The influence of degree of development of the host country and size of a subsidiary on the use of subsidiary initiatives

(1)

The influence of degree of development of the

host country and size of a subsidiary on the use

of subsidiary initiatives

Master thesis International Business and Management

Name: Jeroen Bijsterbosch

Student number: s1764578

Address: Billitonstraat 106, 9715EZ

Telephone number: +31 (0)6 15147529

E-mail: j.bijsterbosch@student.rug.nl

Supervisor:

Dr. H.J. Drogendijk

Co-Assessor:

Drs. H.L. Faber

Faculty of Economics and Business

University of Groningen

(2)

2

ABSTRACT

Subsidiaries worldwide are constantly looking for new opportunities for internal, local and global application, often even without headquarters involvement. Many subsidiaries take initiatives in order to execute these opportunities. Because not all subsidiaries have the same opportunities, these initiatives differ in many aspects like how and where they are taken. This study is the first that tries to explain differences in subsidiary initiatives between subsidiaries on a global level, using data from subsidiaries located in 22 different countries. This is done on the basis of degree of development of the host country and size of a subsidiary. No significant relationships were found between these two concepts and subsidiary initiatives. Further research is necessary in order to determine how these concepts influence subsidiary initiatives.

Keywords: external initiatives, internal initiatives, degree of development of the host country,

(3)

3

ACKNOWLEDGEMENTS

In this section, I would like to thank everyone who helped me to complete this master thesis. The last five months have been a very informative and unique experience.

First of all, I would like to thank my supervisor dr. H.J. Drogendijk for supervising me and all the useful feedback she has provided during the past five months. Furthermore, I would like to thank all the participating subsidiary managers for their time and effort. I would also like to express my appreciation towards my family, and in particular to my dad, who helped me a lot with retrieving the data. And last but not least, I would like to thank my friends for all the valuable discussions and feedback.

(4)

4

LIST OF TABLES AND FIGURES

TABLE 1: Subsidiaries per country 21

TABLE 2: Cronbach’s Alpha’s 24

TABLE 3: Descriptive statistics and Intercorrelations 26 TABLE 4: Summary results for hypotheses 28 TABLE 5: Factor Analysis: Total Variance Explained 28 TABLE 6: Factor Analysis: Component Matrix 29

(6)

6

1. INTRODUCTION

This study is concerned with subsidiary initiatives, which are proactive, autonomous and risk-taking activities that originate outside the home country in a foreign subsidiary of a Multinational Corporation (MNC) and are initiated by actors in the subsidiary (Schmid, Dzedek and Lehrer, 2014). My interest in studying subsidiary initiatives arises from two reasons. Firstly, subsidiary initiative outcomes have been investigated much less in the environmental context than in the organizational context (Schmid et al., 2014). Most studies focus on subsidiary initiatives in relation to organizational factors like degree of autonomy and headquarters attention but not in relation to environmental factors (Schmid et al., 2014), like degree of available resources. Birkinshaw (1997) states that a successful subsidiary initiative is an ‘entrepreneurial process’ that begins with the identification of an opportunity and culminates in the commitment of resources to that opportunity. Because of the large differences in development and thereby the available resources between countries worldwide, it can be expected that subsidiaries located in different countries take different initiatives. But the empirical evidence supporting this statement is limited. Secondly, the current literature, with respect to subsidiary initiatives, is focused on relatively small subsidiaries (Schmid et al., 2014). The differences in initiatives between small and large subsidiaries have also been studied too little. The aim of this study is to provide better insights into the relationships between the degree of development of the host country and size of a subsidiary and the types of initiatives they take.

(7)

7

is most important. Another factor is the so-called ‘herd mentality,’ which causes companies to focus on markets that competitors have identified. This is a clear indication that corporate executives are not able to divide their attention equally between all their subsidiaries. Executive attention (also referred to as selective and focused attention) reflects the individual’s capacity to select relevant information and to ignore irrelevant information in order to facilitate the production of appropriate responses (Posner and Rothbart, 2007). Birkinshaw et al. (2007) argue that executive attention is a scarce resource, because executives do not have the time and energy to process all the information they receive. Management scholars agree that managers and executives are confronted with far more information than they can handle, and so they have to be selective in those aspects of the environment that enter their consciousness (Cyert and March, 1963; March and Simon, 1958; Mintzberg, 1973). Birkinshaw et al. (2007) argue that it is important for subsidiaries to receive enough executive attention, because too little attention can result in missed opportunities and decisions by talented employees to leave. Their research found that initiative taking is one way subsidiaries use to gain visibility in a global company.

Another concept that the literature frequently links with subsidiary initiatives is autonomy. Autonomy can be defined as a discretion or degree of freedom that the subsidiary has to pursue its own independent agenda that may or may not be endorsed by parent headquarters (Raziq, Borini & Perry, 2014). Autonomy is frequently taken to be important in enabling a subsidiary to enhance their performance and play a larger role for their parent company than that given originally. This in turn raises the question of how subsidiaries obtain increased autonomy. One suggestion is that autonomy can increase as subsidiaries challenge their subordinate position by taking initiatives (Birkinshaw, Hood & Young, 2005). Initiative taking can enhance a subsidiary’s technical, managerial and marketing expertise and reduce the parent company’s ability and motivation to monitor a subsidiary as it engages in activity outside the expertise of the head office (Johnson & Medcof, 2002; O’Donnell, 2000). Ambos et al. (2010) found that subsidiary initiatives have a direct effect on subsidiary autonomy, but the caveat is that initiatives also evoke headquarters monitoring, which in turn decreases the subsidiary’s autonomy.

(8)

8

enhanced learning of the entire MNC (Schmid et al., 2014). It seems clear, why the study of subsidiary initiatives constitutes a growing literature in the International Business field (Verbeke, Tavares-Lehman and Van Tulder, 2011). But why is it important to study initiatives in MNC subsidiaries rather than in the parent company? This is because the concept of subsidiary initiative is important in advancing our understanding of how MNEs evolve, because it explains how change can occur from within – through the entrepreneurial efforts of individuals a long way from the corporate headquarters (Ambos et al., 2010). Another reason is that despite the compelling logic for tapping into local markets through the subsidiary network (Bartlett and Ghoshal, 1986), many corporations appear to neglect the creative potential of their subsidiaries (Birkinshaw, 1997). For instance, local subsidiary management plays a crucial role in the scanning and detection of opportunities (Boojihawon, Dimitratos & Young, 2007). And a final and very important reason to have a closer look to subsidiary initiatives is that Ambos and Birkinshaw (2010) show that subsidiaries undertaking entrepreneurial initiatives perform better, both financially and managerially, when compared to other units of the MNC.

(9)

9

same conditions. Scott and Gibbons (2009) state that there is little research-based practical guidance as to the strategies and tactics subsidiary’s managers can adopt to ensure their subsidiary’s survival and to compete for further investment. Taking initiatives can help subsidiaries to develop, partly by helping subsidiary managers respond to threats and opportunities that they have identified in the internal (within MNE) or external (host country) environments or both (Birkinshaw et al., 2005). This leads to the following main research question:

What is the influence of degree of development of the host country and size of a subsidiary on the use of subsidiary initiatives?

The study is organized as follows. First, a classification of subsidiary initiatives is given followed by an overview of the current literature regarding subsidiary initiatives in relation to the degree of development of the host country and size of a subsidiary. Second, the research methodology is described and details about the data collection method and data analysis are provided. Third, the results from the study are discussed. The study closes with a discussion and conclusion of the findings.

2. THEORETICAL BACKGROUND

This section provides an overview of the current literature regarding the main concepts in this study, degree of development of the host country, subsidiary size and subsidiary initiatives.

2.1 CLASSIFICATION OF SUBSIDIARY INITIATIVES

(10)

10

internal initiatives. External initiatives refer to initiative opportunities that are identified outside the boundaries of the corporation and which develop through interactions with local customers, suppliers or other stakeholders and entities (Birkinshaw, 1998). On the other hand, internal initiatives represent opportunities recognized within the boundaries of the corporation and emerge through interactions of subsidiary managers with other actors of the MNC corporate system (Birkinshaw, 1998). Schmid et al. (2014) come up with another criterion which is relevant with respect to external initiatives, namely the ‘locus of application’. This criterion refers to the aim of developing products or services for customers in local or global markets. In summary, three different subsidiary initiatives can be defined: local market initiatives, global market initiatives and internal market initiatives (see figure 1).

Figure 1: three types of initiative (Birkinshaw, 1997)

(11)

11

2.2 DEGREE OF DEVELOPMENT OF THE HOST COUNTRY AND SUBSIDIARY INITIATIVES

Wan and Hoskisson (2003) point out the importance of home country environments and recognize that their levels of environmental munificence vary. Despite the fact that their study is focused on corporate diversification strategies, it describes in detail the differences between home country environments worldwide. They define ‘munificence’ in terms of factors, the more tangible production factors, such as physical infrastructure and natural resources, available to firms, and institutions, the less tangible forms of support, such as judiciary efficiency, that facilitate transactions. They state that firms draw on their home country environments’ factors and institutions in producing goods and services and in exchanging inputs and outputs with others. Factors are used to produce goods or services, whereas institutions are used for the exchange of inputs and outputs with other firms. The environment’s opportunity set is determined by production factors and institutions, and firms seek to capture the profitable opportunities defined by the opportunity set. Birkinshaw (1997) states that a successful subsidiary initiative is an ‘entrepreneurial process’ that begins with the identification of an opportunity and culminates in the commitment of resources to that opportunity. It can be expected that firms, which are located in an environment where production factors are widely available and production institutions are clearly defined, have a better chance to implement a successful subsidiary initiative because of the more available resources. Birkinshaw et al. (1998) argue that enhanced resources lead to an increase in subsidiary initiative. This does not mean that the number of identified opportunities in less munificent environments will necessarily be lower but the available resources to execute these opportunities will probably be lower.

(12)

12

Asian nations. Subsidiary CEOs are constantly urged to contribute beyond their core mandate and to move their subsidiary’s activities up the value chain and to be innovative and entrepreneurial. The encouragement to be innovative and entrepreneurial helps subsidiaries with the initiative taking process. Birkinshaw, Hood & Jonsson (1998) argue that initiative taking requires leadership and an entrepreneurial culture within the subsidiary organization, with these attributes in turn helping a subsidiary to acquire specialized resources, identify initiative taking opportunities, gain international responsibilities and attract HQ attention. Opposed to subsidiaries in developed countries, subsidiaries located in developing countries have little knowledge base because of their role definitions, relatively shorter period of existence, and also due to the lower level of technological advancement of the local environment (Kumar and Demir, 2013). The better available resources, like a higher level of knowledge base, in combination with the pressure of having their operations relocated suggest that subsidiaries located in developed countries will undertake more initiatives compared to those located in developing countries. Especially the lower level of technological advancement of the local environment suggests that it will be harder to implement external initiatives (both local and global initiatives) in developed countries. On top of that, several studies have shown that subsidiary initiatives lead to enhanced subsidiary resources and capabilities (Schmid et al., 2014). This suggests that subsidiaries initiatives lead to even better resources which could be used for future initiatives. So, the differences between developed and developing countries are likely to increase.

(13)

13

transferring the same practices to different locations (Kumar and Demir, 2013). Another example is the ‘irony of alliancing’ as discussed by Eisenhardt and Schoonhoven (1996). This means that firms must have resources in order to get resources. So when subsidiaries want to collaborate with competitors, suppliers, customers or regulatory bodies in the local environment they need resources in order to receive resources from these entities. Birkinshaw et al. (1998) propose that the level of specialized resources in the subsidiary is positively associated with subsidiary initiative and they found support for this proposition. Besides that, Birkinshaw (1997) argues that the success of an initiative is highly dependent on the commitment of resources to that initiative. This is another indication that it will be easier to execute a successful initiative in more developed environments because of the more available resources.

Based on the main argument that in developed countries the environmental munificence and thereby the number of available resources is higher, which is important in order to implement successful initiatives (like collaborations with external entities), the following can be hypothesized:

Hypothesis 1A: Subsidiaries located in more developed countries undertake more external initiatives compared to subsidiaries located in developing countries.

(14)

14

(developing country), Edwards, Ahmad and Moss (2002) also found that the more a subsidiary is integrated in the MNC’s global operations, the more the subsidiary is controlled.

The main argument, that the number of available resources in more developed countries is higher, which is important in order to implement successful initiatives, in combination with the findings of Raziq et al. (2014) leads to the following hypothesis:

Hypothesis 1B: Subsidiaries located in more developed countries undertake more internal initiatives compared to subsidiaries located in developing countries.

2.3 SIZE AND SUBSIDIARY INITIATIVES

(15)

15

countries. These advantages, in turn, allow these subsidiaries to seek out and exploit opportunities in foreign markets. Also Birkinshaw et al. (1998) note that subsidiary growth brings with it an increase in resources and a corresponding reduction in parent control. As subsidiary size increases, the developing resources in the subsidiary shift the balance in the power asymmetry between HQ and subsidiary more towards the subsidiary (Johnston and Menguc, 2007). All these arguments suggest that the bigger the size of the subsidiary, the more resources are available. More resources lead to less parent control and therefore to more autonomy, which could be used to execute more initiatives.

But there are also clear indications that growing size has a downside with respect to subsidiary initiatives. Subsidiaries differ in the market they serve, their competence levels and level of knowledge (Birkinshaw & Pedersen, 2009). Based upon these differences, subsidiaries gain different levels of importance for the parent MNE. Larger subsidiaries in general carry more resources and become important for MNEs (Penrose, 1995) and in turn these subsidiaries are likely to become more integrated within the MNE network and have less autonomy than more isolated subsidiaries (Hedlund, 1981). Ambos, Andersson & Birkinshaw (2010) argue that initiative taking causes parent companies to reduce subsidiary autonomy and increase their monitoring of subsidiaries to control against future deviations from programmed activities. Also Johnston and Menguc (2007) state that increasing size leads to increasing coordination complexity within the subsidiary. The increased information flows and the ensuing expansion in volume and complexity of decision-making require increasing managerial input from HQ. Internal initiatives require a high level of MNE integration which frequently can mean a low level of subsidiary autonomy (Birkinshaw, 1997). In contrast, external initiative taking typically requires high subsidiary autonomy and is positively related to subsidiary autonomy in return (Raziq et al., 2014).

(16)

16

Johnston and Menguc (2007) found that, while the subsidiary is relatively small, increasing subsidiary size will correlate with increasing resources in the subsidiary and a consequent increase in subsidiary autonomy. This positive linear relationship persists until an inflection point is reached and subsidiary autonomy begins to decline. They argue that this is due to increasing subsidiary size bringing increasing coordination complexity, a need for greater inputs of managerial experience and expertise, and growing interdependence between the subsidiary and the rest of the corporation. This suggests that, to a certain point, increasing subsidiary size will lead to both more external and internal initiatives because of the increasing available resources and autonomy. But after a certain point, when a subsidiary is oversized, headquarters coordination will be higher which will lead to less autonomy in order to take external and internal initiatives.

In line with the arguments and conclusions of Johnston and Menguc (2007), this leads to the following hypotheses:

Hypothesis 2A: At lower levels of subsidiary size there is a significant positive relationship between subsidiary size and the number of external and internal initiatives.

Hypothesis 2B: As the level of subsidiary size increases a significant negative relationship between subsidiary size and the number of external and internal initiatives develops.

3. METHODOLOGY

Because this study tries to identify the existence and strength of relations between the variables discussed in the previous section, quantitative research is appropriate in order to answer the formulated hypotheses (Thomas, 2004, p.205). In this section, the methodology is presented.

3.1 SAMPLING METHOD

(17)

17

was selected to overcome potential communication problems. The underlying logic was that it is easier to convince subsidiaries to participate in the study when they are approached through headquarters. With the permission of headquarters, it can be expected that subsidiaries are more willing to participate. Based on database Orbis, which contains data of 125 million organizations worldwide, there are more than 19.000 organizations matching this criterion. Due the fact that subsidiary initiatives are closely linked with attention (Birkinshaw et al. 2007) and this concept is more relevant in organizations with many subsidiaries, a second criterion was selected. It can be expected that large organizations have more subsidiaries compared to small organizations, so only organizations with over 500 employees in total were selected. Based on database Orbis, there are 725 organizations matching both criteria with at least one foreign subsidiary.

3.2 DATA COLLECTION METHOD

(18)

18

be possible that someone else who has access to the mailbox of the manager filled out the survey (Thomas, 2004, p.123).

Three different methods are used in order to collect as many e-mail addresses as possible. First, on the internet was searched for the e-mail addresses of the foreign subsidiary managers of the selected organizations no matter what management position. Because this is time-consuming, this was done for about 100 organizations that are randomly selected from the sample. It was important to determine whether the information on the website of the organizations and Orbis was up-to-date because it could be possible that managers are moved to other companies. This is verified on both the national and international website of the organizations and the LinkedIn profiles of the foreign subsidiary managers to make sure that the information was reliable. In this way, the e-mail addresses of 167 subsidiary managers from 10 different organizations worldwide were found.

Second, an e-mail was sent to the general e-mail address of 20 large Dutch organizations with the question whether they wanted to participate in the study. Because a lot of data in Orbis was outdated or not available, this method was tried in order to gather many e-mail addresses through headquarters. Unfortunately, this has not led to more e-mail addresses. Only a few organizations have sent a response. They thanked for the interest in their organization, but they conduct research rather internally.

The third method was to check whether there were some personal links between the management of the selected organizations and my personal network. A personal connection is managing partner at BDO Arnhem and he could provide me with the e-mail addresses of 55 managers working in the Netherlands for one of the selected organizations. They have all received an mail (see Appendix II) with the question if they could provide me with the e-mail addresses of their foreign subsidiary managers and if they wanted to ask them to fill out the survey. Eventually, 23 managers have sent a response of which 15 managers were willing to cooperate. These managers have given the e-mail address of 39 subsidiary managers and asked them to fill out the survey.

(19)

19

potential respondents was 206. Finally, 41 subsidiary managers have completed the survey, which is a response rate of 19,9%. However, four surveys did not match both criteria, so the total number of useful surveys was 37.

As mentioned before, it was important that cooperation did not take too much time (Thomas, 2004, p. 119). A total of 18 questions were asked and all these questions were displayed at the same page, so the respondents could see the total length of the survey in one glance. The survey was structured in the following way. The first five questions were general questions about the respondent and the subsidiary. The next ten questions were about the different types of initiatives. The last three questions were about the relation between headquarters and subsidiary (see Appendix I).

3.3 VARIABLES

In this section, all variables involved in this study are explained. Details are provided about why these variables are used and how these are measured.

3.3.1 DEPENDENT VARIABLES

(20)

20

ranked from (1) never to (5) plentifully. In order to generate more variation in the data, the decision was made to use a 6-points Likert scale. This means that there was no ‘neutral’ position on the Likert scale. An advantage is that a respondent is forced to make a choice (negative or positive). The fact that a respondent needs to make a choice could also be a disadvantage, especially when the respondent is the subsidiary manager for only a few months. He or she could have no idea how frequently certain initiatives are taken by the subsidiary during the last ten years but he or she is forced to choose.

3.3.2 INDEPENDENT VARIABLES

The independent variables in this study were degree of development of the host country and size. Degree of development was measured in the following way. In the survey, respondents were asked in which country is your unit located? Then, on the basis of the IMF Advanced Economies List (2015, p.150) it was determined whether the subsidiary is located in a developed or developing country. A dummy variable was created where a developed country is labelled with 0 and a developing country with 1. The distinction between advanced and developing countries by IFM is based on three main criteria:

1. Per capita income level 2. Export diversification

3. Degree of integration into the global financial system.

(21)

21 Table 1: Subsidiaries per country

Developed countries Number of subsidiaries Belgium Czech Republic Finland France Germany Greece Italy Spain Sweden United Kingdom 7 1 2 3 3 1 1 1 1 1 Total 21

Developing countries Number of subsidiaries Azerbaijan Brazil Chile China Colombia Kazakhstan Poland Romania Russia Turkey

United Arab Emirates Venezuela 1 1 1 1 1 1 3 1 1 1 3 1 Total 16

(22)

22

most countries a part-time employee but there is no reason to assume that this employee cannot take initiatives or has no network to contribute. So, size was measured by the total number of employees in the subsidiary.

3.3.3 CONTROL VARIABLES

To avoid misspecification, a number of control variables were incorporated. As described in the introduction section, subsidiary initiatives are frequently linked with attention and autonomy. Birkinshaw et al. (2007) argue that it is important for subsidiaries to receive enough executive attention, because too little attention can result in missed opportunities. But what is enough attention? Attention is a relative concept. Subsidiary managers could perceive the same amount of headquarters’ attention in a different way. Therefore, a control variable about ‘perceived attention’ was added to the survey and measured on the basis of the following statement: my unit receives enough attention from headquarters. This statement could provide better insight into the underlying reasons why subsidiaries take initiatives. Birkinshaw et al. (2007) argue that taking initiatives is one way subsidiaries could use to gain more attention and this statement controls whether subsidiaries perceiving less headquarters’ attention take more or less initiatives.

Autonomy is frequently taken to be important in enabling a subsidiary to enhance their performance and play a larger role for their parent company than that given originally (Birkinshaw, Hood & Young, 2005). It is about the degree of freedom that the subsidiary has to pursue its own independent agenda (Raziq et al., 2014). But like attention, also autonomy is a relative concept. So, a statement about ‘perceived autonomy’ was added to the survey: my unit is able to make decisions without headquarters’ involvement. Since a certain degree of autonomy is needed to take initiatives, it can be expected that differences in perceived autonomy could lead to differences in initiative taking. This statement was added in order to control for this variation.

(23)

23

Besides attention and autonomy, a third variable was added to the survey. Birkinshaw et al. (2007) also state that in global organizations, subsidiaries that play pivotal roles in the success of the overall business have no trouble getting attention. This suggests that, because of the differences in attention, there could be a difference in the use of subsidiary initiatives between subsidiaries with a large market and small market. It can be expected that larger product/service markets receive more headquarters attention compared to smaller markets. This adds an additional perspective to the concept of size. Since market size is a relative concept, a question about the ‘perceived market size’ was added. Respondents were asked the following question: what is the average size of your market compared to those of other units in your company? The aim of this question was to control whether there is a difference in initiative taking between subsidiaries with a small and large market. This question was asked on a scale ranked from 1 to 3, where (1) small, (2) normal and (3) large.

3.4 DATA ANALYSIS

In this section, the constructs were checked for their validity and reliability. In addition, the method of analysis is explained.

3.4.1 VALIDITY AND RELIABILITY

The first step in analysing the data was the check whether all responses matched both criteria of the sampling method. All the responses were from organizations with headquarters located in the Netherlands and had over 500 employees in total, so all the responses matched both criteria. But one of the cooperating organizations has sent the survey to both its domestic and foreign subsidiaries managers, which has led to four responses from Dutch subsidiaries. Since subsidiary initiatives are only concerned with foreign subsidiaries (Schmid et al., 2014), these responses needed to be removed from the data set. Therefore, 4 of the 41 responses were removed from the data set, so the final number of respondents was 37.

(24)

24

Alpha reliability test was executed in order to determine whether the individual questions of each initiative all measure the same construct (Connelly, 2011). This test helps to find questions which are not helping to measure the construct and are not interrelated with the other questions. The Cronbach’s Alpha’s of both constructs are shown in table 2.

Table 2: Cronbach’s Alpha’s

Construct Cronbach’s Alpha Number of items

External Initiatives 0,607 6

Internal Initiatives 0,778 4

In general, a Cronbach’s Alpha of at least 0.7 is the criterion used to establish an acceptable level of reliability. However, the recommended minimum Cronbach’s Alpha for exploratory studies is 0.6 (Hassad, 2009). This means that both constructs are reliable with a Cronbach’s Alpha higher than 0.6. The Cronbach’s Alpha of external initiatives is at the minimum of 0.6. However, this is considered good enough because of the low number of respondents and the exploratory character of this study. In addition, when one of six items is removed, the Cronbach’s Alpha becomes lower rather than higher.

3.4.2 METHOD OF ANALYSIS

Now is determined that the internal reliability of both constructs is good enough, the hypothesized relationships can be tested. Two multiple regressions are conducted in order to answer hypotheses 1A, 1B and 2A. There has been chosen for this type of analysis because regression analysis can give clear insights into direct relationships between two or more variables (Thomas, 2004). The aim of this study is to identify the influence of degree of development of the host country and size of a subsidiary on internal and external initiatives, respectively. The technique of multiple regression can show how influential each independent variable is (Thomas, 2004, p. 212). For this reason, multiple regression analysis is considered a suitable type of analysis for the purpose of this study.

(25)

25

(Van Voorhis and Morgan, 2007; Thomas, 2004, p.108). According to Norusis (2008) a sample size should be 10 to 20 responses for each independent variable. Because this study already contains two independent variables and the total number of responses is only 37, the decision was made to add only one control variable to the multiple regressions. Autonomy is used as control variable, because it has a higher variance and standard deviation compared to attention and market size.

Hypothesis 2B is answered on the basis of two different regressions. To measure whether a significant negative relationship between size and external and internal initiatives develops when subsidiary size increases, two additional multiple regressions are necessary to conduct. These two regressions are almost the same as the other two regressions. The only difference is that the squared term of subsidiary size is added as independent variable in both regressions. This squared term of subsidiary size is needed because in order to detect whether this hypothesized relation exist two important criteria must be satisfied (Johnston and Menguc, 2007): (1) the linear main effect of subsidiary size must be positively associated with the number of external and internal initiatives; and (2) the squared-term of subsidiary size must be negative and significantly associated with the number of external and internal initiatives.

The results are discussed in the following section. A confidence interval of 95% is used for all analyses in this study.

4. RESULTS

In this section, the empirical results are discussed. First, some general characteristics and descriptive statistics of the sample are given, followed by the correlations of the variables. Second, the hypotheses are tested using the multiple regression analysis. Finally, some additional research is conducted and the overall conclusion of the results is provided.

4.1 GENERAL CHARACTERISTICS AND DESCRIPTIVE STATISTICS

(26)

26

countries and 16 subsidiaries located in developing countries. The average size of all subsidiaries is 133,68. The smallest subsidiary has 4 employees, while the biggest subsidiary has 700 employees. The Pearson Correlation test is used in order to find the correlations of the variables. These correlations answer whether there is a relationship between two variables and if this relationship is positive or negative. The descriptive statistics and correlations of the variables are summarized in table 3.

Table 3: Descriptive statistics and Intercorrelations Descriptive Statistics Intercorrelations

Variables Mean Std. deviation 1 2 3 4 5 6 7

1. External Initiatives 2. Internal Initiatives 3.Developed/Developing 4. Size 5. Autonomy 6. Attention

7. Perceived Market Size

3,6982 3,5068 0,43 133,68 4,89 5,65 1,89 0,80204 1,11879 0,502 175,255 1,646 1,230 0,809 0,494** -0,012 0,278 0,132 0,124 0,041 0,081 0,059 0,242 0,093 -0,061 -0,091 -0,143 -0,197 0,323 0,172 0,153 0,319 -0,113 0,132 -0,235 ** Correlation is significant at the 0.01 level (2-tailed).

As table 3 shows, the only significant correlation is between both dependent variables, namely external and internal initiatives. No significant positive or negative correlations have been found between dependent and independent variables.

4.2 REGRESSION ANALYSES

Before the results can be discussed, the following needs to be clarified. This study uses the ‘adjusted R square’ for examining the predictive power of the models. Or in other words, how much of variance can be explained by the model. In addition, the ANOVA table is used to provide information regarding the level of significance of each model as a whole.

(27)

27

host country and external initiatives. Hypothesis 2A is partially not supported. Size has no significant positive relation with external initiatives.

The second multiple regression (see Appendix IV) measures the influence of degree of development of the host country and size on internal initiatives. The data results show adjusted R square of -0,012, which means that no variance in internal initiatives can be explained by the independent variables. A negative adjusted R square can be interpreted by 0. The model as a whole is not significant (F=0,863; p>0,05). Also degree of development of the host country (t=0,707; p>0,05) and size (t=0,153; p>0,05) are not significant. There is no significant positive relationship found between degree of development of the host country and internal initiatives. So, H1B needs to be rejected as well. In addition, there is no positive significant relationship found between size and internal initiatives. This means that H2A should be rejected, since there is no significant positive relationship found between size and both types of initiatives.

Almost the same two multiple regressions are conducted in order to test H2B. The only difference is that the squared-term of size is added to both regressions as independent variable. As mentioned in the methodology section, in order to detect whether a significant negative relationship develops between subsidiary size and external and internal initiatives when subsidiary size increases, two important criteria must be satisfied: (1) the linear main effect of size must be positively associated with the number of external and internal initiatives; and (2) the squared-term of size must be negative and significantly associated with the number of external and internal initiatives. The first multiple regression measures both criteria in relation to external initiatives (see Appendix V). The adjusted R square is -0,010, so no variance can be explained by the independent variables. The model as a whole is not significant (F=0,913; p>0,05). Also size and the squared-term of size are not significant. So, both criteria are not supported with respect to external initiatives.

(28)

28

not significant (F=1,575; p>0,05). Hence, H2B should be rejected as well. An overview of the results is provided in table 4.

Table 4: Summary results for hypotheses

Hypothesis Description Result

H1A Degree of development of the host country  External initiatives Rejected H1B Degree of development of the host country  Internal initiatives Rejected H2A Size  External and Internal initiatives Rejected H2B Size  External and Internal initiatives Rejected

4.3 ADDITIONAL RESEARCH: FACTOR ANALYSIS

Because no significant relationships were found on the basis of the regression analyses, some additional research is conducted. According to Thomas (2004, p.212), factor analysis is a technique that can be used to identify the underlying structure within a set of variables. It serves to reduce a large set of variables to a smaller number of ‘factors’ by exploring the inter-correlations among them. In order to examine whether the variance in subsidiary initiatives can be explained by some factors, a factor analysis (varimax rotation) was conducted. This analysis could provide more insights in which items appear to be measuring the same construct. So, all ten items about initiatives were used in order to examine whether there are some factors which explain the variance in subsidiary initiatives.

Table 5: Factor Analysis: Total Variance Explained

Component

Initial Eigenvalues

Extraction Sums of Squared Loadings

Rotation Sums of Squared Loadings Total % of Variance Cumulative % Total % of Variance Cumulative % Total % of Variance Cumulative % 1 3,721 37,215 37,215 3,721 37,215 37,215 2,773 27,730 27,730 2 1,909 19,087 56,302 1,909 19,087 56,302 2,016 20,164 47,894 3 1,307 13,070 69,371 1,307 13,070 69,371 1,770 17,698 65,591 4 1,014 10,141 79,513 1,014 10,141 79,513 1,392 13,921 79,513 5 ,654 6,535 86,048 6 ,462 4,624 90,672 7 ,357 3,574 94,246 8 ,269 2,695 96,941 9 ,170 1,697 98,638 10 ,136 1,362 100,000

(29)

29

In this analysis, only factors with an eigenvalue of 1,00 or higher are used. Table 5 provides more information about all factors found. As table 5 shows, four factors are extracted. These four factors together explain 79,5% of the total variance. In particular, factor 1 and 2 are very strong. Together they explain 56,3% of the total variance. The decision was made to examine these two factors in more detail. Table 6 provides more information about how well all items ‘load’ on each factor.

Table 6: Factor Analysis: Component Matrixa

Component 1 2 3 4 Internal1 ,789 -,024 ,057 ,167 Internal2 ,618 -,005 -,652 -,024 Internal3 ,562 ,209 -,555 ,261 Internal4 ,835 -,043 -,049 -,023 Local1 ,740 -,314 ,090 -,405 Local2 ,276 ,798 -,169 -,255 Local3 ,732 ,016 ,429 -,411 Global1 ,663 -,312 ,414 ,340 Global2 ,324 ,571 ,347 ,589 Global3 -,044 ,838 ,238 -,238

Extraction Method: Principal Component Analysis. a. 4 components extracted.

As table 6 shows, six items are loading well together on factor 1 (marked green). With respect to factor 2, two items are loading well together (marked orange). A closer look at the survey (see Appendix I) leads to some interesting findings with respect to the content of the items, namely:

1) All six items loading on factor 1 are questions about initiatives with respect to product development and product investment.

2) Both items loading on factor 2 are questions about initiatives with respect to collaborations.

(30)

30

if large subsidiaries take more initiatives with respect to collaborations compared to small subsidiaries.

Factor 1 is the dependent variable in the first multiple regression (see Appendix VII). The adjusted R square is 0,068, so 6,8% of the variance can be explained by the independent variables. But the model as a whole is not significant (F=1,808; p>0,05). Factor 2 is the dependent variable in the second multiple regression (see Appendix VIII). The adjusted R square is 0,063, which means that 6,3% of the variance can be explained by the independent variables. Autonomy (t=1,861) has almost a significant positive effect with a p-value of 0,072. But also this model as a whole is not significant (F=1,808; p>0,05). So, the multiple regressions did not found any significant relationship with respect to both factors.

4.4 CONCLUSION OF THE RESULTS

No significant relationships were found between the independent variables and external and internal initiatives. Therefore, all four hypotheses should be rejected. Also the additional factor analysis has not led to new significant relationships. Despite the fact that two factors were found which determine 56,3% of the total variance in initiatives, no significant relationships between the independent variables and both factors were found.

5. DISCUSSION

In this section, the findings of the empirical results are discussed and compared to existing literature. First, a discussion of the findings is provided including an answer to the main question. Then, the limitations of the study are discussed followed by some directions for future research.

5.1 DISCUSSION OF THE FINDINGS

(31)

31

subsidiary initiatives is important, because Ambos and Birkinshaw (2010) show that subsidiaries undertaking entrepreneurial initiatives perform better, both financially and managerially, when compared to other units of the MNC. Studying subsidiary initiatives in this environmental context could lead to valuable insights into why some subsidiaries perform better than others. In addition to the study of Raziq et al. (2014), this study is the first that has tried to explain differences in subsidiary initiatives between subsidiaries in developed and developing countries on a global level. Despite the low number of 37 respondents, subsidiaries located in 22 countries are participating in this study.

The main argument in this study was that subsidiaries located in a more developed environment have better access to resources and therefore are better able to execute initiatives. This statement is not supported by the empirical data. There was no significant positive or negative relationship found between degree of development of the host country and subsidiary initiatives. According to Barney (1991), a firm resource must be valuable, rare and imperfectly imitable and there cannot be strategically equivalent substitutes in order to have the potential to create a sustained competitive advantage. This could be an explanation why subsidiaries located in developed countries, despite the fact that they are likely to have better and more available resources, are not taking more initiatives compared to subsidiaries located in developing countries. If these resources do not create a competitive advantage for subsidiaries in developed countries compared to subsidiaries in developing countries, it is unlikely that these resources cause differences in initiative taking. But because of the low number of respondent in this study, more research is needed in order to explain the role of resources on initiative taking.

(32)

32

generalizability of these findings and to get more insights into the relation between subsidiary size and subsidiary initiatives. Despite the low number of respondents, size and even the squared-term of size was almost significant with respect to internal initiatives. This could be an indication that a more comprehensive study with more respondents could lead to different findings compared to this study.

Now the findings of this study are discussed, it can be tried to answer the main research question. The main research question of this study was:

What is the influence of degree of development of the host country and size of a subsidiary on the use of subsidiary initiatives?

This question remains largely unanswered, due to the fact that no significant positive or negative relationships were found between degree of development of the host country and size of a subsidiary and different types of subsidiary initiatives. Because of the low number of respondents, there is no certainty about what kind of relationship exists between the mentioned concepts. This means that more research is necessary in order to provide a clear answer to this research question.

5.2 LIMITATIONS

(33)

33

be easier. Lastly, the study makes no distinction between different industries. This makes the results also hard to generalize (Thomas, 2004). Some industries are less suitable for initiatives compared to others. Some industries are more dynamic compared to others and also the frequency of initiatives could be different between industries. For instance, the main task of some subsidiaries is product development, so it seems obvious that these subsidiaries will take more initiatives with respect to product development than other subsidiaries.

5.3 DIRECTIONS FOR FURTHER RESEARCH

According to the limitations and the findings of this study, there are several directions for further research. First, this study could be conducted on a larger scale in order to provide more insights into the potential differences between subsidiaries in developed and developing countries. Raziq et al. (2014) found that internal initiatives are less successful with respect to autonomy in large developing economies compared to small and remote developed economies. This suggests that subsidiaries in large developing economies could better focus on external initiatives in order to receive more autonomy but further research is necessary in order to support this claim. When there is an overview of what initiatives are taken by different subsidiaries, future research could focus on which initiatives are most successful in certain situations. This could be valuable information for subsidiary managers, because there is little research-based practical guidance as to the strategies and tactics subsidiary’s managers can adopt to ensure their subsidiary’s survival and to compete for further investment (Scott and Gibbons, 2009).

(34)

34

6. REFERENCES

Ambos, T.C., Birkinshaw, J. (2010). Headquarters’ attention and its effect on subsidiary performance. Management International Review. 50: 449-469.

Ambos, T. C., Andersson, U., & Birkinshaw, J.M. (2010). What are the consequences of initiative-taking in multinational subsidiaries? Journal of International Business Studies. 41(7): 1099–1118.

Barney, J. (1991). Firm resources and sustained competitive advantage. Journal of Management. 17(1): 99-120.

Bartlett, C.A. and Ghoshal, S. (1986). ‘Tap your subsidiaries for global reach’. Harvard Business Review. 64(6): 87-94.

Birkinshaw, J., 1997. Entrepreneurship in multinational corporations: the characteristics of subsidiary initiatives. Strategic Management Journal. 18: 207-229.

Birkinshaw, J., 1998. Corporate entrepreneurship in network organizations: how subsidiary initiative drives internal market efficiency. European Management Journal. 16: 355-364.

Birkinshaw, J., 2000. Entrepreneurship in the Global Firm. Sage, London

Birkinshaw, J.M. and Fry, N. (1998). Subsidiary initiatives to develop new markets. MIT Sloan Management Review. 39(3): 51-61.

Birkinshaw, J. M., Hood, N., & Jonsson, S. (1998). Building firm-specific advantages in multinational corporations: The role of subsidiary initiative. Strategic Management Journal. 19(3): 221–241.

(35)

35

Birkinshaw, J.M., Hood, N., & Young, S. (2005). Subsidiary entrepreneurship, internal and external competitive forces, and subsidiary performance. International Business Review. 14(2): 227–248.

Birkinshaw, J., Bouquet, C. and Ambos, T.C. (2007). Managing Executive Attention in the Global Company. MIT Sloan Management Review. 48(4): 39-45

Birkinshaw, J. M., & Pedersen, T. (2009). Strategy and Management in MNE subsidiaries. In A. M. Rugman (Ed.), The Oxford Handbook of International Business (2nd ed.). L, Leslie waters chair of international business.

Boojihawon, D.K., Dimitratos, P., Young, S. (2007). Characteristics and influences of multinational subsidiary entrepreneurial culture: the case of the advertising sector. International Business Review. 16: 549-572.

Borini, F.M., Fleury, T.L., Fleury, A.C.C., Junior, M.M.O. (2009). The relevance of subsidiary initiatives for Brazilian multinationals. Revista de Administração de Empresas. 49(3): 253-265.

Chiao, Y.C., Yu, C.M.J., Li, P.Y., Chen, Y.C. (2008). Subsidiary size, internationalization, product diversification, and performance in an emerging market. International Marketing Review. 25(6): 612 – 633.

Connelly, L.M. (2011). Cronbach’s Alpha. Medsurg nursing. 20(1): 45.

Cyert, R.M. and March, J.G. (1963), A Behavioral Theory of the Firm, Prentice Hall, Englewood

Cliffs, NJ.

Delany, E. (2000). Strategic development of the multinational subsidiary through subsidiary initiative-taking. Long Range Planning. 33: 220-244.

(36)

36

Eisenhardt, K.M. and Schoonhoven, C.B. (1996) Resource-based View of Strategic Alliance Formation: Strategic and Social Effects in Entrepreneurial Firms. Organization Science. 7(2): 136-148.

Hamprecht, J. and Schwarzkopf, J. (2014). Subsidiary Initiatives in the Institutional Environment. Management International Review. 54: 757-778.

Hassad, R. (2009). Development and validation of a teaching practice scale (tiss) for instructors of introductory statistics at the college level. International Association for Statistical Education. 1-8.

Hedlund, G. (1981). Autonomy of Subsidiaries and Formalization of Headquarters– Subsidiary Relationships in Swedish MNCs. In L. Otterbeck (Ed.), The Management of Headquarters: Subsidiary Relationships in Multinational Corporations (pp. 25–78). Gower: Aldershot.

IMF, 2015. World Economic Outlook April 2015 [online]. Washington: IMF.

Johnston, S. and Menguc, B. (2007). Subsidiary size and the level of subsidiary autonomy in Multinational Corporations: a quadratic model investigation of Australian Subsidiaries. Journal of International Business Studies. 38(5): 787-801.

Johnson, W. H. A., & Medcof, J. W. (2002). Entrepreneurial behaviour in the MNC: an extended agency theory analysis of the parent-subsidiary relationship and subsidiary initiative. The International Journal of Entrepreneurship and Innovation Management. 2(2/3): 186–203.

Kumar, N. and Demir, R., (2013). Managerial attention and antecedents of knowledge source exploitation in MNCs. Critical perspectives on international business. 9(3): 271- 300.

Luomala, H.T., Kumar, R., Singh, J.D., Jaakkola, M. (2014). When an intercultural business negotiation fails: comparing the emotions and behavioural tendencies of individualistic and collectivistic negotiators. Group Decision and Negotiation. 24(3): 537-561.

(37)

37

Mintzberg, H. (1973), The Nature of Managerial Work, Harper & Row, New York, NY.

Norusis, M.J. (2008). PASW Statistics 18 Guide to Data Analysis. Upper Sadle River, New York.

Ocasio, W. (1997). Towards an attention-based view of the firm. Strategic Management Journal. 18: 187-206.

O’Donnell, S. W. (2000). Managing foreign subsidiaries: Agents of headquarters, or an interdependent network? Strategic Management Journal. 21(5): 525–548.

Penrose, E. T. (1995). The Theory of the Growth of the Firm (3rd ed.). Oxford: Blackwell.

Raziq, M.M., Borini, F.M. and Perry, M. (2014). Subsidiary initiatives and subsidiary autonomy: Evidence from New Zealand and Brazil. International Entrepreneurship and Management Journal. 10(3): 589-605.

Schmid, S., Dzedek, L.R. and Lehrer, M. (2014). From Rocking the Boat to Wagging the Dog: A Literature Review of Subsidiary Initiative Research and Integrative Framework. Journal of International Management. 20: 201-218.

Scott, P.S. and Gibbons, P.T. (2009). How subsidiaries are battling to survive and grow. Strategy & Leadership. 37(4): 43-47

Song, M., Im, S., Van der Bij, H. and Song, L.Z. (2011). Does strategic planning enhance or impede innovation and firm performance. Product Innovation Management. 28: 503-520.

Thomas, A.B. (2004) Research Skills for Management Studies, Routledge, London. ISBN 0-415-26899-0.

(38)

38

Verbeke, A., Tavares-Lehman, A.T. and Van Tulder, R. (2011). Entrepreneurship in the Global Firm. Emerald Group, Bingley.

(39)

39

7. APPENDICES

APPENDIX I

Survey

APPENDIX II

Mail to organizations in personal network

(40)

40 APPENDIX I

Please note that in this survey, I refer to the “company” as the whole corporation, while “unit” refers to your own specific local company or subsidiary of a multinational

company.

General questions:

1.1 What is your position in the unit? (general manager, financial manager or other management function)

1.2 In which country is your unit located?

1.3 What is the number of employees in your unit? 1.4 In which industry is your unit active?

1.5 What is the average size of your market compared to those of other units in your company? (scale 1 to 3, 1=small 2=normal 3=large).

Initiatives:

In the last 10 years, to what extent has your unit’s operations engaged in the following activities? (all questions on a Likert scale 1 to 6 (1 = never, 2=very rarely, 3=rarely, 4=occasionally, 5=frequently and 6 = very frequently).

Internal initiatives:

2.1 The creation of new business activities within the company 2.2 Transfer of production process to your country

2.3 The promotion of redistribution of existing corporate assets or resources to your unit such that they are more efficiently deployed

2.4 Attraction of new corporate investments in R&D or manufacturing in your unit Local initiatives:

2.5 Enhancements to existing products or services

2.6 Collaborations with local parties (like governments and suppliers)

2.7 Development of a new product, market or process through opportunities that are first identified in your unit’s market

Global initiatives:

2.8 Development of new products/services locally and sold internationally 2.9 Expansion of existing international responsibility

2.10 Collaborations with parties outside your market Relation headquarters – subsidiaries

To what extent do you agree with the following statements (questions on a Likert scale 1 to 7, 1= strongly disagree, 2= disagree, 3= somewhat disagree, 4= neither disagree nor agree, 5= somewhat agree, 6= agree and 7= strongly agree)

3.1 My unit receives enough attention from headquarters

3.2 My unit is able to make decisions without headquarters’ involvement

(41)

41 APPENDIX II

Geachte heer/mevrouw,

Via mijn vader Jan Bijsterbosch (Partner BDO Arnhem) heb ik uw mail adres verkregen, vandaar dat ik u deze mail stuur. Momenteel ben ik bezig met mijn afstudeer scriptie en ben ik op zoek naar grote Nederlandse bedrijven met vestigingen in het buitenland. Ik doe onderzoek naar de relatie tussen het hoofdkantoor en haar buitenlandse vestigingen. De studie wordt uitgevoerd onder toezicht van dr. H.J. Drogendijk van de faculteit Economie en Bedrijfskunde van de Rijksuniversiteit Groningen. Ik ben op zoek naar managers (maakt niet uit welke management functie) binnen die buitenlandse vestigingen die een survey willen invullen over hun relatie met het hoofdkantoor. De survey bestaat uit 18 zeer gerichte vragen (geen bedrijfsgevoelige informatie) en is binnen vijf minuten in te vullen. Uiteraard volledig anoniem. Ik heb begrepen dat uw bedrijf buitenlandse vestigingen heeft. Mijn vraag is of u mij zou willen helpen aan de mail adressen van buitenlandse managers in uw bedrijf? En eventueel zou willen vragen of zij de survey in willen vullen? Ik heb hun mail adres nodig, zodat ik ze een persoonlijke link kan sturen. Het is namelijk in het belang van het onderzoek dat een manager de survey invult en niet iemand daarbuiten. In de bijlage heb ik de opzet van de survey meegestuurd, zodat u weet welke vragen er gesteld worden. Ik hoop van u te horen en uiteraard voorzie ik u graag van meer informatie.

Bij voorbaat dank voor uw reactie en uw tijd.

Met vriendelijke groet,

(42)

42

APPENDIX III (H1A AND H2A  EXTERNAL INITIATIVES)

Model Summary Model R R Square Adjusted R Square Std. Error of the Estimate Change Statistics R Square Change F Change df1 df2 Sig. F Change 1 ,292a ,085 ,002 ,80110 ,085 1,028 3 33 ,393

a. Predictors: (Constant), Autonomy, Developed/Developing, Size

ANOVAa

Model Sum of Squares df Mean Square F Sig.

1 Regression 1,980 3 ,660 1,028 ,393b

Residual 21,178 33 ,642

Total 23,158 36

a. Dependent Variable: ExternalInitiatives

b. Predictors: (Constant), Autonomy, Developed/Developing, Size

Coefficientsa Model Unstandardized Coefficients Standardized Coefficients t Sig. B Std. Error Beta 1 (Constant) 3,303 ,455 7,260 ,000 Developed/Developing ,041 ,269 ,025 ,151 ,881 Size ,001 ,001 ,265 1,565 ,127 Autonomy ,044 ,083 ,090 ,530 ,600

(43)

43

APPENDIX IV (H1B AND H2A  INTERNAL INITIATIVES)

Model Summary Model R R Square Adjusted R Square Std. Error of the Estimate Change Statistics R Square Change F Change df1 df2 Sig. F Change 1 ,270a ,073 -,012 1,12521 ,073 ,863 3 33 ,470

ANOVAa

1 Regression 3,280 3 1,093 ,863 ,470b

Residual 41,781 33 1,266

Total 45,061 36

a. Dependent Variable: InternalInitiatives

b. Predictors: (Constant), Autonomy, Developed/Developing, Size

Coefficientsa Model Unstandardized Coefficients Standardized Coefficients t Sig. B Std. Error Beta 1 (Constant) 2,523 ,639 3,948 ,000 Developed/Developing ,267 ,378 ,120 ,707 ,484 Size ,000 ,001 ,026 ,153 ,879 Autonomy ,173 ,117 ,254 1,482 ,148

(44)

44

APPENDIX V (H2B  EXTERNAL INITIATIVES)

Model Summary Model R R Square Adjusted R Square Std. Error of the Estimate Change Statistics R Square Change F Change df1 df2 Sig. F Change 1 ,320a ,102 -,010 ,80595 ,102 ,913 4 32 ,468

a. Predictors: (Constant), Autonomy, SizeSquare, Developed/Developing, Size

ANOVAa

1 Regression 2,372 4 ,593 ,913 ,468b

Residual 20,786 32 ,650

Total 23,158 36

a. Dependent Variable: ExternalInitiatives

b. Predictors: (Constant), Autonomy, SizeSquare, Developed/Developing, Size

Coefficientsa Model Unstandardized Coefficients Standardized Coefficients t Sig. Collinearity Statistics

B Std. Error Beta Tolerance VIF

1 (Constant) 3,272 ,460 7,120 ,000

Developed/Developing ,009 ,274 ,006 ,032 ,975 ,953 1,049

Size ,003 ,002 ,661 1,230 ,228 ,097 10,312

SizeSquare -2,998E-6 ,000 -,416 -,777 ,443 ,098 10,224

Autonomy ,033 ,085 ,068 ,388 ,700 ,927 1,079

(45)

45

APPENDIX VI (H2B  INTERNAL INITIATIVES)

Model Summary Model R R Square Adjusted R Square Std. Error of the Estimate Change Statistics R Square Change F Change df1 df2 Sig. F Change 1 ,406a ,165 ,060 1,08466 ,165 1,575 4 32 ,205

a. Predictors: (Constant), Autonomy, SizeSquare, Developed/Developing, Size

ANOVAa

1 Regression 7,413 4 1,853 1,575 ,205b

Residual 37,648 32 1,176

Total 45,061 36

a. Dependent Variable: InternalInitiatives

b. Predictors: (Constant), Autonomy, SizeSquare, Developed/Developing, Size

Coefficientsa Model Unstandardized Coefficients Standardized Coefficients t Sig. Collinearity Statistics

B Std. Error Beta Tolerance VIF

1 (Constant) 2,422 ,618 3,916 ,000

Developed/Developing ,164 ,369 ,074 ,446 ,659 ,953 1,049

Size ,006 ,003 ,949 1,828 ,077 ,097 10,312

SizeSquare -9,734E-6 ,000 -,968 -1,874 ,070 ,098 10,224

Autonomy ,137 ,114 ,201 1,200 ,239 ,927 1,079

(46)

46 APPENDIX VII (F1) Model Summary Model R R Square Adjusted R Square Std. Error of the Estimate Change Statistics R Square Change F Change df1 df2 Sig. F Change 1 ,382a ,146 ,068 ,96529892 ,146 1,878 3 33 ,152

ANOVAa

1 Regression 5,251 3 1,750 1,878 ,152b

Residual 30,749 33 ,932

Total 36,000 36

a. Dependent Variable: REGR factor score 1 for analysis 1 b. Predictors: (Constant), Autonomy, Developed/Developing, Size

Coefficientsa Model Unstandardized Coefficients Standardized Coefficients t Sig. B Std. Error Beta 1 (Constant) ,706 ,548 1,288 ,207 Developed/Developing -,659 ,324 -,331 -2,030 ,050 Size ,001 ,001 ,146 ,894 ,378 Autonomy -,109 ,100 -,179 -1,089 ,284

(47)

47 APPENDIX VIII (F2) Model Summary Model R R Square Adjusted R Square Std. Error of the Estimate Change Statistics R Square Change F Change df1 df2 Sig. F Change 1 ,376a ,141 ,063 ,96792466 ,141 1,808 3 33 ,165

ANOVAa

1 Regression 5,083 3 1,694 1,808 ,165b

Residual 30,917 33 ,937

Total 36,000 36

a. Dependent Variable: REGR factor score 2 for analysis 1 b. Predictors: (Constant), Autonomy, Developed/Developing, Size

Coefficientsa Model Unstandardized Coefficients Standardized Coefficients t Sig. B Std. Error Beta 1 (Constant) -1,139 ,550 -2,073 ,046 Developed/Developing ,528 ,325 ,265 1,622 ,114 Size -1,582E-5 ,001 -,003 -,017 ,987 Autonomy ,187 ,100 ,307 1,861 ,072

The influence of degree of development of the host country and size of a subsidiary on the use of subsidiary initiatives