Decision Making

jroero

1.-Renewal_Through_Reorganization.pdf

Home >Business & Finance homework help >Operations Management homework help >Decision Making

OrganizationScience Vol. 20, No. 2, March–April 2009, pp. 422–440 issn 1047-7039 � eissn 1526-5455 � 09 � 2002 � 0422

informs ® doi 10.1287/orsc.1090.0421

Renewal Through Reorganization: The Value of Inconsistencies Between Formal and

Informal Organization

Ranjay Gulati Harvard Business School, Harvard University, Boston, Massachusetts 02163, [email protected]

Phanish Puranam London Business School, University of London, London NW1 4SA, United Kingdom, [email protected]

We develop a theoretical perspective on how inconsistencies between formal and informal organization arising fromreorganization can help create ambidextrous organizations. We argue that under some conditions, the informal orga- nization can compensate for the formal organization by motivating a distinct but valuable form of employee behavior that the formal organization does not emphasize, and vice versa—an effect we label compensatory fit. We illustrate the concept of compensatory fit by drawing on qualitative data from a reorganization at Cisco Systems. We also derive formal boundary conditions for compensatory fit using a simple game theoretic representation. We show that compensatory fit can only work when there is a powerful informal organization already in existence, and when the gains from ambidexterity are substantial. Further, depending on the strength of the informal organization, breakdown in the conditions necessary for compensatory fit may lead to performance declines and further reorganizations.

Key words: organization structure; formal and informal organization; interunit coordination

Reorganization is an important mechanism by which corporations can renew alignment between strategy and organization (Chandler 1962). In principle, the dissolu- tion and reformation of internal organizational bound- aries allows for improved partitioning and re-integration of activity within the firm. Yet there is little doubt that reorganizations also create significant stresses and strains within organizations (Nadler and Tushman 1998, Romanelli and Tushman 1994, Tushman and Romanelli 1985). While the formal organization—the normative social system designed by managers—can be changed relatively rapidly, the informal organization—the emer- gent pattern of social interactions within organizations— may be subject to limits and lags in its adjustment to the new formal organization (Lamont and Williams 1994, Miller and Friesen 1984, Nickerson and Zenger 2002). The resulting inconsistencies between formal and informal organization are widely viewed as unavoidable and significant costs of reorganization (Amburgey et al. 1993, Oxman and Smith 2003) that can even enhance the hazards of organizational mortality (Hannan et al. 2003a, b). A common premise underlies the negative assess-

ments of post-reorganization inconsistencies: that incon- sistency detracts from internal fit between organiza- tional elements. Internal fit refers to a pattern of rein- forcing interactions between organizational elements such that one element enhances the impact on organi- zational performance of another element (Drazin and

van de Ven 1985; Milgrom and Roberts 1995; Miller 1992; Nadler and Tushman 1997; Siggelkow, 2001, 2002).1 If reorganizations inevitably result in incon- sistencies, and inconsistencies inevitably detract from internal fit, then one must simply accept this as the price of corporate renewal through reorganization. But what if the crucial premise—that reinforcing effects between organizational elements can only arise when they are consistent—is untrue?2 In this paper, we present a theory that outlines how the very inconsistencies that arise between the formal and informal organiza- tion after reorganizations can be the basis for corporate renewal through the pursuit of “dualities”—jointly desir- able, but organizationally incompatible objectives (Evans and Doz 1989). Instances of dualities abound in the literature on organizations and strategy—exploration- exploitation, cost reduction-product differentiation, orga- nizational differentiation-integration, and static–dynamic efficiency (Ghemawat and Costa 1993, Lawrence and Lorsch 1967, March 1991), to note a few. Even when they are jointly desirable, the organizational attributes that underlie one pole of the duality are typically seen as being incompatible with those that underlie the other (Evans and Doz 1989). For instance, the incentive and coordination mechanisms needed to support opportunity exploration differ significantly from those needed for exploitation (Brown and Eisenhardt 1997, Burns and Stalker 1961, Ghemawat and Costa 1993, Levinthal and

422

March 1993). Similarly, it is often argued that differ- ences in the way production, marketing, and HR should be organized for cost focus and differentiation-based strategies (Porter 1985) constrains their simultaneous pursuit (Besanko et al. 2000, Ebben and Johnson 2005, Ghemawat and Costa 1993). We propose that reorganizations can help solve the

problem of organizational incompatibility and allow the pursuit of jointly desirable dualities when the result- ing formal organization and the informal organization each emphasize opposing poles of a duality. We term this an instance of “compensatory” fit between the for- mal and informal organization, in which they compen- sate for each other by motivating dissimilar but jointly valuable employee behaviors. This is distinct from what one might call “supplementary fit,” the more familiar instance of superior performance through the formal and informal organization emphasizing the same set of employee behaviors. Of course, not all inconsistency is valuable—it is only so in the specific case when the for- mal and informal organization are inconsistent because they support different poles of a duality. Our argu- ment is that reorganizations can create such a situation under some boundary conditions. Both supplementary and compensatory fits belong to the broad category of complementarities between organizational elements (or synergies), such that the level of one enhances the marginal value of the other (Kogut and Zander 1996, Milgrom and Roberts 1990, Milgrom and Roberts 1995, Siggelkow 2002). However, achieving supplementary fit relies on one element augmenting the other by “pushing in the same direction,” while compensatory fit relies on one element “pushing in a different direction” in order to make up for the weakness of the other. We illustrate the notion of compensatory fit through an

in depth discussion of a reorganization that occurred in 2001 at Cisco Systems, which appeared to have helped this company renew its capability base to pursue the cost-differentiation duality. We then revisit the ideas described in the illustration of the Cisco reorganization in a more rigorous and general theoretical framework. We draw on a simple game theoretic representation to state precisely the conditions under which compensatory fit through organizational inconsistencies can exist, as well as the consequences if these conditions break down. We find that while it is an attractive means to exploit gains from ambidexterity, compensatory fit is not for every company. Our analysis indicates that it can only work when there is a powerful informal organization already in existence, and when the gains from ambidex- terity are substantial, in order to avoid inefficiencies arising from having employees work in an inconsistent organizational architecture. Further, breakdown in the conditions necessary for compensatory fit may lead to performance declines and the need for more reorganiza- tions, depending on the resulting strength of the informal

organization. The formalization also serves to make our underlying assumptions transparent to other scholars for refining or developing further, and allows us to state our results with precision. The approach we take extends the work of other

authors who have analyzed the relationships between for- mal and informal organization after reorganizations, and their joint consequences for performance (e.g., Hannan et al. 2003a, b; Nickerson and Zenger 2002). Distinc- tively, our analysis proposes that the consequences of the post-reorganization relationship between formal and informal organization hinge on external conditions. Regardless of the internal difficulty of organizing to pur- sue dualities, there are external circumstances largely out of the control of the individual firm, when doing both improves performance—when there are “gains from ambidexterity” (Birkinshaw and Gibson 2004, Smith and Tushman 2005, Tushman et al. 2004), and others in which it is better to focus on one of the poles of the duality—when there are “gains from focus” (Porter 1985, Rust et al. 2002). We propose that reorganiza- tion provides one of the ways in which fit with the external environmental conditions (as represented by the gains from ambidexterity or focus) can be achieved— even while generating apparent inconsistency between the formal and informal organizations.

Reorganization at CISCO Systems To preview the logic of our arguments (which we develop in greater detail in the next section), we describe briefly an instance of a reorganization that took place in 2001 at Cisco Systems. Cisco’s products enable voice, video, and data to travel across computer networks and lie at the heart of the Internet and the intranets of most corporate, public, and educational institutions across the world. From 1997 to 2001, Cisco was organized along three semiautonomous lines of business (Figure 1), each focusing on a distinct customer type: Internet ser- vice providers (ISPs), enterprises (large companies), and small- and medium-sized businesses (SMBs). Within this structure, each of the three lines of business devel- oped and marketed its own products to its specific cus- tomer groups, and a complete product line was built for each customer group. Each line of business had its own sales group, which mirrored the separate organizations. This structure enabled the company to build capabilities for being responsive to the idiosyncratic needs of dif- ferent customer types. Justifying this structure, a senior manager said, “I think the market was expanding like a rocket ship, so I think the good part about that organi- zation was to allow the various business units to really focus on their primary customers, who were really driv- ing the requirements.”3

By 2001, however, changes in the external environ- ment led to a reassessment of the existing structure. The

Figure 1 Cisco Systems Before August 2001

Organizational structure (pre-August 2001):

Service provider line of business

Service provider (SP) sales

Enterprise line of business

Small/medium business line of business

Small/medium business sales

Enterprise sales

explosive growth in the sales of hardware supporting the Internet began to slow by 1999–2000. Startup compa- nies’ needs for technology had substantially contributed to the growth of networking equipment companies like Cisco, and their cutbacks were deeply felt across the industry. In addition, many telecom companies had over- spent their IT budgets, and faced with significant excess capacity in their industry, their networking spending came to a halt. Cisco announced its first loss as a public company in the fiscal quarter that ended March 31, 2001 and cut 18% of its payroll (8,500 employees). By the end of the financial year 2001, it had totaled up a loss of about $1 billion. With slowing demand and falling revenues, the nega-

tive aspects of the customer-centric grouping became conspicuous. Under a customer-oriented grouping, re- dundancies in technological development were common,

Figure 2 Cisco Systems After August 2001

Operations engineering

Storage Voice Core routing

Cisco IOS technologies

division (ITD)

Network Mgmnt services

Internet switching

and services

Optical Senior vice president

Access Aggregation Ethernet access

Wireless

Corporate marketing

Commercial Service provider

Enterprise Technology marketing

Chief marketing officer

Customer segment marketing

Chief development officer

August 2001 restructuring

and were perceived as the cost of being responsive to customers. One manager noted both the advantages (cus- tomer responsiveness) and the disadvantage (redundancy in development) in the older structure: “If there was a (customer) problem, we’d get whatever resources were required to fix it and then execute on it, quickly. But the problem was that ten people would be doing the same thing across the company ten times over, at ten times the cost. And they’d get it done quickly, probably in about one tenth the time that we do now, but it was just incredibly inefficient.” On August 23, 2001, the company announced a

major reorganization. Cisco Systems was to be reorga- nized around 11 technology groups (Figure 2). It was expected that the new structure would promote more rapid and cost-effective technical innovation because engineers who formerly worked in separates silos could now exchange ideas, coordinate development, and gener- ate economies through reuse of technological solutions. Although the products were grouped into these 11 tech- nology units, the three sales groups based on customer segment were retained.4 The reorganization was imple- mented relatively quickly over a period of three months (September–November 2001). The stock markets reacted positively to the announce-

ment of the reorganization. However, several analysts

also noted the paradoxical aspect of reorganization towards a primary grouping by technology at a time when the ability to sell integrated solutions to customers was becoming increasingly valuable. “In a time when everyone seems to be so focused on the customer, reor- ganizing around product lines seems a bit strange,” said one analyst at Forrester Research.5 What accounted for these mixed reactions? Decline in demand (and therefore revenues) was not the only factor affecting Cisco’s busi- ness; commoditization and the entry of low-cost com- petitors was also a source of margin pressure. Prominent among the new competitors was Huawei Technologies from China, which exploited lower labor costs to sell comparable products that were priced substantially lower than Cisco’s equipment. Product commoditization was a problem for IT companies in general, which many (including Cisco) were attempting to combat by selling customized solutions through integration across different technologies (Johansson et al. 2003). Yet, if Cisco pursued cost reduction only (at the

expense of customer responsiveness), it would be at a disadvantage to companies like Huawei, whose cost structure would have been difficult to match. On the other hand, given increased price sensitivity, there were limited prospects of improving prices by increasing efforts to sell customized product solutions. Under these circumstances, the new mix of capabilities the company needed was to maintain customer responsiveness while pursuing cost reduction in order to protect margins—a classic instance of gains from ambidexterity in the con- text of the differentiation—cost focus duality. However, the new formal structure, by itself, appeared

unlikely to enable the development of this new mix of capabilities. While conducive to eliminating redundan- cies in technology development, it also placed organi- zational boundaries between engineers who worked on different technologies that would need to be assembled into a solution, as well as between engineers and mar- keting personnel who would need to work together to customize solutions based on customer requirements. Indeed, many senior managers recognized that the new structure made integration across technologies (horizon- tally) and with customers (vertically) a specialist’s job, whereas it was “everyone’s job” in the older structure. As one senior manager put it, “We moved the inflection point back towards engineering. This allows the tech- nology to be used in multiple customer segments but it does put engineers farther away from the customers � � �” and explicitly recognized the challenges of maintaining customer responsiveness within the new structure, while realizing the benefit of improved cost efficiency.

Compensating Effects of Informal Organization The informal organization of Cisco Systems—specifi- cally, a deeply entrenched culture of customer advocacy,

as well as a pattern of unofficial relationships that sur- vived the change in the formal organization—appeared to have helped Cisco Systems maintain customer respon- siveness despite the emphasis of the new formal organi- zation on cost effective technology development. In operational terms, the impact of the culture of cus-

tomer advocacy was that employees undertook activities that improved collaboration across the various organiza- tional units responsible for collectively meeting partic- ular customer requirements, even when such activities were not explicitly in the formal scope of their work (Gulati 2007, Kohli and Jaworski 1990). Describing such collaboration efforts, one marketing manager noted:

Earlier we were in the LOB; of course we had no prob- lem getting engineering to listen, we were part of the same LOB with common P&L and reporting, right? Now we are not � � � they are no longer obligated to be talking to us, and (they) could say, let the bosses deal with issues. But typically that doesn’t happen. Everybody (i.e. engi- neering) knows that by talking to us they could improve the customer comfort with the product, so we still talk. It’s not just with us, if people need to talk to each other across technologies to put together something that’s important for the customer, they will. Because ultimately its about meeting the customer’s needs.

The second manner in which the informal organiza- tion helped to compensate for the weakness of the formal organization was that ties between individuals formerly in the same organizational unit persisted even though these individuals now functioned within different units. These ties that persisted from the older organization were typically those that originated in the formal structure— relationships between engineers and customers formed during design and support stages, between leads of engi- neering teams working on different technologies, and between product marketing and engineering managers. These relationships were primarily work related to begin with. After the reorganization, the work related aspect of these relationships no longer existed—and yet, indi- viduals used these relationships for advice, informa- tion, and even gossip. Respondents described to us the benefits of these ties in terms of the ability to create new projects, solve customer’s problems and write more effective responses to requests for proposals). As one manager in the voice technology group said, “We have been working together for a lot of years,

the engineers and the marketers know each other really well � � � this helps now, and there are a fair number of collaborations that go on across the business unit, even across technology groups. You can always pick up the phone and find someone on a certain project that you might have had a relationship with in the past.” Another stated, “Accountability has changed � � � under the new structure I ought to have been keeping track of every hour I spend helping [employee in other technology group]. Well, he’s still my friend.”

It is important to note that the behaviors motivated by the customer advocacy culture and supported by the informal relationships were inconsistent with the new formal organization—the actions thus emphasized were distinct from the actions emphasized by the formal orga- nization. For instance, direct cross-technology interac- tions by the team leads were inconsistent with the new formal organization, which had created formal integrat- ing mechanisms and specialized roles for vertical and horizontal coordination, such as the solutions engineering group. If the purpose of organizational boundaries is to structure and limit interactions between members (March and Simon 1958, Nadler and Tushman 1997, Thompson 1967), they did not appear to fully meet these objectives in Cisco Systems. Put differently, the informal organiza- tion at Cisco Systems appeared to have made organiza- tional boundaries more permeable. While it is not possible to establish a causal link

between the reorganization and the performance of the firm in this simple illustration featuring one case, it is worth noting that the data is consistent with the hypoth- esis that the company appeared to have achieved signifi- cant cost efficiencies while staying “close to customers.” Post-reorganization, Cisco’s profitability improved sig- nificantly through a combination of stable revenue in a market with declining demand and improved cost effi- ciency (Table 1). Annual sales per employee at Cisco had fallen from $700,000 in 2000 to about $450,000 in 2001 but increased to $534,000 as of April 2002 and was close to $590,000 in mid-2003. Internal data from the company indicates that it also saw improvements in customer sat- isfaction scores as well as profitability from its hundred largest clients in the same period.6

Yet, these beneficial consequences of inconsistencies appeared to have a definite shelf life. Tellingly, for- mal process overlays to achieve cross-technology inte- gration as well as integration with customers were intro- duced in 2004 in response to a growing recognition of increasing difficulties in cross-functional collaboration. A set of “business councils”—cross-functional leader- ship teams—replicated the older grouping by different

Table 1 Cisco Systems: Revenue, Size, Profitability, and Organizational Form Over Time

Revenue Total Net income Formal Year ($ millions) employees ($ millions) organization

1997 6�440 10�728 1�048�7 Grouping by customer group

1998 8�459 14�623 1�350�1 1999 12�154 20�657 2�096�0 2000 18�928 34�617 2�668�0

2001 22�293 38�000 (1,014) Grouping by technology

2002 18�915 35�790 1�893�0 2003 18�878 34�000 3�578�0

Source. Hoovers online.

customer types at least at the senior management level. One of the council chairs explained:

We have the Commercial Business Council, the Enter- prise Business Council, and Service Provider Council. So surprise-surprise, it sounds a lot like the old (pre- reorganization) LOB structure and essentially what it is, is a cross functional group of people, the VP, SVP, and some directors, that are tasked with being the voice of the customer and essentially providing, in a fairly formal and structured manner, feedback from customer advisory councils where we bring in our lead customers to pro- vide strategic direction to product road maps, service road maps, and business capabilities. So they are really the voice of the customer.

While the membership of these business councils was restricted to relatively senior managers and so did not truly replicate the older grouping by customer groups, it nevertheless maintained some degree of coherence across technologies in order to ensure that they collectively met the needs of each type of customer. In sum, there are five points that we wish to high-

light in this account of reorganization at Cisco Systems: (1) Changes in market conditions made the existing struc- ture of the company inappropriate. In particular, in the changed environment Cisco needed to pursue both cost reduction and customer responsiveness, rather than focus on one alone; (2) the new formal structure alone would have been inadequate to support both a capability for cost reduction as well as customer responsiveness; (3) the informal organization was inconsistent with the formal organization for a noticeable period after the reorganiza- tion; (4) this inconsistency was viewed by many man- agers as beneficial in compensating for the weaknesses of the new formal organization—thus allowing a renewed strategy of pursuing both cost focus and customer respon- siveness; and (5) over time, the shadow of the older infor- mal organization began to disappear, exposing the limits of the formal structure.

Origins and Consequences of Organizational Inconsistencies After Reorganizations The example of the reorganization at Cisco Systems raises the intriguing prospect that post-reorganization inconsistencies between formal and informal organiza- tion can enable the pursuit of dualities (by emphasizing distinct poles of the duality). This can create a form of internal fit where organizational elements compensate for each other, creating compensatory fit, instead of supple- menting each other (supplementary fit). However, sev- eral questions immediately surface: Will organizational inconsistencies always arise after reorganizations? Given the pressures towards conformity between the formal and informal organization, how stable are organizational inconsistencies, and what are the consequences of insta- bility? To what extent are the performance benefits of

achieving compensatory fit offset by the costs incurred by the individuals working in an inconsistent organization? We answer these questions in two steps. First, we first

draw on existing literature on the link between formal and informal organization to explain why inconsisten- cies arise after reorganizations. Once we establish the theoretical basis for the occurrence of such inconsisten- cies, we then explain how inconsistencies can help pursue dualities. We draw on a game theoretic representation to explore the boundary conditions under which inconsis- tencies remain stable and have beneficial consequences, as well as the directions in which the organization is likely to change when the boundary conditions are no longer valid.

Why Do Reorganizations Result in Inconsistencies? Formal organization comprises a set of prescribed roles and linkages between roles, for instance as set forth in job descriptions and reporting relationships (Scott 1998). Informal organization refers to the emergent patterns of individual behavior and interactions between individu- als, as well as the norms, values, and beliefs that under- lie such behaviors and interactions (Roethlisberger and Dickson 1939, Smith-Doerr and Powell 2005). How- ever, there is a close link between formal and informal organization—formal organization affects informal orga- nization via its effects on who interacts with whom. Formal organization, by definition emphasizes some

interactions over others. Consider two basic mechanisms of organization design—grouping and linking, which occur at all levels within organizations (Nadler and Tush- man 1997). Grouping is a basic organization design mechanism that collects formal roles together within organizational boundaries, on the basis of either simi- larity or complementarity of the knowledge underlying those roles. The purpose of grouping is to optimize coor- dination by structuring and limiting interactions between members (March and Simon 1958, Thompson 1967, Nadler and Tushman 1997). For instance, pre-2001, Cisco was organized by groupings around customer seg- ments served (enterprise, SMB, and ISP), which brought together complementary roles in technology and market- ing that were linked in their efforts to meet the needs of each segment. Post-2001, the primary grouping was around similar technologies. Linking mechanisms spec- ify vertical and horizontal interactions between (group- ings of) roles. These include reporting and workflow related relationships and mandated periodic communi- cation. Grouping and linking mechanisms may often be reinforced by collocation and interdependent rewards (Wageman 1995). By emphasizing some interactions over others, group-

ing and linking mechanisms can strongly influence the shape of the emergent informal organization. This is because the likelihood of informal tie formation between individuals increases with propinquity and the frequency

of contact (Smith-Doerr and Powell 2005).7 One might say that the formal organization defines the social (and often the physical) spaces within which individ- uals search locally for opportunities to form ties. Fur- ther, because formal groupings and linking mechanisms are organizational structures with their own identifiable boundaries, membership within such boundaries results in internalization of values, norms, and beliefs specific to that membership (Lawrence and Lorsch 1967). Thus, members of a product development unit may be social- ized into an informal engineering subculture with its own values (e.g., technical novelty), beliefs (e.g., about the relative effectiveness of technical solutions), and norms (e.g., assisting colleagues with technical problems) as a consequence of their membership in a unit that engages in product development activity. Despite the existence of these forces towards consis-

tency, there may be significant lags and even permanent limits to the adjustment of informal organization to the formal organization. Prescribed roles may be changed instantaneously by administrative sanction, but the sur- rounding web of informal organizational elements that comes to be associated with the role may persist for some time.8 For instance, as individuals adjust to the new formal organization they find themselves in, pre- existing informal networks weaken with lack of invest- ments of time and energy by individuals building new relationships and adjusting to new roles, but not instan- taneously. Culture, defined by a set of values, norms, and beliefs changes through exposure to new organiza- tional members and organizational tasks, but again, not instantaneously (Becker and Geer 1960). The instances of past relationships (between development team leaders and between team leaders and product marketing man- agers) persisting into the new organization at Cisco Sys- tems, as well as the persistence of the culture of customer advocacy illustrate these points quite well. The adjustment of the informal organization to the new

formal organization may be subject not only to lags (as described above) but also to permanent limits. In particu- lar, the founder’s articulation of the values and mission of the organization is likely to “lock-in” key informal ele- ments such as the values and premises that guide future decision making (Baron et al. 2001). For instance senior managers who had been with Cisco Systems since its founding believed that the culture of customer advocacy was “imprinted” on the organization during its founding years (Stinchcombe 1965). In sum, reorganizations, which involve a near instanta-

neous change in the formal organization, are likely to be followed by periods in which the informal organization emphasizes a set of employee behaviors that were the same as the ones emphasized by the older formal orga- nization, but not the new one. It is certainly possible that an informal organization that is inconsistent with the new formal organization can hamper the achievement of the

espoused goals of the formal organization by distracting employees towards actions that do not improve organiza- tional performance, or even detract from it through sabo- tage (Homans 1950, Roethsliberger and Dickson 1939). However, this is by no means the only possible outcome of inconsistency. Inconsistencies imply that informal and formal organizations encourage distinct employee behav- iors, but these dissimilar behaviors may in fact be jointly more valuable than either alone. We develop this argu- ment in the next section.

Compensatory Fit Through Inconsistency: Boundary Conditions To analyze the boundary conditions for compensatory fit, we follow in the tradition of other scholars who have formally modeled the dynamics of organization design (Hannan et al. 2003a, b; Nickerson and Zenger 2002; Rivkin and Siggelkow 2003; Siggelkow and Levinthal 2003). We develop a simple game theoretic represen- tation of how formal design choices and employee’s choices interact. Our goal is to state general boundary conditions under which instances of compensatory fit like the one we noted at Cisco Systems can exist in a stable manner, and do not extract too high a penalty from the employees who work amidst organizational inconsistencies. We also explore the manner in which organizations are likely to change when the conditions supporting compensatory fit break down. Relying on a formal representation helps us state these conditions and consequences with a high degree of precision, and also exposes our underlying assumptions for critique and fur- ther refinement by other scholars (Lave and March 1993). We present a verbal and graphic account of the model and results in the text of the paper in a self-contained manner—the technically minded reader may like to refer to the appendix for a formal version. We conceptualize a stylized organization consisting

of two actors—a set of employees, and an organiza- tion designer who could represent senior management or the CEO. There are two distinct types of activities that the employees can undertake, which support one each of the dualities A and B. For instance, these could represent activities relating to “being cost efficient” and “being customer responsive” respectively.9 Employees decide their mix of activities (the split between A and B type efforts) based on what activities are being supported by the formal and informal organization. When we say that the formal organization “supports” a particular type of activity, we refer to the creation of formal coordina- tion mechanisms (grouping of activities into administra- tive units, reporting relationships, cross-unit teams, and processes) and incentives (salary, bonus, stock options, career advancement, and job security) that enable and motivate employees to undertake that set of activities (Nadler and Tushman 1997). The rewards from behav- ing consistently with the formal organization thus include

salary, bonus, career advancement, job security, and free- dom from managerial sanctions by superiors. It is important to note that while incentives are a part

of this formal organization, we are conceptualizing the formal organization more broadly to represent everything that the designer does to encourage certain behaviors. For instance, in Cisco Systems post-2001, the reorga- nization created a formal organization that encouraged cost efficiency-seeking behaviors by grouping all related technology developments within common organizational units, at the expense of customer responsiveness. The informal organization also influences the mix

of activities chosen by the employees by “rewarding” employees for behaving in consonance with it—through utility from relationships, influence, conformance to an identity category, and belonging to a cohesive group (Barnard 1938). In Cisco Systems, the informal organi- zation post-2001 continued to encourage behaviors that supported customer responsiveness in this way, even though this was no longer the focus of the formal organization. We depict the above arguments visually in Figure 3.

The x-axis shows the mix of A and B type activities. The left end of the axis corresponds to a focus exclu- sively on A activities, and the right end corresponds to a focus exclusively on B type activities. Intermediate points represent some mix of A and B type activities, with the point E being an even split between the two. An “�” type formal (or “A” type informal organization) simply means one that encourages activity A over activ- ity B, through the appropriate combination of grouping and linking mechanisms. An “organizational architecture” denotes the combina-

tion of choices made by the designer and the employee. Thus, the architecture (�� A) is consistent since the employees emphasize A type activity and so does the formal organization. In contrast the architecture (�� B) is inconsistent, as the employees emphasize B while the formal organization emphasizes A. We can also speak of the formal and informal organization being consistent with each other when both encourage the same type of activity; else, they are inconsistent with each other. We assume that the designer chooses the formal orga-

nization to enhance organizational performance net of the costs of implementing a formal organization, taking into account how the employees are likely to react to

Figure 3 Effects of Formal and Informal Organization on Mix of Activities

All A All B

FormalβFormal α

Informal A InformalB

the chosen formal organization and the existing infor- mal organization. Thus, we assume that the designer’s payoff is organizational performance net of the costs of design and implementation (including wages paid out) of the chosen formal organization. Note that the designer cannot directly influence the informal organiza- tion in our model. Organizational performance may be understood in terms of profitability, market capitalization, operating efficiency, or market share—measures that the designer/top management of the organization may be rea- sonably assumed to care about. Organizational performance depends on the mix of

activities that the employees engage in, as well as the manner in which the two kinds of effort interact. Figure 4 illustrates the two basic types of interactions between the activity types. In this figure, the performance of the orga- nization is plotted on the vertical axis as a function of the mix of Types A and B activities that the employ- ees engage in. For instance, when employees choose a mix of mostly A type activities, then the performance of the organization corresponds to the height of the perfor- mance curve at points to the left of E on the horizontal axis. Between points A and B lie various combinations of the two activity types. We can distinguish two situations. First, consider the

case when the performance of the organization is higher at point A or B than at any intermediate point between them. This means that any combination of the two activ- ities underperforms “pure A” or “pure B” type activ- ity. This describes a case of “gains from focus”—it arises because the two activity types have strong substi- tution effects (doing more of one activity decreases the marginal product of the other activity) between them so that the greater the extent of activity A, the less valuable it is to engage in activity B (and vice versa) (Milgrom and Roberts 1990, 1995).

Figure 4 Interactions Between Activity-Types and Organizational Performance

O rg

an iz

at io

na l

pe rf

or m

an ce

Gains from ambidexterity

All A All BE

Gains from focus

Mix of activities

In contrast, the second situation is one in which the performance of the organization is lower at points A and B than at any other point along the line ACB. This means that any combination of the two activities always domi- nates “pure A” or “pure B” type activity. This is a case of “gains from ambidexterity”—it arises because the two activity types are strongly complementary (doing more of one activity increases the marginal product of the other activity), so that the greater the extent of activity A, the more valuable it is to engage in activity B (and vice versa) (Milgrom and Roberts 1990, 1995).10 Through Figure 4, we wish to emphasize that irrespective of the organiza- tional costs and complexities associated with the simul- taneous pursuit of both poles of a duality, there are situ- ations when it is potentially valuable to do so (i.e. when there are gains from ambidexterity) and others when it is not (i.e. when there are gains from focus) In Table 2, we list three well-known dualities—cost

differentiation, exploration-exploitation, differentiation- integration—as well as conditions under which efforts aimed at pursuing the two poles of each duality are pre- ferable to a focus on one of the poles alone. For instance, consider the duality that was at the heart of the reorgani- zation at Cisco—cost focus versus differentiation (in this case, customer responsiveness). On an efficiency frontier, dividing efforts between enhancing cost efficiency and differentiation is never sensible, because improvements in one can only come at the expense of the other. This is the case of “gains from focus” popularized by Porter through his warning to companies to avoid getting “stuck in the middle” (Porter 1985). In contrast, when mov- ing towards an efficiency frontier, it is better to spread efforts towards both cost reduction and differentiation. This is because the shortest path is typically likely to involve some movement along both cost and differen- tiation dimensions, so that in this case there are “gain from ambidexterity.”11 Some of the early influential work on quality management implicitly espoused such a logic when arguing for the possibility of simultaneously pur- suing quality improvements and cost reductions. Which

Table 2 Dualities: When Is It Better to Be Ambidextrous?

Gains from Gains from Duality ambidexterity when focus when

Cost vs. Firms are far from Firms are on the differentiation the efficiency frontier efficiency frontier

Exploration vs. Moderate levels of Extremes of stability exploitation change (i.e., some or change in the

change and some search environment stability in the search environment)

Differentiation vs. Heterogeneity in Heterogeneity without integration unit-level interdependence,

environments or interdependence coupled with without heterogeneity. interunit interdependence

condition—gains from ambidexterity or gains from focus better describes a setting, we believe, is dependent on contingencies such as those set out in Table 2, rather than necessarily being one or the other (He and Wong 2004, Katila and Ahuja 2002). Given this representation, (for a technical version of

the model described above, please see Appendix §A1) we can now ask under what conditions an inconsistent orga- nizational architecture can exist stably, so that neither the employees nor the designer would seek to alter their choices, given the other’s choices. Put differently, given an informal organization, when would the designer opti- mally choose a formal organization that would result in the employees acting inconsistently with it? The answer could help to understand how the compensatory fit we observed at Cisco Systems worked.

Stability. We outline the intuition for the conditions under which inconsistent architectures can be stable graphically here, though a complete technical presenta- tion may be found in the appendix. Figure 5 makes it clear that the impact of choosing a

formal organization inconsistent with the informal orga- nization is to move the mix of employee activities away from the edges towards the center (E). Further, such a course of action is beneficial when there are gains from ambidexterity—i.e., when organizational performance is higher under some mix of activities than under just activ- ity A or B alone. For instance, given an informal orga- nization that encourages primarily “B” type behavior, as in Figure 5, under gains from ambidexterity, the designer is better off choosing the formal organization �, which keeps the distribution of employee efforts near the center (and therefore at a higher value) rather than formal orga- nization � that would push the mix of activities towards the corner. In contrast, as Figure 6 shows, under gains from focus, given an informal organization (emphasiz- ing B), the designer would pick � in order to push the mix of activities towards the corner (and therefore attain higher organizational performance).12

However, this only covers the designer’s motivations; we also need to understand why the employees would not

Figure 5 Compensatory Fit Through Inconsistent Architecture

All A E

All B

O rg

an iz

at io

na l

pe rf

or m

an ce

FormalA

InformalB (α, B)

Figure 6 Supplementary Fit Through Consistent Architecture

All A All B

O rg

an iz

at io

na l

pe rf

or m

an ce

FormalB

InformalB (β, B)

adjust their mix of activities to attain consistency with the formal organization, since the rewards from the formal organization increase by doing so. If employees prefer to choose a mix of activities that is inconsistent with the formal organization, it must be because the rewards for acting consistently with the informal organization must be large enough to outweigh the gains to employees from acting consistently with the formal organization. Thus, a stable inconsistent organizational architecture

can exist under the conjunction of two circumstances: a powerful informal organization and contingencies that create gains from ambidexterity (see Appendix §A2.1). Further, the informal organization must be “powerful” in the precise sense that it motivates employees to select a mix of activities that is predominantly skewed towards one pole of the duality. Because the resulting incon- sistent architecture generates superior performance, it represents a state of fit between the formal and infor- mal organization. We label this as compensatory fit, as the formal and informal organizations motivate inconsis- tent (distinct) but complementary (jointly value creating) activities. This captures the post-reorganization situation at Cisco Systems—a state of compensatory fit between the formal organization (which emphasized cost effi- ciency) and the informal organization (which emphasized customer responsiveness) given external conditions that made achieving a balance between the two valuable. We state this result formally as our first proposition.

Proposition 1 (P1). A state of compensatory fit through an inconsistent organizational architecture can exist under the conjunction of two conditions: (i) gains from ambidexterity and (ii) an informal organization that strongly motivates behavior consistent with one of the poles of the relevant duality.

Efficiency. In any inconsistent architecture, the employees incur some opportunity costs, because given the emphasis of the informal organization (say B) they would be better off if the designer had chosen a formal organization that was consistent with it (�). Put differently, with an inconsistent architecture, employees miss out on the chance of being rewarded by the formal organization for doing what they would (largely) do

anyway in response to the pressures of the informal organization. Thus the employees would always prefer a formal organization that is consistent with the informal organization. For instance, in post-reorganization Cisco Systems, the employees would very likely have preferred a formal organization that continued to encourage customer responsiveness, given the strong emphasis of the informal organization—both in terms of culture and the existing pattern of relationships—on encouraging such activities (through the customer advocacy culture and the pattern of informal relationships, for instance). We depict this graphically in Figure 7, where the real-

ized organizational architecture is �� B� but employees would in fact prefer (�e� Be�. The designer, however, prefers to choose an inconsistent formal organization, resulting in �� B� if the conditions for compensatory fit we outlined in (P1) are met. Thus, due to gains from ambidexterity, while the designer’s payoff is max- imised by pulling the allocation towards the center, the employee’s payoff is maximized under a formal organiza- tion that pushes the allocation towards the edge. Since the designer does not take into account the employee’s pay- offs from the informal organization, there is a tendency to pull the allocation towards the center to an extent that is more than warranted by joint surplus maximization. This creates a source of inefficiency whenever compensatory fit exists. The value maximization principle (Milgrom and Roberts 1992, p. 36) states that a set of choices is efficient only if it maximizes the total value to all partic- ipants. Inefficiencies can eventually lead to lower perfor- mance (for instance, through lowered employee motiva- tion or turnover), so it is important to assess the condi- tions under which the inefficiencies created by compen- satory fit are minimized. In Figure 7, the efficient architecture denoted by

(�∗� B∗� will lie somewhere between (�� B� and (�e� Be�. Our key result about the efficiency of compensatory fit is that ��∗� B∗� and �� B� converge to each other (i.e., the inefficiency reduces) as the magnitude of the gains from ambidexterity increase. The intuition for this result can

Figure 7 Efficiency of Compensatory Fit

O rg

an iz

at io

na l

pe rf

or m

an ce

FormalA

InformalB

(βE, BE)

(α, β)

(α*, B*)

be explained in terms of Figure 7 (for a technical discus- sion, please see Appendix §A2.2). The relative distance between ��∗� B∗� and �� B� versus between (�∗� B∗� and ��e� Be� indicates the relative emphasis placed on the employee’s and designer’s payoffs in joint value maximization (the shorter the distance, the greater the weight). By definition, the point (�∗� B∗� places exactly the right weight on the designer’s and employee’s payoffs for total value maximization. Now, the greater the gains from ambidexterity (i.e. the higher the peak in the orga- nizational performance curve—compare (P1) and (P2), the greater the designer’s payoff at �� B� while the employee’s payoff is unaffected. Therefore, the greater the magnitude of the gains from ambidexterity, the more the emphasis on the designer’s payoff relative to the employee’s in joint value maximization, so that the effi- cient choice and the designer’s actual choice draw closer.

Proposition 2 (P2). The inefficiencies associated with employees working in an inconsistent architecture in situations of compensatory fit decline with the magnitude of gains from ambidexterity.

Change. To explore how the inconsistent architectures supporting compensatory fit can change when the nec- essary conditions outlined in (P1) break down, we now assume that there are two periods to consider in our model—a design period �t1� in which the designer makes his choice of formal organization, and an adjustment period �t2� in which the informal organization or external contingencies (such as gains from ambidexterity) change. The designer cannot revise choices in the second period (if he could, this would be equivalent to the simple one period model discussed so far), but we assume that a fore- sighted designer can anticipate the impact of the changes when making decisions about the formal organization in the first period. In other words, the designer optimizes the present value of organizational performance (net of any design costs) over the two periods when making his choice in the first period. In contrast a “myopic” designer would ignore the adjustments taking place in the second period (see Appendix §A3). We first consider the case when beginning with

an inconsistent architecture, the informal organization adjusts to the formal organization over time. For instance, as we observed at Cisco Systems, the informal orga- nization supporting customer responsiveness seemed to weaken over time and become increasingly aligned with the formal organization and its emphasis on cost efficiency—so much so that new formal organizational structures (the business councils) had to be put in place to create additional pressures toward customer responsiveness. Figure 8 illustrates the possible consequences of such

changes to the informal organization. A designer with foresight would factor in this change (appropriately dis- counted) into the first period decision making. As a con- sequence, the conditions for compensatory fit become

Figure 8 Dynamics of Compensatory Fit

All A All B

O rg

an iz

at io

na l

pe rf

or m

an ce

Informal t12 A

Informal t 22 A

Et 1 3 t1t

2 3

(β, A)

(α, A) (α, B)

more stringent in the first period—the informal orga- nization would have to be even “stronger” than in the case with no adjustment for the designer to select a for- mal organization inconsistent with the informal one (see Appendix §A3.1). Interestingly, what happens in the next design period

(period 3) depends on the magnitude of the adjustment that takes place in period 2. With moderate adjustment of the informal organization in period 2 (e.g., to the lev- els marked t12�, there is no incentive for reorganization in period 3, as the new optimal architecture in period 3 �t13� is consistent—despite the gains from ambidexterity. With strong adjustment (e.g., to the levels marked t22�, we get a second period that clearly displays misfit (despite consistency between the formal and informal organiza- tion, because the designer would have preferred to set the formal organization inconsistent with the informal organization, if that were possible in this period). This is followed by reorganization towards an inconsistent archi- tecture in period 3 �t23�. The intuition is similar to that for (P1), which states the conditions necessary for a stable inconsistent architecture; unless the informal organiza- tion becomes strong in the opposite direction, one can- not have an inconsistent architecture again (please see Appendix §A3.1.1 for technical details). Adjustment of the informal organization to the formal

organization thus undermines the stability of inconsistent architectures—if the adjustment is moderate, we would expect consistency between formal and informal organi- zation and no further reorganization; if the adjustment is substantial, a period of misfit is followed by a reorganiza- tion to a new inconsistent architecture (and compensatory fit). Thus, assuming unchanging conditions of gains from ambidexterity, adjustment of the informal organization to the formal organization could provide an endogenous explanation for repeated reorganization (Nickerson and Zenger 2002), though this depends critically on the rate of adjustment of the informal organization (i.e., the extent of adjustment in period 2). We next consider how an inconsistent architecture

would be affected if the contingencies change so that the gains from ambidexterity give way to gains from

focus (see Table 2). We find that an inconsistent archi- tecture in the first period implies that there is a limit to the adjustment anticipated by the designer between periods—if there were very large gains from focus in the second period, a foresighted designer will effectively ignore gains from ambidexterity in the first period and select a consistent architecture in the first period itself (see Appendix §A3.2). We also find that although the informal organization

remains unchanged (by assumption), it still plays a criti- cal role in determining what happens in periods 2 and 3. If the strength of the informal organization is above a critical threshold value, we find that the designer has no incentives to reorganize in period 3—an inconsis- tent architecture then prevails despite gains from focus. However, if the informal organization is weaker than this threshold, then the second period clearly displays mis- fit (because the designer would have preferred to set the formal organization consistent with the informal orga- nization, if that were possible in this period), and in period 3 the designer will reorganize towards a formal organization consistent with the informal organization (for instance, compare the difference in organizational performance that would be achieved with (�� B) and �� B� in Figures 5 and 6). Therefore, a strong informal organization also protects against misfit and the need for reorganization when there is a change from gains from ambidexterity to focus (see Appendix §A3.2.1).We sum- marize the key insights from this analysis of change as follows.

Proposition 3 (P3). The need for reorganization fol- lowing the breakdown of conditions for compensatory fit depends on the strength of the informal organization: When gains from ambidexterity prevail, there is no incen- tive to reorganize if the informal organization adjusts only moderately to the formal organization; when gains from focus replace gains from ambidexterity, there is no incentive to reorganize in the presence of a strong infor- mal organization.

Discussion It has, of course, been well-known since the work of Chester Barnard (1939, p. 169) and Herbert Simon (1957) that the informal organization can augment the formal organization by legitimizing its authority and enabling the performance of tasks that can be only par- tially specified through formal organizations (Child and McGrath 2001, Mintzberg 1990, Nadler and Tushman 1998). Our study adds crucial texture to this general insight: the informal organization can enhance the effec- tiveness of the formal organization either by supplement- ing it—in effect acting as “the last mile” that connects the formal organization to employee actions—or by com- pensating for it, by motivating behaviors that are valuable

but not adequately emphasized by the formal organiza- tion. As Nadler and Tushman (1997) note, while informal processes may be outside the direct purview of formal design, “� � � managers should be looking for ways to pro- vide for designs and roles that are consistent with—and that capitalize on these informal processes” (p. 111). Our analysis shows that consistency may not be necessary in order to capitalize on the informal organization. We also contribute to the literature on ambidextrous

organizations (Brown and Eisenhardt 1997, Tushman et al. 2004, Tushman and O’Reilly 1996) in two specific ways. First, we note that gains from ambidexterity may not be ubiquitous—in some cases gains from focus may dominate. We also outline possible contingencies that distinguish between the two situations for some common dualities (Table 2). Put differently, ambidextrous organi- zations may not always be desirable or needed. Second, in this study we have explored the origins of

combinations of formal and informal organization that enable ambidexterity. The classic solutions to achieve ambidexterity have relied on organizational separation— both spatial and temporal. For instance, the common principle underlying “skunkworks” (Christensen 1997) and ambidextrous organizational forms (Tushman and O’Reilly 1996) is the spatial separation of exploration and exploitation processes across different organizational units. The principle of temporal separation is best known in the context of innovation and involves exploration and exploitation in the same organizational unit, but at different times (Brown and Eisenhardt 1997, Puranam et al. 2006). An alternative approach to balancing the conflicting

organizational demands of dualities lies in combina- tions of elements of formal and informal organization into a hybrid arrangement. For instance, Gibson and Birkinshaw (2004) point to such combinations that cre- ate contextual ambidexterity, and Brown and Eisenhardt (1997) describe “semistructures,” which combine a few key elements that promote exploitation with features that support exploration. However, we know little about how these unusual combinations of formal and informal orga- nizational elements arise, apart from the fact that they “must be grown, not assembled at a single point in time” (Brown and Eisenhardt 1997, p. 31), and that there is equifinality—“depending on the administrative heritage of a given business and the values of its leaders, equally valid but slightly different organizational context solu- tions can be created” (Gibson and Birkinshaw 2004, p. 223). This paper suggests that reorganizations, and the resulting organizational inconsistencies they create, can be a means of building such a solution. There are specific ways in which our distinction

between supplementary and compensatory fit extends and generalizes results from recent studies on the link between organizational change and performance. Nickerson and Zenger (2002), for instance, used a

dynamic simulation model to explain why companies seem to undergo frequent reorganizations and move in a pendulum-like fashion between discrete organizational structures that embody centralization and decentraliza- tion. In an insightful analysis, they show that under the assumptions that the formal organization is discrete (it can only support all A or all B, for instance) but the informal organization is continuous and adjusts gradually to changes in the formal organization, then modulating between centralized and decentralized formal organiza- tion may be an efficient way to overcome the discreteness of formal organization choices. Whereas the Nickerson and Zenger argument pertains

to the value of a “continuous” informal organization over- coming the “discreteness” of formal organization, our argument is about how inconsistencies between the for- mal and informal organization can help achieve comple- mentary dualities. In our model, we treat the formal and informal organization symmetrically, in that both affect employee behavior, so that the focus is on how they jointly serve to adjust the mix of employee behaviors (Figure 3).13 Further, since we conceptualize the formal organization more broadly than discrete grouping deci- sions to include linking structures as well (Nadler and Tushman 1997), we do not need to impose a discrete- ness assumption on the formal organization. As a conse- quence, we argue that the inertial nature of the informal organization, which slows adjustment between the formal and informal organization, is only likely to be beneficial under gains from ambidexterity; if gains from focus are dominant, then inconsistency can be harmful. A related study by Hannan et al. (2003a) presents a

formal theory to elucidate why changes to single orga- nizational elements are sometimes observed to increase the hazard of organizational mortality. They argue that change to a single element (such as a division-level incentive system) can set off a cascade of changes lead- ing to a full-blown reorganization, if the initial element being changed is tightly linked (central) to other ele- ments. These cascades of changes arise because each ele- ment must achieve consistency with the elements it is connected to. Hannan et al. argue that during the periods when consistency is being attained, the organization is unable to function effectively. This leads to their conclu- sion that the more central the element that experiences the initial change, the longer the cascade of changes and consequently, the greater the increase in the hazards of mortality. Our analysis suggests that the logic of Hannan et al.

applies most strongly to situations of gains from focus, when organizational inconsistencies can lower perfor- mance, and the direct costs of achieving consistency as well as the opportunity costs incurred during the period of achieving consistency can doubtless increase the haz- ards of mortality. However, our framework suggests that

in the presence of gains from ambidexterity, inconsisten- cies can enhance rather than detract from performance if the formal and informal organizations emphasize the two poles of an underlying duality. Therefore, we expect that the impact of changes to central elements on the hazards of organizational mortality is strongest in the presence of gains from focus; with gains from ambidexterity this effect may be weaker. Since our contribution is primarily conceptual, we rec-

ognize that further research is needed to establish a causal link between organizational inconsistency and per- formance, something that neither a model nor a single case alone can achieve. Despite this, we believe that our study is valuable because it provides a novel and precise way of thinking about the relationships between orga- nizational elements in general, and formal and informal organization in particular. Further, we foresee some spe- cific avenues for research that builds on this study. For instance, our analysis shows that the stability of incon- sistent organizational forms depends on continued limits to the adjustment of the informal organization, as well as a continued state of gains from ambidexterity. We believe this proposition will prove to be a fruitful line of research, as it appears amenable to cross-sectional anal- ysis. For instance, our theory would predict that among a population of firms facing similar conditions of gains from ambidexterity, the strength of the informal organi- zation would predict the likelihood and shape of future reorganizations. Our study is not without limitations, which arise partly

from structural features of the research tools and design we employ—single case studies and formal analysis. The former enjoys the benefits of detailed observation but suffers from limits to generalization (Miles and Huber- man 1994, Yin 1994), while the latter allows for theoret- ical generalization but involves unreal and stark assump- tions about behavior in organizations (Lave and March 1993). However, each methodology also provided unique insights not possible with the other. We never could have reached the stage of developing a formal theory about compensatory fit without the initial insights gleaned from our fieldwork about how inconsistencies between formal and informal organization could be useful. However, the formalization added several new insights that the field- work and verbal theorizing alone may not have gener- ated. For instance, while our research at Cisco certainly helped us see how the informal and formal organization could be inconsistent and yet create value under gains from ambidexterity, the condition that there should be a threshold level of the strength of the informal orga- nization only became clear once we explicitly modeled the cost of implementing the formal organization, as well as the idea of different strengths of the informal organization (P1). The model also helped us to distin- guish the concepts of “consistency” and “fit” from each other very sharply—it helped us see that an inconsistent

architecture can arise even in the absence of gains from ambidexterity—and that gains from focus do not auto- matically lead to a consistent architecture (P3). The dis- tinctions arise because of the costs of design and differ- ences in the marginal product of the two kinds of efforts (see the appendix). It was also fairly intuitive to us from the fieldwork

that there could be winners and losers with compen- satory fit—the designer gained while employees bore the opportunity costs of living and working in an inconsis- tent organization. Analyzing the model helped us see that while inefficiency appeared inevitable with compen- satory fit, as the magnitude of the gains from ambidex- terity increased, the designer’s choices converged to the efficient outcome (P2). Finally, once it came to exploring the dynamics of what would happen when the conditions for compensatory fit broke down, we found the model even more indispensable. To keep track of the effects of changing over time, (a) the gains form ambidexter- ity, (b) adjustment to the informal organization, as well as (c) a rational designer’s anticipation of these changes proved too complicated to do verbally—there were just too many “moving parts” to be able to reach a clear con- clusion. Modeling this problem, however, made it easy to see what would happen (P3).14 We hope that by combin- ing the use of two very different methodologies (quali- tative interview data and formal analysis), we have been able to achieve our own version of “compensatory fit.” We acknowledge that analysis of a formal model is one

among many theory-building techniques, such as induc- tion from field data, verbal explication, or integration across prior literatures (as opposed to empirical theory- testing techniques such as experiments or regression analysis). When working with a formal, as opposed to a verbal model, we are of course subject to the same trade- off between realism and rigor that applies to all models in the social sciences. Formal models such as the one we use in this paper merely highlight the trade-off more sharply—they accentuate the benefits of clearly stating assumptions and the value to using solution concepts (like the subgame perfect Nash equilibrium) that can gen- erate nonintuitive insights (Lave and March 1993). How- ever, the assumptions that underlie models such as ours also appear stark and unrealistic (two players, two poles of a duality, etc.). In our work, we have been guided by other students of organization who have demonstrated that the simplifying assumptions of mathematical models are justified as long as they provide a rigorous basis for improving our understanding of a complex phenomenon, and generate interesting and testable predictions (Abra- hamson and Rosenkopf 1993, Bhattacharya et al. 1998, Bruderer and Singh 1996, Carroll and Harrison 1998, Lave and March 1993, Siggelkow 2002). We hope that we have met these criteria in this paper. While we have focused on reorganization as a mecha-

nism that achieves inconsistency in this study, it is clearly

not the only one. Rotation of employees across formal roles or “mandated subversion” of the formal organiza- tion by managers may be other mechanisms that deserve closer scholarly attention. Yet another direction for exten- sion might be a study of the relative efficacy of formal linking mechanisms such as incentives, cross-functional teams and boundary spanners compared to the “residual” informal linkages created by personal ties and organiza- tional culture in overcoming the discreteness of formal grouping choices. While all these mechanisms are well- known in theory, and also well-studied empirically in iso- lation, comparative analysis remains scarce. Finally, we note that our study suggests some novel

and subtle aspects of the intertemporal relationship between formal and informal organization, and the con- sequences for organizational change and renewal. (Also see Tripsas (2009) and Eggers and Kaplan (2009) for managerial cognition perspectives on organizational change and renewal.) Given lags and limits in the adjust- ment of the informal organization to the formal orga- nization, under gains from ambidexterity, the optimal formal organization capitalizes on yesterday’s informal organization and lays the ground for tomorrow’s. An inconsistent formal organization (achieved through reor- ganization, for instance) leverages the existing informal organization to motivate the mix of behaviors needed to achieve compensating fit. The formal organization itself focuses on the behaviors that are not adequately encour- aged by the informal organization. However, the choice of today’s formal organization also shapes tomorrow’s informal organization (within limits), due to the process of eventual adjustment between the two. Further investi- gation of these intertemporal links in organization design should prove fascinating.

Acknowledgments Authors names are in alphabetical order. We are grateful for useful suggestions and advice during the course of this project from Sourav Bhattacharya, Sumantra Ghoshal, Tobias Kretschmer, Jackson Nickerson, Madan Pillutla, Hayagreeva Rao and Freek Vermeulen. All errors remain our own. Puranam acknowledges funding from the William and Phyllis Mack Center for Technological Innovation at The Wharton School.

Appendix

A1. Model Specification Employees (acting as a unitary agent) are assumed to choose how they will split their total efforts (normalized to 1) between the two kinds of activities. Let x represent B type effort; then 1− x represents the effort spent on A type efforts. The rewards to the employees �w� from choosing any particular mix of activities, depends on both the rewards from the formal (wf � and the informal organization (wi�, as captured in the following equation:

w�x� = wf �x� + wi�x� (1)

where

wf �x� = ( 1 2 + �)x + ( 1

2 − �)�1 − x� and

wi�x� = ( 1 2 + )x + ( 1

2 − )�1 − x��

The parameter � represents the formal organization cho- sen by the designer to support A or B type activities by employees to different extents. Thus, when the designer sets a positive (negative) value of �, this corresponds to a behavior- based contract that encourages employees to do more B (A) than A (B). The parameter ∈ − 1

2 � 1 2 � represents rewards to

the employee from behaving in consonance with the informal organization. Thus, a positive value of indicates the existence of an informal organization that emphasizes B type efforts over A type efforts. To complete specifying the employee’s opti- mization problem, we define a cost of effort function �c� for the employee:

c�x� = x 2

2 + �1 − x�

2 � (2)

The assumption of independent quadratic costs of effort is common in the literature on teams, and agency theory in gen- eral (Baker 2002, Cremer 1990, Gibbons 1998, Prendergast 1999, Siggelkow 2002) and simply captures the notion of increasing marginal disutility of an additional unit of effort for each kind of effort.

We assume that the designer chooses the formal orga- nization to enhance organizational performance, taking into account the costs of creating and maintaining such a formal organization. Organizational performance arises as a function of the mix of activities that the employees engage in, as well as the manner in which the two kinds of effort interact. The out- put function (P � specified below allows us to explicitly capture these possible interactions between the two kinds of efforts as well as the magnitude of the interaction:

P �x� = ax + �1 − x� + �x�1 − x� (3) where � �= 0 is a parameter that describes the nature of inter- action between choices, and a > 0 is a parameter that sets the relative magnitude of the marginal contributions of each kind of effort to output. When the parameter � > 0, this represents a situation of gains from ambidexterity; some split of activity between A and B type efforts is typically better than all A or all B (the exact split depends on a). Conversely there are gains from focus when � < 0—all A or all B is usually better than a mix (which corner is better will depend on the value of a). Finally, to complete specification of the designer’s optimiza- tion problem, we specify the cost of design incurred by the designer for choosing any given formal organization (�).

D�� = d�2 (4) where d > 0. This corresponds to the cost of designing and implementing formal structures, information channels and incentives, which increases with the emphasis placed on encouraging either type of behavior. We will make the tech- nical assumption that d + � > 0 to ensure an interior maxi- mum. This articulates the assumption that the cost of design is nontrivial.

A2. Static Model The equilibrium concept we use is subgame perfect Nash equilibrium. We will assume that the designer moves first to choose the formal organization �∗ to maximize P �x∗� �� − wf �� − D�� by looking forward and anticipating the employee’s action x∗. The employee chooses x∗ to maximize w��∗� x� − c�x�.

Solving the employee’s optimization problem, and keeping in mind that x ∈ 0� 1� we find:

x∗�� =

⎧⎪⎪⎨ ⎪⎪⎩

1 2 + + � if − 1

2 < + � < 1

0 if + � ≤ − 1 2

1 if + � ≥ 1 2 �

(5)

Solving the designer’s optimization problem, for an inter- nal solution for x∗, we find: �∗ = �a − 1 − 2 �1 + ��/ �2�2 + d + ��. We can therefore write:

If � > 0

x∗ = 1 2

+ a − 1 + 2�1 + d� 2�2 + d + �� and

�∗ = a − 1 − 2 �1 + �� 2�2 + d + �� for ∈

[− 1 2 � 1 2

] �

If � < 0.

If −�a + 1 + d + ��

2�1 + d� < < �3 − a + d + ��

2�1 + d� �

x∗ = 1 2

+ a−1+2�1+d� 2�2+d +�� and �

∗ = a−1−2 �1+�� 2�2+d +��

If ≤ −�a + 1 + d + �� 2�1 + d� ,

x∗ = 0 and �∗ = − 1 2 − � (6)

If ≥ �3 − a + d + �� 2�1 + d� �

x∗ = 1 and �∗ = 1 2 − �

Note that 2P �� /� �� = −2�. Thus, when � < 0 moving � in same direction as the informal organization strengthens the marginal effect of the informal organization (supplementary fit). When � > 0 moving � in a direction opposite to the informal organization strengthens the marginal effect of the informal organization (compensatory fit) because 2P ��/� �−�� = 2��

Definitions. An inconsistent organizational architecture is one in which the employee emphasizes the activity other than that emphasized by the formal organization. Thus �� B� denotes the inconsistent organizational architecture in which �∗ < 0, x∗ > 1

2 while �� A� denotes the case where �∗ > 0,

x∗ < 1 2 . A consistent organizational architecture is thus one

in which the employee emphasizes the same activity as that emphasized by the formal organization eg. �� A� or �� B�. Note that an inconsistent architecture can arise even in the absence of gains from ambidexterity, and that a consistent architecture is not automatic with gains from focus.

Compensatory fit arises when an inconsistent organizational architecture is used to harness the gains from ambidexterity (� > 0�. Supplementary fit arises when a consistent organi- zational architecture is used to exploit the gains from focus �� < 0�.

A2.1. Compensatory Fit The condition for the architecture �� B� to generate compen- satory fit is equivalent to stating the condition when �∗ < 0, x∗ > 1

2 given � > 0, > 0.

This is > max��a − 1�/�2�1 + �� 1 − a�/�2�1 + d�� By symmetry, the mirror image architecture �� A� generates compensatory fit if < min��a − 1�/�2�1 + �� 1 − a�/ �2�1 + d��.

The conditions for supplementary fit are �a−1�/�2�1+�� > > �1 − a�/�2�1 + d�� for (�� B) and �a − 1�/�2�1 + �� < < �1 − a�/�2�1 + d�� for (�� A), given � < 0.

The restrictions a ∈ −�� 2 + �� and a ∈ −d� 2 + d� on the marginal rate of technical substitution between A and B type efforts ensure that − 1

2 ≤ �a − 1�/�2�1 + �� ≤ 1

2 and

− 1 2

≤ �1 − a�/�2�1 + d�� ≤ 1 2 .

Thus, compensatory fit requires that the informal organi- zation be above a threshold level of strength. Further, the marginal contributions of each type of effort to output should not be too different.

A2.2. Efficiency of Compensatory Fit We denote the efficient choice of formal organization as �e

�e =arg max �

P ��−wf ��−D��+wf ��+wi��−c�x��

�e = a−1−2 � 2�1+d +��

�e −�∗ = a−1+2�1+d� 2�1+d +��2+d +��

⇒ ��e −�∗� is strictly decreasing in � �� >0�. Thus, the difference between the efficient and actual (compen- satory fit) formal organization decreases in the magnitude of the gains from ambidexterity.

A3. Dynamics

A3.1. Adjustment of Informal Organization We now model a two-stage version of the problem. In the first stage, given an informal organization 1 the designer chooses a formal organization � and the employee chooses an alloca- tion x1∗ given �� 1�. In the second stage, the informal orga- nization adjusts and takes a new value 2 and the employee chooses a new allocation x2∗ given �� 2�. A foresighted designer therefore sets � taking into account both these stages, and we call his optimal choice � . We consider a simple adjust- ment mechanism for the informal organization.

n+1��n� = {

k�n if �n n < 0

n if �n n ≥ 0

} 0 < k < 1� (7)

The designer’s problem is to choose � to maximize 15

� �� ≡ �1�x1�� 1�� + �2�x2�� 2�� given

x1�� 1� = 1 2 + � + 1

2�� = k� ⇒ x2�� 2�� = 1 2 + � + k��

We find � =arg max� � = ��2+k��a−1�−2�1+�� 1�/ �2�4+2�d +k�+� 1+�1+k�2��. By way of contrast, the for- mal organization selected by a “myopic” designer (who is blind to the second period adjustment of the informal organi- zation) is �∗ = ��a − 1� − 2�1 + �� 1�/�2�2 + d + ��.

We first analyze the conditions for compensatory fit in the first period.

A3.2. Compensatory Fit with Adjustment of the Informal Organization

As before, the condition for the architecture (�� B� to generate compensatory fit is the condition when � < 0, x

∗ > 1 2 given

� > 0, 1 > 0. This is

1 > max {

�a − 1��2 + k� 2�1 + ��

�1 − a��2 + k� 2z

} � where

z = 1 + 2�1 + d� + 2k + ��1 + k�2 > 0� By symmetry, the conditions for the architecture �� A� to generate compensatory fit in period 1 is

1 < min {

�a − 1��2 + k� 2�1 + ��

�1 − a��2 + k� 2z

} �

The restrictions a ∈ �1 − �� + 1�/3� 1 + �� + 1�/3� and a ∈ �1−z/�2+k��1+z/�2+k�� on the marginal rate of techni- cal substitution between A and B type efforts ensure that 1

2 ≥

�a − 1��2 + k�/�2�� + 1�� ≥ − 1 2 and 1

2 ≥ �1 − a��2 + k�/z

≥ − 1 2 �

Proof. Suppose 1>0, then � <0⇒ 1> �a−1��2+k�/ �2�1+��. If a ≤ 1, this condition always holds, since the right-hand side is nonpositive. Now consider a > 1.

a ≤ 4 + � 3

⇒ a ≤ 1 + � + 1 3

⇒ 3�a − 1� ≤ � + 1 ⇒ �a − 1��2 + k� ≤ � + 1 ⇒ �a − 1��2 + k�

2�1 + �� ≤ 1 2

�

Similarly,

a ≥ 1 − � + 1 3

⇒ �a − 1��2 + k� 2�� + 1� ≥ −

1 2

�

Thus, given a ∈ �1 − �� + 1�/3� 1 + �1 + ��/3� we can be sure that there are admissible values of 1 such that 1

2 ≥

�a − 1��2 + k�/�2�� + 1�� ≥ − 1 2 �

Similarly, x1∗ > 1 2 if 1 > �1 − a��2 + k�/�2z� If a > 1, this

condition always holds, since the right-hand side is nonposi- tive. Now consider a ≤ 1. If

a > 1 − z/�2 + k� ⇒ z �2 + k� ≥ 1 − a

⇒ 1 2

≥ �1 − a��2 + k� 2z

�

By symmetry,

a ≤ 1 + z �2 + k� ⇒ −

1 2

≤ �1 − a��2 + k� 2z

�

Thus, given a ∈ �1 − z/�2 + k�� 1 + z/�2 + k�� we can be sure that there are admissible values of 1 such that 1

2 ≥

�1 − a��2 + k�/�2z� ≥ − 1 2 .

Note that the conditions for the formal and informal orga- nization to be inconsistent with each other are more strin- gent when the adjustment of the informal organization is anticipated—both in terms of the strength of the informal orga- nization, as well as the limit on the difference in marginal contribution of both types of effort. Therefore if � is incon- sistent with the informal organization, so is �∗

A3.1.1. What Happens in Period 3? Suppose we have architecture �� B� in the first period, then

⇒ 1 = > 0 and �1 < 0 ⇒ 2 = k�1 < 0 (from v).

Given �1 < 0, there is a misfit (in the sense that designer would ideally have preferred �1 > 0 for the second period) if 2 < �a − 1��2 + k�/�2�1 + ��.

3 = 2 < 0 (from v)� In period 3, � will be set again as �3 = ��2 + k��a − 1� − 2�1 + �� 3�/�2�4 + 2�d + k�� + ��1 + �1 + k�2��. This will result in an inconsistent organizational architecture �� A� if

3 < min {

�a − 1��2 + k� 2�1 + ��

�1 − a��2 + k� 2z

} �

Therefore, if the informal organization had adjusted strongly to the formal organization in period 2, there would be a mis- fit in period 2, and a reorganization in period 3 to a formal organization inconsistent with the informal organization. Else, there is a consistent architecture in period 3 despite gains from ambidexterity. Thus, whether there is a need for further reor- ganization after the adjustment of the informal organization depends critically on the rate of adjustment of the informal organization to the formal organization (i.e., the amount of adjustment in period 2).

A3.2.1. Switching Regimes: From Ambidexterity to Focus. Next, we analyze the case where starting from an inconsistent architecture, the informal organization remains unchanged, but � changes sign from the first stage to the second—a switch from a regime of gains from ambidexterity to one of gains from focus. We will assume that the switch in � occurs according to the following mechanism:

�2 = −j�1� j > 0� (8) As before, the designer cannot change the formal organization in the second stage. Taking into account both these stages, a foresighted designer therefore sets � and we call his optimal choice ��. The designer’s problem is to choose �� to maximize

�t �� 1� �2� ≡ �1�x1�� 1� + �2�x2�� 2� given

x1�� = x2�� = 1 2 + � + �

The condition for internal solution is now d + �t > 0 where �t = ��1 + �2�/2 = �1�1 − j�/2

We find �� = arg max� �t = ��a − 1� − 2�1 + �t ��/ �2�2 + d +2�t ��. In contrast a myopic designer would choose: �∗ = ��a − 1� − 2�1 + �1��/�2�2 + d + 2�1��

A3.3. Compensatory Fit with Switch in Ambidexterity/Focus Regimes

The condition for the architecture �� B� to generate compen- satory fit in period 1 (�� < 0� x

∗ > 1 2 given �1 > 0, > 0� is

> max��a − 1�/�2�1 + �t �� 1 − a�/�2�1 + d��. By symmetry, the mirror image architecture �� A� gen-

erates compensatory fit if < min��a − 1�/�2�1 + �t �� 1 − a�/�2�1 + d��. The restrictions a ∈ −�t � 2 + �t � and a ∈ −d� 2 + d� on the marginal rate of technical substitu- tion between A and B type efforts ensure that − 1

2 ≤ �a − 1�/

�2�1 + �t �� ≤ 1 2 and − 1

2 ≤ �1 − a�/�2�1 + d�� ≤ w1/2.

Note that the first part of the constraint for �� B� can be rewritten as j < �1 − a + �2 + �1� �/��1 � or as a con- straint on the magnitude of the gains from focus in the second period. Thus, there is a limit to the adjustment that is possi- ble between periods—if there are very large gains from focus in the second period, a foresighted designer will effectively ignore gains from ambidexterity in period 1 and select a con- sistent architecture.

A3.3.1. What Happens in Period 3? Suppose we have architecture �� B� in the first period.

⇒ > 0 and �� < 0� If < �a − 1�/�2�1 − j�1��, there is a misfit in the second period (in the sense that designer would ideally have preferred �� > 0 for the second period), and in period 3 the designer will reorganize towards a formal organization consistent with the informal organization. Therefore, a strong informal organiza- tion also protects against misfit and the need for reorganization when there is a change from gains from ambidexterity to focus. If > �a − 1�/�2�1 − j�1��, then the inconsistent architecture continues in period 3.

Endnotes 1This is distinct from external fit, which refers to the optimal alignment of an organization’s strategic choices with its envi- ronment. See Siggelkow (2001) and Miller (1992) for useful discussions of the distinction between the two notions of fit. 2For instance, the “McKinsey 7S” is a popular practitioner framework that embodies the belief in the importance of the formal and informal organization being consistent with each other in motivating the same kind of employee behavior. 3All quotes are drawn from our fieldwork involving interviews with 53 managers in various parts of the company, with the majority (36) drawn from the engineering groups, which was the locus of the reorganization. The interviews were conducted between December 2001 and March 2005. 4In a further refinement in May 2002, the company consoli- dated the 11 technology groups into 8, in order to put greater emphasis on routing and switching, the company’s core busi- ness. However, the basic principle of grouping engineering by technology instead of customer segment did not change, but was reinforced through this refinement. 5Analyst reactions drawn from the following sources: Net- work Briefing Daily, August 24, 2001, “Cisco scraps corporate structure, centralizes engineering”; InfoWorld, George Chidi, August 24, 2001, “Cisco’s reorganization cannot unpop the bubble”; InformationWeek, September 3, 2001, “Cisco rein- vents itself to extend its market reach”; Tele.com, September 3,

2001, “Rejiggered Cisco targets the telecom market by way of the enterprise.” 6Secondary data on post reorganization performance drawn from: Fortune, January 20, 2003, “The 100 best places to work in America”; Company Press Release, FY2002 Results; Mercury News, June 23, 2002, Jennifer Files, “Cisco leans to adjust to new reality”; see also the VAR satisfaction sur- vey reported in VARBusines.com by Hailey Lynne McKeefry “Cisco Systems By a Nose” http://www.varbusiness.com/ sections/98pages/204prod.asp. 7Empirical evidence on the emergence of social networks (and in particular on the effects of formal prescribed patterns of interaction on this emergence) is scarce. However, there is evi- dence that shows significant overlaps between formal author- ity relationships and informal networks of communication and influence (Ibarra 1993, Krackhardt 1990, Smith-Doerr and Powell 2005). It therefore appears plausible that formally spec- ified roles lead to informal communication and influence ties and perhaps ultimately to friendship ties (in particular, see Smith-Doerr and Powell 2005, pp. 384–385). 8These gradual changes to the informal organization may be a consequence of the fact that the processes underlying the emergence of informal organization—socialization, tie forma- tion, learning, and preference transformation—are themselves gradual. 9Though in our setting, employees do face two different kinds of tasks, we do not emphasize the multitasking problem famil- iar to agency theorists—our focus is not the optimal incentive contract between the designer and employees, but rather focus on how the choices of the two interact. 10Figure 5 can be generated by a number of production func- tions y = f �A� B� with budget constraints A + B = k that share the properties (a) of an internal stationary point and (b) whether the stationary point is a maxima or minima determined by the sign of the cross-partial derivative yA� B. 11As an additional rationale, it is preferable to engage in a com- bination of both activities if the costs of efforts towards each activity are convex. 12To keep the graphical exposition simple, we ignore any dif- ferences in the marginal rate of technical substitution between the two efforts (so that E lies at the center), as well as the costs of design (so that the designer’s actions are driven only by the shape of the performance function). The technical analysis in the appendix, of course, accounts for both these factors. 13Thus, in our model the informal organization plays no role in terms of creating variety that helps to adapt to changing circumstances—its function is primarily one of keeping the employee’s allocation of effort closer to a point preferred by the organization designer. 14Having obtained the results, of course, we have worked hard to make them intuitive and accessible without the need for reverting to the formalization, but this should not underplay the value of the formalization in generating the insight in the first place. 15Note that we ignore any discounting of the second period— explicitly modeling discounting does not alter our results qualitatively.

References Abrahamson, E., L. Rosenkopf. 1993. Institutional and competitive

bandwagons: Using mathematical modeling as a tool to explore. Acad. Management Rev. 18(3) 487–513.

Amburgey, T. A., D. Kelley, W. P. Barnett. 1993. Resetting the clock: The dynamics of organizational change and failure. Admin. Sci. Quart. 38 51–73.

Baker, G. 2002. Distortion and risk in optimal incentive contracts. J. Human Resources 37(4) 728–751.

Barnard, C. 1938. The Functions of the Executive. Harvard University Press, Cambridge, MA.

Baron, J. N., M. T. Hannan, M. D. Burton. 2001. Labor pains: Change in organizational models and employee turnover in young, high- tech firms. Amer. J. Sociol. 106(4) 960–1013.

Becker, H. S., B. Geer. 1960. Latent culture: A note on the theory of latent social roles. Admin. Sci. Quart. 5(2) 304–313.

Besanko, D., D. Dranove, M. Shanley. 2000. Economics of Strategy. John Wiley and Sons, New York.

Bhattacharya, R., T. M. Devinney, M. M. Pillutla. 1998. A formal model of trust based on outcomes. Acad. Management Rev. 23(3) 459–472.

Birkinshaw, J., C. Gibson. 2004. Building ambidexterity into an orga- nization. MIT Sloan Management Rev. 45(4) 47–55.

Brown, S. L., K. M. Eisenhardt. 1997. The art of continuous change: Linking complexity theory. Admin. Sci. Quart. 42(1) 1–34.

Bruderer, E., J. V. Singh. 1996. Organizational evolution, learning, and selection: A genetic-algorithm-based model. Acad. Man- agement J. 39(5) 1322–1349.

Burns, T., G. M. Stalker. 1961. The Management of Innovation. Tavistock, London.

Carroll, G. R., J. R. Harrison. 1998. Organizational demography and culture: Insights from a formal model and simulation. Admin. Sci. Quart. 43(3) 637–667.

Chandler, A. D. 1962. Strategy and Structure. MIT Press, Cambridge.

Child, J., R. G. McGrath. 2001. Organizations unfettered: Organiza- tional form in an information-intensive economy. Acad. Man- agement J. 44(6) 1135–1148.

Christensen, C. M. 1997. The Innovator’s Dilemma. HBS Publishing, Cambridge, MA.

Cremer, J. 1990. Common knowledge and the coordination of eco- nomic activities. M. Aoki, B. Gustaffson, O. E. Williamson, eds. The Firm as a Nexus of Treaties. Sage Publishers, London.

Drazin, R., A. H. van de Ven. 1985. Alternative forms of fit in contingency theory. Admin. Sci. Quart. 30(4) 514–539.

Ebben, J., A. C. Johnson. 2005. Efficiency, flexibility or both? Evi- dence linking strategy to performance in small firms. Strategic Management J. 26 1249–1259.

Eggers, J. P., S. Kaplan. 2009. Cognition and renewal: Compar- ing CEO and organizational effects on incumbent adaptation to technical change. Organ. Sci. 20(2) 461–477.

Evans, P., Y. Doz. 1989. The dualistic organization. A. Laurent, ed. Human Resource Management in International Firms. Macmil- lan, Basingstoke, UK.

Ghemawat, P., J. E. Costa. 1993. The organizational tension between static and dynamic efficiency. Strategic Management J. 14(8) 59–73.

Gibbons, R. 1998. Incentives in organizations. J. Econom. Perspec- tives 12(4) 115–132.

Gibson, C. B., J. Birkinshaw. 2004. The antecedents, consequences, and mediating role of organizational ambidexterity. Acad. Man- agement J. 47(2) 209–226.

Gulati, R. 2007. Silo busting—How to execute on the promise of customer focus. Harvard Bus. Rev. 85(5) 98–108.

Hannan, M. T., L. Polos, G. R. Carroll. 2003a. Cascading organiza- tional change. Organ. Sci. 14(5) 463–482.

Hannan, M. T., L. Polos, G. R. Carroll. 2003b. The fog of change: Opacity and asperity in organizations. Admin. Sci. Quart. 48(3) 399–432.

He, Z.-L., P.-K. Wong. 2004. Exploration vs. exploitation: An empirical test of the ambidexterity hypothesis. Organ. Sci. 15(4) 481–494.

Homans, G. C. 1950. The Human Group. Harcourt, Brace & Co., New York.

Ibarra, H. 1993. Network centrality, power, and innovation involve- ment: Determinants of technical and administrative roles. Acad. Management J. 36(3) 471–501.

Johansson, J. E., C. Krishnamurthy, H. Schlissberg. 2003. Solving the solutions problem. McKinsey Quart. 3 116–125.

Katila, R., G. Ahuja. 2002. Something old, something new: A longi- tudinal study of search behavior and new product introduction. Acad. Management J. 45(6) 1183–1194.

Kogut, B., U. Zander. 1996. What firms do? Coordination, identity, and learning. Organ. Sci. 7(5) 502–518.

Kohli, A. K., B. J. Jaworski. 1990. Market orientation—The con- struct, research propositions, and managerial implications. J. Marketing 54(2) 1–18.

Krackhardt, D. 1990. Assessing the political landscape: Structure, cognition, and power in organizations. Admin. Sci. Quart. 35(2) 342–369.

Lamont, B. T., R. J. Williams. 1994. Performance during “M-form” reorganization and recovery time: The effects of prior strategy and. Acad. Management J. 37(1) 153–166.

Lave, C. A., J. G. March. 1993. An Introduction to Models in the Social Sciences. University Press of America, Lanham, MD.

Lawrence, P. R., J. W. Lorsch. 1967. Organizations and Environment: Managing Differentiation and Integration. Harvard University Press, Cambridge, MA.

Levinthal, D. A., J. G. March. 1993. The myopia of learning. Strate- gic Management J. 14(8) 95–112.

March, J. G. 1991. Exploration and exploitation in organizational learning. Organ. Sci.: J. Inst. Management Sci. 2(1) 71–87.

March, J. G., H. A. Simon. 1958. Organizations. Wiley Press, New York.

Miles, M. B., A. M. Huberman. 1994. Qualitative Data Analysis. SAGE Publications, London.

Milgrom, P., J. Roberts. 1990. The economics of modern manufac- turing: Technology, strategy, and organization. Amer. Econom. Rev. 80(3) 511–528.

Milgrom, P., J. Roberts. 1992. Economics, Organization and Man- agement. Prentice Hall, Englewood Cliffs, NJ.

Milgrom, P., J. Roberts. 1995. Complementarities and fit: Strat- egy, structure, and organizational change in manufacturing. J. Accounting Econom. 19(2/3) 179–208.

Miller, D. 1992. Environmental fit versus internal fit. Organ. Sci. 3(2) 159–178.

Miller, D., P. Friesen. 1984. Organizations: A Quantum View. Prentice Hall, Englewood Cliffs, NJ.

Mintzberg, H. 1990. The design school: Reconsidering the basic premises of strategic management. Strategic Management J. 11(3) 171–195.

Nadler, D. A., M. L. Tushman. 1997. Competing by Design: The Power of Organizational Architecture. Oxford University Press, Oxford, UK.

Nadler, D. A., M. L. Tushman. 1998. Competing by design. Executive Excellence 15(8) 12–13.

Nickerson, J. A., T. R. Zenger. 2002. Being efficiently fickle: A dynamic theory of organizational choice. Organ. Sci.: J. Inst. Management Sci. 13(5) 547–566.

Oxman, J. A., B. D. Smith. 2003. The limits of structural change. MIT Sloan Management Rev. 45(1) 77–82.

Porter, M. E. 1985. Competitive Advantage. Free Press, New York.

Prendergast, C. 1999. The provision of incentives in firms. J. Econom. Literature XXXVII 7–63.

Puranam, P., H. Singh, H. Zollo, 2006. Organizing for innovation: Managing the coordination-autonomy dilemma in technology acquisitions. Acad. Management J. 49(2) 263–280.

Rivkin, J. W., N. Siggelkow. 2003. Balancing search and stabil- ity: Interdependencies among elements of organizational design. Management Sci. 49(3) 290–311.

Romanelli, E., M. L. Tushman. 1994. Organizational transformation as punctuated equilibrium: An empirical test. Acad. Manage- ment J. 37(5) 1141–1666.

Roethlisberger, F. J., W. J. Dickson. 1939. Management and the Worker. Harvard University Press, Cambridge, MA.

Rust, R., C. Moorman, P. Dickson. 2002. Getting return on quality: Revenue expansion, cost reduction, or both? J. Marketing 66(4) 7–24.

Scott, W. 1998. Organizations: Rational, Natural, and Open Systems, 4th ed. Prentice Hall, Upper Saddle River, NJ.

Siggelkow, N. 2001. Change in the presence of fit: The rise, the fall, and the renaissance of liz claiborne. Acad. Management J. 44(4) 838–857.

Siggelkow, N. 2002. Misperceiving interactions among complements and substitutes: Organizational consequences. Management Sci. 48(7) 900–916.

Siggelkow, N., D. A. Levinthal. 2003. Temporarily divide to con- quer: Centralized, decentralized, and reintegrated organizational approaches to exploration and adaptation. Organ. Sci.: J. Inst. Management Sci. 14(6) 650–669.

Simon, H. A. 1957. Administrative Behavior. Macmillan, New York.

Smith, W. K., M. L. Tushman. 2005. Managing strategic contra- dictions: A top management model for managing innovation streams. Organ. Sci. 16(5) 522–536.

Smith-Doerr, L., W. W. Powell. 2005. Networks and economic life. R. Swedberg, ed. The Handbook of Economic Sociology. Russel Sage Foundation, New York, 379–402.

Stinchcombe, A. 1965. Social structure and organizations. J. March, ed. Handbook of Organizations. Rand McNally, Chicago.

Thompson, J. D. 1967. Organizations in Action. McGraw Hill, New York.

Tripsas, M. 2009. Technology, identity, and inertia through the lens of “The Digitial Photography Company.” Organ. Sci. 20(2) 441-460.

Tushman, M., E. Romanelli. 1985. Organizational evolution: A meta- morphosis model of convergence and evolution. B. Staw, ed. Research in Organizational Behavior, Vol. 7. JAI Press, Greenwich, CT, 171–222.

Tushman, M., W. Smith, R. Wood, G. Westerman, C. A. O’Reilly. 2004. Innovation streams and ambidextrous organizational designs: On building dynamic capabilities. Working paper.

Tushman, M. L., C. A. O’Reilly. 1996. Ambidextrous organizations: Managing evolutionary and revolutionary change. California Management Rev. 38(4) 8–30.

Wageman, R. 1995. Interdependence and group effectiveness. Admin. Sci. Quart. 40(1) 145–180.

Yin, R. K. 1994. Case Study Research: Design and Methods. Sage, Newbury Park, CA.

OrganizationScience Vol. 20, No. 2, March–April 2009, pp. 478–480 issn 1047-7039 � eissn 1526-5455 � 09 � 2002 � 0478

informs ® doi 10.1287/orsc.1090.0425

About Authors

Rajshree Agarwal (“Strategic Renewal of Organi- zations”) is the John Georges Professor of Technol- ogy Management and Strategy at the University of Illi- nois at Urbana-Champaign. She received her Ph.D. in economics from the State University of New York at Buffalo. Her research interests focus on the implica- tions of entrepreneurship and innovation for industry and firm evolution, and her recent work examines employee mobility and entrepreneurship as sources of knowledge diffusion. Address: College of Business, University of Illinois at Urbana-Champaign, 350 Wohlers Hall, Cham- paign, IL 61820; e-mail: [email protected]. David Benson (“Corporate Venture Capital as a Win-

dow on New Technologies: Implications for the Per- formance of Corporate Investors When Acquiring Star- tups”) is a Ph.D. candidate of strategy at the Stephen M. Ross School of Business, University of Michigan. His research focuses on the role of established firms in funding, collaborating with, and shaping the direction of entrepreneurial startups. Address: Marriott School of Management, Brigham Young University, Provo, UT 84602; e-mail: [email protected]. Laurence Capron (“Selection Capability: How Capa-

bility Gaps and Internal Social Frictions Affect Internal and External Strategic Renewal”) is an associate pro- fessor of strategy at INSEAD and Research Director of the INSEAD-Wharton alliance. Her research interests include mergers and acquisitions, alliances, corporate strategy, and modes of resource acquisition. Address: INSEAD, Boulevard de Constance, 77305 Fontainebleau, France; e-mail: [email protected]. Saikat Chaudhuri (“Integrating Acquired Capabili-

ties: When Structural Integration is (Un)necessary”) is an assistant professor of management at The Whar- ton School of the University of Pennsylvania. He earned his doctorate in business administration from Harvard University. His research encompasses techno- logical innovation, mergers and acquisitions, outsourc- ing, and organizational adaptation in dynamic environ- ments. Address: Management Department, The Whar- ton School, University of Pennsylvania, 2000 Steinberg Hall-Dietrich Hall, 3620 Locust Walk, Philadelphia, PA 19104; e-mail: [email protected]. J. P. Eggers (“Cognition and Renewal: Comparing

CEO and Organizational Effects on Incumbent Adapta- tion to Technical Change”) is an assistant professor of management and organizations at the Stern School of Business, New York University. He received his Ph.D.

from The Wharton School of the University of Penn- sylvania. His research interests include the dynamics of emerging technological markets and managerial decision making under uncertainty. Address: 44 West Fourth Street, Suite 7-150, New York University, New York, NY 10012; e-mail: [email protected]. Ranjay Gulati (“Renewal Through Reorganization:

The Value of Inconsistencies Between Formal and Infor- mal Organization”) is the Michael Ludwig Nemmers Distinguished Professor of Strategy and Organizations at the Kellogg School of Management, Northwestern Uni- versity. He received his Ph.D. in organizational behavior from Harvard University. His research interests include the dynamics of social networks, with a focus on the antecedents and consequences of social structure on economic exchange relationships between firms. He has written extensively on the impact of networks on entrepreneurial firms and has recently published a book entitled Managing Network Resources: Alliances, Affil- iations, and Other Relational Assets. His research has been published or is forthcoming in Administrative Sci- ence Quarterly, Managerial and Decision Economics, Strategic Management Journal, California Management Review, and Strategic Management Journal. Address: Harvard Business School, Soldiers Field Park, Boston, MA 02163; e-mail: [email protected]. Constance E. Helfat (“Strategic Renewal of Orga-

nizations”) is the J. Brian Quinn Professor of Technology and Strategy at the Tuck School of Busi- ness, Dartmouth College. Her research focuses on firm capabilities, including technological innovation and firm adaptation and change. She also has conducted research on corporate executives, including women executives. Address: Tuck School of Business, Dart- mouth College, 100 Tuck Hall, Hanover, NH 03755; e-mail: [email protected]. Sarah Kaplan (“Cognition and Renewal: Comparing

CEO and Organizational Effects on Incumbent Adap- tation to Technical Change”) is an assistant professor of management at The Wharton School of the Uni- versity of Pennsylvania. She received her Ph.D. from Sloan School of Management, Massachusetts Institute of Technology and is a coauthor of the book, Cre- ative Destruction. Her research seeks to unpack the role of cognitive framing in technology evolution, het- erogeneous organizational response to discontinuities, and strategy making inside organizations. Address: The Wharton School, University of Pennsylvania, 3620

478

Locust Walk, Suite 2019, Philadelphia, PA 19104, e-mail: [email protected]. Hann Earl Kim (“Innovation and Strategic Renewal

in Mature Markets: A Study of the Tennis Racket Indus- try”) is an associate professor of strategy and innovation at the KDI School of Public Policy and Management in Seoul, Korea. He received his Ph.D. from The Wharton School of the University of Pennsylvania and served on the faculty of Australian Graduate School of Manage- ment, University of New South Wales before joining the KDI School. His research interests include the interac- tion between supply and demand of innovation in indus- try evolution and the emergence and operation of large business groups. Address: KDI School of Public Policy and Management, 87 Hoegiro, Dongdaemun-gu, Seoul 130-868, Korea; e-mail: [email protected]. Anne Marie Knott (“Firm R&D Behavior and Evolv-

ing Technology in Established Industries”) is an assistant professor of strategy at Olin Business School, Washing- ton University in St. Louis. She received her Ph.D. from the University of California, Los Angeles. Her research primarily focuses on firm strategies and market condi- tions where firms operating in their own self-interest generate innovation and sustainable economic growth. Address: Olin Business School, Washington University in St. Louis, Campus Box 1133, 1 Brookings Drive, St. Louis, MO 63130-4899; e-mail: [email protected]. Will Mitchell (“Selection Capability: How Capability

Gaps and Internal Social Frictions Affect Internal and External Strategic Renewal”) is the J. Rex Fuqua Pro- fessor of International Management at Duke University’s Fuqua School of Business. He studies business dynam- ics, investigating how businesses change as the envi- ronments in which firms compete change and, in turn, how the business changes contribute to ongoing corpo- rate success or failure. Address: Fuqua School of Busi- ness, Duke University, 1 Towerview Drive, Durham, NC 27708; e-mail: [email protected]. Johannes M. Pennings (“Innovation and Strategic

Renewal in Mature Markets: A Study of the Tennis Racket Industry”) is Marie and Joseph Melone Pro- fessor of Management and Sociology at The Whar- ton School of the University of Pennsylvania. He received his doctorate from the University of Michi- gan in 1973. His present research revolves around innovation and technological trajectories in the imag- ing sector addressing recombinative capabilities and the propensity for market incumbents to forge break- through innovations together with the strategic effects of interfirm networks derived from proprietary knowl- edge through a firm’s position as high-status broker. Address: Department of Management, The Wharton School, University of Pennsylvania, Philadelphia, PA 19104; e-mail: [email protected]. Hart E. Posen (“Firm R&D Behavior and Evolving

Technology in Established Industries”) is an assistant

professor of strategy at the Stephen M. Ross School of Business, University of Michigan. He received his Ph.D. in strategy from The Wharton School of the Uni- versity of Pennsylvania in 2005. Prior to his return to graduate school, he was an entrepreneur in both the technology and retailing sectors. His research addresses the barriers facing firms in profiting from innova- tion. In particular, his research has addressed ques- tions related to the lifespan of advantage from R&D investments, incentives to make such investments, and the survival/failure consequences of alternative inno- vation strategies (e.g., exploration versus exploitation). His research has been published in the Strategic Man- agement Journal, Administrative Science Quarterly, and Organization Science. Address: Ross School of Busi- ness, University of Michigan, 701 Tappan Street, Ann Arbor, MI 48109; e-mail: [email protected]. Phanish Puranam (“Integrating Acquired Capabil-

ities: When Structural Integration is (Un)necessary” and “Renewal Through Reorganization: The Value of Inconsistencies Between Formal and Informal Orga- nization”) is an associate professor of strategic and international management and also Co-Director of the Aditya Birla India Centre at London Business School. His research interests include organization design and interorganizational coordination, both between firms as well as within firms. Address: London Busi- ness School, Regent’s Park, London NW1 4SA, UK; e-mail: [email protected]. Carlo Salvato (“Capabilities Unveiled: The Role

of Ordinary Activities in the Evolution of Product Development Processes”) is an associate professor of strategic management at Bocconi University, Milan. He received a Ph.D. from Bocconi University and a second Ph.D. from Jönköping International Busi- ness School, Sweden. His research interests include dynamic capabilities, strategic change, and perfor- mance learning. Address: Management Department, Bocconi University, Via Röntgen 1, 20136 Milan, Italy; e-mail: [email protected]. Harbir Singh (“Integrating Acquired Capabilities:

When Structural Integration is (Un)necessary”) is the Mack Professor of Management, Vice Dean for Global Initiatives, and Co-Director of the Mack Center for Tech- nological Innovation at The Wharton School of the Uni- versity of Pennsylvania. He received his Ph.D. from the University of Michigan. His research includes strategies for corporate acquisitions, corporate governance, joint ventures, management buyouts, and corporate restruc- turing. Address: Management Department, The Whar- ton School, University of Pennsylvania, 2000 Steinberg Hall-Dietrich Hall, 3620 Locust Walk, Philadelphia, PA 19104; e-mail: [email protected]. Mary Tripsas (“Technology, Identity, and Inertia:

Through the Lens of ‘The Digital Photography Com- pany’ ”) is an associate professor of business admin- istration at the Harvard Business School. She earned

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

<< /ASCII85EncodePages false /AllowTransparency false /AutoPositionEPSFiles true /AutoRotatePages /None /Binding /Left /CalGrayProfile (Dot Gain 20%) /CalRGBProfile (sRGB IEC61966-2.1) /CalCMYKProfile (U.S. Web Coated \050SWOP\051 v2) /sRGBProfile (sRGB IEC61966-2.1) /CannotEmbedFontPolicy /Error /CompatibilityLevel 1.3 /CompressObjects /Tags /CompressPages true /ConvertImagesToIndexed true /PassThroughJPEGImages true /CreateJDFFile false /CreateJobTicket false /DefaultRenderingIntent /Default /DetectBlends true /DetectCurves 0.0000 /ColorConversionStrategy /CMYK /DoThumbnails false /EmbedAllFonts true /EmbedOpenType false /ParseICCProfilesInComments true /EmbedJobOptions true /DSCReportingLevel 0 /EmitDSCWarnings false /EndPage -1 /ImageMemory 1048576 /LockDistillerParams false /MaxSubsetPct 100 /Optimize true /OPM 1 /ParseDSCComments true /ParseDSCCommentsForDocInfo true /PreserveCopyPage true /PreserveDICMYKValues true /PreserveEPSInfo true /PreserveFlatness true /PreserveHalftoneInfo false /PreserveOPIComments true /PreserveOverprintSettings true /StartPage 1 /SubsetFonts true /TransferFunctionInfo /Apply /UCRandBGInfo /Preserve /UsePrologue false /ColorSettingsFile () /AlwaysEmbed [ true ] /NeverEmbed [ true ] /AntiAliasColorImages false /CropColorImages true /ColorImageMinResolution 300 /ColorImageMinResolutionPolicy /OK /DownsampleColorImages true /ColorImageDownsampleType /Bicubic /ColorImageResolution 300 /ColorImageDepth -1 /ColorImageMinDownsampleDepth 1 /ColorImageDownsampleThreshold 1.50000 /EncodeColorImages true /ColorImageFilter /DCTEncode /AutoFilterColorImages true /ColorImageAutoFilterStrategy /JPEG /ColorACSImageDict << /QFactor 0.15 /HSamples [1 1 1 1] /VSamples [1 1 1 1] >> /ColorImageDict << /QFactor 0.15 /HSamples [1 1 1 1] /VSamples [1 1 1 1] >> /JPEG2000ColorACSImageDict << /TileWidth 256 /TileHeight 256 /Quality 30 >> /JPEG2000ColorImageDict << /TileWidth 256 /TileHeight 256 /Quality 30 >> /AntiAliasGrayImages false /CropGrayImages true /GrayImageMinResolution 300 /GrayImageMinResolutionPolicy /OK /DownsampleGrayImages true /GrayImageDownsampleType /Bicubic /GrayImageResolution 300 /GrayImageDepth -1 /GrayImageMinDownsampleDepth 2 /GrayImageDownsampleThreshold 1.50000 /EncodeGrayImages true /GrayImageFilter /DCTEncode /AutoFilterGrayImages true /GrayImageAutoFilterStrategy /JPEG /GrayACSImageDict << /QFactor 0.15 /HSamples [1 1 1 1] /VSamples [1 1 1 1] >> /GrayImageDict << /QFactor 0.15 /HSamples [1 1 1 1] /VSamples [1 1 1 1] >> /JPEG2000GrayACSImageDict << /TileWidth 256 /TileHeight 256 /Quality 30 >> /JPEG2000GrayImageDict << /TileWidth 256 /TileHeight 256 /Quality 30 >> /AntiAliasMonoImages false /CropMonoImages true /MonoImageMinResolution 1200 /MonoImageMinResolutionPolicy /OK /DownsampleMonoImages true /MonoImageDownsampleType /Bicubic /MonoImageResolution 1200 /MonoImageDepth -1 /MonoImageDownsampleThreshold 1.50000 /EncodeMonoImages true /MonoImageFilter /CCITTFaxEncode /MonoImageDict << /K -1 >> /AllowPSXObjects false /CheckCompliance [ /None ] /PDFX1aCheck false /PDFX3Check false /PDFXCompliantPDFOnly false /PDFXNoTrimBoxError true /PDFXTrimBoxToMediaBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXSetBleedBoxToMediaBox true /PDFXBleedBoxToTrimBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXOutputIntentProfile () /PDFXOutputConditionIdentifier () /PDFXOutputCondition () /PDFXRegistryName () /PDFXTrapped /False /Description << /CHS <FEFF4f7f75288fd94e9b8bbe5b9a521b5efa7684002000410064006f006200650020005000440046002065876863900275284e8e9ad88d2891cf76845370524d53705237300260a853ef4ee54f7f75280020004100630072006f0062006100740020548c002000410064006f00620065002000520065006100640065007200200035002e003000204ee553ca66f49ad87248672c676562535f00521b5efa768400200050004400460020658768633002> /CHT <FEFF4f7f752890194e9b8a2d7f6e5efa7acb7684002000410064006f006200650020005000440046002065874ef69069752865bc9ad854c18cea76845370524d5370523786557406300260a853ef4ee54f7f75280020004100630072006f0062006100740020548c002000410064006f00620065002000520065006100640065007200200035002e003000204ee553ca66f49ad87248672c4f86958b555f5df25efa7acb76840020005000440046002065874ef63002> /DAN <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> /DEU <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> /ESP <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> /FRA <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> /ITA <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> /JPN <FEFF9ad854c18cea306a30d730ea30d730ec30b951fa529b7528002000410064006f0062006500200050004400460020658766f8306e4f5c6210306b4f7f75283057307e305930023053306e8a2d5b9a30674f5c62103055308c305f0020005000440046002030d530a130a430eb306f3001004100630072006f0062006100740020304a30883073002000410064006f00620065002000520065006100640065007200200035002e003000204ee5964d3067958b304f30533068304c3067304d307e305930023053306e8a2d5b9a306b306f30d530a930f330c8306e57cb30818fbc307f304c5fc59808306730593002> /KOR <FEFFc7740020c124c815c7440020c0acc6a9d558c5ec0020ace0d488c9c80020c2dcd5d80020c778c1c4c5d00020ac00c7a50020c801d569d55c002000410064006f0062006500200050004400460020bb38c11cb97c0020c791c131d569b2c8b2e4002e0020c774b807ac8c0020c791c131b41c00200050004400460020bb38c11cb2940020004100630072006f0062006100740020bc0f002000410064006f00620065002000520065006100640065007200200035002e00300020c774c0c1c5d0c11c0020c5f40020c2180020c788c2b5b2c8b2e4002e> /NLD (Gebruik deze instellingen om Adobe PDF-documenten te maken die zijn geoptimaliseerd voor prepress-afdrukken van hoge kwaliteit. De gemaakte PDF-documenten kunnen worden geopend met Acrobat en Adobe Reader 5.0 en hoger.) /NOR <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> /PTB <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> /SUO <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> /SVE <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> /ENU (Use these settings to create Adobe PDF documents best suited for high-quality prepress printing. Created PDF documents can be opened with Acrobat and Adobe Reader 5.0 and later.) >> /Namespace [ (Adobe) (Common) (1.0) ] /OtherNamespaces [ << /AsReaderSpreads false /CropImagesToFrames true /ErrorControl /WarnAndContinue /FlattenerIgnoreSpreadOverrides false /IncludeGuidesGrids false /IncludeNonPrinting false /IncludeSlug false /Namespace [ (Adobe) (InDesign) (4.0) ] /OmitPlacedBitmaps false /OmitPlacedEPS false /OmitPlacedPDF false /SimulateOverprint /Legacy >> << /AddBleedMarks false /AddColorBars false /AddCropMarks false /AddPageInfo false /AddRegMarks false /ConvertColors /ConvertToCMYK /DestinationProfileName () /DestinationProfileSelector /DocumentCMYK /Downsample16BitImages true /FlattenerPreset << /PresetSelector /MediumResolution >> /FormElements false /GenerateStructure false /IncludeBookmarks false /IncludeHyperlinks false /IncludeInteractive false /IncludeLayers false /IncludeProfiles false /MultimediaHandling /UseObjectSettings /Namespace [ (Adobe) (CreativeSuite) (2.0) ] /PDFXOutputIntentProfileSelector /DocumentCMYK /PreserveEditing true /UntaggedCMYKHandling /LeaveUntagged /UntaggedRGBHandling /UseDocumentProfile /UseDocumentBleed false >> ] >> setdistillerparams << /HWResolution [2400 2400] /PageSize [612.000 792.000] >> setpagedevice