Manag Int Rev (2010) 50:513–532 DOI 10.1007/s11575-010-0044-1 R e s e a r c h Art i c l e
Regional Management as a System A Longitudinal Case Study Rebecca Piekkari · Phillip C. Nell · Pervez N. Ghauri
Abstract: We apply information processing theory to investigate the evolution of regional management. Based on a longitudinal case study of a European multinational we capture the overall structural development of the firm and within this context, we conduct an in-depth analysis of two regions and their regional management centres. We conceptualise regional management as a system of differentiated centres. The notion that headquarters can be found in a single location is replaced by the view that headquarters’ activities are spatially distributed within a system. Over time, this multi-level system adapts to changing information processing requirements.
0 0
0
Keywords: Regional management · System · Information processing · Regional headquarters
Received: 26.02.2008 / Revised: 08.09.2009 / Accepted: 20.10.2009 / Published online: 15.07.2010 © Gabler-Verlag 2010 Prof. R. Piekkari () Department of Management, School of Economics, Aalto University, Helsinki, Finland e-mail:
[email protected] Ass. Prof. P. C. Nell Center for Strategic Management and Globalization, Copenhagen Business School, Frederiksberg, Denmark Prof. P. N. Ghauri Department of Management, School of Social Science and Public Policy, King’s College London, London, United Kingdom
514
R. Piekkari et al.
Introduction While regional structures are considered increasingly important in today’s regionallystructured world, this topic has attracted relatively limited scholarly attention (Enright 2005a, b; Ghemawat 2003; Paik and Sohn 2004; Schütte 1996; Rugman 2005; Rugman and Verbeke 2004). Previous work on regional structures has been conducted at the corporate level of the organization, that is the “macro level” (Egelhoff 1982, 1988; Stopford and Wells 1972; Wolf and Egelhoff 2002) and at the unit level of individual regional management centres (Daniels 1987; Lasserre 1996; Schütte 1996; Williams 1967; Yeung et al. 2001). In this paper, we define regional management centres as organizational units concerned with, and involved in, the control, coordination or integration of activities of one or more subsidiaries in the region (Schütte 1996). These two streams of literature are, however, surprisingly disconnected, and thus call for a more holistic analysis of regional management. The first stream on the macro-level has produced important insights into the circumstances under which multinational corporations (MNCs) introduce regional structures or “area divisions” (Egelhoff 1982, 1988; Stopford and Wells 1972; Wolf and Egelhoff 2002). Since the regional level is often introduced to improve the information processing capacity of the firm, Egelhoff (1982, 1988) has successfully used this very perspective to advance our understanding of regional management. While being theoretically rigorous, Egelhoff’s (1982, 1988) work, however, does not recognize internal differences between regions or regional centres.1 Nor does he explain the adaptation processes firms undergo when matching information processing capacities with requirements since his research is mostly cross-sectional. The second stream, in turn, offers some insight into these matters. For example, it has identified different types of regional management centres and investigated the challenges associated with implementing them (see e.g., Aoki and Tachiki 1992; Daniels 1986). Nevertheless, this stream is rather managerial and descriptive in its approach. Due to its strong focus on individual units, particularly the regional headquarters, it fails to acknowledge interdependencies between several regional management centres within one firm (for exceptions, see Lehrer and Asakawa 1999; Enright 2005a). Therefore, the present study aims to bridge both streams of literature at their intersection. We advance the understanding of regional management by combining the strengths of the rigorous information processing approach with the more fine-grained perspective on individual units. Coupled with our longitudinal methodology, we are able to reach beyond previous work and contribute with theoretical insights about regional management. Based on our findings, we suggest a novel conceptualization of regional management as a system. Our findings indicate that a regional management system does provide information processing capacity. It consists of a set of differentiated regional management centres which adapts over time to internal and external contingencies. Hence, we find that matching information processing requirements with capacities is a complex process of evolution and adaptation rather than of instant fit. We also find that our concept of a regional management system allows us to reinterpret the implementation problems of regional organizations which we found in the case company. Contrary to previous information processing logic, we view such challenges as triggers of adaptation processes rather than inefficiencies. We conclude with a discussion of the role of headquarters as an
Regional Management as a System
515
initiator of adaptation processes and emphasize the increasing geographical distribution of headquarters’ activities within the management system. The remainder of the paper is organised as follows. We proceed by reviewing the literature on regional management. We then detail our longitudinal case study and discuss our findings both chronologically as well as thematically. In the conclusion, we provide insights at the intersection of the two streams of research. Theoretical Background One important approach to studying MNC structures is information processing theory. The theory was, to a large extent, developed by Galbraith (1969, 1973, 1977), Thompson (1967), and by Egelhoff (1982, 1988) who applied it to the MNC. In this perspective, organizations, specific coordination mechanisms, or individuals are seen as possessors of information processing capacity. They gather data, transform it into information, communicate it and finally store information (Galbraith 1973). At the firm level, it has been suggested that external and internal contingencies require certain levels and types of information processing. For example, firm growth and internationalization escalate the requirements for information processing due to increased complexity (Egelhoff 1991; Thihanyi and Thomas 2005; Wolf and Egelhoff 2002). It is also argued that firms perform well when information processing requirements fit the information processing capacities of the organizational structure (Egelhoff 1991). One of the MNC structures investigated by Egelhoff (1982) at the macro level was the regional structure which is usually associated with the use of regional headquarters. In contrast to other structures, such as the product division structure, regional structures are assumed to yield high strategic and tactical information processing capacity for company, country and product matters within the region (Egelhoff 1982). They have been found to work well under contingencies of high levels of foreign manufacturing and R&D, low to medium foreign product diversity, and high levels of product modifications (Egelhoff 1982). Furthermore, regional structures are supposed to relieve corporate levels of information overload once the company becomes large and complex (Wolf and Egelhoff 2001). To this end, the information processing theory has produced valuable insights into the regional dimension of MNCs. However, it has some important limitations which have only partly been resolved by the unit-level research. The macro-level approach has produced a picture of a uniform area structure and has thus neglected any internal differences between regions or regional management centres. Inter-regional differences are important to study since they might reflect diverse information processing requirements and even lead to differentiated regional arrangements (Rugman 2005). To this point, the unit-level literature has already identified a wide range of regional management centres (Erramilli 2003; Enright 2005a, b; Lasserre 1996). More importantly, two similar centres might still yield different information processing capacities. Schütte (1997) and Paik and Sohn (2004) report, for example, that two regional headquarters may differ in this respect because their relationships with corporate headquarters vary. Here, the problem of information processing theory lies in its treatment of information processing capacity
516
R. Piekkari et al.
as an abstract construct. This leads to the assumption that a particular structure always yields the same information processing capacity. However, while the unit-level literature provides a more fine-grained view on regional management, this stream focuses strongly on individual units only. The few exceptions in this stream are Enright (2005a) and Lehrer and Asakawa (1999), who point out that firms might use different types of regional management centres simultaneously. For example, Lehrer and Asakawa (1999) consider the interplay between units and report that over time the importance of smaller, more flexible and specialised regional units and virtual arrangements increased at the expense of fullblown regional headquarters in Japanese firms. Another limitation of prior research is that it has primarily been cross-sectional, thus producing only limited insight into how headquarters initiate adaptation processes and how a close fit is achieved. As an exception, Lasserre (1996) used his typology to develop an evolutionary model of regional headquarters, suggesting that they start off being very entrepreneurially oriented. Over time, when subsidiaries become more mature, regional headquarters take over more coordinating or integrative functions, and finally disappear when they perform neither important coordinative nor entrepreneurial functions. Yet, Forsgren (2008) is sceptical about whether corporate headquarters possess sufficient knowledge to re-design information processing capacities so that they match the requirements. More generally, the unit-level literature demonstrates firms’ (in)ability to quickly and effectively implement or dissolve regional structures. For example, Aoki and Tachiki (1992) describe divisional rivalries and Daniels (1986) and Schütte (1996) point to resistance from strong subsidiaries. Paik and Sohn (2004) even argue that regional management centres may seldom be as beneficial as expected even if they are implemented in situations that require regional structures. In essence, these findings cast doubts on the assumption in the macro level literature that structural adaption to certain contingencies, in other words structural fit, will lead to a fit between realised information processing requirements and realized capacities. In view of the extant literature, we develop the following broad research questions: How does regional management evolve over time? What explains this evolution? Case Study Methodology To address our research question, we adopt a longitudinal case study approach which allows us to obtain rich data and explain holistically the complex and dynamic phenomenon of regional management (Ghauri 2004). We conducted theoretical sampling at two levels (Fletcher and Plakoyiannaki 2011)— the overall company, and the regional level. First, the company should be in a relatively mature stage of internationalization with a large number of foreign subsidiaries. Under such circumstances, the question of if and how to set up regional management becomes relevant (Egelhoff 1988). Second, it should have operations in at least two regions to allow cross-regional comparison. Third, it should have a regional structure comprising not only regional headquarters but also other types of centres. When we started the first round of data collection in 1993, Kone, a publicly listed company and one of the most international firms in Finland, met these criteria. Kone’s total sales at that time amounted
Regional Management as a System
517
to FIM 10.8 billion, with approximately 21,000 employees worldwide (Kone Annual Report 1993). Foreign sales accounted for 95% of total sales; personnel outside Finland had reached 85% of total employees worldwide. Kone’s business operations concerned elevators and escalators in all major markets in Europe, Asia and North America. From a structural perspective, Kone had a global matrix, but was strongly dominated by global product divisions. It was not the only company that met our selection criteria, but given that the first author had established a trusting relationship with their top management, a prerequisite for undertaking a processual study was met. The sampling of the regions covered by Kone was based on our desire to contrast and compare operations in different regional contexts (see e.g., Fletcher and Plakoyiannaki 2011). Based on suggestions in recent literature, we assumed that different regions require different information processing (Rugman 2005). We selected two of the firm’s five regions, Northern Europe and Asia-Pacific, which were very different at the start of data collection in 1993. The mature subsidiaries in Northern Europe represented the oldest foreign markets of the corporation. The distant region of Asia-Pacific was relatively new for Kone, with rather young subsidiaries. It offered significant growth potential at a time when elevator markets in Europe were stagnating. The yearly average growth rate in Asia-Pacific exceeded 60% in the 1970s and 1980s and was still strongly positive in the 1990s; the lowest growth rate being 17% in Hong Kong.2 In addition, strong volatility could be observed, for example, in Singapore where the market declined more than 30% in one year. In Northern Europe, growth rates for new elevators were slightly positive during the 1970s and 1980s, with subsidiary growth ranging from 0.5% to 4.5%. Yet in the 1990s, growth rates in the region became negative, except for Denmark, which continued to show modest growth. Our research design allowed us to study the overall structural development of Kone (i.e. at the macro level) and, within this context, to conduct an in-depth analysis of two regions and their regional management centres (i.e. the unit level). The study included a total of eight regional management centres (three in Northern Europe and five Asia-Pacific): Two regional headquarters, one sub-regional headquarters and five regional coordination centres for engineering of new elevators, modernisation services and marketing. Data Collection and Analysis During fieldwork between 1993 and 2008, retrospective and real-time data were collected at several points in time to capture process dynamics. The first round of data collection was conducted in Northern Europe during 1993–94. The second round took place during 1994–1995 and primarily concerned Asia-Pacific and three other Kone regions. The third round, in 2002, involved interviewing current and former CEOs of Kone. The final interview round, interviews with regional heads and regional managers, was carried out in 2008. The multiple points of data collection allowed us to capture the structural evolution at Kone during four decades, between 1968 and 2008. The major source of data was open-ended interviews; 70 persons were interviewed in total; 28 represented Northern Europe, 35 Asia-Pacific, and seven persons represented corporate headquarters. Unlike much of the earlier research, our data represented three organizational levels: The corporate (7), regional (29) and local subsidiary levels (34). Most key informants had been per-
518
R. Piekkari et al.
sonally involved in making strategic decisions concerning regional management or were directly affected by it. For example, seven of the eight Kone CEOs during 1968–2008 were interviewed. We acknowledge that the collection of historical information entails a risk of retrospective bias and memory decay. However, we used triangulation to contrast and compare the data from key informants (Huber and Power 1985). We also compared our findings across multiple units of the MNC (Marschan-Piekkari et al. 2004) and secondary sources of data such as annual reports, organizational charts, and articles in company magazines and newspapers in order to mitigate the concerns of retrospective bias (Langley 1999). As the study progressed, we also discussed our emerging findings with the representatives of the company. Data analysis was intertwined with data collection and earlier research (Coffey and Atkinson 1996). The interviews were tape-recorded (except one), transcribed and notes were taken and written up after the interviews (Ghauri 2004). All interview transcripts were returned to the interviewees for factual verification and interview summaries were written after each field visit. Key company informants were also asked to comment on these summaries and on the final draft of the case analysis (Ghauri and Grønhaug 2005; Patton 1990). Guided by the insights from the macro and the unit-level literature, we grouped our data into broad categories such as ‘motives to introduce a new structure’, ‘types and responsibilities of regional management centres’, ‘performance and survival of regional management centres over time’, and ‘structural adaptation process’. The insight of the systemic perspective on regional management was highly inductive which led us to further scrutinise the literature. Refining and adding new categories rather than pre-specifying them is inherent in qualitative content analysis (Ryan and Bernard 2000). We conducted a systematic comparison of empirical patterns across the eight regional management centres, between the two regions, and over time. These patterns are explored in the following section. Case Narrative Kone’s approach to regional management evolved between 1968 and 2008 from informal to formal constellations of regions and from a relatively homogenous regional organization to a more differentiated one. Based on our analysis, we identified four phases of evolution (see Table 1). Phase I from 1968 to 1988: Increasing Requirements for Regional Management In 1968, which marks the starting point of our analysis, Kone acquired Asea-Graham in Sweden. As one of the former CEOs explained, this company operated under a matrix structure which influenced Kone’s choice of a similar structure. In the 1960s and 1970s, Kone’s top management steered its foreign subsidiaries in Europe through close personal relationships. To illustrate, one of the CEOs with a tenure of more than 15 years recalled:
Transnational— simultaneous strive for global integration and local responsiveness
Phase IV 2004–2008 Matrix: Products, functions and regions
Matrix: Products, functions and regions
Regional adaptation Matrix: Products, and global integration functions and regions
Global integration
1989–1995
Phase III 1996–2003
Phase II
Table 1: Four phases of regional management at Kone Phases Corporate strategy Formal structure at the corporate level Phase I 1968–1988 Adaptation Matrix: Products and functions Regional organization
Towards the end of the phase: loose informal regional organization consisting of a federation of independent companies Global product divisions Rapid build-up of regional resources with broad competencies in marketing, finance, product development and human resources • Regional headquarters • Regional coordination centres • Sub-regional headquarters No representation of the regions on the executive board Global product divisions Very lean regional management centres and global functions Virtual setup with part-time regional roles for local managers No representation of the regions on the executive board Abolishment of regional coordination centres Matrix: Products, funcRe-strengthening of regional headquarters in tions and regions terms of increasing the number of functions No dominating structural and employees but less than in Phase 2 dimension Representation of the regions on the executive board Dissolution of the Northern European region (merged with Central Europe) Regional headquarters and local units replicating the matrix structure at the macro level
Dominant structural dimension Country organizations, acquired foreign firms
Regional Management as a System 519
520
R. Piekkari et al.
The old management style in Kone […] followed primarily the idea of walking around which, of course, resulted in a hell of a lot of travelling […]. During the growth period, subsidiary managers were seldom invited to Finland, rather we went there to get to know the company and the managing director. We pottered around factory floors and offices and tried to build these relationships […]. [The management style] was very spontaneous […]. However, such an informal management style became difficult to maintain since Kone continued to expand internationally. The number of foreign subsidiaries rose from the initial four units in 1968 to 98 units in 1988; all of which reported to the same CEO (company annual reports). At the same time, the geographical dispersal increased from four countries to 37 countries by 1988. As one CEO recalled, during the most intensive period of acquisitions the company became “a federation of different companies” rather than being one unified company with common practices, procedures and values. While a corporate-wide planning, budgeting and reporting system was introduced in each unit, it proved to be insufficient to achieve coordination of the acquired units. Another CEO explained that “Kone headquarters was relatively weak” and, as a regional manager put it, the acquired units had “a lot of autonomy”. Moreover, there was limited communication between the local subsidiaries. In order to ensure continuous top management attention to all operating units, a team of consultants was invited to Kone to investigate the possibilities of forming informal “lift cultures” in the regions where Kone operated. A “lift culture” was defined as a group of countries characterised by a combination of similar demand patterns for different lift types, buyer behaviour, construction practices, and the competitive situation of the market (company documents). For example, Asia-Pacific was considered a distinct “lift culture” in terms of market requirements. As a regional top manager in Hong Kong put it: We are […] developing our own products for the region […] in some cases totally new concepts to suit what is required here, which is different to Europe and different to America […]. During the first 20 years of our analysis, Kone went through an intensive growth period through foreign acquisitions. The company consisted of a large number of local markets which had a lot in common at a regional level. Phase II from 1989 to 1995: The Struggle to Implement Regional Management At the start of the second phase in 1989, the regional headquarters were established. Consequently, the matrix structure, which had been based on informal “lift cultures”, was reinforced and institutionalised. Under this formal matrix, each region was headed by a regional director and his3 staff who formed the regional headquarters and had full profit and loss responsibility for the respective region. These formal regions closely followed the original “lift cultures” and covered Northern, Central and Southern Europe, AsiaPacific and Americas. During this phase, substantial resources were allocated to regions. This is because Kone introduced regional coordination centres and sub-regional headquarters below regional headquarters. At that time, the number of regional management staff was at its maximum.
Regional Management as a System
521
Manufacturing and product development were largely controlled by the region. Rather than focusing on individual country markets, the corporate strategy aimed to reap benefits of regional integration and remain regionally responsive. One manager said that “we decided to integrate [by] […] streamlining production systems”. A regional manager with long tenure confirmed that in 2008 “there is not much local adaptation now, but there was in the mid-1990s”. Despite this, the Kone matrix was still dominated by product divisions, and regional managers had to fight for regional responsiveness. There were interviewees at all organizational levels who considered the implementation of regional management challenging. For example, a top manager from one regional engineering centre in Northern Europe described the situation as if “the Berlin wall had been built” between them and local subsidiaries. Another local subsidiary manager in Norway was unhappy about being dependent on a regional engineering centre outside Norway. “We have to say [to our final customer]: ‘One moment, please, we will ask our competence centre in Finland!’”. In situations where they perceived the formal supply processes to have failed, subsidiaries in Northern Europe frequently by-passed the regional centres and approached the global manufacturing units directly. Frictions were also mentioned in the relationships with headquarters. A regional manager in Asia-Pacific explained in 1994, five years after the introduction of formal regions, how corporate staff sometimes intervened in local operational matters and by-passed regional headquarters: There are some very strange things happening. Although we are responsible for Asia-Pacific, […] [corporate headquarters] does not take a consolidated picture of what we are saying to them. They still go down to the level underneath and look in detail at what is happening in all of the [local] companies. Our data show that despite the strategic rationale to introduce regional management, the implementation process was demanding. Phase III from 1996 to 2003: Downsizing of Regional Management and Global Integration The third phase emerged in the late 1990s when the regional structure was stripped down and manufacturing as well as product development were largely moved to the corporate level. By 2000, the regions were fully separated from the supply chain and factories with the purpose of focusing on the “sales side of the business”. Regional coordination centres were abolished and the regional headquarters were organised almost “virtually”, with some part-time directors responsible for the entire region. For example, in Northern Europe regional responsibilities were handled by granting local managers dual part-time roles. While the structure was lean and cost-efficient, it only allowed for limited regional adaptation. This was a direct response to an economic downturn, increasing cost pressures, and a shift towards pursuing a global integration strategy.
522
R. Piekkari et al.
Phase IV from 2004 to 2008: Re-enforcement of Regional Management In the fourth phase (2004–2008), the regional resources were increased slightly but remained rather lean. A full-time regional director was appointed for Asia-Pacific and, for example, some functional tasks of the regional engineering centre, which existed in the mid-1990s in Hong Kong, were reintroduced by appointing one director and his assistant. Following the global matrix structure, he reported to both the global head for this particular business segment and to the regional head of Asia-Pacific. When interviewed in November 2008, he commented that the “matrix structure exists all over Kone today”. The scope of responsibility of the regional organization in Asia-Pacific, however, was reduced as some activities, such as the global spare parts that had previously belonged to the regional organization, were moved to the corporate level. Contrary to the overall principle of controlling manufacturing globally, some of the factories in China were managed locally due to the distinctiveness and growth potential of the market, reflecting a transnational strategy. In this phase, a balance was reached between the regional and product dimensions of the matrix when regional heads were accepted on the executive board. Hence, in 2008, Kone’s matrix consisted of three global product division heads and four regional heads, all of equal status and represented on the executive board. The board covered only four regions since at the end of the period of our analysis, the region for Northern Europe was dissolved and merged with Central Europe to form a bigger entity. One regional manager explained that this restructuring was carried out to achieve better cost effectiveness, transfer best practices, and enhance learning within the region. Also, the number of shared customers between countries had increased, which in turn increased the interdependencies between Central and Northern Europe. Only then did the new regional headquarters become the “principal connection in the communication between the global corporate level and the local frontline units”. Findings and Discussion In this section we present our key findings and discuss them by intertwining empirical observations with theoretical insights. This form of reporting was chosen in order to better represent the iterative process of theorising from case data (Golden-Biddle and Locke 2007; Piekkari et al. 2009). We will now move from the chronological case narrative to discussing the theoretical insights we found. Regional Management as an Adaptive, Multi-level System In order to explain how and why regional management evolved in Kone over time in the way it did, we combined the insights from the two streams of literature. What we found could be labelled a multi-level system of regional management, which we identified at the intersection of the macro and micro-level literature. This system consists of differentiated centres which are adapted over time. The information processing theory at the macro-level gave us the concepts and tools to explain why and when the system
Regional Management as a System
523
of regional management was changed. However, relying solely on this literature would have produced a narrow view of Kone’s regional structure as a uniform arrangement within the corporation. By turning to the unit-level literature we were able to appreciate the differences between individual centres and across regions. Finally, the longitudinal case methodology allowed us to capture the adaptation process that Kone underwent. It revealed that although Kone had a regional structure in place throughout the period of analysis, the regional dimension changed over time. The number and type of units shifted and so did the overall regional headcount, the scope of regional responsibility in terms of functions, and internal status compared to the other structural dimensions (see Table 1). In the following, we adopt a systemic perspective to regional management and explain the adaptation process at the level of individual regional centres, the entire region and the Kone corporation as a whole. The Level of Regional Management Centres Kone complemented regional headquarters with functionally, geographically and hierarchically differentiated centres that were jointly responsible for the coordination task within and across regions. The company invested heavily in building up regional management centres and their number rose quickly in both regions, especially in Phase II. For example, Kone introduced a regional engineering centre in Hong Kong in 1994 by reallocating local resources to the regional level through staff transfers. The role of this centre was to act as a bridge between globally managed manufacturing units and local sales subsidiaries in their respective countries. In addition, a regional coordination centre for engineering new elevators was set up in Taiwan, where new manufacturing facilities had been acquired. Its scope of activities was rather narrow since it coordinated the supply chain of one particular product within a sub-region, and therefore there was no need to relocate expatriates. With these two coordination centres, Kone attempted to facilitate and channel communication by formalising information flow, reducing duplication and cutting costs. A local subsidiary manager commented on the supply process for new elevators; “the local company only deal[s] with regional engineering centres which…take care of the other relationships”. During Phase II, a similar principle for organising regional operations had already been applied in Northern Europe. A regional coordination centre for engineering of new elevators was established in Finland in 1990, and in 1993 a corresponding centre for maintenance and modernisation services was formed in Sweden. In the early 1990s, key global manufacturing units were located in Italy, which meant that the regional centres in Northern Europe had to coordinate the supply process beyond their own region, and this required respective information processing (see Fig. 1). Our analysis reveals that alongside the corporate headquarters, regional centres themselves might engage in cross-regional coordination and occupy an important role in horizontal information processing. This is a novel finding since Egelhoff (1991) claims that the corporate headquarters is the only structural mechanism that coordinates activities between regions. In Phase II, the Asian high-rise elevator market was booming and required a stronger presence and better coordination in the region. Hence, Kone introduced a special marketing centre for high-rise projects that was responsible for the Asia-Pacific region. There
524
R. Piekkari et al.
Organisational elements and units outside the region
Extra-regional manufacturing units
Car factory Austria
Door factory Germany
Control panel factory Finland
Organisational elements and units inside the region
Regional engineering centres
Local sales subsidiaries
Kone Sweden Regional Engineering Centre for New Elevators in Northern Europe Regional Modernisation and Maintenance Engineering Centre Northern Europe
Kone Norway
Kone Denmark
Kone Finland
Fig. 1: Example of a regional coordination centre
was a need to give a local face to the prestigious high-rise projects and improve Kone’s bargaining power in the region. The marketing centre was organised virtually with the regional manager based in Hong Kong and his staff physically scattered across four countries—Singapore, Malaysia, Thailand and the Philippines. As Kone continued to grow in the region through acquisitions, a sub-regional headquarters for the small ASEAN countries was established in Hong Kong in 1994. This was headed by a British regional manager who shared the office with the regional headquarters. The reorganization was implemented to provide better support for the sub-region and to channel and aggregate sub-regional information to the regional level and to relieve the regional headquarters from some information processing. However, profit-and-loss responsibility remained with the regional headquarters. The small but important subregional headquarters was later moved to Singapore to ensure closeness to the countries in the sub-region. Our longitudinal analysis reveals that the number of regional management centres and regions belonging to the system was adjusted over time and the regional coordination centres as well as the regional headquarters of Northern Europe did not survive the entire
Regional Management as a System
525
period of analysis. The reported bypassing of regional centres suggests that the regional level did not fully free up corporate resources from operative information processing. Apparently, neither corporate headquarters’ managers nor subsidiary managers wanted to relinquish influence to the regions (see e.g., Daniels 1986; Ghemawat 2005; Morrison and Roth 1992; Schütte 1996). This is in line with earlier contributions arguing that during the implementation of regional headquarters, powerful subsidiaries may resist regional coordination (see e.g., Aoki and Tachiki 1992). As a consequence, the performance of the regional management system in Kone can be considered rather weak in relation to the costs that these additional management layers imply. Lasserre (1996) argues that regional management centres are likely to disappear when they become “administrators” without important entrepreneurial or integrative functions. Alternatively, from a systemic perspective we would argue that individual centres disappear when they are inefficient, that is when their contribution to realized information processing capacity is smaller than their costs. The regional centres in Kone became increasingly inefficient when a strong informal network existed. Moreover, implementation problems inhibited the unfolding of regional information processing capacity, rendering the entire structure fragile. This is a novel finding since much of the earlier information processing research is cross-sectional and suffers from survival bias. Furthermore, in his criticism of the information processing theory, Forsgren (2008) suggests that the problems stemming from implementing a structure fall outside the scope of Egelhoff’s theory. In contrast, we argue that by adopting a systemic perspective, implementation problems do have a role to play in the theory because they mark the start of new adaptation processes towards matching information processing requirements with corresponding capacities. In essence, our case data reveal a process similar to the one described by Lehrer and Asakawa (1999) who point out that Asian firms increasingly “unbundle” their regional headquarters into distinct regional functions and units. We extend this view by arguing that individual units may generate specific information processing capacity in response to particular information processing requirements. Collectively, we posit, these centres build the information processing capacity of the region. Our case study shows that a certain misfit between information processing capacities and requirements does not automatically lead to removal of the regional dimension altogether, as suggested by the macro level literature (Egelhoff 1991). Instead, only parts of the structure such as individual centres may be abolished, thereby reinforcing our conceptualisation of regional management as a system. The Regional Level and the Differences between Northern Europe and Asia-Pacific While macro-level information processing theory explains the introduction of a regional structure with inter-regional differences (Egelhoff 1988; Wolf and Egelhoff 2002), our unit-level analysis shows that these differences might also lead to structural differentiation internally. Notwithstanding some similar adjustment processes in Northern Europe and Asia-Pacific, the differences between them had an important impact on the structural arrangements.
526
R. Piekkari et al.
One important difference between the regions was that staff in Northern Europe and in Asia-Pacific had different relationships with corporate headquarters which influenced how information was processed. The history of Northern Europe is based on the acquisition of the Swedish firm Asea-Graham, which also provided Kone with ownership of its subsidiaries in Denmark and Norway. In addition, as compared to Asia-Pacific, the regional headquarters for Northern Europe was responsible for a slightly narrower range of activities. This was because corporate and regional headquarters were located in the same premises in Finland for most of the period of analysis. Local staff maintained longrunning, direct and often informal communication lines with global units and headquarters, and key positions were staffed by long-serving Finnish managers. Over the years, staff in Northern Europe had developed a broad network of personal relationships with units outside regional coordination centres and they possessed the critical ‘know-who’ information for operating effectively in Kone. As a Finnish expatriate on assignment in the Swedish local subsidiary explained: I have the advantage that I don’t need to call solely the pre-determined contact person but I can call the factories and manufacturing units directly. The contrary applied to Asia-Pacific where the persistence of cultural, geographic and language distances, the lack of social integration, and strong economic growth strengthened the position of the regional organization and made it more independent from corporate headquarters in terms of information processing. Originally, Kone became involved in the region as a result of acquiring a UK-based lift company—Marryat and Scott—in 1978, which had operations in Hong Kong, Singapore, Malaysia and the Philippines. Subsequently, Kone expanded into Australia which was a key market in the region and located their regional headquarters in Sydney. The regional headquarters was moved to Hong Kong in 1993 due to the growth potential of the Asian elevator market. British, Australian and Finnish senior managers were sent on expatriate assignments to the AsiaPacific region to hold key positions. Thus, the organizational heritage was different for the two regions. In Asia-Pacific, despite some problems in the early years, local subsidiary staff usually turned to their respective regional centres with their queries rather than by-passed them. Given the short tenure of Asia-Pacific units with Kone and the geographical distance, local subsidiary staff had fewer opportunities to establish personal relationships with colleagues outside Asia-Pacific. As a regional manager in Hong Kong described the challenge of maintaining relationships from Hong Kong where the regional headquarters was located: It’s like we are at the end of the tail of the dog and that’s the head over there [in Europe]. The regional head of Asia-Pacific, who had previously worked at headquarters in Finland, shared this view: My ability to be in touch with […] the [corporate] centre had a certain momentum from the fact that I had been there. Now I find […] that I have very little idea about what those guys [at corporate headquarters] are up to. I worked quite hard to try to
Regional Management as a System
527
keep them involved in what I am doing and trying to understand what they are trying to do. I have actually now almost given that up. This manager did not know the new CEO personally, and he and his staff directed their attention to internal development of the region with the intention to have, as a regional manager put it, a “self-supporting region”. Staff were also increasingly trained locally or regionally rather than being sent to the corporate training centre in Finland. Our case study shows that the different external and internal contingencies led to different regional management systems in the two regions in Kone, while the macro-level structure remained the same over our period of analysis. We observed an expansive, more autonomous regional system in the turbulent, distant region of Asia-Pacific, and a contracting regional system in Northern Europe—the region characterised by less turbulence, stagnating sales, proximity to the headquarters, and a strong informal network. In Phase IV, the Northern European region was dissolved and all regional management centres merged with the Central Europe region. We also found a greater number of centres (five in total, including a sub-region) in Asia-Pacific, compared to three centres in Northern Europe. Hence, we join Rugman (2005) who suggests that regions differ in their information processing requirements and we show that this can lead to differentiation of regional management systems providing information processing capacity. The Macro Level and Kone Corporate Structure Overall, our case study confirms the information processing logic at the macro level of the case company (Egelhoff 1991). Kone’s regional dimension was introduced under conditions of strong international growth and significant differences between regional elevator markets. These contingencies were rather persistent throughout the period of analysis. Under such conditions, the regional level was introduced to ease the information overload at corporate headquarters and establish a better link between local subsidiary management and Kone corporate headquarters. As a CEO explained: In fact, the area management has two hats; they represent us towards their subsidiaries and the subsidiaries towards us. A cross-sectional macro-level approach alone would have characterised Kone’s structure as a global matrix with a regional dimension. Our longitudinal case study reveals, however, subtle changes in the regional dimension and an adaptation process that took place in Kone. In Phase II, Kone followed an apparent strategy of regional adaptation, but the regions were still not represented on the executive board. Despite the fact that Kone had formally adopted a matrix structure with a regional dimension, the global product divisions dominated decision-making (see Table 1). From this perspective, there was evidence of misfit on the macro level. Especially in this period, we observed a high intensity of implementation problems as documented, for example in the form of the bypassing of regional headquarters and coordination centres mainly in Northern Europe. In response to these structural inefficiencies, the headquarters nearly abolished the regional organization in subsequent years, despite the persisting pressure to operate with strong regions. While Phase III is consistent in terms of Kone having a global integration
528
R. Piekkari et al.
strategy supported by a very lean regional organization (see Table 1), towards the end of this period, the constant differences between regions called for reinforcement of the regional management structure. This pendulum movement marked the start of Phase IV. The capacity to process regional information in Kone changed over time. For example, while the regional resources were considerably smaller in Phase IV than in Phase II, the information processing capacity had de facto increased with the acceptance of regional managers on the executive board in 2005. This is in line with Paik and Sohn (2004) who suggest that effective regional management requires full representation at the corporate level. The formal position between the Kone top managers of global product divisions, global functions, and regions was levelled. As one regional manager described, placing the regions at this “higher level in the organization” meant that “[t]he area director’s voice is louder and better heard” than previously. In addition, communication was better “aligned” between global and regional levels. Membership of the executive board was crucial for regional headquarters. They turned into a two-way information nexus and facilitated communication at Kone. A regional head described his role as the “phone book and contact guy” in the region and added that they would all be an “unnecessary overhead if the locals could deal efficiently with the globals”. Moreover, this process was supported through rotating staff, because managers who had previously had global roles were now given regional roles in Asia-Pacific. In sum, our findings shed light on the fundamental question of whether corporate headquarters is able to achieve fit (see, also Forsgren 2008). We found evidence that headquarters have substantial problems in perceiving and implementing fit as documented for example, by the misfit between strategy and structure, the implementation problems and the bypassing behaviour which persisted for a number of years. It took roughly 15 years before the regional dimension was granted equal status on the top management board. Adaptation to misfit and underperformance of individual units only happened after a considerable time lag. However, once the Kone headquarters learnt about the challenges associated with implementing the regional structure, it became a key player in initiating adaptation processes towards new fit. Our analysis shows that the adaptation of the regional management system occurred on the level of individual regional centres, entire regions and the corporation as a whole which brought about considerable complexity. Thus, we argue that the value-added of corporate headquarters lies in initiating adaptation processes and in achieving coordination between regional elements of the structural system. The longitudinal case study depicts a continuous adaptation process rather than an achievement of instant fit. Conclusion and Suggestions for Future Research The purpose of our case study was to explain the evolution of regional management and the interplay between different regional management centres in Kone, a European multinational corporation. For this purpose, we bridged two important research streams: The information processing literature, which has examined the macro level of organizations (Egelhoff 1988, 1991; Wolf and Egelhoff 2002), and the unit-level literature, which has focused primarily on the location, set-up, and evolution of individual centres, in particular
Regional Management as a System
529
regional headquarters (Paik and Sohn 2004; Schütte 1996). By integrating both perspectives in a longitudinal investigation, we arrived at an intermediary level of analysis and conceptualized regional management as a system. The systemic view of regional management is novel. It consists of several regional elements which adapt to information processing requirements on the unit, regional and macro level of the organization. Such a view suggests that complex adaptation processes may occur simultaneously. Our findings show that while parts of the Kone structure were altered and some even abolished, the company did not give up its underlying structure, the matrix with a regional dimension. Hence, the systemic view explicitly acknowledges the interdependencies between different regional elements which have been neglected in prior research but which would offer a promising area for future work. Our study echoes recent work on the changing role of headquarters, which is becoming increasingly multifaceted (Barner-Rasmussen et al. 2010; Birkinshaw et al. 2006). Lehrer and Asakawa (1999) argue that the internal and external complexity facing large multinational corporations lead them to unbundle their (regional) headquarters’ activities and reallocate them to different units or even single managers, residing in different locations (see also, Birkinshaw et al. 2006). Consequently, it is increasingly difficult to identify a single unit or location of headquarters. The Kone executive board consists of members based in five countries which challenges the traditional definition of corporate headquarters. Recent research conceptualizes headquarters as a set of dispersed activities (Birkinshaw et al. 2006) and questions where the nexus of corporate control resides (Barner-Rasmussen et al. 2010). Indeed, the spatial dispersal and the split-up of headquarters’ activities render it challenging to assume that headquarters have the knowledge and understanding to design the organizational structure of the firm. Based on our study, we would argue that the value-added of corporate headquarters may lie in initiating adaptation processes and processing information at the apex of the organization. However, the interdependent and complex nature of the structural system suggests that headquarters may face problems in initiating and achieving fit. We believe that the systemic view of regional management complements current theorizing about headquarters and invites fresh approaches to investigating regional management. Acknowledgements: We would like to thank Perttu Kähäri, Mark Lehrer, Denice Welch and Lawrence Welch for their helpful comments on earlier versions of this manuscript as well as the audience of a faculty seminar at Copenhagen Business School for constructive feedback. We appreciate the guidance and persistence of the guest editors, Björn Ambos and Volker Manhke, during the review process. Above all, we are grateful to the interviewees at Kone who shared with us their experiences from the field.
Endnotes 1 We acknowledge that information processing theory has been used at many different levels of analysis. In international business, Ambos (2002) and Luo (2006) represent such examples for the subsidiary level. However, when it comes to analyzing regional management structures, information processing theory has, to our knowledge, only been used at the macro level.
530
R. Piekkari et al.
2 A large part of the growth in Asia-Pacific was due to acquisitions made at that time. 3 All regional managers were male.
References Ambos, B. (2002). Internationales Forschungs- und Entwicklungsmanagement. Wiesbaden: Gabler. Aoki, A., & Tachiki, D. (1992). Overseas Japanese business operations: The emerging role of RHQ. RIM Pacific Business and Industries, 24(1), 28–39. Barner-Rasmussen, W., Piekkari, R., Scott-Kennel, J., & Welch, C. (2010). Commander-in-chief or absentee landlord? Key perspectives on headquarters in multinational corporations. In U. Anderson & U. Holm (Eds.), Managing the contemporary multinational: The role of headquarters (pp. 85–105). Cheltenham: Edward Elgar. Birkinshaw, J., Braunerhjelm, P., Holm, U., & Terjesen, S. (2006). Why do some multinational corporations relocate their headquarters overseas? Strategic Management Journal, 27(7), 681–700. Coffey, C., & Atkinson, P. (1996). Making sense of qualitative data. Thousand Oaks: Sage. Daniels, J. D. (1986). Approaches to European regional management by large U.S. multinational firms. Management International Review, 26(2), 27–34. Daniels, J. D. (1987). Bridging national and global marketing strategies through regional operations. International Marketing Review, 4(3), 29–45. Egelhoff, W. G. (1982). Strategy and structure in multinational corporations: An informationprocessing approach. Administrative Science Quarterly, 27(3), 435–458. Egelhoff, W. G. (1988). Strategy and structure in multinational corporations: A revision of the Stopford and Wells Model. Strategic Management Journal, 9(1), 1–14. Egelhoff, W. G. (1991). Information-processing theory and the multinational enterprise. Journal of International Business Studies, 22(3), 341–368. Enright, M. J. (2005a). Regional management centres in the Asia-Pacific. Management International Review, 45, Special Issue, 69–82. Enright, M. J. (2005b). The roles of regional management centres. Management International Review, 45, Special Issue, 83–102. Erramilli, M. K. (2003). Regionalization of multinationals: Implications for research in international Marketing. In S. C. Jain (Ed.), Handbook of research in international marketing (pp. 81–98). Cheltenham: Edward Elgar. Fletcher, M., & Plakoyiannaki, E. (2011 forthcoming). Case study selection: Key issues and challenges for international business researchers. In R. Piekkari & C. Welch (Eds.), Rethinking the case study: Towards greater pluralism in international business and international management research. Cheltenham: Edward Elgar. Forsgren, M. (2008). Theories of the multinational firm: A multidimensional creature in the global economy. Cheltenham: Edward Elgar. Galbraith, J. R. (1969). Organization design: An information-processing perspective. Cambridge: Ballanger. Galbraith, J. R. (1973). Designing complex organizations. Reading: Addison-Wesley. Galbraith, J. R. (1977). Organization design. Reading: Addison-Wesley. Ghauri, P. N. (2004). Designing and conducting case studies in international business research. In R. Marschan-Piekkari & C. Welch (Eds.), Handbook of qualitative research methods for international business (pp. 109–124). Cheltenham: Edward Elgar. Ghauri, P. N., & Grønhaug, K. (2005). Research methods in business studies: A practical guide (3rd ed.). London: Financial Times Prentice Hall.
Regional Management as a System
531
Ghemawat, P. (2003). Semiglobalization and international business strategy. Journal of International Business Studies, 34(2), 138–152. Ghemawat, P. (2005). Regional strategies for global leadership. Harvard Business Review, 83(12), 98–108. Golden-Biddle, K., & Locke, K. (2007). Composing qualitative research (2nd ed.). Thousand Oaks: Sage. Huber, G. P., & Power, D. J. (1985). Retrospective reports of strategic-level managers: Guidelines for increasing their accuracy. Strategic Management Journal, 6(2), 171–180. Kone Annual Report. (1993). www.kone.com Langley, A. (1999). Strategies for theorizing from process data. Academy of Management Review, 24(4), 691–710. Lasserre, P. (1996). RHQ: The spearhead for Asia Pacific markets. Long Range Planning, 29(1), 30–37. Lehrer, M., & Asakawa, K. (1999). Unbundling European operations: Regional management and corporate flexibility in American and Japanese MNCs. Journal of World Business, 34(3), 267–286. Luo, Y. (2006). Autonomy of foreign R&D units in an emerging market: An information-processing perspective. Management International Review, 46(3), 349–378. Marschan-Piekkari, R., Welch, C., Penttinen, H., & Tahvanainen, M. (2004). Interviewing in the multinational corporation: Challenges of the organisational context. In R. Marschan-Piekkari & C. Welch (Eds.), Handbook of qualitative research methods for international business (pp. 244–263). Cheltenham: Edward Elgar. Morrison, A., & Roth, K. (1992). The regional solution: An alternative to globalization. Transnational Corporations, 1(2), 37–55. Paik, Y., & Sohn, J. H. D. (2004). Striking a balance between global integration and local responsiveness: The case of toshiba corporation in redefining RHQ’s role. Organizational Analysis, 12(4), 347–359. Patton, M. Q. (1990). Qualitative evaluation and research methods (2nd ed.). Beverly Hills: Sage. Piekkari, R., Welch, C., & Paavilainen, E. (2009). The case study as disciplinary convention: Evidence from international business journals. Organizational Research Methods, 12(3), 567–589. Rugman, A. M. (2005). The regional multinationals. Cambridge: Cambridge University Press. Rugman, A. M., & Verbeke, A. (2004). A perspective on regional and global strategies of multinational enterprises. Journal of International Business Studies, 35(1), 3–18. Rugman, A. M., & Verbeke, A. (2005). Towards a theory of regional multinationals: A transaction cost economics approach. Management International Review, 45(1), 5–17. Ryan, G. W., & Bernard, H. R. (2000). Data management and analysis methods. In N. K. Denzin & Y. S. Lincoln (Eds.), Handbook of qualitative research (2nd ed., pp. 769–802). Thousand Oaks: Sage. Schütte, H. (1996). RHQ of multinational corporations (PhD Dissertation, Universität St. Gallen, 1996). Schütte, H. (1997). Strategy and organisation: Challenges for European MNCs in Asia. European Management Journal, 15(4), 436–445. Stopford, J., & Wells, L. T. (1972). Managing the multinational enterprise. New York: Basic Books. Thihanyi, L., & Thomas W. B. (2005). Information-processing demands and the multinational enterprise: A comparison of foreign and domestic Earnings Estimates. Journal of Business Research, 58(3), 285–292. Thompson, J. D. (1967). Organizations in action. New York: Basic Books. Williams, C. R. (1967). Regional management overseas. Harvard Business Review, 45(1), 87–91.
532
R. Piekkari et al.
Wolf, J., & Egelhoff, W. G. (2001). Strategy and structure: Extending the theory and integrating the research on national and international firms. Schmalenbach Business Review, 53(2), 117–139. Wolf, J., & Egelhoff, W. G. (2002). A re-examination and extension of international strategy-structure theory. Strategic Management Journal, 23(2), 181–189. Yeung, H. W.-C., Poon, J., & Perry, M. (2001). Towards a regional strategy: The role of regional headquarters of foreign firms in Singapore. Urban Studies, 38(1), 157–183.
