility for the overall strategic develop-ment and financial control of the firm (see Figure 13.8).

Underpinning the organization is a belief that the value creation activities of eachproduct division should be coordinated by that division worldwide. Thus, the world-wide product divisional structure is designed to help overcome the coordination prob-lems that arise with the international division and worldwide area structures (see theManagement Focus on Abbott Laboratories for a detailed example). This structure pro-vides an organizational context that enhances the consolidation of value creation ac-tivities at key locations necessary for realizing location and experience curveeconomies. It also facilitates the transfer of core competencies within a division’s world-wide operations and the simultaneous worldwide introduction of new products. Themain problem with the structure is the limited voice it gives to area or country man-agers, since they are seen as subservient to product division managers. The result canbe a lack of local responsiveness, which, as we saw in Chapter 12, can be a fatal flaw.

Global Matrix Structure

Both the worldwide area structure and the worldwide product divisional structure havestrengths and weaknesses. The worldwide area structure facilitates local responsive-


Figure 13.6

The International StructuralStages Model

Source: Adapted from John M.Stopford and Louis T. Wells, Strategyand Structure of the MultinationalEnterprise (New York: Basic Books,1972).

ForeignProductDiversity

Foreign Sales as a Percentage of Total Sales

Alternate Pathsof Development

WorldwideProductDivision Global Matrix

("Grid")

AreaDivisionInternational

Division

Figure 13.7

A Worldwide AreaStructure

Headquarters

North AmericanArea

Latin AmericanArea

EuropeanArea

Middle Eastern-African Area

Far EastArea





ness, but it can inhibit the realization of location and experience curve economies andthe transfer of core competencies between areas. The worldwide product divisionstructure provides a better framework for pursuing location and experience curveeconomies and for transferring core competencies, but it is weak in local responsive-ness. Other things being equal, this suggests that a worldwide area structure is moreappropriate if the firm’s strategy is multidomestic, while a worldwide product divi-sional structure is more appropriate for firms pursuing global or international strate-gies. However, as we saw in Chapter 12, other things are not equal. As Bartlett andGhoshal have argued, to survive in some industries, firms must adopt a transnationalstrategy. That is, they must focus simultaneously on realizing location and experiencecurve economies, on local responsiveness, and on the internal transfer of core compe-tencies (worldwide learning).10

Many firms have attempted to cope with the conflicting demands of a transnationalstrategy by using a matrix structure (see Figure 13.2). In the classic global matrix struc-ture, horizontal differentiation proceeds along two dimensions: product division andgeographic area (see Figure 13.9). The philosophy is that responsibility for operatingdecisions pertaining to a particular product should be shared by the product division


Figure 13.8

A Worldwide ProductDivisional Structure

Functional Units Functional Units

Headquarters

WorldwideProduct Groupor Division A

WorldwideProduct Groupor Division B

WorldwideProduct Groupor Division C

Area 1

(Domestic)

Area 2

(International)

Figure 13.9

A Global Matrix StructureHeadquarters

Area 1 Area 2 Area 3

ProductDivision A

ProductDivision B

ProductDivision C

Manager HereBelongs to Division B and Area 2





and the various areas of the firm. Thus, the nature of the product offering, the mar-keting strategy, and the business strategy to be pursued in Area 1 for the products pro-duced by Division A are determined by conciliation between Division A and Area 1management. It is believed that this dual decision-making responsibility should enablethe firm to simultaneously achieve its particular objectives. In a classic matrix struc-ture, giving product divisions and geographical areas equal status within the organiza-tion reinforces the idea of dual responsibility. Individual managers thus belong to twohierarchies (a divisional hierarchy and an area hierarchy) and have two bosses (a di-visional boss and an area boss).

The reality of the global matrix structure is that it often does not work anywherenear as well as the theory predicts. In practice, the matrix often is clumsy and bureau-cratic. It can require so many meetings that it is difficult to get any work done. Theneed to get an area and a product division to reach a decision can slow decision mak-ing and produce an inflexible organization unable to respond quickly to market shiftsor to innovate. The dual-hierarchy structure can lead to conflict and perpetual powerstruggles between the areas and the product divisions, catching many managers in themiddle. To make matters worse, it can prove difficult to ascertain accountability in thisstructure. When all critical decisions are the product of negotiation between divisionsand areas, one side can always blame the other when things go wrong. As a managerin one global matrix structure, reflecting on a failed product launch, said to the author,“Had we been able to do things our way, instead of having to accommodate those guysfrom the product division, this would never have happened.” (A manager in the prod-uct division expressed similar sentiments.) The result of such finger-pointing can bethat accountability is compromised, conflict is enhanced, and headquarters loses con-trol over the organization.

In light of these problems, many transnational firms are now trying to build “flexi-ble” matrix structures based more on firmwide networks and a shared culture and vi-sion than on a rigid hierarchical arrangement. Dow Chemical, profiled in theaccompanying Management Focus, is one such firm. Within such companies the in-formal structure plays a greater role than the formal structure. We discuss this issuewhen we consider informal integrating mechanisms in the next section.

|Integrating Mechanisms

In the previous section, we explained that firms divide themselves into subunits. Nowwe need to examine some means of coordinating those subunits. One way of achiev-ing coordination is through centralization. If the coordination task is complex, how-ever, centralization may not be very effective. Higher-level managers responsible forachieving coordination can soon become overwhelmed by the volume of work re-quired to coordinate the activities of various subunits, particularly if the subunits arelarge, diverse, and/or geographically dispersed. When this is the case, firms look to-ward integrating mechanisms, both formal and informal, to help achieve coordination.In this section, we introduce the various integrating mechanisms that internationalbusinesses can use. Before doing so, however, let us explore the need for coordinationin international firms and some impediments to coordination.

Strategy and Coordination in the International Business

The need for coordination between subunits varies with the strategy of the firm. Theneed for coordination is lowest in multidomestic companies, is higher in interna-tional companies, higher still in global companies, and highest of all in transnationalcompanies. Multidomestic firms are primarily concerned with local responsiveness.Such firms are likely to operate with a worldwide area structure in which each areahas considerable autonomy and its own set of value creation functions. Since eacharea is established as a stand-alone entity, the need for coordination between areas isminimized.






The need for coordination is greater in firms pursuing an international strategy andtrying to profit from the transfer of core competencies and skills between units at homeand abroad. Coordination is necessary to support the transfer of skills and product of-ferings between units. The need for coordination is also great in firms trying to profitfrom location and experience curve economies; that is, in firms pursuing global strate-gies. Achieving location and experience economies involves dispersing value creationactivities to various locations around the globe. The resulting global web of activities

447

447M

an

ag

em

en

t Fo

cu

s

Dow Chemical’s Matrix Structure

www.mhhe.com/hill

overseeing business sectors about which plants shouldbe built and where. In short, the structure didn’t work.Instead of abandoning the structure, however, Dow de-cided to see if it could be made more flexible.

Dow’s decision to keep its matrix structurewas prompted by its move into the pharma-ceuticals industry. The company realized thatthe pharmaceutical business is very differ-ent from the bulk chemicals business. Inbulk chemicals, the big returns come fromachieving economies of scale in production.This dictates establishing large plants in keylocations from which regional or global marketscan be served. But in pharmaceuticals, regulatoryand marketing requirements for drugs vary so muchfrom country to country that local needs are far moreimportant than reducing manufacturing costs throughscale economies. A high degree of local responsive-ness is essential. Dow realized its pharmaceutical busi-ness would never thrive if it were managed by thesame priorities as its mainstream chemical operations.

Accordingly, instead of abandoning its matrix, Dowdecided to make it more flexible so it could better ac-commodate the different businesses, each with itsown priorities, within a single management system. Asmall team of senior executives at headquarters nowhelps set the priorities for each type of business. Af-ter priorities are identified for each business sector,one of the three elements of the matrix—function,business, or geographic area—is given primary au-thority in decision making. Which element takes thelead varies according to the type of decision and themarket or location in which the company is compet-ing. Such flexibility requires that all employees under-stand what is occurring in the rest of the matrix.Although this may seem confusing, Dow claims thisflexible system works well and credits much of itssuccess to the quality of the decisions it facilitates.

Source: “Dow Draws Its Matrix Again, and Again, and Again,”The Economist, August 5, 1989, pp. 55–56.

A handful of major players compete head tohead around the world in the chemical in-dustry. These companies are Dow Chemicaland Du Pont of the United States, GreatBritain’s ICI, and the German trio of BASF,Hoechst AG, and Bayer. The barriers to thefree flow of chemical products between na-tions largely disappeared in the 1970s. Thisalong with the commodity nature of mostbulk chemicals and a severe recession in theearly 1980s ushered in a prolonged period of

intense price competition. In such an environment,the company that wins the competitive race is the onewith the lowest costs, and in recent years the clearwinner has been Dow.

Dow’s managers insist that part of the credit mustbe placed at the feet of its much maligned “matrix”organization. Dow’s organizational matrix has threeinteracting elements: functions (e.g., R&D, manufac-turing, marketing), businesses (e.g., ethylene, plas-tics, pharmaceuticals), and geography (e.g., Spain,Germany, Brazil). Managers’ job titles incorporate allthree elements—for example, plastics marketing man-ager for Spain—and most managers report to at leasttwo bosses. The plastics marketing manager in Spainmight report to both the head of the worldwide plasticsbusiness and the head of the Spanish operations. Theintent of the matrix was to make Dow operations re-sponsive to both local market needs and corporate ob-jectives. Thus, the plastics business might be chargedwith minimizing Dow’s global plastics productioncosts, while the Spanish operation might be chargedwith determining how best to sell plastics in the Span-ish market.

When Dow introduced this structure, the resultswere less than promising; multiple reporting channelsled to confusion and conflict. The large number ofbosses made for an unwieldy bureaucracy. The overlap-ping responsibilities resulted in turf battles and a lack ofaccountability. Area managers disagreed with managers





must be coordinated to ensure the smooth flow of inputs into the value chain, thesmooth flow of semifinished products through the value chain, and the smooth flowof finished products to markets around the world.

The need for coordination is greatest in transnational firms, which simultaneously pur-sue location and experience curve economies, local responsiveness, and the multidirec-tional transfer of core competencies and skills among all of the firm’s subunits (referred toas global learning). As in global companies, coordination is required to ensure the smoothflow of products through the global value chain. As in international companies, coordi-nation is required for ensuring the transfer of core competencies to subunits. However, thetransnational goal of achieving multidirectional transfer of competencies requires muchgreater coordination than in international firms. In addition, transnationals require co-ordination between foreign subunits and the firm’s globally dispersed value creation ac-tivities (e.g., production, R&D, marketing) to ensure that any product offering andmarketing strategy is sufficiently customized to local conditions.

Impediments to Coordination

Managers of the various subunits have different orientations, partly because they havedifferent tasks. For example, production managers are typically concerned with pro-duction issues such as capacity utilization, cost control, and quality control, whereasmarketing managers are concerned with marketing issues such as pricing, promotions,distribution, and market share. These differences can inhibit communication betweenthe managers. Quite simply, these managers often do not even “speak the same lan-guage.” There may also be a lack of respect between subunits (e.g., marketing managers“looking down on” production managers, and vice versa), which further inhibits thecommunication required to achieve cooperation and coordination.

Differences in subunits’ orientations also arise from their differing goals. For exam-ple, worldwide product divisions of a multinational firm may be committed to costgoals that require global production of a standardized product, whereas a foreign sub-sidiary may be committed to increasing its market share in its country, which will re-quire a nonstandard product. These different goals can lead to conflict.

Such impediments to coordination are not unusual in any firm, but they can be par-ticularly problematic in the multinational enterprise with its profusion of subunits athome and abroad. Differences in subunit orientation are often reinforced in multina-tionals by the separations of time zone, distance, and nationality between managers ofthe subunits.

For example, until recently the Dutch company Philips had an organizationcomprising worldwide product divisions and largely autonomous national organiza-tions. The company has long had problems getting its product divisions and na-tional organizations to cooperate on such things as new product introductions.When Philips developed a VCR format, the V2000 system, it could not get itsNorth American subsidiary to introduce the product. Rather, the North Americanunit adopted the rival VHS format produced by Philip’s global competitor, Mat-sushita. Unilever experienced a similar problem in its detergents business. Theneed to resolve disputes between Unilever’s many national organizations and itsproduct divisions extended the time necessary for introducing a new product acrossEurope to several years. This denied Unilever the first-mover advantage crucial tobuilding a strong market position.11

Formal Integrating Mechanisms

The formal mechanisms used to integrate subunits vary in complexity from simple di-rect contact and liaison roles, to teams, to a matrix structure (see Figure 13.10). In gen-eral, the greater the need for coordination, the more complex the formal integratingmechanisms need to be.12

Direct contact between subunit managers is the simplest integrating mechanism.By this “mechanism,” managers of the various subunits simply contact each other






whenever they have a common concern. Direct contact may not be effective if themanagers have differing orientations that act to impede coordination, as pointed outin the previous subsection.

Liaison roles are a bit more complex. When the volume of contacts between sub-units increases, coordination can be improved by giving a person in each subunit re-sponsibility for coordinating with another subunit on a regular basis. Through theseroles, the people involved establish a permanent relationship. This helps attenuate theimpediments to coordination discussed in the previous subsection.

When the need for coordination is greater still, firms tend to use temporary orpermanent teams composed of individuals from the subunits that need to achievecoordination. They are typically used to coordinate product development and in-troduction, but they are useful when any aspect of operations or strategy requires thecooperation of two or more subunits. Product development and introduction teamsare typically composed of personnel from R&D, production, and marketing. The re-sulting coordination aids the development of products that are tailored to consumerneeds and that can be produced at a reasonable cost (design for manufacturing).

When the need for integration is very high, firms may institute a matrix structure,in which all roles are viewed as integrating roles. The structure is designed to facilitatemaximum integration among subunits. The most common matrix in multinationalfirms is based on geographical areas and worldwide product divisions. This achieves ahigh level of integration between the product divisions and the areas so that, in the-ory, the firm can pay close attention to both local responsiveness and the pursuit of lo-cation and experience curve economies.

In some multinationals, the matrix is more complex still, structuring the firm intogeographical areas, worldwide product divisions, and functions, all of which report di-rectly to headquarters. Thus, within a company such as Dow Chemical (see the Man-agement Focus) each manager belongs to three hierarchies (e.g., a plastics marketingmanager in Spain is a member of the Spanish subsidiary, the plastics product division,and the marketing function). In addition to facilitating local responsiveness and loca-tion and experience curve economies, such a matrix fosters the transfer of core compe-tencies within the organization. This occurs because core competencies tend to residein functions (e.g., R&D, marketing). A structure such as Dow’s facilitates the transferof competencies existing in functions from division to division and from area to area.

However, as discussed earlier, such matrix solutions to coordination problems inmultinational enterprises can quickly become bogged down in a bureaucratic tanglethat creates as many problems as it solves. Matrix structures tend to be bureaucratic,inflexible, and characterized by conflict rather than the hoped-for cooperation. As inthe case of Dow Chemical, for such a structure to work it needs to be somewhat flex-ible and to be supported by informal integrating mechanisms.

Figure 13.10

Formal IntegratingMechanisms


Direct Contact

Liaison Roles

Teams

Matrix Structures

Increasing Complexityof Integrating Mechanism





Informal Integrating Mechanism: Management Networks

In attempting to alleviate or avoid the problems associated with formal integratingmechanisms in general, and matrix structures in particular, firms with a high need forintegration have been experimenting with an informal integrating mechanism: man-agement networks that are supported by an organization culture that values teamworkand cross-unit cooperation.13 A management network is a system of informal contactsbetween managers within an enterprise.14 The great strength of a network is that itcan be used as a nonbureaucratic conduit for knowledge flows within a multinationalenterprise.15 For a network to exist, managers at different locations within the orga-nization must be linked to each other at least indirectly. For example, Figure 13.11shows the simple network relationships between seven managers within a multina-tional firm. Managers A, B, and C all know each other personally, as do Managers D,E, and F. Although Manager B does not know Manager F personally, they are linkedthrough common acquaintances (Managers C and D). Thus, we can say that ManagersA through F are all part of the network, and also that Manager G is not.

Imagine Manager B is a marketing manager in Spain and needs to know the solu-tion to a technical problem to better serve an important European customer. ManagerF, an R&D manager in the United States, has the solution to Manager B’s problem.Manager B mentions her problem to all of her contacts, including Manager C, and asksif they know of anyone who might be able to provide a solution. Manager C asks Man-ager D, who tells Manager F, who then calls Manager B with the solution. In this way,coordination is achieved informally through the network, rather than by formal inte-grating mechanisms such as teams or a matrix structure.

For such a network to function effectively, however, it must embrace as many man-agers as possible. For example, if Manager G had a problem similar to manager B’s, hewould not be able to utilize the informal network to find a solution; he would have toresort to more formal mechanisms. Establishing firmwide networks is difficult, and al-though network enthusiasts speak of networks as the “glue” that binds multinationalcompanies together, it is far from clear how successful firms have been at building com-panywide networks. Two techniques being used to establish networks are informationsystems and management development policies.

Firms are using their computer and telecommunications networks to provide thephysical foundation for informal information systems networks.16 Electronic mail,videoconferencing, and high-speed data systems make it much easier for managersscattered over the globe to get to know each other. Without an existing network ofpersonal contacts, however, worldwide information systems are unlikely to meet afirm’s need for integration.

Firms are using their management development programs to build informal net-works. Tactics include rotating managers through various subunits on a regular basis sothey build their own informal network and using management education programs to

Figure 13.11

A Simple ManagementNetwork


A

B

F

EG

C D





bring managers of subunits together in a single location so they can become ac-quainted. Both of these tactics are used at Unilever to build its informal managementnetwork (see the opening case for details).

Management networks by themselves may not be sufficient to achieve coordinationif subunit managers persist in pursuing subgoals that are at variance with firmwidegoals. For a management network to function properly—and for a formal matrix struc-ture to work, also—managers must share a strong commitment to the same goals. Toappreciate the nature of the problem, consider again the case of Manager B and Man-ager F. As before, Manager F hears about Manager B’s problem through the network.However, solving Manager B’s problem would require Manager F to devote consider-able time to the task. Insofar as this would divert Manager F away from his own regu-lar tasks—and the pursuit of subgoals that differ from those of Manager B—he may beunwilling to do it. Thus, Manager F may not call Manager B, and the informal net-work would fail to provide a solution to Manager B’s problem.

To eliminate this flaw, organization’s managers must adhere to a common set ofnorms and values that override differing subunit orientations.17 In other words, thefirm must have a strong organizational culture that promotes teamwork and coopera-tion. When this is the case, a manager is willing and able to set aside the interests ofhis own subunit when doing so benefits the firm as a whole. If Manager B and Man-ager F are committed to the same organizational norms and value systems, and if theseorganizational norms and values place the interests of the firm as a whole above theinterests of any individual subunit, Manager F should be willing to cooperate withManager B on solving her subunit’s problems.

Summary

The message contained in this section is crucial to understanding the problems of man-aging the multinational firm. Multinationals need integration—particularly if they arepursuing global, international, or transnational strategies—but it can be difficult toachieve due to the impediments to coordination we discussed. Firms traditionally havetried to achieve coordination by adopting formal integrating mechanisms. These do notalways work, however, since they tend to be bureaucratic and do not necessarily addressthe problems that arise from differing subunit orientations. This is particularly likelywith a complex matrix structure, and yet, a complex matrix structure is required for si-multaneously achieving location and experience curve economies, local responsive-ness, and the multidirectional transfer of core competencies within the organization.The solution to this dilemma seems twofold. First, the firm must try to establish an in-formal management network that can do much of the work previously undertaken by aformal matrix structure. Second, the firm must build a common culture. Neither ofthese partial solutions, however, is easy to achieve.18

A major task of a firm’s leadership is to control the various subunits of the firm—whetherthey be defined on the basis of function, product division, or geographic area—to ensuretheir actions are consistent with the firm’s overall strategic and financial objectives. Firmsachieve this with various control and incentive systems. In this section, we first reviewthe various types of control systems firms use to control their subunits. Then we brieflydiscuss incentive systems. Then we will look at how the appropriate control and incen-tive systems vary according to firms’ international strategies.

|Types of Control Systems

Four main types of control systems are used in multinational firms: personal controls,bureaucratic controls, output controls, and cultural controls. In most firms, all four areused, but their relative emphasis varies with the strategy of the firm.


Control Systems and Incentives





Personal Controls

Personal control is control by personal contact with subordinates. This type of controltends to be most widely used in small firms, where it is seen in the direct supervisionof subordinates’ actions. However, it also structures the relationships between man-agers at different levels in multinational enterprises. For example, the CEO may use agreat deal of personal control to influence the behavior of his or her immediate sub-ordinates, such as the heads of worldwide product divisions or major geographic areas.In turn, these heads may use personal control to influence the behavior of their sub-ordinates, and so on down through the organization. For example, Jack Welsh thelongtime CEO of General Electric who retired in 2001, had regular one-on-one meet-ings with the heads of all of GE’s major businesses (most of which are international).He used these meetings to “probe” the managers about the strategy, structure, and fi-nancial performance of their operations. In doing so, he essentially exercised personalcontrol over these managers and, undoubtedly, over the strategies that they favored.

Bureaucratic Controls

Bureaucratic control is control through a system of rules and procedures that directsthe actions of subunits. The most important bureaucratic controls in subunits withinmultinational firms are budgets and capital spending rules. Budgets are essentially a setof rules for allocating a firm’s financial resources. A subunit’s budget specifies withsome precision how much the subunit may spend. Headquarters uses budgets to influ-ence the behavior of subunits. For example, the R&D budget normally specifies howmuch cash the R&D unit may spend on product development. R&D managers knowthat if they spend too much on one project, they will have less to spend on other proj-ects, so they modify their behavior to stay within the budget. Most budgets are set bynegotiation between headquarters management and subunit management. Headquar-ters management can encourage the growth of certain subunits and restrict the growthof others by manipulating their budgets.

Capital spending rules require headquarters management to approve any capital ex-penditure by a subunit that exceeds a certain amount (at GE, $50,000). A budget al-lows headquarters to specify the amount a subunit can spend in a given year, andcapital spending rules give headquarters additional control over how the money isspent. Headquarters can be expected to deny approval for capital spending requeststhat are at variance with overall firm objectives and to approve those that are con-gruent with firm objectives.

Output Controls

Output controls involve setting goals for subunits to achieve and expressing thosegoals in terms of relatively objective performance metrics such as profitability, pro-ductivity, growth, market share, and quality. The performance of subunit managers isthen judged by their ability to achieve the goals.19 If goals are met or exceeded, sub-unit managers will be rewarded. If goals are not met, top management will normallyintervene to find out why and take appropriate corrective action. Thus, control isachieved by comparing actual performance against targets and intervening selectivelyto take corrective action. Subunits’ goals depend on their role in the firm. Self-contained product divisions or national subsidiaries are typically given goals for prof-itability, sales growth, and market share. Functions are more likely to be given goalsrelated to their particular activity. Thus, R&D will be given product developmentgoals, production will be given productivity and quality goals, marketing will be givenmarket share goals, and so on.

As with budgets, goals are normally established through negotiation between sub-units and headquarters. Generally, headquarters tries to set goals that are challengingbut realistic, so subunit managers are forced to look for ways to improve their opera-tions but are not so pressured that they will resort to dysfunctional activities to do so(such as short-run profit maximization). Output controls foster a system of “manage-






ment by exception,” in that so long as subunits meet their goals, they are left alone. Ifa subunit fails to attain its goals, however, headquarters managers are likely to ask sometough questions. If they don’t get satisfactory answers, they are likely to interveneproactively in a subunit, replacing top management and looking for ways to improveefficiency.

Cultural Controls

Cultural controls exist when employees “buy into” the norms and value systems of thefirm. When this occurs, employees tend to control their own behavior, which reducesthe need for direct supervision. In a firm with a strong culture, self-control can reducethe need for other control systems. We shall discuss organizational culture later. Mc-Donald’s actively promotes organizational norms and values, referring to its fran-chisees and suppliers as partners and emphasizing its long-term commitment to them.This commitment is not just a public relations exercise; it is backed by actions, in-cluding a willingness to help suppliers and franchisees improve their operations by pro-viding capital and/or management assistance when needed. In response, McDonald’sfranchisees and suppliers are integrated into the firm’s culture and thus become com-mitted to helping McDonald’s succeed. One result is that McDonald’s can devote lesstime than would otherwise be necessary to controlling its franchisees and suppliers.

|Incentive Systems

Incentives refer to the devices used to reward appropriate employee behavior. Manyemployees receive incentives in the form of annual bonus pay. Incentives are usuallyclosely tied to the performance metrics used for output controls. For example, settingtargets linked to profitability might be used to measure the performance of a subunit,such as a global product division. To create positive incentives for employees to workhard to exceed those targets, they may be given a share of any profits over above thosetargeted. If a subunit has set a goal of attaining a 15 percent return on investment andit actually attains a 20 percent return, unit employees may be given a share in the prof-its generated in excess of the 15 percent target in the form of bonus pay. We shall re-turn to the topic of incentive systems in Chapter 18 when we discuss human resourcestrategy in the multinational firm. For now, however, several important points need tobe made.

First, the type of incentive used often varies depending on the employees and theirtasks. Incentives for employees working on the factory floor may be very different fromthe incentives used for senior managers. The incentives used must be matched to thetype of work being performed. The employees on the factory floor of a manufacturingplant may be broken into teams of 20 to 30 individuals, and they may have their bonuspay tied to the ability of their team to hit or exceed targets for output and product qual-ity. In contrast, the senior managers of the plant may be rewarded according to met-rics linked to the output of the entire operation. The basic principle is to make surethe incentive scheme for an individual employee is linked to an output target that heor she has some control over and can influence. The individual employees on the fac-tory floor may not be able to exercise much influence over the performance of the en-tire operation, but they can influence the performance of their team, so incentive payis tied to output at this level.

Second, the successful execution of strategy in the multinational firm often re-quires significant cooperation between managers in different subunits. For example,as noted earlier, some multinational firms operate with matrix structures where acountry subsidiary might be responsible for marketing and sales in a nation, while aglobal product division might be responsible for manufacturing and product devel-opment. The managers of these different units need to cooperate closely with eachother if the firm is to be successful. One way of encouraging the managers to cooper-ate is to link incentives to performance at a higher level in the organization. Thus,






the senior managers of the country subsidiaries and global product divisions might berewarded according to the profitability of the entire firm. The thinking here is thatboosting the profitability of the entire firm requires managers in the country sub-sidiaries and product divisions to cooperate with each other on strategy implementa-tion and linking incentive systems to the next level up in the hierarchy encouragesthis. Most firms use a formula for incentives that links a portion of incentive pay tothe performance of the subunit in which a manager or employee works and a portionto the performance of the entire firm, or some other higher-level organizational unit.The goal is to encourage employees to improve the efficiency of their unit and to co-operate with other units in the organization.

Third, the incentive systems used within a multinational enterprise often have tobe adjusted to account for national differences in institutions and culture. Incentivesystems that work in the United States might not work, or even be allowed, in othercountries. For example, Lincoln Electric, a leader in the manufacture of arc weldingequipment, has used an incentive system for its employees based on piecework rates inits American factories (under a piecework system, employees are paid according to theamount they produce). While this system has worked very well in the United States,Lincoln has found that the system is difficult to introduce in other countries. In somecountries, such as Germany, piecework systems are illegal, while in others the prevail-ing national culture is antagonistic to a system where performance is so closely tied toindividual effort. For further details, see the accompanying Management Focus.

Finally, it is important for managers to recognize that incentive systems can haveunintended consequences. Managers need to carefully think through exactly what be-havior certain incentives encourage. For example, if employees in a factory are re-warded solely on the basis of how many units of output they produce, with no attentionpaid to the quality of that output, they may produce as many units as possible to boosttheir incentive pay, but the quality of those units may be poor.

|Control Systems, Incentives, and Strategy

in the International Business

The key to understanding the relationship between international strategy, control sys-tems, and incentive systems is the concept of performance ambiguity.

Performance Ambiguity

Performance ambiguity exists when the causes of a subunit’s poor performance are notclear. This is not uncommon when a subunit’s performance is partly dependent on theperformance of other subunits; that is, when there is a high degree of interdependencebetween subunits within the organization. Consider the case of a French subsidiary ofa U.S. firm that depends on another subsidiary, a manufacturer based in Italy, for theproducts it sells. The French subsidiary is failing to achieve its sales goals, and the U.S.management asks the managers to explain. They reply that they are receiving poor-quality goods from the Italian subsidiary. So the U.S. management asks the managersof the Italian operation what the problem is. They reply that their product quality isexcellent—the best in the industry, in fact—and that the French simply don’t knowhow to sell a good product. Who is right, the French or the Italians? Without more in-formation, top management cannot tell. Because they are dependent on the Italiansfor their product, the French have an “alibi” for poor performance. U.S. managementneeds to have more information to determine who is correct. Collecting this informa-tion is expensive and time consuming and will divert attention away from other issues.In other words, performance ambiguity raises the costs of control.

Consider how different things would be if the French operation were self-contained,with its own manufacturing, marketing, and R&D facilities. The French operationwould lack a convenient alibi for its poor performance; the French managers wouldstand or fall on their own merits. They could not blame the Italians for their poor sales.






455

455M

an

ag

em

en

t Fo

cu

s

Organizational Culture and Incentive Systems at Lincoln Electric

www.mhhe.com/hill

boost productivity, for doing so influences their levelof pay. Lincoln’s factory workers have been able toearn a base pay that often exceeds the average man-ufacturing wage in the area by more than 50 per-cent and receive pay bonus on top of this that ingood years could double their base pay. De-spite high employee compensation, theworkers are so productive that Lincoln hasa lower cost structure than its competitors.

While this organizational culture and setof incentives works well in the UnitedStates, where it is compatible with the indi-vidualistic culture of the country, it did nottranslate easily into foreign operations. In the1980s and early 1990s, Lincoln expanded aggressivelyinto Europe and Latin America, acquiring a number oflocal arc welding manufacturers. Lincoln left localmanagers in place, believing that they knew local con-ditions better than Americans. However, the localmanagers had little working knowledge of Lincoln’sstrong organizational culture and were unable or un-willing to impose that culture on their units, which hadtheir own long-established organizational cultures.Nevertheless, Lincoln told local managers to introduceits incentive systems in acquired companies. They fre-quently ran into legal and cultural roadblocks. In manycountries, piecework is viewed as an exploitive com-pensation system that forces employees to work everharder. In Germany, where Lincoln made an acquisi-tion, it is actually illegal. In Brazil, a bonus paid for morethan two years becomes a legal entitlement! In manyother countries, both managers and workers were op-posed to the idea of piecework. Lincoln found thatmany European workers valued extra leisure morehighly than extra income and were not prepared towork as hard as their American counterparts. Many ofthe acquired companies were also unionized, and thelocal unions vigorously opposed the introduction ofpiecework. As a result, Lincoln was not able to repli-cate the high level of employee productivity that it hadachieved in the United States, and its expansion pulleddown the performance of the entire company.

Sources: J. O’Connell, “Lincoln Electric: Venturing Abroad,” Har-vard Business School Case, # 9-398-095, April 1998, andwww.lincolnelectric.com.

Lincoln Electric is one of the leading compa-nies in the global market for arc weldingequipment. Lincoln’s success has beenbased on extremely high levels of employeeproductivity. The company attributes its pro-ductivity to a strong organizational cultureand an incentive scheme based on piece-work. Lincoln’s organizational culture datesback to James Lincoln, who in 1907 joinedthe company that his brother had establisheda few years earlier. Lincoln had a strong re-

spect for the ability of the individual and believed that,correctly motivated, ordinary people could achieve ex-traordinary performance. He emphasized that Lincolnshould be a meritocracy where people were rewardedfor their individual effort. Strongly egalitarian, Lincolnremoved barriers to communication between “work-ers” and “managers,” practicing an open-door policy.He made sure that all who worked for the companywere treated equally; for example, everyone ate in thesame cafeteria, there were no reserved parking placesfor “managers,” and so on. Lincoln also believed thatany gains in productivity should be shared with con-sumers in the form of lower prices, with employees inthe form of higher pay, and with shareholders in theform of higher dividends.

The organizational culture that grew out of JamesLincoln’s beliefs was reinforced by the company’s in-centive system. Production workers receive no basesalary but are paid according to the number of piecesthey produce. The piecework rates at the company en-able an employee working at a normal pace to earn anincome equivalent to the average wage for manufac-turing workers in the area where a factory is based.Workers have responsibility for the quality of theiroutput and must repair any defects spotted by qualityinspectors before the pieces are included in the piece-work calculation. Since 1934, production workershave been awarded a semiannual bonus based onmerit ratings. These ratings are based on objective cri-teria (such as an employee’s level and quality of out-put) and subjective criteria (such as an employee’sattitudes toward cooperation and his or her depend-ability). These systems give Lincoln’s employees an in-centive to work hard and to generate innovations that





The level of performance ambiguity, therefore, is a function of the interdependence ofsubunits in an organization.

Strategy, Interdependence, and Ambiguity

Now let us consider the relationship among international strategy, interdependence,and performance ambiguity. In multidomestic firms, each national operation is astand-alone entity and can be judged on its own merits. The level of performance am-biguity is low. In an international firm, the level of interdependence is somewhathigher. Integration is required to facilitate the transfer of core competencies and skills.Since the success of a foreign operation is partly dependent on the quality of the com-petency transferred from the home country, performance ambiguity can exist.

In global firms, the situation is still more complex. Recall that in a pure global firmthe pursuit of location and experience curve economies leads to the development of aglobal web of value creation activities. Many of the activities in a global firm are in-terdependent. A French subsidiary’s ability to sell a product does depend on how wellother operations in other countries perform their value creation activities. Thus, thelevels of interdependence and performance ambiguity are high in global companies.

The level of performance ambiguity is highest of all in transnational firms. Transna-tional firms suffer from the same performance ambiguity problems that global firms do.In addition, since they emphasize the multidirectional transfer of core competencies,they also suffer from the problems characteristic of firms pursuing an internationalstrategy. The extremely high level of integration within transnational firms implies ahigh degree of joint decision making, and the resulting interdependencies createplenty of alibis for poor performance. There is lots of room for finger-pointing intransnational firms.

Implications for Control and Incentives

The arguments of the previous section, along with the implications for the costs ofcontrol, are summarized in Table 13.1. The costs of control can be defined as theamount of time top management must devote to monitoring and evaluating subunits’performance. This is greater when the amount of performance ambiguity is greater.When performance ambiguity is low, management can use output controls and a sys-tem of management by exception; when it is high, managers have no such luxury.Output controls do not provide totally unambiguous signals of a subunit’s efficiencywhen the performance of that subunit is dependent on the performance of anothersubunit within the organization. Thus, management must devote time to resolvingthe problems that arise from performance ambiguity, with a corresponding rise in thecosts of control.

Table 13.1 reveals a paradox. We saw in Chapter 12 that a transnational strategyis desirable because it gives a firm more ways to profit from international expansionthan do multidomestic, international, and global strategies. But now we see that dueto the high level of interdependence, the costs of controlling transnational firms arehigher than the costs of controlling firms that pursue other strategies. Unless there issome way of reducing these costs, the higher profitability associated with a transnational


Table 13.1

Interdependence,Performance Ambiguity,and the Costs of Controlfor the Four InternationalBusiness Strategies

Performance

Strategy Interdependence Ambiguity Costs of Control

Multidomestic Low Low Low

International Moderate Moderate Moderate

Global High High High

Transnational Very high Very high Very high





strategy could be canceled out by the higher costs of control. The same point, al-though to a lesser extent, can be made with regard to global firms. Although firmspursuing a global strategy can reap the cost benefits of location and experience curveeconomies, they must cope with a higher level of performance ambiguity, and thisraises the costs of control (in comparison with firms pursuing an international or mul-tidomestic strategy).

This is where control systems and incentives come in. When we survey the systemsthat corporations use to control their subunits, we find that irrespective of their strat-egy, multinational firms all use output and bureaucratic controls. However, in firmspursuing either global or transnational strategies, the usefulness of output controls islimited by substantial performance ambiguities. As a result, these firms place greateremphasis on cultural controls. Cultural control—by encouraging managers to want toassume the organization’s norms and value systems—gives managers of interdependentsubunits an incentive to look for ways to work out problems that arise between them.The result is a reduction in finger-pointing and, accordingly, in the costs of control.The development of cultural controls may be a precondition for the successful pursuitof a transnational strategy and perhaps of a global strategy as well.20 As for incentives,the material discussed earlier suggests that the conflict between different subunits canbe reduced and the potential for cooperation enhanced, if incentive systems are tiedin some way to a higher level in the hierarchy. When performance ambiguity makes itdifficult to judge the performance of subunits as stand-alone entities, linking the in-centive pay of senior managers to the entity to which both subunits belong can reducethe resulting problems.

We defined processes as the manner in which decisions are made and work is performedwithin the organization.21 Processes can be found at many different levels within anorganization. There are processes for formulating strategy, processes for allocating re-sources, processes for evaluating new product ideas, processes for handling customerinquiries and complaints, processes for improving product quality, processes for evalu-ating employee performance, and so on. Often, the core competencies or valuableskills of a firm are embedded in its processes. Efficient and effective processes can lowerthe costs of value creation and add additional value to a product. For example, theglobal success of many Japanese manufacturing enterprises in the 1980s was based inpart on their early adoption of processes for improving product quality and operatingefficiency, including total quality management and just-in-time inventory systems. To-day, the competitive success of General Electric can in part be attributed to a numberof processes that have been widely promoted within the company. These include thecompany’s six-sigma process for quality improvement, its process for “digitalization” ofbusiness (using corporate intranets and the Internet to automate activities and reduceoperating costs), and its process for new idea generation, referred to within the com-pany as “workouts,” where managers and employees get together for intensive sessionsover several days to identify and commit to ideas for improving productivity.

An organization’s processes can be summarized by means of a flow chart, which il-lustrates the various steps and decision points involved in performing work. Manyprocesses cut across functions, or divisions, and require cooperation between individ-uals in different subunits. For example, product development processes require em-ployees from R&D, manufacturing, and marketing to work together in a cooperativemanner to make sure new products are developed with market needs in mind and de-signed in such a way that they can be manufactured at a low cost. Because they cutacross organizational boundaries, performing processes effectively often requires theestablishment of formal integrating mechanisms and incentives for cross-unit cooper-ation (see above).


Processes





A detailed consideration of the nature of processes and strategies for process im-provement and reengineering is beyond the scope of this book. However, it is impor-tant to make two basic remarks about managing processes, particularly in the contextof an international business.22 The first is that in a multinational enterprise, manyprocesses cut not only across organizational boundaries, embracing several differentsubunits, but also across national boundaries. Designing a new product may require thecooperation of R&D personnel located in California, production people located inTaiwan, and marketing located in Europe, America, and Asia. The chances of pullingthis off are greatly enhanced if the processes are embedded in an organizational cul-ture that promotes cooperation between individuals from different subunits and na-tions, if the incentive systems of the organization explicitly reward such cooperation,and if formal and informal integrating mechanisms are used to facilitate coordinationbetween subunits.

Second, it is particularly important for a multinational enterprise to recognize thatvaluable new processes that might lead to a competitive advantage can be developedanywhere within the organization’s global network of operations. New processes maybe developed by a local operating subsidiary in response to conditions pertaining to itsmarket. Those processes might then have value to other parts of the multinational en-terprise. For example, in response to competition in Japan and a local obsession withproduct quality, Japanese firms were at the leading edge of developing processes for to-tal quality management (TQM) in the 1970s. Because few American firms had Japa-nese subsidiaries at the time, they were relatively ignorant of the trend until the 1980swhen high-quality Japanese products began to make big inroads into the UnitedStates. An exception to this generalization was Hewlett-Packard, which had a verysuccessful operating company in Japan, Yokogwa Hewlett-Packard (YHP). YHP was apioneer of the total quality management process in Japan and won the prestigiousDeming Prize for its achievements in improving product quality. Through YHP,Hewlett-Packard learned about the quality movement ahead of many of its U.S. peersand was one of the first Western companies to introduce TQM processes into itsworldwide operations. Not only did Hewlett-Packard’s Japanese operation give thecompany access to a valuable process, but the company also transferred this knowledgewithin its global network of operations, raising the performance of the entire company.The ability to create valuable processes matters, but it is also important to leveragethose processes. This requires both formal and informal integrating mechanisms suchas management networks.

Chapter 3 applied the concept of culture to nation-states. Culture, however, is a socialconstruct ascribed to societies, including organizations.23 Thus, we can speak of orga-nizational culture and organizational subculture. The basic definition of culture re-mains the same, whether we are applying it to a large society such as a nation-state ora small society such as an organization or one of its subunits. Culture refers to a systemof values and norms that are shared among people. Values are abstract ideas about whata group believes to be good, right, and desirable. Norms mean the social rules andguidelines that prescribe appropriate behavior in particular situations. Values andnorms express themselves as the behavior patterns or style of an organization that newemployees are automatically encouraged to follow by their fellow employees. Al-though an organization’s culture is rarely static, it tends to change relatively slowly.

|How Is Organizational Culture Created and Maintained?

An organization’s culture comes from several sources. First, there seems to be wideagreement that founders or important leaders can have a profound impact on an or-


Organizational Culture





ganization’s culture, often imprinting their own values on the culture.24 This was cer-tainly the case with Lincoln Electric where the values of James Lincoln became thecore values of Lincoln Electric (see the Management Focus). Another famous exam-ple of a strong founder effect concerns the founder of the Japanese firm Matsushita,Konosuke Matsushita, whose almost Zen-like personal business philosophy was codi-fied in the “Seven Spiritual Values” of Matsushita that all new employees still learntoday. These values are (1) national service through industry, (2) fairness, (3) harmonyand cooperation, (4) struggle for betterment, (5) courtesy and humility, (6) adjust-ment and assimilation, and (7) gratitude. A leader does not have to be the founder tohave a profound influence on organizational culture. Jack Welsh is widely creditedwith having changed the culture of GE, primarily by emphasizing when he first becameCEO a countercultural set of values, such as risk taking, entrepreneurship, steward-ship, and boundaryless behavior. It is more difficult for a leader, however forceful, tochange an established organizational culture than it is to create one from scratch in anew venture.

Another important influence on organizational culture is the broader social cultureof the nation where the firm was founded. In the United States, for example, the com-petitive ethic of individualism looms large and there is enormous social stress on pro-ducing winners. Many American firms find ways of rewarding and motivatingindividuals so that they see themselves as winners.25 The values of American firms of-ten reflect the values of American culture. Similarly, the cooperative values found inmany Japanese firms have been argued to reflect the values of traditional Japanese so-ciety, with its emphasis on group cooperation, reciprocal obligations, and harmony.26

Thus, although it may be a generalization, there may be something to the argumentthat organizational culture is influenced by national culture.

A third influence on organizational culture is the history of the enterprise, whichover time may come to shape the values of the organization. In the language of histo-rians, organizational culture is the path-dependent product of where the organizationhas been through time. For example, Philips NV, the Dutch multinational, long oper-ated with a culture that placed a high value on the independence of national operat-ing companies. This culture was shaped by the history of the company. During WorldWar II, Holland was occupied by the Germans. With the head office in occupied ter-ritories, power was devolved by default to various foreign operating companies, suchas Philips subsidiaries in the United States and Great Britain. After the war ended,these subsidiaries continued to operate in a highly autonomous fashion. A belief thatthis was the right thing to do became a core value of the company.

Decisions that subsequently result in high performance tend to become institu-tionalized in the values of a firm. In the 1920s, 3M was primarily a manufacturer ofsandpaper. Richard Drew, who was a young laboratory assistant at the time, came upwith what he thought would be a great new product; a glue-covered strip of paper,which he called “sticky tape.” Drew saw applications for the product in the automo-bile industry, where it could be used to mask parts of a vehicle during painting. He pre-sented the idea to the company’s president, William McKnight. An unimpressedMcKnight suggested that Drew drop the research. Drew didn’t; instead he developedthe “sticky tape” and then went out and got endorsements from potential customers inthe auto industry. Armed with this information, he approached McKnight again. Achastened McKnight reversed his position and gave Drew the go-ahead to start devel-oping what was to become one of 3M’s main product lines—sticky tape—a business itdominates to this day.27 From then on, McKnight emphasized the importance of giv-ing researchers at 3M free rein to explore their own ideas and experiment with prod-uct offerings. This soon became a core value at 3M and was enshrined in the company’sfamous “15 percent rule,” which stated that researchers could spend 15 percent of thecompany time working on ideas of their own choosing. Today, new employees are of-ten told the Drew story, which is used to illustrate the value of allowing individuals toexplore their own ideas.






Culture is maintained by a variety of mechanisms. These include: (1) hiring andpromotional practices of the organization, (2) reward strategies, (3) socializationprocesses, and (4) communication strategy. The goal is to recruit people whose valuesare consistent with those of the company. Lincoln Electric, for example, hires indi-viduals who are very self-reliant, which is necessary in the company’s individualisticculture. To further reinforce values, a company may promote individuals whose be-havior is consistent with the core values of the organization. Merit review processesmay also be linked to a company’s values, which further reinforces cultural norms.Thus, at Lincoln Electric, the merit review process rewards people for behavior that isconsistent with the attainment of high productivity.

Socialization can be formal, such as training programs for employees that educatethem in the core values of the organization. Informal socialization may be friendly ad-vice from peers or bosses or may be implicit in the actions of peers and superiors to-ward new employees. As for communication strategy, many companies with strongcultures devote a lot of attention to framing their key values in corporate mission state-ments, communicating them often to employees, and using them to guide difficult de-cisions. Stories and symbols are often used to reinforce important values (e.g., theDrew and McKnight story at 3M).

|Organizational Culture and Performance

in the International Business

Management authors often talk about “strong cultures.”28 In a strong culture, almostall managers share a relatively consistent set of values and norms that have a clear im-pact on the way work is performed. New employees adopt these values very quickly,and employees that do not fit in with the core values tend to leave. In such a culture,a new executive is just as likely to be corrected by his subordinates as by his superiorsif he violates the values and norms of the organizational culture. Firm’s with a strongculture are normally seen by outsiders as having a certain style or way of doing things.Lincoln Electric, profiled in the Management Focus, is an example of a firm with astrong culture. Lincoln’s organizational culture places a high value on individualachievements, meritocracy, and egalitarian behavior. Unilever, profiled in the openingcase, is another example of a firm with a strong culture. Unilever places a high value onsociability, cooperation, and consensus-building behavior among its employees.

Strong does not necessarily mean good. A culture can be strong but bad. The cultureof the Nazi Party in Germany was certainly strong, but it was most definitely not good.Nor does it follow that a strong culture leads to high performance. One study foundthat General Motors had a “strong culture,” but it was a strong culture that discour-aged lower-level employees from demonstrating initiative and taking risks, which theauthors argued was dysfunctional and led to low performance at GM.29 Also, a strongculture might be beneficial at one point, leading to high performance, but inappropri-ate at another time. The appropriateness of the culture depends on the context. In the1970s and early 1980s, when IBM was performing very well, several management au-thors sang the praises of its strong culture, which among other things placed a highvalue on consensus-based decision making.30 These authors argued that such adecision-making process was appropriate given the substantial financial investmentsthat IBM routinely made in new technology. However, this process turned out to be aweakness in the fast-moving computer industry of the late 1980s and 1990s. Consensus-based decision making was slow, bureaucratic, and not particularly conducive to cor-porate risk taking. While this was fine in the 1970s, IBM needed rapid decision makingand entrepreneurial risk taking in the 1990s, but its culture discouraged such behav-ior. IBM found itself outflanked by then-small enterprises such as Microsoft and Com-paq Computer.

One academic study concluded that firms that exhibited high performance over aprolonged period tended to have strong but “adaptive cultures.” According to this






study, in an adaptive culture most managers care deeply about and value customers,stockholders, and employees. They also strongly value people and processes that cre-ate useful change in a firm.31 While this is interesting, it does reduce the issue to a veryhigh level of abstraction; after all, what company would say that it doesn’t care deeplyabout customers, stockholders, and employees? A somewhat different perspective is toargue that the culture of the firm must match the rest of the architecture of the orga-nization, the firm’s strategy, and the demands of the competitive environment, for su-perior performance to be attained. All these elements must be consistent with eachother. Lincoln Electric provides another useful example. Lincoln competes in a busi-ness that is very competitive, where cost minimization is a key source of competitiveadvantage. Lincoln’s culture and incentive systems both encourage employees to strivefor high levels of productivity, which translates into the low costs that are critical forLincoln’s success.

The Lincoln example also demonstrates another important point for internationalbusinesses: A culture that leads to high performance in the firm’s home nation may not beeasy to impose on foreign subsidiaries! Lincoln’s culture has clearly helped the firm toachieve superior performance in the U.S. market, but this same culture is very “Amer-ican” in its form and difficult to implement in other countries. The managers and em-ployees of several of Lincoln’s European subsidiaries found the culture to be alien totheir own values and were reluctant to adopt it. The result was that Lincoln found itvery difficult to replicate in foreign markets the success it has had in the United States.Lincoln compounded the problem by acquiring established enterprises that alreadyhad their own organizational culture. Thus, in trying to impose its culture on foreignoperating subsidiaries, Lincoln had to deal with two problems: how to change the es-tablished organizational culture of those units, and how to introduce an organizationalculture whose key values might be alien to the values held by members of that society.These problems are not unique to Lincoln; many international businesses have to dealwith exactly the same problems.

The solution Lincoln has adopted is to establish new subsidiaries, rather than ac-quiring and trying to transform an enterprise with its own culture. It is much easier toestablish a set of values in a new enterprise than it is to change the values of an estab-lished enterprise. A second solution is to devote a lot of time and attention to trans-mitting the firm’s organizational culture to its foreign operations. This was somethingLincoln originally omitted. Other firms make this an important part of their strategyfor internationalization. When MTV Networks opens an operation in a new country,it initially staffs that operation with several expatriates. The job of these expatriates isto hire local employees whose values are consistent with the MTV culture and to so-cialize those individuals into values and norms that underpin MTV’s unique way of do-ing things. Once this has been achieved, the expatriates move on to their nextassignment, and local employees run the operation. A third solution is to recognizethat it may be necessary to change some aspects of a firm’s culture so that it better fitsthe culture of the host nation. For example, many Japanese firms use symbolic behav-ior, such as company songs and morning group exercise sessions, to reinforce coopera-tive values and norms. However, such symbolic behavior does not go down well inWestern cultures, where it is seen as odd, so many Japanese firms have not used suchpractices in Western subsidiaries.

The need for a common organizational culture that is the same across a multina-tional’s global network of subsidiaries probably varies with the strategy of the firm.Shared norms and values can facilitate coordination and cooperation between indi-viduals from different subunits.32 A strong common culture may lead to goal congru-ence and can attenuate the problems that arise from interdependence, performanceambiguities, and conflict among managers from different subsidiaries. As noted ear-lier, a shared culture may help informal integrating mechanisms such as managementnetworks to operate more effectively. As such, a common culture may be of greatervalue in a multinational that is pursuing a strategy that requires cooperation and






coordination between globally dispersed subsidiaries. This suggests that it is more im-portant to have a common culture in firms employing a transnational strategy than amultidomestic strategy, with global and international strategies falling between thesetwo extremes.

In Chapter 12, we identified four international business strategies: multidomestic, in-ternational, global, and transnational. So far in this chapter we have looked at severalaspects of organization architecture, and we have discussed the interrelationships be-tween these dimensions and strategies. Now it is time to synthesize this material (seeTable 13.2).

|Multidomestic Firms

Firms pursuing a multidomestic strategy focus on local responsiveness. Table 13.2shows that multidomestic firms tend to operate with worldwide area structures,within which operating decisions are decentralized to functionally self-containedcountry subsidiaries. The need for coordination between subunits (areas and coun-try subsidiaries) is low. This suggests that multidomestic firms do not have a highneed for integrating mechanisms, either formal or informal, to knit together differ-ent national operations. The lack of interdependence implies that the level of per-formance ambiguity in multidomestic concerns is low, as (by extension) are the costsof control. Thus, headquarters can manage foreign operations by relying primarilyon output and bureaucratic controls and a policy of management by exception. In-centives can be linked to performance metrics at the level of country subsidiaries.Since the need for integration and coordination is low, the need for commonprocesses and organization culture is also quite low. Were it not for the fact thatthese firms are unable to profit from the realization of location and experience curve


Synthesis: Strategy and Architecture

Table 13.2

A Synthesis of Strategy, Structure, and Control Systems

Strategy

Structure

and Controls Multidomestic International Global Transnational

Vertical Decentralized Core competency Some Mixed differentiation centralized; rest centralized centralized

decentralized and decentralized

Horizontal Worldwide area Worldwide Worldwide Informal matrixdifferentiation structure product product

division division

Need for Low Moderate High Very highcoordination

Integrating None Few Many Very manymechanisms

Performance Low Moderate High Very highambiguity

Need for cultural Low Moderate High Very highcontrols





economies, or from the transfer of core competencies, their organizational simplic-ity would make this an attractive strategy.

|International Firms

Firms pursuing an international strategy attempt to create value by transferring corecompetencies from home to foreign subsidiaries. If they are diverse, as most of themare, these firms operate with a worldwide product division structure. Headquarters typ-ically maintains centralized control over the source of the firm’s core competency,which is most typically found in the R&D and/or marketing functions of the firm. Allother operating decisions are decentralized within the firm to subsidiary operations ineach country (which in diverse firms report to worldwide product divisions).

The need for coordination is moderate in such firms, reflecting the need to trans-fer core competencies. Thus, although such firms operate with some integratingmechanisms, they are not that extensive. The relatively low level of interdependencethat results translates into a relatively low level of performance ambiguity. Thesefirms can generally get by with output and bureaucratic controls and with incentivesthat are focused on performance metrics at the level of country subsidiaries. The needfor a common organizational culture and common processes is not that great. An im-portant exception to this is when the core skills or competencies of the firm are em-bedded in processes and culture, in which case the firm needs to pay close attentionto transferring those processes and associated culture from the corporate center tocountry subsidiaries. Overall, although the organization of international firms is morecomplex than that of multidomestic firms, the increase in the level of complexity isnot that great.

|Global Firms

Firms pursuing a global strategy focus on the realization of location and experiencecurve economies. If they are diverse, as most of them are, these firms operate with aworldwide product division structure. To coordinate the firm’s globally dispersed webof value creation activities, headquarters typically maintains ultimate control overmost operating decisions. In general, global firms are more centralized than enterprisespursuing a multidomestic or international strategy. Reflecting the need for coordina-tion of the various stages of the firms’ globally dispersed value chains, the need for in-tegration in these firms also is high. Thus, these firms tend to operate with an array offormal and informal integrating mechanisms. The resulting interdependencies canlead to significant performance ambiguities. As a result, in addition to output and bu-reaucratic controls, global firms tend to stress the need to build a strong organizationalculture that can facilitate coordination and cooperation. They also tend to use incen-tive systems that are linked to performance metrics at the corporate level, giving themanagers of different operations a strong incentive to cooperate with each other to in-crease the performance of the entire corporation. On average, the organization ofglobal firms is more complex than that of multidomestic and international firms.

|Transnational Firms

Firms pursuing a transnational strategy focus on the simultaneous attainment of loca-tion and experience curve economies, local responsiveness, and global learning (themultidirectional transfer of core competencies or skills). These firms may operate withmatrix-type structures in which both product divisions and geographic areas have sig-nificant influence. The need to coordinate a globally dispersed value chain and totransfer core competencies creates pressures for centralizing some operating decisions(particularly production and R&D). At the same time, the need to be locally respon-sive creates pressures for decentralizing other operating decisions to national opera-tions (particularly marketing). Consequently, these firms tend to mix relatively high






degrees of centralization for some operating decisions with relative high degrees of de-centralization for other operating decisions.

The need for coordination is particularly high in transnational firms. This is re-flected in the use of an array of formal and informal integrating mechanisms, includ-ing formal matrix structures and informal management networks. The high level ofinterdependence of subunits implied by such integration can result in significant per-formance ambiguities, which raise the costs of control. To reduce these, in addition tooutput and bureaucratic controls, transnational firms need to cultivate a strong cultureand to establish incentives that promote cooperation between subunits.

|Environment, Strategy, Architecture, and Performance

Underlying the scheme outlined in Table 13.2 is the notion that a “fit” between strat-egy and architecture is necessary for a firm to achieve high performance. For a firm tosucceed, two conditions must be fulfilled. First, the firm’s strategy must be consistentwith the environment in which the firm operates. We discussed this issue in Chapter12 and noted that in some industries a global strategy is most viable, in others an in-ternational or transnational strategy may be most viable, and in still others a multido-mestic strategy may be most viable (although the number of multidomestic industriesis on the decline). Second, the firm’s organization architecture must be consistent withits strategy.

If the strategy does not fit the environment, the firm is likely to experience signif-icant performance problems. If the architecture does not fit the strategy, the firm is alsolikely to experience performance problems. Therefore, to survive, a firm must strive toachieve a fit of its environment, its strategy, and its organizational architecture. Youwill recall that we saw the importance of this concept in the opening case. Philips NV,the Dutch electronics firm, provides another illustration of the need for this fit. Forreasons rooted in the history of the firm, Philips operated until recently with an orga-nization typical of a multidomestic enterprise in which operating decisions were de-centralized to largely autonomous foreign subsidiaries. Historically, electronicsmarkets were segmented from each other by high trade barriers, so an organizationconsistent with a multidomestic strategy made sense. However, by the mid-1980s, theindustry in which Philips competed had been revolutionized by declining trade barri-ers, technological change, and the emergence of low-cost Japanese competitors thatutilized a global strategy. To survive, Philips needed to adopt a global strategy itself.The firm recognized this and tried to adopt a global posture, but it did little to changeits organizational architecture. The firm nominally adopted a matrix structure basedon worldwide product divisions and national areas. In reality, however, the nationalareas continued to dominate the organization, and the product divisions had littlemore than an advisory role. As a result, Philips’ architecture did not fit the strategy,and by the early 1990s Philips was losing money. It was only after four years of wrench-ing change and large losses that Philips was finally able to tilt the balance of power inits matrix toward the product divisions. By 1995, the fruits of this effort to realign thecompany’s strategy and architecture with the demands of its operating environmentwere beginning to show up in improved financial performance.33

Multinational firms periodically have to alter their architecture so that it conforms tothe changes in the environment in which they are competing and the strategy they arepursuing. To be profitable, Philips NV had to alter its strategy and architecture in the1990s so that both matched the demands of the competitive environment in the elec-tronics industry, which had shifted from a multidomestic to a global industry. While adetailed consideration of organizational change is beyond the scope of this book, a few


Organizational Change





comments are warranted regarding the sources of organization inertia and the strate-gies and tactics for implementing organizational change.

|Organizational Inertia

Organizations are difficult to change. Within most organizations, there are stronginertia forces. These forces come from a number of sources. One source of inertia isthe existing distribution of power and influence within an organization.34 Thepower and influence enjoyed by individual managers is in part a function of theirrole in the organizational hierarchy, as defined by structural position. By definition,most substantive changes in an organization require a change in structure and, byextension, a change in the distribution of power and influence within the organiza-tion. Some individuals will see their power and influence increase as a result of or-ganizational change, and some will see the converse. For example, in the 1990s,Philips NV increased the roles and responsibilities of its global product divisions anddecreased the roles and responsibilities of its foreign subsidiary companies. Thismeant the managers running the global product divisions saw their power and in-fluence increase, while the managers running the foreign subsidiary companies sawtheir power and influence decline. As might be expected, some managers of foreignsubsidiary companies did not like this change and resisted it, which slowed the paceof change. Those whose power and influence are reduced as a consequence of orga-nizational change can be expected to resist it, primarily by arguing that the changemight not work. To the extent that they are successful, this constitutes a source oforganizational inertia that might slow or stop change.

Another source of organizational inertia is the existing culture, as expressed in normsand value systems. Value systems reflect deeply held beliefs, and as such, they can bevery hard to change. If the formal and informal socialization mechanisms within an or-ganization have been emphasizing a consistent set of values for a prolonged period, andif hiring, promotion, and incentive systems have all reinforced these values, then sud-denly announcing that those values are no longer appropriate and need to be changedcan produce resistance and dissonance among employees. For example, Philips NV his-torically placed a very high value on local autonomy. The changes of the 1990s implieda reduction in the autonomy enjoyed by foreign subsidiaries, which was counter to theestablished values of the company and thus resisted.

Organizational inertia might also derive from senior managers’ preconceptionsabout the appropriate business model or paradigm. When a given paradigm has workedwell in the past, managers might have trouble accepting that it is no longer appropri-ate. At Philips, granting considerable autonomy to foreign subsidiaries had workedvery well in the past, allowing local managers to tailor product and business strategyto the conditions prevailing in a given country. Since this paradigm had worked sowell, it was difficult for many managers to understand why it no longer applied. Con-sequently, they had difficulty accepting a new or business model and tended to fallback on their established paradigm and ways of doing things. This made change diffi-cult, for it required managers to let go of long-held assumptions about what workedand what didn’t work, which was something many of them couldn’t do.

Institutional constraints might also act as a source of inertia. National regulationsincluding local content rules and policies pertaining to layoffs might make it difficultfor a multinational to alter its global value chain. As with Unilever (see the openingcase), a multinational might wish to take control for manufacturing away from localsubsidiaries, transfer that control to global product divisions, and consolidate manu-facturing at a few choice locations. However, if local content rules (see Chapter 5) re-quire some degree of local production and if regulations regarding layoffs make itdifficult or expensive for a multinational to close operations in a country, a multina-tional may find that these constraints make it very difficult to adopt the most effectivestrategy and architecture.






|Implementing Organizational Change

Although all organizations suffer from inertia, the complexity and global spread ofmany multinationals might make it particularly difficult for them to change their strat-egy and architecture to match new organizational realities. Yet at the same time, thetrend toward globalization in many industries has made it more critical than ever thatmany multinationals do just that. In industry after industry, declining barriers to cross-border trade and investment have led to a change in the nature of the competitive en-vironment. Cost pressures have increased, requiring multinationals to respond bystreamlining their operations to realize economic benefits associated with location andexperience curve economies and with the transfer of competencies and skills withinthe organization. At the same time, local responsiveness remains an important sourceof differentiation. To survive in this emerging competitive environment, multina-tionals must not only change their strategy, but they must also change their architec-ture so that it matches strategy in discriminating ways. The basic principles forsuccessful organizational change can be summarized as follows: (1) unfreeze the orga-nization through shock therapy, (2) move the organization to a new state throughproactive change in the architecture, and (3) refreeze the organization in its new state.

Unfreezing the Organization

Because of inertia forces, incremental change is often no change. Those whose poweris threatened by change can too easily resist incremental change. This leads to the bigbang theory of change, which maintains that effective change requires taking bold ac-tion early to “unfreeze” the established culture of an organization and to change thedistribution of power and influence. Shock therapy to unfreeze the organization mightinclude the closure of plants deemed uneconomic or the announcement of a dramaticstructural reorganization. It is also important to realize that change will not occur un-less senior managers are committed to it. Senior managers must clearly articulate theneed for change so employees understand both why it is being pursued and the bene-fits that will flow from successful change. Senior managers must also practice whatthey preach and take the necessary bold steps. If employees see senior managerspreaching the need for change but not changing their own behavior or making sub-stantive changes in the organization, they will soon lose faith in the change effort,which will flounder as a result.

Moving to the New State

Once an organization has been unfrozen, it must be moved to its new state. Movementrequires taking action—closing operations; reorganizing the structure; reassigning re-sponsibilities; changing control, incentive, and reward systems; redesigning processes;and letting people go who are seen as an impediment to change. In other words, move-ment requires a substantial change in the form of a multinational’s organization ar-chitecture so that it matches the desired new strategic posture. For movement to besuccessful, it must be done with sufficient speed. Involving employees in the changeeffort is an excellent way to get them to appreciate and buy into the needs for changeand to help with rapid movement. For example, a firm might delegate substantial re-sponsibility for designing operating processes to lower-level employees. If enough oftheir recommendations are then acted on, the employees will see the consequences oftheir efforts and consequently buy into the notion that change is really occurring.

Refreezing the Organization

Refreezing the organization takes longer. It may require that a new culture be estab-lished, while the old one is being dismantled. Thus, refreezing requires that employeesbe socialized into the new way of doing things. Companies will often use managementeducation programs to achieve this. At General Electric, where longtime CEO JackWelsh instituted a major change in the culture of the company, management educa-






tion programs were used as a proactive tool to communicate new values to organiza-tion members. On their own, however, management education programs are notenough. Hiring policies must be changed to reflect the new realities, with an empha-sis on hiring individuals whose own values are consistent with that of the new culturethe firm is trying to build. Similarly, control and incentive systems must be consistentwith the new realities of the organization, or change will never take. Senior manage-ment must recognize that changing culture takes a long time. Any letup in the pres-sure to change may allow the old culture to reemerge as employees fall back intofamiliar ways of doing things. The communication task facing senior managers, there-fore, is a long-term endeavor that requires managers to be relentless and persistent intheir pursuit of change. One striking feature of Jack Welsh’s two-decade tenure at GE,for example, is that he never stopped pushing his change agenda. It was a consistenttheme of his tenure. He was always thinking up new programs and initiatives to keeppushing the culture of the organization along the desired trajectory.


Chapter Summary

This chapter identified the organizational architecturethat can be used by multinational enterprises to manageand direct their global operations. A central theme ofthe chapter was that different strategies require differentarchitectures; strategy is implemented through architec-ture. To succeed, a firm must match its architecture toits strategy in discriminating ways. Firms whose archi-tecture does not fit their strategic requirements will ex-perience performance problems. It is also necessary forthe different components of architecture to be consis-tent with each other. The following points were made inthe chapter:

1. Organizational architecture refers to the totalityof a firm’s organization, including formal organi-zational structure, control systems and incentives,processes, organizational culture, and people.

2. Superior enterprise profitability requires threeconditions to be fulfilled: the different elementsof a firm’s organizational architecture must be in-ternally consistent, the organizational architec-ture must fit the strategy of the firm, and the strat-egy and architecture of the firm must beconsistent with competitive conditions prevail-ing in the firm’s markets.

3. Organizational structure means three things:the formal division of the organization into sub-units (horizontal differentiation), the locationof decision-making responsibilities within thatstructure (vertical differentiation), and the es-tablishment of integrating mechanisms.

4. Control systems are the metrics used to measurethe performance of subunits and make judgmentsabout how well managers are running thosesubunits.

5. Incentives refer to the devices used to reward ap-propriate employee behavior. Many employees

receive incentives in the form of annual bonuspay. Incentives are usually closely tied to the per-formance metrics used for output controls.

6. Processes refer to the manner in which decisionsare made and work is performed within the or-ganization. Processes can be found at many dif-ferent levels within an organization. The corecompetencies or valuable skills of a firm are of-ten embedded in its processes. Efficient and ef-fective processes can help to lower the costs ofvalue creation and to add additional value to aproduct.

7. Organizational culture refers to a system of val-ues and norms that is shared among employees.Values and norms express themselves as the be-havior patterns or style of an organization thatnew employees are automatically encouraged tofollow by their fellow employees.

8. There are four main dimensions of organizationalstructure: vertical differentiation, horizontal dif-ferentiation, integration, and control systems.

9. Firms pursuing different strategies must adopt adifferent architecture to implement those strate-gies successfully. Firms pursuing multidomestic,global, international, and transnational strate-gies all must adopt an organizational architec-ture that matches their strategy.

10. While all organizations suffer from inertia, thecomplexity and global spread of many multina-tionals might make it particularly difficult forthem to change their strategy and architectureto match new organizational realities. At thesame time, the trend toward globalization inmany industries has made it more critical thanever that many multinationals do just that.






Critical Discussion Questions

1. “The choice of strategy for a multinational firmmust depend on a comparison of the benefits ofthat strategy (in terms of value creation) with thecosts of implementing it (as defined by organiza-tional architecture necessary for implementa-tion). On this basis, it may be logical for somefirms to pursue a multidomestic strategy, others aglobal or international strategy, and still others atransnational strategy.” Is this statement correct?

2. Discuss this statement: “An understanding of thecauses and consequences of performance ambigu-ity is central to the issue of organizational designin multinational firms.”

3. Describe the organizational architecture a transna-tional firm might adopt to reduce the costs ofcontrol.

4. What is the most appropriate organizational ar-chitecture for a firm that is competing in an in-dustry where a global strategy is most appropriate?

5. If a firm is changing its strategy from an interna-tional to a transnational strategy, what are themost important challenges it is likely to face inimplementing this change? How can the firmovercome these challenges?

Clo

sin

g C

as

e Organizational Change at Royal Dutch/Shell

The Anglo-Dutch company Royal Dutch/Shellis the world’s largest nonstate-owned oil com-pany with activities in more than 130 countriesand 1997 revenues of $128 billion. From the1950s until 1994, Shell operated with a “matrixstructure” invented for it by McKinsey, a man-agement consulting firm that specializes in orga-nizational design. Under this matrix structure,

the head of each operating company reported to twobosses. One boss was responsible for the geo-graphical region or country in which the op-erating company was based, while the otherwas responsible for the business activity thatthe operating company was engaged in(Shell’s business activities included oil ex-ploration and production, oil products,chemicals, gas, and coal). Thus, for example,the head of the local Shell chemical com-pany in Australia reported both to the headof Shell Australia and to the head of Shell’sentire chemical division, who was based inLondon. Both bosses had equal influenceand status within the organization.

This matrix structure had two very visibleconsequences at Shell. First, because each operatingcompany had two bosses to satisfy, decision making typ-ically followed a pattern of consensus building, with dif-ferences of perspective between country (or regional)heads on the one hand and the heads of business divi-sions on the other being worked out through debate. Al-though this process could be slow and cumbersome, itwas seen as a good thing in the oil industry where mostbig decisions are long-term ones that involve substantial

capital expenditures and where informed debate be-tween different viewpoints can clarify the pros and consof issues, rather than hinder decision making. Second,because the decision-making process was slow, it was re-served for only the most important decisions (such asmajor new capital investments). The result was substan-tial decentralization by default to the heads of the indi-vidual operating companies, who were largely left aloneto run their own operations. This decentralization

helped Shell respond to local differences ingovernment regulations, competitive condi-tions, and consumer tastes. Thus, for exam-ple, the head of Shell’s Australian chemicalcompany was given the freedom to deter-mine pricing practices and marketing strat-egy in the Australian market. Only if Shellwished to undertake a major capital invest-ment, such as building a new chemical plant, would the consensus-building decision-making system be invoked.

As desirable as this matrix structureseemed to many, in 1995 Shell announced aradical plan to dismantle it. The primaryreason given by top management for the

shift was continuing slack demand for oil and weak oilprices, which had put pressure on Shell’s profit margins.Although Shell had traditionally been among the mostprofitable oil companies in the world, in the early 1990sits relative performance began to slip as other oil compa-nies, such as Exxon, adapted more rapidly to a world oflow oil prices by sharply cutting overhead costs and con-solidating production in efficient facilities. Consolidat-ing production at these companies often involved serving





the world market from a smaller number of large-scale re-fining facilities and shutting down smaller facilities. Incontrast, Shell still operated with a large head office,which was required to effect coordination within Shell’smatrix structure, and substantial duplication of oil andchemical refining facilities across operating companies,each of which typically developed the facilities requiredto serve its own market.

In 1995, Shell’s senior management realized thatlowering operating costs required a sharp reduction inhead office overhead and, where appropriate, theelimination of unnecessary duplication of facilitiesacross countries. To achieve these goals, top execu-tives decided to reorganize the company along divi-sional lines. Shell now operates with five main globalproduct divisions—exploration and production, oilproducts, chemicals, gas, and coal. Each operating com-pany reports to whichever global division is the mostrelevant. Thus, the head of the Australian chemical op-eration now reports directly to the head of the globalchemical division. The thinking is that this will increase the power of the global chemical division andenable that division to eliminate any unnecessary dupli-cation of facilities across countries. Eventually, produc-tion may be consolidated in larger facilities that serve anentire region, rather than a single country, thereby en-abling Shell to reap greater scale economies.

The country (or regional) chiefs remain but theirroles and responsibilities are reduced. Now their primaryresponsibility is coordination between operating compa-nies within a country (or region) and relations with thelocal government. There is a solid line of reporting andresponsibility between the heads of operating companiesand the global divisions and only a dotted line between

the heads of operating companies and country chiefs.Thus, for example, the ability of the head of Shell Aus-tralia to shape the major capital investment decisions ofShell’s Australian chemical operation was substantiallyreduced as a result of these changes. Furthermore, thesimplified reporting system reduced the need for a largehead office bureaucracy, and Shell trimmed the workforce at its London head office by 1,170, driving downShell’s cost structure.

Sources: “Shell on the Rocks,” The Economist, June 24, 1995,pp. 57–58; D. Lascelles, “Barons Swept out of Fiefdoms,” FinancialTimes, March 30, 1995, p. 15; C. Lorenz, “End of a Corporate Era,”Financial Times, March 30, 1995, p. 15; R. Corzine, “Shell DiscoversTime and Tide Wait for No Man,” Financial Times, March 10, 1998,p. 17; and R. Corzine, “Oiling the Group’s Wheels of Change,”Financial Times, April 1, 1998, p. 12.

Case Discussion Questions

1. What were the benefits of the matrix structure atShell? What were the drawbacks? Did the matrixstructure fit the environment of the global oil andchemical industries in the 1980s?

2. What shift occurred in Shell’s operating environmentin the 1990s? How did this shift affect the financialperformance of the firm? What does this suggest aboutthe fit between strategy and architecture?

3. What kind of structure did Shell adopt in 1995? Inwhat ways did the architecture of Shell’s organiza-tion after 1995 differ from that before 1995?

4. Comment on the fit between operating environ-ment, strategy, and organizational architecture atShell after the 1995 reorganization. Did the changelead to enhanced fit?

1. D. Naidler, M. Gerstein, and R. Shaw, Organi-zation Architecture (San Francisco: Jossey-Bass,1992).

2. G. Morgan, Images of Organization (BeverlyHills, CA: Sage Publications, 1986).

3. The material in this section draws on John Child,Organizations (London: Harper & Row, 1984).

4. Allan Cane, “Microsoft Reorganizes to MeetMarket Challenges,” Financial Times, March 16,1994, p. 1.

5. For research evidence that is related to this is-sue, see J. Birkinshaw, “Entrepreneurship in theMultinational Corporation: The Characteris-tics of Subsidiary Initiatives,” Strategic Manage-ment Journal 18 (1997), pp. 207–29, and J.

Birkinshaw, N. Hood, and S. Jonsson, “Build-ing Firm Specific Advantages in MultinationalCorporations: The Role of Subsidiary Initia-tives,” Strategic Management Journal 19 (1998),pp. 221–41.

6. For more detail, see S. M. Davis, “Managing andOrganizing Multinational Corporations,” inC. A. Bartlett and S. Ghoshal, TransnationalManagement (Homewood, IL: Richard D. Irwin,1992).

7. A. D. Chandler, Strategy and Structure: Chap-ters in the History of the Industrial Enterprise(Cambridge, MA: MIT Press, 1962).

8. Davis, “Managing and Organizing Multina-tional Corporations.”

Notes







9. J. M. Stopford and L. T. Wells, Strategy and Struc-ture of the Multinational Enterprise (New York:Basic Books, 1972).

10. C. A. Bartlett and S. Ghoshal, Managing acrossBorders (Boston: Harvard Business SchoolPress, 1989).

11. Guy de Jonquieres, “Unilever Adopts a CleanSheet Approach,” Financial Times, October 21,1991, p. 13.

12. See J. R. Galbraith, Designing Complex Organi-zations (Reading, MA: Addison-Wesley, 1977).

13. See Bartlett and Ghoshal, Managing across Bor-ders, and F. V. Guterl, “Goodbye, Old Matrix,”Business Month, February 1989, pp. 32–38.

14. M. S. Granovetter, “The Strength of WeakTies,” American Journal of Sociology 78 (1973),pp. 1360–80.

15. A. K. Gupta, and V. J. Govindarajan, “Knowl-edge Flows within Multinational Corpora-tions,” Strategic Management Journal 21, no. 4(2000), pp. 473–96.

16. For examples, see W. H. Davidow and M. S.Malone, The Virtual Corporation (New York:Harper Collins, 1992).

17. W. G. Ouchi, “Markets, Bureaucracies, andClans,” Administrative Science Quarterly 25(1980), pp. 129–44.

18. For some recent empirical work that addressesthis issue, see T. P. Murtha, S. A. Lenway, andR. P. Bagozzi, “Global Mind Sets and CognitiveShift in a Complex Multinational Corpora-tion,” Strategic Management Journal 19 (1998),pp. 97–114.

19. C. W. L. Hill, M. E. Hitt, and R. E. Hoskisson,“Cooperative versus Competitive Structures inRelated and Unrelated Diversified Firms,” Or-ganization Science 3 (1992), pp. 501–21.

20. Murtha, Lenway, and Bagozzi, “Global MindSets.”

21. M. Hammer and J. Champy, Reengineering theCorporation (New York: Harper Business, 1993).

22. T. Kostova, “Transnational Transfer of Strate-gic Organizational Practices: A Contextual Per-spective,” Academy of Management Review 24,no. 2 (1999), pp. 308–24.

23. E. H. Schein, “What Is Culture?” in P. J. Frostet al., Reframing Organizational Culture (New-bury Park, CA: Sage, 1991).

24. E. H. Schein, Organizational Culture and Lead-ership, 2nd ed. (San Francisco: Jossey-Bass,1992).

25. G. Morgan, Images of Organization (BeverlyHills, CA: Sage, 1986).

26. R. Dore, British Factory, Japanese Factory (Lon-don: Allen & Unwin, 1973).

27. M. Dickson, “Back to the Future,” FinancialTimes, May 30, 1994, p. 7.

28. See J. P. Kotter and J. L. Heskett, Corporate Cul-ture and Performance (New York: Free Press,1992) and M. L. Tushman and C. A. O’Reilly,Winning through Innovation (Boston: HarvardBusiness School Press, 1997).

29. Kotter and Heskett, Corporate Culture andPerformance.

30. The classic song of praise was produced by T.Peters and R. H. Waterman, In Search of Excel-lence (New York: Harper & Row, 1982). Ironi-cally, IBM’s decline began shortly after Petersand Waterman’s book was published.

31. Kotter and Heskett, Corporate Culture andPerformance.

32. Bartlett and Ghoshal, Managing across Borders.33. See F. J. Aguilar and M. Y. Yoshino, “The Philips

Group: 1987,” Howard Business School Case,388-050, 1987; “Philips Fights Flab,” The Econ-omist, April 7, 1990, pp. 73–74; and R. Van deKrol, “Philips Wins Back Old Friends,” Finan-cial Times, July 14, 1995, p. 14.

34. J. Pfeffer, Managing with Power: Politics and In-fluence within Organizations (Boston: HarvardBusiness School Press, 1992).

International business ch13_view

Business

Transcript of International business ch13_view