The Determinants of Subsidiary Autonomy in European Multinational Corporations
Vo Van Dut
Research Master in International Economics and Business Faculty of Economics and Business, University of Groningen
JULY 2009
The Determinants of Subsidiary Autonomy in European Multinational Corporations
Vo Van Dut s1544969
Research Master in International Economics and Business Faculty of Economics and Business, University of Groningen
Supervisors:
Dr Gjalt de Jong
Prof. dr Sjoerd Beugelsdijk
Acknowledgements
I would firstly like to acknowledge my supervisors Dr Gjalt de Jong and Prof. dr Sjoerd Beugelsdijk, who have not only provided continuous guidance, considerable support and valuable feedback for this thesis, but have also assisted me with their considerable practical knowledge and experience and have provided kind help since I began my studies in Groningen. My debt to them is greater than I can possibly fully acknowledge.
I would like to deeply thank to Dr Bart Los – cordinator of the Research Mater Programme – who has adviced me many things during my following this program.
I am grateful to the NPT Program me – the cooperation between NUFFIC, the University of Groningen, and the School of Economics and Business Administration (SEBA) of the University of Can Tho – for their financial support which gave me the opportunity to study in the Netherlands. I would like to thank the staff of the International Relations Office, Groningen University greatly, especially Gonn y Lakerveld, Anita Veltmaat and Wiebe Zijlstra.
Finally, I would like to express my special thanks to my parents, sisters and brothers for their support and great encouragement while I was living far from home.
Groningen, 20 July 2009
Vo Van Dut
The Determinants of Subsidiary Autonomy in European Multinational Corporations
ABSTRACT
This study employs OLS, Poisson and negative binomial regression s to analyse the factors determining the degree of subsidiary autonomy granted by multinational corporations (MNCs). The secondary data used in this analysis is derived from Amadeus, the sample consists of 263 European subsidiaries of 18 MNCs in 25 European countries. The empirical findings show that subsidiary autonomy is associated with the centralized strategies applied by parent firms. Parent company size, subsidiary age, and the extent of subsidiary ownership are positively associated with de gree of subsidiary autonomy. It is interesting that the absolute sizes of a subsidiary and its autonomy have an inverted U-shaped relationship; while degree of
parent company’s internationalization, economic distance,geographic distance and individualism are negatively associated with degree of subsidiary autonomy. The study, however, fails to find any effect of product diversity, institutional quality in the host country, degree of interdependence, power distance, masculinity and the relatedness between the host and home countries on the degree of subsidiary autonomy.
Key words: autonomy, subsidiary characteristics, parent company characteristics, institutional
environment, economic distance, geographic distance, and culture.
Table of Contents
1. INTRODUCTION ... 1
2. THEORETICAL BACKGROUND AND HYPOTHESES ... 4
2.1 Parent-company characteristics ... 5
2.2 The degree of institutional qu ality in the host country ... 8
2.3 Subsidiary characteristics ... 8
2.4 Geographic, economic and cultural distance ... 11
2.3 Control variables ... 18
3. THE ECONOMETRIC MODEL ... 19
4. DATA AND IMPLEMENTATION ... 21
4.1 Variables and measures ... 21
4.2 Sample and data ... 25
5. EMPIRICAL RESULTS ... 26
5.1 Description ... 26
5.2 Correlations ... 30
5.3 Findings ... 30
6. CONCLUSIONS ... 35
6.1 Key findings ... 35
6.2 Litmitations and further research ... 41
REFERENCES ... 44
APPENDIX ... 51
1. INTRODUCTION
The past decade has been characterized by profound changes and advance s in technology as well as increased globalization. As a result, there have been dramatic shifts in the way businesses are organized and how they compete. These rapid changes in the nature of global competition have c aused international managers and international management researchers alike to search for new ways to frame problems and answer questions about how to manage multinational corporations (MNCs) most effectively (O’Donnell, 2000). The role of the MNC subsidiary therefore continues to change in response to globalization, and the evolution of subsidiary roles can be taken one step further than previously realized (Birkinshaw et al., 1998). Several studies have pointed out that the relationship between the parent company and its foreign affiliates has become more complicated and sometimes conflictory (Garnier et al., 1979; Enright, 2000; Edwards et al., 2002). Such authors have further asserted that in modern multiunit business firms, the division of decision -making authority between the headquarters and the various operational units responds to a complex set of factors. This issue is particularly true in MNCs that operate affiliates in several foreign countries in sometimes widely different contexts. Thus, many of the issues raised are highly relevant to research today, especially with the management of the multinational corporation subsidiary emerging as a distinct field of investigation (see for example, Birkinshaw, 2001; Paterson and Brock, 2002; Young and Tavar es, 2004).
Jarillo and Martínez (1990) find that subsidiary strategies shift toward closer integration with their parent company. However, the evolution of the subsidiary’s strategy varies across subsidiaries.
For example, some MNCs allow their subsidiaries a great deal of independence while others assume control of the subsidiaries’ activities (White and Poynter, 1984). In addition, Gnan and Songini (1995) and Dirks (1995) rightly argue that Japanese firms allow subsidiaries little decision -making freedom in the early stages of development, while there has been a significant relaxation of this position in recent years. Conversely, Blaine (1994) found that German -owned subsidiaries may have lost some decision - making power. Eventually, the scope and content of decision-making might differ between subsidiaries depending on their MSC’s country of origin (Taggart, 1997). Therefore, this study aims to answer the question: ‘What factors determine the degree of autonomy of European multinational corporation subsidiaries’.
This study is an expanded research from a previous study performed in 2007. The previous
study was conducted with the same research question as a guideline, but with largely different
aspects. The previous study investigated the effects on the degree of subsidiary autonomy of, for
example, the overall strategic approach of the MNC, parent company size, the degree of
institutional embeddedness , subsidiary size, subisidary ownership, and age of subsidiary. As will be explained, this study includes these variables that were significant and thus important explanatory variables for the autonomy of subsidiaries. The present study adds a number of variables to the equation, too, investigates new aspects of autonomy, namely internationalization, interdependence, and geographic, economic and cultural distance . In the previous study, we found significant results on all variables but the degree of product diversification and the relatedness between home and host countries. Thus, the previous study has been an inspiration to us in many ways, but in no way should this study be considered as a somewhat adjusted copy.
According to Björkman (2003), variations in the decision-making authority of a subsidiary can be explaned by characteristics of the parent company, characteristics of the subsidiary, characteristics of the local institutional environment. However, the results of previous studies on the effect of parent- company characteristics on subsidiary autonomy have been mixed and no clear understanding has been developed, while the impact of subsidiary characteristics on their autonomy is clearer and shows a little more consistency than parent-company characteristics (see Varblane et al., 2005). This result was presented by Young and Tavares (2004) ; however, their study is confined to a description of all the relevant studies in the field (Young and Tavares, 2004). Moreover, there has been much less analysis concerning the effect of the local institutional environment on subsidiary autonomy, especially concerning the host-country role in providing opportunities for subsidiar ies to increase their autonomy (see for example, Soskice, 1999; Varblane et al., 2005; Geppert and Williams, 2006). In addition, previous research on national cultural distance and subsidiary co ntrol issues has yielded inconsistent and often contradictory empirical evidence. Some scholars found that increased cultural distance is associated with a greater level of control (Davidson and McFeteridge, 1985; Root, 1987; Boyacigiller, 1990; Pan, 1996; Erramilli and Rao, 1993; Padmanabhan and Cho, 1996; Anand and Delios, 1997;
Wilkimson et al., 2008); while others found increased cultural distance is associated with a lesser level
of control (Alpander, 1976; Anderson and Gatignon, 1986; Gatignon and And erson, 1988; Kogut and
Singh, 1988; Kim and Hwang, 1992). On the other hand, some scholars link cultural distance with the
need for cooperative arrangements (Gatignon and Anderson, 1988); Kogut and Singh, 1988; Erramilli
and Rao, 1993). Others submit that cultural distance problems can best be addressed with strong
hierarchical control (Shane, 1994; Padmanabhan and Cho, 1996; Erramilli et al., 1997). Thus, little
agreement has been forthcoming with respect to the influence of cultural on control issues (Wil kinson
et al., 2008). From our standpoint, no study has yet investigated the combined impact of the four factors
mentioned above on the degree of subsidiary autonomy using empirical evidence, especially in relation
to European MNCs. Thus, to answer the research question above, the present study has derived the following investigative questions:
- Do the characteristics of the parent company influence the degree of subsidiary autonomy?
- Does the institutional environment of the host country affect the degree o f subsidiary autonomy?
- Is subsidiary autonomy affected by its own characteristics?
- Does the distance between the home and host countr ies impact the degree of subsidiary autonomy?
There are several reasons why this study investigates what factors determine the degree of subsidiary autonomy within MNCs, especially in European MNCs. First, the degree of subsidiary autonomy was substantial across European MNCs during the early 1980s. However, the majority of the studies on the topic only focused on one or two specific European countries (for example, Hedlund, 1981; Gate and Egelhoff, 1986; Vachani, 1999; Varblane, 2005; Johnston, 2005). Second, the previous studies only look at one of the four factors mentioned above. By combining several aspects of the four factors to investigate the degree of subsidiary autonomy, this study enhances the established literature because it analyses data from most MNCs in several different European countries. Third, the characteristics of the parent company and the subsidiary, the institutional environment as well as the economic, geographic and economic distance between the home and host countries are expected to have a large impact on the degree of subsidiary autonomy since these factors directly influence the relationship between the parent company and the subsidiary. Fourth, in this study the degree of autonomy is measured as ‘count data’, which is a significantly different method to that employed by previous studies.
The first contribution of this paper to the established lit erature is that we use cross-sectional data from 263 subsidiaries of 18 MNCs in 25 European countries to study the effect of MNC characteristics, subsidiary characteristics and the institutional environment of the host country on the degree of subsidiary autonomy. Secondly, taking the methods used to measure the degree of subsidiary autonomy in previous studies (for example, Hedlund, 1981; Garnier, 1982; Vachani, 1999; Varblane, 2005; Johnston, 2005) we combine the measurement method (see Taggart and Hood, 1999), and our present data in an econometric model to obtain econometric evidence on the effect of the four aspects mentioned above on the degree of subsidiary autonomy.
The remainder of the paper is organized as follows. Section 2 will discuss the main theoretical
background and hypotheses related to the present research issues , while Section 3 will present the
econometric model. Following this, the data employed in this study as well as variable selection will be
described in Section 4. In Section 5, we will analyse the empirical results and examine which factors determine the degree of subsidiary autonomy granted by MNC s in an econometric model. Finally, we will discuss our findings before providing our conclusions and remarks, in addition to future res earch implications.
2. THEORETICAL BACKGROUND AND HYPOTHESES
Goehle (1980) states that ‘one way to assess the decision -making authority of a subsidiary manager would be to consider the width of his “zone of discretion” in making programmed or basic decisions as well as non-programmed or routine decisions’ (1980: 20). The author suggests that the width of the particular discretionary zone in which a manager makes decisions is affected by factors in the corporation itself as well as by environmental factors surrounding the particular subsidiary organization. The study then proposed that these influences be classified into two groups of characteristics, corporate and subsidiary, which appear to have a major effect on the degree of decision-making authority available to the local subsidiary manager. Hedlund’s study (1981) of the autonomy of subsidiaries and the formalization of headquarters -subsidiary relationships in Swedish MNCs is also based on these two aspects. In addition, characteristics of the local environment such as the national business system and national regulatory institutions also play a critical role in determining the bargaining power of local managers (for example, Hall and Soskice, 1999; Whitley, 1999; Edwards et al. 2002; Johnston, 2005: 98) . Furthermore, Birkinshaw and Hood (1998), and Fenton-O’Creevy (2008) who view subsidiary autonomy as determined not only by the relationship of the head office to its subsidiaries, but also by the nature of the local institutional environment in which the subsidiary operates. On the other hand, geographic, economic and cultural distance between home and host countries are also considered as the control mechanisms of parent firm with respect to subsidiaries (Garnier, 1982; Harzing, 1999; 200 0;
Harzing et al., 2002; 2003; Varblane et al., 2005; Harrison et al., 1994; Vachani, 1999; Wilkinson et al., 2008). As such, there are four main elements of the literature relevant to the present study:
the characteristics of the parent company, the local institutional e nvironment (national business
system), the characteristics of the subsidiary, and geographic, economic and cultural distance
between home and host countries .
2.1 Parent-company characteristics
According to Garnier (1982), Pahlberg (1996), Johnston (2005), and Geppert and Williams (2006) , as well as organizational theorists (Ghoshal and Westney, 1993; Czarniawska, 2006) , the characteristics of the MNC system which appear to have a major influence on the degree of subsidiary autonomy include: (1) corporate organizational structure, (2) degree of product diversification , (3) parent- company size., and (4) degree of parent firm’s internationalization. Each of these will be discussed below.
Centralized/decentralized strategy of a multinational group Soskice (1999) states that at the end of
1980s the production regimes of most advanced economies fell in to one of two main patterns. The first, referred to as ‘coordinated market economies’ (CMEs), includes Germany, Switzerland, most northern European countries an d, in a different form, Japan and South Korea. The other pattern, named ‘liberal market economies’ (LMEs), consists of the Anglo-Saxon economies and Ireland.
Soskice (1999) indicates that the characteristics of manufacturing in CME s are based on diversified quality production (DQP). The MNC subsidiaries in these countries across the world have a local rather than a global focus and thus are less subject to centralized control which impairs their ability to respond to local market pressures (Fenton et al., 2003). For example, German MNCs have recently embarked on a cautious internationalization process but still follow a ‘local responsiveness’
strategy of local differentiation among their foreign subsidiaries. Thus, Geppert and Williams (2006)
argue that headquarter management representatives in German y emphasize that subsidiaries
worldwide have relative autonomy in running their own operations. Moreover, the study by Lane
(1989) shows that German and Japanese MNCs allocate more resources and responsibilities as well
as organizational and financial autonomy to their subsidiaries to develop networks in host countries
similar to those existing in German and Japanese industry. In short, international corporations which
are in favour of imposing decentralized str ategies on their subsidiaries, such as German, Japanese
and Swedish MNCs, tend to respect the autonomy of local subsidiaries. In contrast, MNCs in Anglo-
Saxon economies like to impose their standardized global strategies on their subsidiaries. Divergent
interests and the local power resources of key subsidiary managers and employee representative
bodies are played down or ignored (Geppart and Williams, 2006). Also, the study by Geppart and
Williams (2006) provides examples which show that US and UK MNCs favour central control and
standardized strategies for their subsidiaries. In fact, these MNCs prefer their subsidiaries to serve the
global rather than just the domestic market, and thus there is a powerful incentive for them to centralize control (Fenton et al., 2003). On this basis, we come to the first hypothesis:
Hypothesis 1: An MNC located in a CME with negotiated, decentralized strategies would grant more autonomy to its subsidiaries than an MNC located in an LME with centralized, imposed global strategies.
Degree of product diversification of the parent company Picard (1977) predicted that MNC
decentralization is positively associated with product diversity. Support for this perspective comes from Gate and Egelhoff (1986) who also found that subsidiary autonomy was positively correlated with product diversity. Garnier (1982) conducted a study related to decision-making autonomy in the foreign affiliates of US multinational corporations by investigating 42 subsidiaries in France and 102 in Mexico. The findings pointed out that the degree of autonomy of subsidiaries in France was positively correlated with product diversification , while the number of products negatively affected the degree of decision-making in subsidiaries in Mexico. Vachani ’s results (1999) support the latter case for related product diversification, but found that the unrelated product diversification positively influences the degree of autonomy. Indeed, there are conflicting arguments concerning the relationship between product divers ity and a subsidiary’s degree of decision-making authority.
However, most organizational theorists assert that both internal and external complexity – product diversity, product modification differences between subsidiaries, extent of foreign acquisition
are associated with decentralization. Moreover, the greater the degree of product diversification of MNCs, the more the subsidiary management by MNCs becomes complex and more difficult to control, enabling their subsidiaries to assume more autonomy. This can be summarized with the following hypothesis:
Hypothesis 2: Subsidiaries of MNCs with higher product diversification will have higher autonomy.
Size of the parent company Organizational theorists argue that the greater the parent -company size,
the greater the autonomy of the local managers. To prove this, Child (1973) conducted research on a
sample of 82 British businesses. The results indicated that decentralization is positively correlated
with size. Supporting this perspective, Goehle (1980) and Hedlund (1981) have also found that the
size of the parent company is positively correlated with the degree of subsidiary autonomy. Gate and
Egelhoff (1986) examined this issue again by investigating 50 large US, UK and European MNCs.
The most recent empirical study by Männik et al. (2004) emphasize s that large firms grant more autonomy to their subsidiaries. The authors point out that the size of MNCs positively affects the degree of decision-making by the subsidiary. Thus, we propose the following hypothesis :
Hypothesis 3: The parent-company size is positively associated with the degree of subsidiary autonomy.
Degree of internationalization of parent firm In the existing internationalization literature, it has
been shown that sustaining successful internati onal business relationships is highly advantageous to internationalization (Welch and Luostarinen, 1993). Because this is international trade and collaboration is an exchange process that begins with exchange of information having do with the matching of a product or foreign technology with a perceived market need at home or abroad (Liang and Parkhe, 1997). Internationalization can increase a firm’s sales volume by expanding sales into new foreign markets, and can reduce risks by setting up operations in di versified geographic markets.
International expansion also serves as a learning opportunity for firms . In order to obtain these aims, parent firms expand subsidiaries into more countries. They usually grant a higher level of autonomy to their subsidiaries so that the subsidiaries can respond to local needs quickly (Chiao et al., 2008).
In addition, Garnier (1982) argued the more degree of internationalization of MNC group , the smaller dependence on each local affiliate (the lower the risk), and hence the higher the general autonomy. Then, the author investigated 42 subsidiaries in France and 102 subsidiaries in Mexico.
The empirical results contradicted with this argument.
According to internationalization theory (Johanson and Vahl ne, 1977), internationalization also helps parent firms develop alliances with foreign businesses (Zhou et al., 2007). Through such alliances, firm can gain access to new technologies, complementary resources, and various forms of institutional support (Wan and Hoskisson, 2003). Based on these viewpoints, we argue that internationalization can increase profitability by creating economies of scale or scope, allowing firm access potential market opportunities (Zahra et al., 2002; Zhou et al., 2007) . To avoid the risk and exploit the economies of scale or scope in foreign markets; parent firms need to take care more their subsidiaries. Thus, we propose the following hypothesis:
Hypothesis 4: The degree of internationalization of a parent firm is negatively associated with the
degree of subsidiary autonomy.
2.2 The degree of institutional quality in the host country
CMEs such as Germany, Japan, Switzerland and northern European countries have a highly integrated national business system (NBS) whose key characteristic is that major in stitutions are more interdependent. For example, they have inter -linkages between national infrastructure, corporate strategy and firm behavio ur as a result of institutional complementarities (Hall and Soskice, 2001). The ‘strategic interaction’ is reflect ed by ‘dense networks linking the managers and technical personnel inside a company to their counterparts in other firms ’ (Hall and Soskice, 2001:
23) and ‘an internal structure of the firm based on collaborative and cooperative modes of action ’ (Hall and Soskice, 2001: 24). Moreover, these economies have developed enterprise -based unions in which labour union and government agencies have very strong influences on firms, such as participating in firm decision-making (Dore, 2000). Therefore, MNCs may face se veral difficulties in implementing global practices in subsidiaries located in these countries (Geppert and Williams, 2006).
However, on the other hand, the LMEs of Anglo-Saxon countries and Ireland have low-level integrated NBSs (Whitley, 1999). They have a low level of commitment and cooperation between firms and between employers and employees , and a high level of mobility of operations. The main characteristics of these LMEs are a lack of integration or systematic coordination of activities, absence of legal constraints on management’s use of labour resource s and weak rights of employee representative bodies. Hence, MNCs are easily able to apply global strategy in subsidiaries located in these economies (Geppert and Williams, 2006). As a result, we propose the following hypothesis:
Hypothesis 5: The degree of institutional quality in the host-country environment negatively affects the autonomy level of a subsidiary granted by an MNC.
2.3 Subsidiary characteristics
Subsidiary age The research shows age to be a determinant of the degree of decision -making
authority of a subsidiary. Most studies assert that after several years of operation subsidiaries are
allowed more autonomy than those with little experience. Stopford and Wells (1972) first cited an
example in support of this contention: ‘well established subsidiaries are allowed considerably more
flexibility in short-term financial management than are the new subsidiaries ’ (Stopford and Wells,
1972: 94). This result was also demonstrated with empirical evidence. Alsegg (1971) also found the
age of the subsidiary to be an effective factor, suggesting that subsidiaries which have long been
dependent on the multinational firm will have well-established connections with local stakeholders and extensive local experience. In such cases the risk of granting greater autonomy seems to be low.
Thus, older subsidiaries are more autonomous than subsidiaries which have had a shorter affiliation with their foreign parent company. This argument was confirmed by Youssef (1975), whose results found that this factor increases with the increasing age of the subsidiaries. Similarly, the decentralization of personal decision -making is interpreted by Welge (1981) to suggest that older subsidiaries develop managers who are familiar with local conditions. Consequently, they can participate in the decision-making process to a higher degree. The empirical research findings of Garnier (1982) and Gate and Egelhoff (1986) also suggest that centralization is negatively correlated with the age of subsidiaries due to older subsidiaries having more experience, thus presenting less risk to the parent company and enjoying greater autonomy. A recent empirical study by Taggart and Hood (1999) emphasizes once more that the degree of subsidiary autonomy is positively correlated with its age. Finally, all the findings of the papers mentioned have again been confirmed by Young and Tavares (2004) with their descriptive research. Hence :
Hypothesis 6: Subsidiary age is positively associated with the degree of its autonomy.
Extent of ownership of subsidiary Welge (1981) conducted a study in six German MNCs within the
chemical industry and fifteen foreign manufacturing subsidiaries. The findings indicated that the relationship between ownership and dec ision-making dimensions is positive. This result seems to support the expectation of a positive correlation between ownership and coordination intensity empirically substantiated by Alsegg (1971) , Youssef (1975) and Picard (1977). In addition, Singh (1981) showed that the decision-making authority of a local manager within the host country increases as a result of the latter’s increasing ownership demand. Garnier (1982) also asserted that the higher the extent of local ownership of foreign subsidiaries, the lower is the foreign affiliates’
dependence on the parent company. The findings of Gates and Egelhoff (1986) confirmed that the extent of local ownership of a subsidiary positively affects the extent of autonomy granted by MNCs.
On the other hand, Becker-Ritterspach (2006) argues that the parent company will make demands
and intervene in the affairs of local partners with full ownership in India , while Birkinshaw (1996)
states that ‘the literature on the world product mandate primarily arose in Canada wher e the level of
ownership is sensitive to the needs for subsidiaries to take responsibilities beyond their own borders ’
(1996: 469). Birkinshaw and Hood (2000) subsequently found that high local ownership of a
subsidiary is significantly and negatively asso ciated with its autonomy, using a sample of 229
subsidiaries in Canada, Scotland and Sweden. In spite of several oppos ing arguments, from our standpoint, if a subsidiary holds the absolute equity majority in its unit, this leads to sufficient power to have de facto control and decision -making freedom. Hence, the following hypothesis :
Hypothesis 7: The greater the local ownership of a subsidiary, the greater the degree of autonomy granted.
Subsidiary size A survey of the previous literature shows that the relationship between subsidiary
size and autonomy is not straightforward. Some studies found a negative association between size and autonomy while others identified a positive relationship (Young and Tavares, 2004). Young and Tavares also cited Alymer (1970), Peccei and Warner (1976) and Picard (1977) in support of this argument, though all failed to find a significant relationship. Similarly, while suggesting that the manager of a larger subsidiary may exert more influence in marketing decisions, Goehle also failed to find a significant relationship (Goehle, 1980). In addition, Garnier (1982) found that the larger the affiliates, the more difficult it is for the parent company to directly control their operations. A local firm that is relatively large in comparison to the owning group will be in a better position to obtain a high degree of decision-marking autonomy. Harzing (1999) also wishes to emphasize this, relying on individual survey respondents from a fairly small number of organizations (a total of 25 respondents respectively from UK and Dutch companies, 16 from German, 14 from French, 14 from Swiss, 13 from US and 11 from Swedish companies). The empirical finding indicated that subsidiary size positively affects its autonomy. Conversely, Gate and Eg elhoff (1986) indicated that there is a mixed relationship between relative subsidiary size and its autonomy , but a positive relationship in terms of absolute size. Taggert and Hood (1999) tested distinct proxies for subsidiary size, namely employment and sales. Their empirical findings established a positive association between employment and autonomy which was not significant, and a negative relationship between sales and autonomy. Moreover, larger subsidiaries usually have more resources and are less dep endent on parent firms because they are able to develop relationship with a greater number of external organizations, affording them access to a variety of resources (Preffer and Salancik, 1978).
Therefore, a subsidiary having greater resources and capabilities are usually associated with greater autonomy (Prahalad and Doz, 1981; Johnston and Menguc, 2007; Chiao et al., 2008).
Another explanation, by Hedlund (1981), is that as a small subsidiary builds up its resources,
it becomes less strongly tied to the MNC and its autonomy increases (phase 1). However, when a
subsidiary becomes larger its role within the MNC becomes greater and the parent company
increasingly controls its decision-making authority (phase 2). This argument is confirmed once more by the empirical investigations by Johnston (2005), Johnston and Menguc (2007). Indeed, subsidiary size consists of two types, namely relative size and absolute size, with previous research usually measuring size using data such as sales, number of employees or net assets. However, Johnston’s study only mentions absolute size in terms of the number of employees because the relative size only reflects gross value not value added. Value added reflects the real value of the subsidiary; this value contributes to the whole added value of the parent company. The subsidiaries with the greatest relative size may be the most vertically integrated within the MNC and therefore the least autonomous. This does not reflect variations in autonomy as size changes. Thus, we propose the following hypothesis:
Hypothesis 8: The degree of subsidiary autonomy and its absolute size will be an inverted U- shaped relationship.
2.4 Geographic, economic and cultural distance
Geographic distance Previous studies argued that geographic distance has significantly associated
with the control of a parent firm with respect to its subsidiaries. Hedlund (1981) argued geographic distance to have an opposing effect on subsidiary autonomy is that h eadquarters managers would lack information on unfamil iar environments and, therefore, allow subsi diary managers greater autonomy. In addition, multinationals with higher geographic distance between home and host countries can create opportunities for benefits from economies of scale, ‘spill -over’ effects such as those in advertising (Daniels & Radebaugh, 1998), lower costs of technology transfer within a region (Vachani, 1991), lower costs and greater operational flexibility from using JIT principles in product shipments among subsidiaries within a region. Su ch companies have an opportunity to reduce coordination costs (Grant, 1987) made possible by physical and cultural proximity among a group of countries and by similarities in their level of economic development (Vachani, 1991).
Companies that choose to tak e advantage of this opportunity to reduce coordination costs will
probably find it easier to do so by decentralizing decision -making, since subsidiaries are more
likely to have the information necessary for taking advantage of regional synergy. Thus , Vachani
(1999) hypothesized that multinationals with higher geographic distance have higher subsidiary
autonomy. The empirical results showed that this hypothesis was supported and statistically
significant. In addition, based on an international mail survey co vering 169 subsidiaries of MNCs headquartered in the USA, Japan, the UK, Germany, France and the Netherlands, Harzing and Noorderhaven (2006) investigate the impact of geographical distance on the role that subsidiaries play in the MNC network and the way they are managed. They asserted that headquarters to grant a higher level of autonomy to subsidiaries as the geographical distance makes centralized decision - making less appropriate and less feasible . Thus, we propose the following hypothesis:
Hypothesis 9: The higher geographic dista nce between the home and host counties , the higher degree of subsidiary autonomy granted.
Economic distance Most of previous studies had no much attention to t he economic development
level between the home and host countries on subsidiary autonomy, as had other factors.
Nevertheless, it could be an important determinant of the control exercised by the parent firm. The
literature (e.g., Edwards et al., 2002) show ed that the more developed is the country in which the
subsidiary is located – in the sense of demand, existence of potential sourcing partners and level of
the national innovation system – the higher is the likelihood that the subsidiary could develop an
extensive external network, improve different capacities, and therefore gain more autonomy. In
this line, Männik et al. (2004) asserted that the autonomy ratio of the foreign subsidiaries of MNC s
might have a strong relations to the economic development lev el between home and host countries
since the role of a subsidiar y varies according to such contingencies as the local environment, the
structural context imposed by the parent, and the entrepreneurial capacity of management
(Birkinshaw and Hood, 2000). The more developed the host country the more responsibility could
be given to the local unit of the multinational company . In addition, Hoskisson et al. (2000) argued
that the emerging and transforming markets are economically fast -growing and structurally
volatile. Consequently, the external networks of subsidiaries in t hese countries are quickly
changing, providing bases for much more rapid change in the capacities and also in their role in
internal (corporate) networks . Thus, Männik et al. (2004) and Varblane et al. (2005) have
hypothesized that subsidiaries located in more developed countries will have a higher degree of
autonomy than subsidiaries situated in less developed countries. They found a significant and
positive relationship between the economic development level of the country in which a subsidiary
is located and the subsidiary autonomy. Thus, we suggest the following hypothesis :
Hypothesis 10: The economic difference between the home and host count ries are above, a parent company grants less autonomy to its subsidiaries .
Cultural distance According to Balgia and Jaeger (1984), MNCs use two types of control over
their subsidiaries: bureaucratic and cultural control. Bureaucratic control utilizes an extensive set of rules, regulations and procedures that clearly limit the subsidiaries’ role and autonomy. Cult ural control utilizes a set of shared values and norms for work processes , behaviors and the like (Ouchi, 1980; Balgia and Jaeger , 1984). Cultural control may be achieved by placing a number of expatriates who may either directly control subsidiary operati ons by acting as mini headquarters or indirectly control the subsidiary based on socialization (Harzing , 2001). Thus, the number of expatriates present in a subsidiary reflects the level of cultural control a parent firm wants to exert on that subsidiary (Konopaske et al., 2002; Colakoglu and Kaligui ri, 2008).
Although the amount of cultural distance between headquarters and the subsidiary in question might be an important factor influencing the type of control mechanism that is used, very little previous research is available in this field (Harzing, 1999). Rosenzweig and Singh (1991) hypothesize a positive relationship between cultural distance and the reliance on formal mechanisms of control. Rosenzweig and Singh do not define what they mean by formal con trol, but Harzing (1999) assumed that it comprises both the bureaucratic formalised control and the output control category. However, Brott (1984) and Herbert (1999) consider how culture defies the ability of managers to generalize strategies. Differences in culture not only bring into question to what extent conclusions are to be generalized, but also create a certain bias in the development of the tools of analysis (Adler, 1983). It is thus not satisfactory to ta ke theories developed in the US , for example, and just apply them to foreign countries. On the other hand, their applicability abroad should not simply be ignored (Paterson and Brock, 2002) . Thus, the previous studies have produced conflicting findings: one stream of research argues that when cultu ral distance is greater firms increase their level of control; while the other stream suggests that greater cultural distance is associated with a loosening of control (Wilkinson et al., 2008) .
Hofstede (1980: 25) defined culture as: ‘the collective progr amming of the mind which distinguishes the members of one human group from those of another. Culture in this sense is a system of collectively held values and applies to both national and corporate cultures’. He described the content of mental programmes a s values and in a survey on employee attitudes of the world-wide subsidiaries of IBM, Hofstede (1980) found four norm values which he termed the
‘dimensions’ of culture. Later research on these dimensions showed that for the countries of Asia it
was necessary to add another dimension, as the people of these cultures have a different philosophy concerning life fulfillment. It is along these dimensions that Hofstede (1980) found that cultures in separate countries differ. Hofstede’s (1980) notion of culture h as been widely used in literature and it has been the source of several interesting results. Nevertheless, his dimensions have also been the subject of critique, mainly because conclusions were drawn from data of a single multinational and the time-span since the research, leading to the development of other frameworks making the concept of culture more tangible. One of them is the GLOBE project, which is an extension of the Hofstede (1980) framework. In recent years a group of over 150 researchers conceptu alized and developed measures of nine cultural dimensions. This could, thus, be seen as a significant addition to the previous framework. However, so far, it has not been as widely used in literature as the Hofstede (1980) framework. Moreover, Leung et al. (2005) have argued that although subsequent work on culture has led to some clarification and refinement, the Hof stede dimensions are robust.
In this study, we consider subsidiary autonomy as the central issue . Therefore, we need to address in which way culture might influence the degree of autonomy possessed by subsidiaries. In line with other literature ( Harrison et al., 1994) the abstract term of culture will be made more tangent through the use of Hofstede’s culture dimensions in order to relate it to autonomy. As argued above, the extensive utilization and recognition of the framework proposed by Hofstede (1980), makes it a useful tool for analysis between both countries and diffe rent bodies of literature.
Hofstede (1980) analyzed survey data on work -related values obtained between 1967 and 1973 from more than 117,000 IBM employees working in 40 different countries, and found that four statistically-independent dimensions explained the inte r-country variation in employee responses to his survey questions. He labeled these four dimensions ‘power distance’, ‘uncertainty avoidance’ ,
‘individualism’, and ‘masculinity’, and assigned each country in his sample a score on them that
varied between 0 and 100. It was argued that in order to generalize for global use, an extra
dimension had to be added that corresponded with the Asian cultures, but to a much lesser extent to
the Western cultures (Hofstede and Bond, 1988). This dimension is known as ‘sho rt-term versus
long-term orientation’ or ‘Confucian dynamism’. Over time, the validity of these dimensions has
been confirmed by many studies (e.g., Van Oudenhoven, 2001 ; for an overview of ear lier
replications, see Søndergaard, 1994), suggesting that they can reliably be used to classify countries
according to their national cultures and to determine the cultural dist ance between them. For the
purposes of the present study, which range no further than the boundaries of Europe, this fifth
dimension is considered to be of no added value (Harrison et al., 1994). Moreover, the data of the
fifth dimension in many countries is not yet available. Therefore, this dimension wi ll not be included in the model.
Power distance
Power distance refers to the extent to which people believe and accept that power and s tatus are distributed unequally and accept an unequal distribution of power as the proper way for social systems to be organized. In organizations, power distance influences the amount of formal hierarchy, the degree of centralization and the amount of in decision making. Companies in high power distance countries tend to be more centralize and have less employee participation in decision making (Newman and Nollen, 1996). There are societies that are characterized by a norm value that inequalities between its members should be minimized. The existing hiera rchies in these societies and organizations within these are only for administrative convenience (Hofstede, 1980:
122). These societies have a low score on the power distance (PD) scale. In low PD societies subordinates and superiors regard each other as e quivalent people, who have equal rights (Hofstede, 1980: 122), and subordinates expect to be consulted on actions or decisions that affect them (Hofstede, 1980: 110). By contrast, other societies are characterized by the acceptance of inequality and the institutionalization of hierarchy - a hierarchy that reflects and underpins inequality. In these high PD societies, superiors are supposed to perform a leading role, to make decisions autocratically and paternalistically, and subordinates are commonly fearsom e and unwilling to disagree with their superiors (Hofstede , 1980: 110). Because power distance is directly associated with the acceptance (high PD) or rejection (low PD) of the unequal distribution of power, Hofstede (1980 : 107) argues that high PD will be associated with greater centralization in organizational design, and low PD with greater decentralization (Harrison et al., 1994).
Therefore, we propose the following hypothesis:
Hypothesis 11: Subsidiaries located in a country with high score on power distance receive less autonomy from a parent firm than subsidiaries located in a country with low score on power distance.
Individualism
Individualism (IDV) and its opposite collectivism refer to the degree to which a society emphasizes
the role of the individual as opposed to that of the group (Hofstede, 1980). In individualist
countries (high IDV), the ties between individuals are very loose. Everybody is expected to look
after his own interests and at most the interest of immediate family. Newman and Nollen (1996) indicated that in organizations, individualism is manifested as autonomy, individual responsibility for results, and individual -level rewards. By contrast, in a collectivistic society (low IDV), the ties between individuals are very strong. Peo ple are born into collectivities or groups and people are supposed to look after the interest of that entire group (Hofstede, 1980).
An individualist culture is typified by a focus on the individual as having a self -identity which is both unique and whole, and by an emphasis on the individual's interests and attributes, rather than on the interest of the group(s) to which individuals may belong. Conversely, in a collectivist culture, a person is seen as whole only when considered in terms of a group affilia tion (Harrison et al., 1994). The dimension of individualism is essential because it includes elements of autonomy, initiative, freedom and challenge. These elements are also being considered in the decentralization decision. As Hofstede (1980: 160) puts it, in high IDV societies it is important to have freedom and challenge in jobs; managers aspire to leadership and variety; it is more important for managers to have considerable autonomy; and individual initiative is socially encouraged, as compared to low IDV cultures where such initiative is socially not desirable.
Because decentralization is many times seen as an organizational mechanism promoting these attributes (e.g. Harrison et al., 1994), we propose the following hypothesis:
Hypothesis 12: Subsidiaries situated in countries scoring high on individualism will be granted more autonomy by the MNC than subsidiaries in low individualism countries .
Masculinity
Masculinity (MAS) and its counterpart femininity refer to the extent to which a society emphasizes traditional masculine values such as competitiveness, assertiveness, achievement, ambition, and high earnings, as opposed to feminine ones such as nurturing, helping others, putting relationships with people before money, not showing off, and minding the quality of life.
Feminine cultures define relatively overlapping social roles for the sexes, in which neither men
nor women need to be ambitious or competitive (Hofstede, 1984: 390). In other words, it
represents the degree to which people prefer value s of success and competition over modesty and
concern for others and therefore it can also be seen as ego enhancement versus relationship
enhancement. In this sense, it could be argued that that the ego enhancement country, so a low
MAS score, would opt fo r decentralization of decision -making, as it creates an extra feeling of
importance and power. By contrast, a country scoring high on MAS would tend to centralize
decision-making, thereby bonding with other important individuals and groups. Therefore, we propose the following hypothesis:
Hypothesis 13: Subsidiaries of MNCs originating from h igh MAS cultures will have less autonomy than subsidiaries of MNCs from low MAS cultures.
Uncertainty avoidance
Uncertainty avoidance (UA) refers to the extent to whic h people are threatened by uncertain, unknown, or unstructured situations . This dimension refers to how a society confronts the uncertainty of the future. A high UA society maintains a shared belief that ‘the uncertainty inherent in life is felt as a continuous threat that must be fought’ (Hofstede, 1980: 184) by ‘adhering to strict laws and rules, and safety and security measures ’ (Hofstede and Bond, 1988: 11). A low UA society is one in which life's uncertainty ‘ is more easily accepted and each day is taken as it comes’
(Hofstede, 1980: 184). In the organizations of a culture with high UA, the use of technology, rules,
and rituals are widely applied to impose order and predictability on an otherwise uncertain
environment. UA is, therefore, an indicator of the degree to which a culture values a sense of
control (Barr and Glynn, 2003). Moreover, Hofstede states that, ‘uncertainty avoidance leads to an
escape from ambiguity’ ( Hofstede 2001: 148). Therefore, Hofstede’s conceptualization of UA
embeds the notion of control. In this sense, it can be argued that control is to be seen as
formalization or standardization of rules and decision -making processes in order to decrease
uncertainty. Consequently, or inversely, decentralization of decision -making would increase
uncertainty for the parent company. In addition, Barr and Glynn (2003) related the Hofstede (1980)
cultural dimensions to controllability. They studied cultural variations in the strategic issue labels
of threats and opportunities. In a survey of 276 Am erican and international respondents, they
investigated the sensitivity of issue attributes that discriminate between threat and opportunity. The
results showed that the cultural value of uncertainty avoidance had a significant effect: Compared
to low UA cultures, individuals from high UA cultures were significantly more sensitive to
controllability in perceiving strategic issues. However, other cultural value dimensions
(individualism, masculinity, power distance) did not have similar effects . Furthermore, Wilkinson
et al. (2008) conducted research to explore the diminishing effect of cultural distance on subsidiary
control issues. They employed Hofstede’s cultural dimensions in a regression model to see whether
distance affects the control exercised by hea dquarters for Japanese MNCs. The strongest effect was
found in the ‘uncertainty avoidance’ dimension. With this line in mind, we propose that:
Hypothesis 14: Subsidiaries originating from low UA cultures receive more autonomy from a parent firm than subsidiaries from high UA cultures.
2.3 Control variables
The relatedness of the home and the host countries It is defined in terms of the level of similarity
between the business environment in the parent company’s country of origin and the country where the subsidiary is located. In fact, if this similarity level is high, the head office managers of MNCs are able to use their knowledge to control foreign subsidies , while head offices depend on the local knowledge of foreign subsidiary managers in operating a local business where the similarity is low (for example, Caves, 1982; Erramilli and Rao, 1990; Edwards et al., 2002) . Moreover, the external environment and the host-country environment determine the role of the MNC subsidiary, including its autonomy (O’Donnell, 2000; Benito et al., 2003). Thus, we would expect that if the home and host countries have similar business environments, the autonomy of the subsidiary will be low, if there is little similarity the autonomy will be high.
The degree of interdependence It refers to the condition in which one subsidiary or subunit of the
MNC relies on another subunit’s activities or inputs in order to perform its role effectively
(Garnier, 1982; O’Donnel, 2000)
.According to Garnier (1982), degree of interdependence is
evaluated through two aspects: (i) the percentage of the local affiliates’ total sales that are going to
the parent or to the MNC group, and (ii) the percentage of its purchases coming from the parent or
the group. In both cases, Garnier (1982) found that the higher the percentage, the higher degree of
interdependence, and the lower the affiliate’s autonomy. Then, O’Donnel (2000) confirmed this
result by investigating 255 headquarters–foreign subsidiary pairs in U.S. MNCs . In addition,
Mirchanani and Lederer (2004) showed that intra-company purchases can create a critical type of
dependence in a multinational corporation, and are an important tool with which the parent can
exploit its multinationality in order to compete. They can lead to higher agency cost s and to
reduced autonomy for the subsidiary. This is because subsidiary -level decisions have greater
consequences for the firm as a whole when interdependence is high. Therefore, we expect that the
degree of interdependence increases, the subsidiary auton omy increases.
3. THE ECONOMETRIC MODEL
For the following reasons, we use multilinear regression, Poisson regression and a negative binomial model to analyse the determinants of the degree of subsidiary autonomy within MNCs . First, our dependent variable is constructed as a count variable (the number of decisions per subsidiary in a given year). Second, most previous research uses OLS estimation while the dependent variable of the present study is a count variable, thus we want to examine whether or not there are differences in the results of the three models.
The first model examines the OLS estimate with the following specification:
Y = β0 +β1X1 + β2X2+ β3X3 + β4X4+ β5X5+ β6X6+ β7X7 + β8X8 + β9X9 +
β10X10 + β11X11+ β12X12 + β13X13 + β14X14 + β15X15 + β16X16+ ε
(1)
where Y = the degree of subsidiary autonomy
X
1= the overall strategic approach of the parent company X
2= the degree of product diversification of the parent company X
3= the size of the parent company
X
4= the degree of internationalization
X
5= the degree of institutional embeddedness of the subsidiary in the host country X
6= the age of the subsidiary
X
7= the extent of ownership of the subsidiary X
8= the size of the subsidiary
X
9= the geographic distance between the home and host countries X
10= the economic distance between the home and host countries X
11= the power distance of the host country
X
12= the individualism of the host country
X
13= the uncertainty avoidance of the host country X
14= the masculinity of the host country
X
15= the relatedness of the home and the host countries X
16= the degree of interdependence
β
0- β
16are unknown parameters to be estimated
The least square estimations are the Best Linear Unbiased Estimators (BLUE) only if the six following assumptions hold (Hill et al. 200 8; Wooldrige, 2001):
MR1: y
t= β
0+ β
iX
it+ e
ti = 1 10; t = 1 263 MR2: E(y
t) = β
0+ β
iX
it E(e
t) = 0
MR3: var(y
t) = var(e
t) =
2(No heteroscedasticity) MR4: cov(y
t, y
s) = cov(e
t, e
s) = 0 (No autocorrelation)
MR5: The values of X
tare not random and are not exact linear functions of the other explanatory variables.
MR6: y
t~ N(β
0+ β
iX
it,
2) e
t~ N(0,
2)
If in our linear regression model one of these assumptions is violated and we use the least squared estimator to estimate the unknown coefficients , then the least squares estimator is still a linear regression and unbiased estimator, but (1) it is no longer BLUE; (2) the standard errors usuall y computed for the least squares estimator are incorrect ; and (3) confidence intervals and hypothesis tests that use these standard errors may be misleading (Hill et al., 2008; Greene, 2002).
The second model examines the quasi -maximum likelihood estimator (QMLE) that the degree of subsidiary autonomy is granted by the parent company. We use Poisson regression with the following Poisson distribution (Hill et al. , 2008; Wooldridge, 2001; Greene, 2002):
) !
( Y
Y e f
Y i
; Y = 0, 1, 2, 3…10 (2a)
where f(Y
i) denotes the probability that the variable Y takes non -negative integer values. Y is the number of decisions undertaken by a subsidiary .
Y! is Y factorial with Y! = Y x (Y-1) x (Y- 2) x (Y-3) x … x 2 x 1. is the mean number of decisions, for all subsidiaries, with:
= exp(βi 0
+ β
1X
i1+ β
2X
i2+ …. + β
16X
i16) (2b)
Assumption of Poisson regression: var(Y) = (its variance is the same as its mean value).
We have a Poisson regression model with the following specification:
Y
i= e
β0+ β 1Xi1+ β 2X
i2+…. + β 16X
i16
+ ε
i(3)
However, the negative binomial model is a generalization that is often used when the Poisson regression assumption that var(Y) = = E(Y) is unacceptable.
The third model also uses a quasi-maximum likelihood estimator that the degree of subsidiary autonomy is granted by the parent company. We use a negative binomial regression model with the following distribution function:
N
i
i i
i i
i m x Y m x Y Y
Y l
1
2 2
2 2
2 ( , )) ( 1/ )log(1 ( , )) log ( 1/ ) log( !) log (1/ )
log(
,
(4) where: m(x
i,
) = E(y
i|x
i,
)
is a variance parameter to be jointly estimated with conditional mean parameter2
.
The negative binomial distribution is often used when there is overdispersion in the data, so that v(x
i,)>m(x
i,
), if the following moment conditions hold:
m(x
i,
) = E(y
i|x
i,
) = exp(x
i’) var(y
i|x
i,
) = m(x
i,
)(1+
m(x2 i,
)) If we maximize the negative binomial log likelihood, given above, for fixed
, we obtain the2QMLE of the conditional mean parameters. This QML estimator is consistent even if the conditional distribution of Y is not negative binomial, provided that it is correctly specified (Wooldridge, 2001).
4. DATA AND IMPLEMENTATION
4.1 Variables and measures