Subsidiary financial performance and headquarter attention
BY CHENSHENG KONG S2842475 Supervisor HAMMAD UL HAQ Co-assessor RIAN DROGENDIJK June 21st 2017University of Groningen Faculty of Economics and Business MSc International business and management
Master Thesis
Abstract
Headquarter attention could be viewed as a significant support for subsidiaries. However, since headquarter attention could not equally distribute, subsidiaries in the network might compete with each other, in order to gain more attention from their headquarter. Financial performance, served as a numeral indicator of performance, might impact the attention received from headquarter. Subsidiary initiatives, functioned as an important aspect of impression management, might play a role during the attention distribution. This study presents the influence of subsidiary financial performance on the attention received from headquarter and the moderating role of subsidiary initiatives. Through a qualitative method, this study focused on the Chinese subsidiaries whose headquarters are located in the foreign countries. The general arguments are that financial performance has a positive linear influence on attention received from headquarters, and that subsidiary initiatives strengthen this relationship. The results proved the positive linear relationship between financial performance and attention from headquarters. But the results of subsidiary initiatives are shown to be opposite, which means subsidiary initiatives are helpless for subsidiaries in the competition of attention. Hence, in the reality, subsidiary managers should be aware of the negative influence of initiatives taking and be cautious to take such an action.
Table of contents
Abstract ... 3 Table of contents ... 5 1. Introduction ... 7 2. Literature review ... 10 2.1 Theoretical background ... 102.2 Main concept and hypotheses ... 14
3. Methodology ... 21 3.1 Research design ... 21 3.2 Snowball sampling ... 22 3.3 Data collection ... 24 3.4 Data analysis ... 26 3.5 Measurements ... 29 4. Result ... 32 4.1 Factor analysis ... 32
4.2 Descriptive & Correlation ... 33
4.3 Regression analysis ... 35
5. Conclusion ... 37
5.1 Discussion and Contribution ... 37
5.2 Managerial application ... 39
5.3 Limitation and further research ... 40
5.4 Reflection ... 42
References ... 44
Appendix 1: Survey information ... 50
1. Introduction
According to the attention-based theory, corporations are initially attention-distributed systems (Simon, 1947; Ocasio, 1997). In the case of multinational enterprises, it became the fact that headquarters distribute attention to their subsidiaries. A few literature emphasized that the attention from headquarters is a significant support to their subsidiaries, not only for current performance but also for future development (Dutton, 1997; Haas & Hansen, 2001). Similarly, from the positive perspective, Bouquet and Birkinshaw (2008a: 579) defined headquarter attention as “the extent to which a parent company recognizes and gives credit to a subsidiary for its contribution to the MNE as a whole.” However, it always happens that some subsidiaries receive considerable attention from headquarters, which would result in great development and performance. Subsidiaries that receive less attention from their headquarters would perform badly and lose opportunities for the parent company (Bouquet & Birkinshaw, 2008a). Headquarters are unable to equally divide the attention to all the units for two main reasons. On the one hand, the attention of headquarters is limited (Ocasio, 1997). On the other hand, the nature of MNEs is geographically dispersed (Bouquet, 2005). This leads to the diversity of languages, cultures and institutional contexts. The transfer of corporate’s strategy is various among subsidiaries in different markets. Meanwhile, the operation in the global area is complex (Ambos & Birkinshaw, 2010). Subsidiaries might have different functions and different strategies regarding the local environments, which in turn, results in the obstacles for headquarter to give equal attention completely. For headquarters, it comes to the question that how to distribute the limited attention to sub-units. For subsidiaries, the question is how could subsidiaries gain more attention.
of subsidiary (Bouquet & Birkinshaw, 2008a).
However, both “weight” and “voice” are affected by internal elements of the subsidiary, such as subsidiary achievement and performance. Considering the difficulties to measure non-financial performance and achievement, this study will only discuss the influence of financial performance. Whether financial performance has an influence on attention received from headquarters might be explained from two dimensions. On the one hand, the higher level of financial performance naturally leads to more contribution of the unit to the whole company, since multinational enterprises are in the end profit pursuing organizations (Ocasio, 1997). Thus, subsidiaries with better financial performance could gather more attention (Ghoshal & Barlett, 1990). On the other hand, according to studies on impression management, individuals are more willing to present themselves in front of others when their performance is better (Bolino, Kacmar, Turnley & Gilstrap, 2008). For instance, in a job interview, interviewee attempts to show up and create a good image. Similarly, it might be applied to an intra-organizational setting. In order to improve the image from headquarters and thus gain more attention, subsidiary managers with great financial performance would like to take some proactive actions. Therefore, the main
research question of this study is how does subsidiary financial performance impact attention received from headquarter. Taking self-presentation into account,
managers’ pro-active action (i.e. Subsidiary initiatives) might be another important reason to explain the attention distribution (Ambos, Andersson and Birkinshaw, 2008). Thus, a second research question is whether the pro-active action of managers
has a moderating influence on the relationship between financial performance and attention.
Birkinshaw (2010) used a sample of 283 subsidiaries located in Australia, Canada, and the United Kingdom. This study adopted a different sample focusing on the foreign-owned subsidiaries in China. In addition, the discussion on how subsidiary performance affects attention received from its’ headquarter has practical meanings. It helps subsidiary managers to understand the influence of performance on attention and further develop a detailed understanding of how they could gain more attention for the unit future path.
In order to answer the main research question, a quantitative research approach is used. One of the main reasons for applying quantitative method is that this study aims at finding a pattern between headquarter and subsidiary and generalizing it to a group of similar MNEs (Lincoln & Guba, 1985). Besides, the conceptual model in this research is built on combined theories: attention-based theory and impression management. The population of this study is a large amount of MNEs with a subsidiary in China. Thus, the sample in this study is a group of Chinese subsidiaries. Due to the time limitation and actual situation, I finally decided to operationalize the sample selection via snowball sampling method rather than random selection. For data collection, I adopted a survey to get necessary information since survey provides a high level of anonymity, and it is easy to operate (Polit & Beck, 2010).
2. Literature review
This chapter contains theoretical background, main concepts and hypotheses formulation. In the first section, theoretical background, I discussed MNE background and impression management, which are combined to build up the relationship between subsidiary financial performances and headquarter attention. The second section consists of main concepts and hypotheses formulation. At the end of this chapter, the conceptual model (figure 1) is presented. It shows the positive influence of subsidiary financial performance on the attention from its headquarter. The subsidiary initiative plays a moderating role in the relationship.
2.1 Theoretical background
Multinational enterprises
MNE are companies that own or control the production of goods or services in at least two countries (Pitelis & Sugden, 2000; Caves, 2007). It consists of headquarter and sub-units (Dunning & Lundan, 2008). Typically, headquarters functions as controllers in one country (home country). Other facilities are located worldwide. Some other academics emphasized the role of subsidiaries in the whole company. MNE could be viewed as “a network of semi-autonomous subsidiary in different sovereign jurisdictions” (Birkinshaw & Ridderstrale, 1999: 154).
subsidiaries has influence on the company strategy (Barlett & Ghoshal, 1986; Birkinshaw, Hood & Jonsson, 1998). For instance, subsidiaries also help their headquarters to learn local knowledge and therefore make a better decision of internalization. Generally, headquarters are considered as an important unit for administration, controlling and monitoring (Ciabuschi, Dellestrand & Holm, 2012). They control the development of the whole companies as well as the strategies (international strategy). Birkinshaw, Braunerhjelm and Terjesen (2006) figured out different types of headquarters, such as corporate headquarters, divisional or business unit headquarters, functional headquarters and regional headquarters. In this study, headquarter refers to the corporate headquarter in the home country. Birkinshaw et al. (2006) differentiated corporate headquarter by its legal domicile and formal responsibility of the whole operations.
As corporate headquarters are responsible for planning the development of the company, headquarter managers have to make decisions for the company. This refers to a multi-step process for making choices (Robbins & Judge, 2007). Generally, this process includes definition of the problems, identification of alternatives, evaluation, implementation and monitoring (Nigel, 1998). However, decision-making in the origination is biased. The process of such decision-making is based on two assumptions: first, decision makers have all the information needed; second, decision makers have the time, cognitive ability, and resources to evaluate (Simon, 1959, 1976). Thus, the problems during decision-making process mainly relate to the limitation of managers abilities, and limitation on information and knowledge. First, human beings are capability limited. Even if managers have all the information or knowledge, the capability of analyzing complex data or problems could not be ensured. Second, Denrell, Arvidsson and Zander (2004) pointed out that the biggest challenge for headquarters is how to evaluate. They argued that headquarter managers lack knowledge about local market when making decisions. This means headquarter managers do not have all information and knowledge needed during the decision making process.
Impression management
The concept of impression management came from the concept of “the presentation of self” by Goffman (1959). The initial idea of impression management is image change at an individual level. There are two types of players. One is “actor” and the other is “audience”. Standing on the different perspectives, undoubtedly, “actor” and “audience” have different images on the “actor”. In order to reach the desired image, “actor” takes specific behaviors (Bolino et al., 2008). In this way, impression management could be viewed as “efforts to create, maintain, protect or otherwise alter an image” (Bolino et al., 2008: 1080). The research about IM could be mainly divided into two dimensions, at the individual level and at the organizational level (Bolino et al., 2008).
that individuals will take IM behaviors (figure 1). On the left, people with very bad performance tend to seek excuses and forgiveness from others (Greenberg, 1996). Thus, people use IM behaviors to “hide” or more accurately reduce the damage of the bad performance. On the right, people with good performance (e.g., skilled, educated etc.) are more willing to present themselves in front of others. In such situation, the image from other is, most of the time, worse than the real level of an individual. All others saw might be only a tip of an iceberg. Thus, to get the desired image, individuals are reasonable to take actions.
Developed from the social psychology, Bolino et al., (2008) also discussed impression management at the organizational level. Organizational impression management refers to “any actions that is intentionally designed and carried out to influence an audience’s perceptions of the organization” (2008: 1095; Elsbach, Sutton & Principe, 1998). In my case, organizational impression management occurred within MNE network. Subsidiaries take extra actions to change the headquarters’ perceptions. Mohamed, Gardner and Paolillo (1999) characterized organizational impression management behaviors in to a 2x2 matrix (figure 2). Two dimensions include direct tactics versus indirect tactics and assertive tactics versus defensive tactics. The former dimension refers to techniques of managing relevant information about the units. Direct tactics relate to techniques to present information about the subsidiaries themselves including achievements and capabilities. Indirect tactics refers to techniques of managing activities that subsidiaries engaged in. These activities might show the subsidiaries’ current status, organizational culture, vision and so on. The later dimension explains two purposes of taking proactive actions in order to
FIGURE 1:
change image from headquarter. Similar to the individual level, the aim of assertive tactics is to enhance images. Organization takes defensive tactics to protect the image from being damaged by threatens (Mohamed et al., 1999).
Source: Mohamed et al. 1999
2.2 Main concept and hypotheses
Attention
The concept of attention is board and uncertain. In the field of psychology, James (1890) stated: “Everyone knows what attention is. It is the taking possession by the mind in clear and vivid form of one out of what seem several simultaneous objects or trains of thought” (James, 1890: 403). Applied this psychological concept in the business fields, attention is related to the consciousness of managers (Dutton, Walton & Abrahamson, 1989; Fiske & Taylor, 1984; Bouquet & Birkinshaw, 2008a). Simon (1947) pointed out that using consciousness of individual managers to explain firm behavior is limited. He enlarged the scope of attention from individual to organization. He stated that both attentional capacity of humans and the structural influences of organizational on individual firm behavior or decision-making in organization affect firm behavior (Ocasio, 1997; Simon, 1947). Based on Simon’s work about attention, Ocasio (1997) further explored the structuring of attention in organizations and focused on how firms distribute and regulate the attention of decision makers.
Based on work by Simon (1947), Ocasio (1997) developed attention-based view to different organizational behaviors (Palmié, Lingens, & Gassmann, 2016). He stated
FIGURE 2:
that “firms as systems of structurally distributed attention” (1997:189). As I already mentioned the definition of attention by Ocasio (1997), he claimed that decision making by top managers is dependent on the issues top managers pay attention to. The concept of attention is on the cognitive level in the field of psychology and lacks application to business management. Hence, classification or categorization of attention concept is unsystematic. Despite this, Ocasio (2011) distinguished three categories of attention-focus researches on the organizational level. But he claimed that existing literature might not completely match these categories. The categories include “attention perspective, attention engagement, and attention selection” (Ocasio, 2011: 1288). Attention perspective relates to the organizational strategy, implying organizational mission and vision in practice. Attention engagement is associated with sustained attention headquarter spent on a certain unit. Attention selection refers to headquarters’ decision of attention distribution. Since attention is limited within MNE, headquarters have to decide which sub-unit gains more attention and which one gains less (Ocasio, 2011; Haq, Drogendijk & Holm, 2017). As Haq (2017:113) discussed that attention perspective and attention engagement are linked, and attention selection is the outcome of this interaction. There is no doubt that, from the perspective of headquarters, attention perspective and attention engagement are more helpful to explain the success of organizational behaviors that how some MNE could successfully control or re-allocate resources. Given the initial goal of this study focusing on the perspective of subsidiaries, I concentrated on the attention selection, the output. Standing on the subsidiary side, attention is not viewed as an approach to control everything simultaneously. Instead, the purpose of a subsidiary is to get support (e.g. financial support) from headquarter. As mentioned by Ocasio (2011), when headquarters face a dispersed and complex situation of all subsidiaries, headquarters tend to select some certain subsidiaries. Hence, subsidiaries have to consider which factors may impact headquarters’ decision and gain more attention.
Birkinshaw (2008a) stated that it is hard to differentiate negative headquarter attention from headquarter control and monitoring.
Positive attention is more popular among studies about attention (Ambos & Birkinshaw, 2010). The idea of positive attention is introduced by Bouquet and Birkinshaw (2008a). They explained that this positive attention is a vital support to subsidiary development in the future. It is consistent with the idea that attention from headquarter is resources for subsidiaries. In their study, Bouquet and Birkinshaw (2008a) broke attention into three types: relative attention, supportive attention and visible attention. Relative attention is associated with the weight of attention a subsidiary received compared to the other units in the corporation (Dutton, 1997; Jones & Baumgartner, 2005). Supportive attention refers to the level of headquarter supporting the subsidiary development (Bouquet & Birkinshaw, 2008a; Luo, 2003). Based on the interviews of Bouquet and Birkinshaw (2008a), they found that subsidiaries would be limited in many fields if they lack of support from headquarter. Both these two types of attention are internal attention, while the last one, visible attention, is considered to be external attention from social media and stakeholders. It refers to the degree of exposure of specific subsidiary to the public (Bouquet & Birkinshaw, 2008a; Cho & Hambrick, 2006). For instance, the achievement of the subsidiary is expressed in annual report of the parent company.
Subsidiary financial performance
It seems to be understandable that subsidiary performance has a positive influence on attention received from headquarter. Subsidiary performance contains two parts, which are financial performance and non-financial performance (Ambos & Birkinshaw, 2010). The discussion in this study is mainly focusing on financial performance for two reasons. First, financial performance is a direct indicator of the operation in the local context. However, considering the differences within MNEs, I used a subjective measurement in the survey. Second, non-financial performance involves in too many fields including corporate social responsibility, managerial performance and some others. These dimensions are, most of the time, hard to measure and compare. For instance, even within the network, the managerial performance of one unit is depending on the realistic context, which means the managerial performance might be fit the local context but might not be the best within the network.
Overall, the hypothesis is formulated as follows:
H1: The better subsidiary performs in the local market, the more attention it will receive from its headquarter.
Impression management within MNE network
As I have already discussed in the theoretical background section, impression management came from individual behavior and then is applied to the organizational setting. Thus, in the MNE context, the IM behavior is divided into two dimensions: the individual level and the organizational level. At the individual level, the main reason why subsidiary managers are active to take impression management behaviors is to obtain desirable outcomes (Leary & Kowalski, 1990). Subsidiary managers tend to take IM behaviors in order to get valuable results such as cash reward, job promotion, extra training opportunities and so on. In this sense, managers are more likely to seek for a positive achievement of the subsidiary to present. However, Greenberg (1996) argued that people with poor performance might use IM tactics to seek for forgiveness. Managers in the poorly performed subsidiary might also take IM actions as a way of getting rid of punishment.
At the organizational level, according to Bolino et al (2008), subsidiary top management teams could also use IM to impact headquarter image of its subsidiary. In their paper, organizational impression management (OIM) is introduced using a matrix (Mohamed et al., 1999). Assertive versus defensive tactics are able to explain the behavior of top management team. Similar to the individual action, assertive tactics are used to improve the image while the defensive tactics are used for responding threaten or damages (Mohamed et al., 1999). Well-performing subsidiaries go with the assertive tactics, while on the contrary, poorly performed subsidiaries would attempt to use OIM to get attention from headquarter. The aims are reacting to the bad situation and changing the current position.
Subsidiary initiatives
Following the definition of Birkinshaw and Ridderstrale (1999: 151), initiative taking refers to “discrete, proactive undertaking that advances a new way for the MNC to use or expand its resources”. The idea of subsidiary initiatives is retrieved from “issue selling” (Dutton & Ashfold, 1993). In simple words, subsidiary initiatives are entrepreneurial activities spontaneously taken by subsidiary managers in the foreign market. Headquarters play the role of buyers and the subsidiaries are the sellers. Ambos et al (2008) argued that subsidiary initiatives add value to the subsidiary from a positive perspective. Some other studies have shown that the autonomous behaviors have significant effects on the role of subsidiary (Birkinshaw, Hood & Josson, 1998; Ghoshal, 1986; Rugman & Verbeke, 2001).
Such kinds of actions could be viewed as an approach to IM behavior. According to the study by Birkinshaw and Bouquet (2008a), the proactive actions of subsidiary are two aspects. The actions taken by the subsidiary managers are mandatory, which are profile building. The actions that are voluntary and to some extent are out of strategy of headquarters are considered to be initiatives taking. This paper only discussed the voluntary actions. Profile building is excluded in this study mainly because the impact of profile building is mainly based on the subsidiary achievement (refers to financial performance in this study). In another word, in this study, subsidiary initiative is considered to be a proxy of impression management.
issues are more active and therefore get more attention from headquarter. In addition, Bouquet and Birkinshaw (2008b) discussed the role of initiatives for low-power actors on changing status in network. They argued that for sub-units with limited legitimacy or those with inappropriate behaviors in the past, initiatives could help them create opportunities to regain considerable control (Bouquet & Birkinshaw, 2008b).
However, initiatives do not always result in pleasant outcomes. As Ambos et al (2010:1103) stated that subsidiary initiatives are always not accepted by headquarter. Dutton et al (1997) discussed the negative side of initiatives taking. They argued that the initiatives taking might cause the damage of current image of the subsidiary, since these types of activities are, most of the time, opposite with the initial goal of headquarter or are seen to be out of control (Dutton et al., 1997; Schulz, 2001). It highly depends on the acceptance (tolerance) of headquarter, headquarter willingness to listen and the communication skill between subsidiary and headquarter (Schulz, 2001).
Therefore, the hypothesis is presented as follows:
H2: Subsidiary initiatives strengthen the positive relationship between financial performance and attention from its headquarter.
Subsidiary financial performance
3. Methodology
In this section, the research design of this study is presented. It follows by sample section, which explains the method of snowball sampling. Next, data collection captures the process of collecting data via survey. It follows with data analysis section. Lastly, the measurements of each variable are discussed.
3.1 Research design
I used a quantitative research method in this study. Although there are few studies focusing on subsidiary and headquarter attention, it has not been generalized to a larger population among multinational companies. The main goal of this study is to find a pattern and to reach generalization of such pattern among subsidiaries. To reach this goal, a quantitative method is appropriated to adopt in this research. Lincoln and Guba (1985) viewed the most common quantitative method as the nomothetic generalization. The first step of quantitative method is to identify the population to which researchers wish to generalize (Polit & Beck, 2010). Such population should have common, clearly defined characteristics. Then, an amount of participants is selected from such population. According to the central limit theorem, if the sample is sufficiently large from a population, the mean of the sample will be approximately equal to the mean of the population (Rice, 1995). Therefore, the normal distribution pattern followed by the sample is approximately the same as the distribution of the entire population (Rice, 1995). I attempted to find a pattern between subsidiary financial performance and the attention it received from its’ headquarter and tried to extend this pattern to a large amount of MNEs. Thus, it is appropriated to adopt a quantitative method in this study.
all acquired from the prior literature (See Appendix 1).
Thirdly, in a quantitative research, the data could be analyzed fairly and statistically, which is more convincible and reliable than qualitative method (Castellan, 2010). The data are collected via survey. There are two reasons to apply survey. Firstly, comparing to interview and case study, a survey is easier and faster to apply (Polit & Beck, 2010). Secondly, since this research is sensitive regarding the performance and some pro-active actions in the subsidiaries, survey with high anonymity is suitable in this study (Polit & Beck, 2010).
3.2 Snowball sampling
I applied snowball sampling for reaching the respondents in this study. The idea of snowball sampling is from the multiplicity sampling introduced by Sirken (1970). Rothbart, Fine and Sudman (1982) expanded it to the “snowball” sampling. Voicu (2009) suggested that this method is recommended when the population is hidden or hard to reach. The general strategy of snowball sampling is that selecting the initial sample of respondents. Then, the initial respondents are requested to recommend their peers to participate in the study. The respondents on the 2nd layer are also asked to recruit other respondents that they knew. In such a way, the response rate is guaranteed (Voicu & Babonea, 2011).
willing to answer the survey (Atkinson & Flint, 2001; Faugier & Sargeant, 1997). Third, the topic of this study relates to financial performance and some pro-active activities within the subsidiaries. To some extent, those sorts of information involve company privacies. The sensitivity of such survey is actually severe. Unless based on personal connection and trust, people are more likely to refuse to answer the survey (Biernacki & Waldorf, 1981). Even though I emphasized in my cover letter that the confidentiality is ensured, many managers still refused to take part in my study. Forth, snowball sampling is an effective method of getting access to the relevant person (Faugier & Sargeant, 1997). Relevant people, in this study, refer to subsidiary manager. Unlike normal population, those managers are hard to get accessed through common way of sampling. On the one hand, it is because subsidiary managers are small and unique groups in the society. This means the possibility to reach such a manager on the street is very low. On the other hand, contact information of managers is not easy to access. Most of time, such information is viewed as personal privacy.
Source: Voicu & Babonea (2011)
Source: Voicu & Babonea (2011)
Source: Voicu & Babonea (2011)
Figure 4:
Linear version of snowball method
Figure 5:
Non-discriminative exponential version of snowball method.
Figure 6:
Discriminative exponential version of snowball method
There are three versions of snowballing. The first one (figure 4) is a linear method
that respondents on the 1st layer recommend a single referral. The second version
(figure 5) is that the initial respondents introduce potential participants as many as possible within their personal network. It is named as non-discriminative exponential method. I adopted this version of the method in the study. The third method is similar to the second version, named as discriminative exponential method (figure 6). The difference is that the researcher would select one of the referrals recommended from respondents in random (Voicu & Babonea, 2011).
The main reason why I chose the second version is that there is time limitation. The second version of the method is the fastest way to reach as many as responses in a short period. The first version is easy to end up, if one of the respondents refused to recruit acquaintance. As to the third version, it takes some time to determine which referral to go with and which referral to abandon. Besides, in order to choose a more suitable referral, the time spending on the getting information of referrals might be long.
3.3 Data collection
In order to collect data, I mailed survey (Appendix 2) to the top management teams of few selected subsidiaries. The management team includes office managers, country managers, country financial managers, and personally relevant to management. These people have an overview of their subsidiaries and know details about the connection
with headquarters. I consider them as the 1st layer. Respondents are provided three
options to participate the survey: 1) they could print the survey, respond to the questions and then return a scanned copy; 2) they could respond the survey by directly answer the questions in the PDF version; 3) they could also finish the online survey accessed by a link in the email. The confidentiality of contact information and
company information is promised. Then, I asked managers in the 1st layer to send the
survey to people who are working at a foreign-owned subsidiary among their friends,
as the 2nd layer. When I get contact with managers from the 2nd layer, I further
provided an introduction of my study and asked whether they would like to help me
forward it to their acquaintances (3rd layer). As I already mentioned in the last section,
the snowball of relevant people grew fast in this way. However, there is a possibility
that a few people in the 2nd or 3rd layers are not the top management team. I tried my
best to lower this possibility. When I requested them to introduce managers among their network, I emphasized the criteria of my target sample.
In addition, I took some special steps to increase the response rate. Firstly, I included a short introduction about “attention” and “performance” in the email to attract the interest of participants. Secondly, I also promised them to be given access to a summative report of my results and findings. Thirdly, in order to get attention from participants, I sent a reminder to non-respondents three days after the initial mailing. And a second reminder was sent to the rest of non-respondents after three days (Boyd, 2002).
that 18 companies are owned dependently by a foreign company. 8 subsidiary units were joint ventures, captured approximately 26%. The number of acquisition and merger companies is 3, while the Greenfield companies were less, only 2. On average, the subsidiaries have operated for 15 years. The number of employees ranges from 10 to 5000, with a mean of 662.
I think there are several reasons that cause such a low response rate. Firstly, because the information needed in my research is, to some extent, relates to the privacy of the company, it is understandable that managers refused to participate in my study. Secondly, Chinese managers do not have such awareness to participate in a student study. For example, Harzing (1996) has proven that the response rate in Hong Kong is much lower than the rate in other countries. She (1996) also argued that the geographical distance between sender and respondents might be a reason. Another reason that might lead to the low response rate is the language. Since the survey is in English, it costs some time to read and understand for Chinese managers. Thus, most Chinese managers lost interest to read. Thirdly, the time is limited in such a study, especially population and samples are hard to reach and time consuming, for instance, due to the busy schedules of managers. It might be possible to collect over 50 responses in half a year. Fourth, the online survey might not be an effective method to do a quantitative research. The main advantage of the survey is that it guaranteed the anonymity of respondents. The online survey, however, has a drawback as the IP address of respondents is trackable.
3.4 Data analysis
Missing data
The first step was checking whether there is missing data. The missing data is inevitable, since the participants are allowed to skip or stop the survey at anytime they want. After calculating the mean of each variable, I found that one respondent did not answer question 9 (Independent variable: Subsidiary financial performance) and three did not answer question 11 (Dependent variable: subsidiary initiatives). Considering the importance of these data and the nature of MCAR, the missing data was imputed through mean imputation method (Enders, 2010). The advantage of applying mean imputation method is that the sample size is maintained. However, the validity is lowered. I calculated the average of response scores of the item and impute the average value on the missing data (Eekhout & De Boer, 2012).
Factor analysis
In order to make sure that the questionnaire is reliable, factor analysis goes as follows. I conducted an exploratory factor analysis for independent variable (subsidiary performance), dependent variables (attention including relative, supportive and structural) and moderator (subsidiary initiatives). It aims at summarizing the data and identifying the structure of variables. For subsidiary financial performance and subsidiary initiatives, the analysis indicated whether items under such variable statistically represent the variable. As to the attention, according to the conceptual model, it was divided into three dimensions. They represent the integrated concept of headquarter attention while having different aspects for their each dimension. The analyzing processes were first examining if items under three dimensions stand for themselves. Next, I tested whether the three dimensions represent the integrated concept of attention. All the factor analyses were being done in the SPSS. I chose principal component analysis. I also contained an orthogonal rotation in the analysis. The initial idea of loading rotation is to figure out the correlation between items. Moreover, orthogonal rotation has varimax, quartimax, and equamax (IBM SPSS statistics, 2017). Since the varimax is the most widely used method, I conducted this method.
Descriptive statistics and correlation matrix
including its mean and distribution. As to the correlation matrix, correlation coefficient whether and to what extent that two variables are related. Since all the data were analyzed in the SPSS, there are two types of correlation analyses, which were widely used in the literature. Pearson correlation is often used for the variables that are expected to have a linear relationship (Economicsnetwork, 2017). In another word, in the Pearson method, one variable changed, the related variable would increase or decrease at the same time. Spearman rank-order correlation is another common method. But it is utilized for a monotonic relationship in which variables are changed together but not at a certain rate (Corder & Foreman, 2014). Another difference between these two methods is the type of variables. Spearman correlation tests the variables, which are based on rank order (Wayne, 1990). Pearson method examines nominal variables. In my study, the relationship between financial performance and attention from headquarter is assumed to be positive linear (Economicsnetwork, 2017). Besides, the variables are either scale or nominal. Thus, I chose Pearson correlation matrix to follow.
Regression model
I conducted a linear regression to model the relationship between IV, DV and moderator. The regression model helps to analyze the model as a whole. However, there is one thing should be mentioned: because the number of responses is not sufficient to operate (N=31), I delete control variables in the regression analysis. Therefore, I cannot draw a conclusion whether my hypotheses are supported or not. The results are only used for obtaining a tendency.
Before processing the regression model, I firstly standardized the independent variable (financial performance, simplified as FP) and moderator (subsidiary initiatives, simplified as SI). Then the interaction item “FPxSI” was created and calculated by “zscore_financial performance” multiplied by “zscore_subsidiary initiatives”.
3 were developed for testing the moderating influence. Thus, the information in model 3 would reveal the interaction effects between financial performance and subsidiary initiatives on attention.
3.5 Measurements
Independent variable and dependent variable are mainly based on primary data collecting via the survey (Appendix 2). Independent variable refers to subsidiary financial performance. Moderator is subsidiary initiatives. The dependent variable refers to attention. Attention is broken into three parts, which is consistent with the existing literature. I took subsidiary size, subsidiary age and institutional distance as control variables.
Independent variables
Subsidiary performance: I measured this construct using a four-item scale from Ambos and Birkinshaw (2010). However, in this thesis, financial performance is focused mainly because managerial performance is less visible and, moreover, is incomparable due to the different institutional and industrial context between subsidiaries. The measure for financial performance is based on asking respondents to indicate (‘1= much lower’ to ‘7= much higher’) how their companies performed (Question 8 in the survey) in terms of a) return on investment, b) annual profit, c) productivity, d) cash flow from operation (Ambos & Birkinshaw, 2010; Birkinshaw, Hood & Young, 2005).
Moderator
manufacturing attracted by Dutch management” (Birkinshaw & Bouquet, 2008a).
Dependent variable
I measured this construct through three types of attention developed by Bouquet and Birkinshaw (2008a) and Ocasio (1997). Bouquet and Birkinshaw broke attention into three types corresponding to the relative, supportive and visible. However, the visible attention is excluded since it is associated with external attention from the public and stakeholders. Instead, I included structural attention developed by using insights from the attention-based view of the firm (Ocasio, 1997). Relative attention used a three-item scale ranging from 1 (“much lower”) to 7 (“much higher”). Respondents were asked to evaluate the attention from its headquarter compared to the following units: a) emerging markets within the company (e.g., subsidiaries with high performance within the company), b) similar sized markets in the region, c) similar sized markets in other parts of the world (Bouquet & Birkinshaw, 2008a). I measured supportive attention (Question 10 a-c in the survey) by asking respondents to indicate the extent they agree with the following statements (“1=strongly disagree; 7=strongly agree”): a) “Corporate headquarter provides cash bonuses and career opportunities to our employees”; b) “Corporate headquarter wants to learn more about our local market and products”; c) “The headquarter helps diffuse our best practices across the company’s global network” (Bouquet & Birkinshaw, 2008a). Structural attention (Question 10 d-f in the survey) was operationalized with three items based on the work of Ocasio (1997). The items are: d) “We frequently take part in meetings with the headquarters’ managers in the company”; e) “We frequently take part in corporate strategy meetings with the headquarters’ managers”; f) “The headquarters’ managers frequently visit our subsidiary” (“1= strongly disagree; 7= strongly agree”).
Control variables
following factors as control variables: subsidiary size, subsidiary age and institutional distance.
Subsidiary size: Most of the time, large size subsidiaries have an administrative priority to get more attention from headquarter (Birkinshaw & Bouquet, 2008a). Thus, I used employee number to indicate the size of a subsidiary in general.
Subsidiary age: It should be noticed that early stage subsidiaries might get considerable attention from headquarter even though they perform poorly. As in the survey, I asked respondents to indicate the year of subsidiary establishment. The age of subsidiary was thus calculated via the equation:
Subsidiary age= current year – the year of establishment.
Formal institutional distance: Because institutional distance is the key challenges for headquarter and subsidiary relationship, I constructed this factors. I used Global competitiveness index 2016 for measuring regulative and normative distance.
4. Result
This chapter consists of three main parts, following the process of analysis. I first finished the factor analysis to ensure the validity of the items. Next, I presented descriptive statistics including control variables, independent variable and moderator. However, considering the final number of responses is less than 50, I decided not to include the control variables in the regression model.
4.1 Factor analysis
As I mentioned in the methodology, I first tested the independent variable and moderator. Presenting in table 1, these two variables are undoubtedly fit in this study. Under the financial performance, the reliability had a Cronbach’s alpha of 0.895, which is higher than 0.7 recommended by Nunnally (1978). Thus, all four items (ROI, annual profit, productivity and cash flow) are considered to be satisfactory to present financial performance. As to subsidiary initiatives, the construct of five items (new product, new bids, new international business, enhancement of product lines and in R&D investment) is proven to be satisfied (Cronbach’s alpha: 0.908>0.7)
Table 1： Factor analysis Cronbach’s Alpha Performance 0.895 Subsidiary Initiatives 0.908 Relative Attention 0.898 Supportive Attention 0.902 Structural Attention 0.920 Table 2:
Factor analysis: Attention Integrated
Attention Component 1 Component 2
As to the dependent variable, I took two steps. Firstly, I tested whether items under each dimension are suitable. Table 1 shows the results. All items under relative attention (Cronbach’s alpha: 0.898), supportive attention (0.902), and structural attention (0.920) are considered to be satisfactory, since they are all above 0.7. In the second step, I examined whether all three dimensions with their items are appeared to be one factor (attention). Table 2 indicates that there are at least two factors under attention. Items under relative attention are extracted as one factor (component 2). Two items (headquarter learning willingness & diffusion of best practice) from supportive attention and all three items from structural attention are showed to be one factor (component 1). There might be a possibility that these two variables are to some extent related to each other. In this situation, I step back to the survey design. The questions asked about the willingness of parent company to learn and to diffuse good practice in the subsidiary (Supportive attention) might relate to the questions asked about real actions under structural attention. The logic here is the more willingly headquarter would like to learn and diffuse the “best” practice from a subsidiary, the more frequently headquarter would like to connect or communicate with the subsidiary. Thus, in the following text, the structural attention will be combined into supportive attention and overall named as “supportive attention”. In addition, it is very interesting to notice that one supportive item (cash bonuses and career opportunities) is shown to neither completely be factor 1 nor 2. It is a mixed item (0.588 with component 1; 0.611 with component 2). Thus, I dropped this item when consolidating the overall attention. The overall attention was calculated as the average of relative attention and supportive attention.
4.2 Descriptive & Correlation
Another significant correlation is between financial performance and subsidiary initiatives. It shows that an increase in subsidiary financial performance is significantly related to an increase in subsidiary initiatives (r=0.4758, p<0.01). Besides, Consistent with my expectation, attention is relatively high correlated to both financial performance (r=0.812) and subsidiary initiatives (r=0.561). In addition, it might be interesting to look at the negative correlation (r=-0.213) between attention and interaction (FPxSI).
4.3 Regression analysis
Table 4 and table 5 reported the details information of regression model. Table 4
showed the results when attention is all combined. The 1st model presented the results
of financial performance only, and model 2 and model 3 added subsidiary initiatives as a moderator. Table 5 revealed the results of attention under each dimension (relative attention and supportive attention). Model 4-6 showed the information about the independent variable, moderator and relative attention as the dependent variable. Model 7-9 gave information when supportive attention served as the dependent variable.
(model 6) and supportive attention (model 9), the results showed that there is significant influence of subsidiary initiatives on the relationship between financial performance and attention. This negative interaction on relative attention was significant in model 6 (B=-0.536, p<0.05). In model 9, the interaction of financial performance and subsidiary initiatives is negative on supportive attention. The result
was significant at the level of 0.1 (B=0.022).
Table 4:
Regression model: Attention Integrated Attention
Model 1 Model 2 Model 3
(Constant) 3.905 3.905 4.033
Financial performance (FP) 1.158*** 1.004*** 0.871***
Subsidiary Initiatives (SI) 0.323* 0.481**
FP*SI -0.279 R2 0.660 0.700 0.678 R2 adjusted 0.648 0.728 0.698 a p<0.1; **p<0.05; ***p<0.01 b N=31 Table 5:
Regression model: Attention Separate
Relative Attention Supportive Attention
Model 4 Model 5 Model 6 Model 7 Model 8 Model 9
(Constant) 3.855 3.855 4.101 3.955 3.968 3.965
Financial performance (FP) 1.135*** 1.022 0.766*** 1.181*** 0.987*** 0.975***
Subsidiary Initiatives (SI) 0.237 -0.539* 0.410* 0.422*
FP*SI -0.536** -0.022*
R2 0.470 0.486 0.564 0.563 0.615 0.615
R2 adjusted 0.452 0.449 0.515 0.548 0.588 0.572
a p<0.1; **p<0.05; ***p<0.01 b N=31
5. Conclusion
This chapter contains four parts. The first section is contribution and discussion. And then managerial advices about how to deal with headquarters and how to get more attention are discussed. It follows limitations and further study. The last section is reflection. It reveals a failure of data collection in my early step of doing this study.
5.1 Discussion and Contribution
The tendency between financial performance and attention is positive linear. I already discussed two explanations that why financial performance plays an important role in the attention distribution. First, within MNE, financial profit is the main contribution of a subsidiary to the network. Initially, MNE is a profit-maximizing organization with facilities in the foreign countries. Although the role of subsidiaries is proven to be more valuable than “marker access”, the primary mission of subsidiaries is to sell products or services in the local markets. Thus, it is reasonable to say that subsidiaries with better financial performance contribute more to the whole company, and thus, they deserve more attention. Second, financial performance is a reliable statistical indicator about the subsidiaries’ performance and the local markets. Information asymmetry exists between headquarters and subsidiaries. Therefore, to some extent, the reports or information provided by subsidiary managers are less reliable. Hence, the headquarters tend to make attention decision depending on financial performance.
Second, it is interesting to notice that the moderating influence of subsidiary initiative on the linear relationship between financial performance and attention is negative. This means subsidiary initiatives weaken the relationship between financial performance and attention. The results are not consistent with the hypotheses. I expected that the subsidiary initiatives might strengthen the relationship between financial performance and attention.
measured appropriately. This study adopted the measurements of Birkinshaw et al. (1998). They used five-item scales in the survey, which are mainly linked to existing products or services of the company. They argued that if the initiatives were not aligned with the organizational strategies, initiatives activities would be considered as self-behavior of subsidiaries. Thus, the initiatives could hardly be accepted by headquarters (1998: 236). Hence, they viewed subsidiary initiative activities as extra opportunities for existing products or services. This, however, excluded the entrepreneur activities, which result in entirely new products or services. Besides, the subsidiary initiatives are not only outcomes but also the process of starting such an action and reporting it. Bishop, Simsarian and O’Neill (2011) discussed that issue-selling moves are introduced by Dutton et al (2001), and they included different patterns: packaging, involvement, and process. Packaging moves are related to the plan and proposal of an initiative action. Involvement moves include gathering information from different relevant groups. Process means the formality to plan and preparation of an entrepreneur action. Hence, an entire action of initiatives should include all the moves rather than only considering the outcomes. In addition, this study only conducted initiatives based on the local markets and then exploited to other units. However, subsidiary initiatives could also involve in the global market. Birkinshaw (1997) distinguished two types of subsidiary initiatives: local market initiatives and global market initiatives. The former activities are based on the local products or market needs and subsequently expanded in the whole network globally (1997: 212). This perspective is similar to the items utilized in this study. The later activities are associated with suppliers and customers around the world. Unfortunately, Birkinshaw (1997) stated that the process of global initiatives is unclear, and there is no literature evidence of the global market initiatives. It is still worthy to take global market initiatives into account.
entrepreneur activities reported by subsidiaries. They argued that there is a so-called “corporate immune system” in the company. They (1990: 150) stated that headquarter would prefer to reject an initiative with potential value (type I error) rather than take an innovative action with potential damage (type II error). Lastly, the sample of Chinese subsidiaries might be another aspect for such results. Haq et al. (2017) indicated that emerging markets (such as China) lack attention from headquarters, because headquarters lack knowledge and interest about local context. Subsidiaries in emerging markets are facing communication obstacles with its headquarter due to the large linguistic distance and cultural distance. Haq et al (2017) stated that large cultural distance between headquarters and subsidiaries causes the unfamiliarity with each other. On the headquarter perspective, headquarter managers do not have full information or knowledge about the local market. On the perspective of subsidiaries in emerging market, they lack strategic knowledge of the organizational plan (Haq, 2017: 172). Thus, the unfamiliarity increases the cost of communication and enhances the risk of misunderstanding. Subsidiaries in emerging market might not package and share their initiatives activities to their headquarters in the appropriated expression. Consequently, the headquarter might not completely understand and then lose interest, and subsidiaries in such situation cannot get considerable attention.
5.2 Managerial application
subsidiary that does not share its initiative activities might receive more attention from its headquarter. Third, subsidiary managers should be cautious to take such an entrepreneur action. To large extent, the influence of initiatives is unclear and it is depending on the acceptance of headquarter. Most of the time, headquarters tend to resist those activities from subsidiaries, in order to avoid potential damage for the companies (Haq, 2017). Besides, headquarters are willing to see a balanced power distributed within the network (Haq, 2017). Initiatives might bring subsidiaries more power in the network, which in turn might lead to the unbalanced situation that a certain unit becomes so powerful that headquarters can hardly control it. Managers are strongly advised to make sure to what degree that the headquarters accept such entrepreneur activities. Besides, they should take an early step of communication before actual reporting initiatives. For instance, they might submit an antecedent report about local knowledge and analysis outweighing potential advantages and limitations about such an activities. If possible, face-to-face meetings with headquarter managers about the initiatives might help headquarter managers to understand the idea and the local context, which in turn, might increase attention from headquarter.
5.3 Limitation and further research
There are several limitations that should be acknowledged. First, the main limitation of this study is that there are no actual findings in this study. The number of the sample (n=31) is not sufficient. Theoretically, the least number of doing a survey is over 50 responses and the suitable size of responses to conclude is over 100 (MacCallum, Widaman, Zhang & Hong, 1999). The insufficient response number leads to the problem that the control variables could not be included in the regression model. The lack of control variables in the regression model reduces the reliability of the results. Thus, I could not strongly conclude that the results of regression model support or reject hypotheses (1&2).
of the sample in this study. On the one hand, the target population is too large to identify the actual number. On the other hand, the responses are all based on the respondents’ referrals, which could be shortly ended or continuously developed. Secondly, there is limited control over this method. The interference only occurs at the beginning of the data collection (Voicu & Babonea, 2011), which means I can only control over the initial respondents. Since the referrals are not randomly selected, the control over the sample is weak or even void. Thirdly, there is a possibility that
the respondents on the 2nd and 3rd layers have the same view and characteristics with
the initial respondents. The referrals came from the personal network of the initial respondents. Thus, the final respondents could only represent a small group of the whole population (Voicu, 2009).
Third, quantitative method is not the most suitable method in this study. As I mentioned in the introduction, this study filled the gap that there is no research studing how financial performance impacts the subsidiary attention received from headquarter. Thus, the aim of this study is exploring the new relationship. Quantitative research is mostly used for examining the relationship using different data or practically examining hypotheses that are strongly literature based (Castellan, 2010). Besides, the nature of quantitative method and survey leads to the fact that the detailed information inside the company might be missed (Castellan, 2010). Questions in the survey are selected but they might not cover all aspects of performance and attention. Therefore, this research could only explore the relationship between performance and attention in a general idea and leave the details and story beyond for future study. In addition, since the survey keeps the anonymity, detailed information inside the company might be missed. This point might be easily overcome by a qualitative study because its main advantage is to go inside of the case and to dig out the behind story.
efforts on how to control it in order to maximize the profits around the world (Ocasio, 2011). Secondly, this study only focused on the financial performance, since the managerial performance is hard to measure. A future study could take the managerial performance into account. Ambos and Birkinshaw (2010) discussed the influence of attention on performance including managerial performance. Thus, the reverse impact of managerial performance (e.g. corporate social responsibility) could be an interesting direction to study. Thirdly, as it is not included in this study, some other factors might have a significant influence on the attention, such as cultural distance, linguistic distance and industrial competition. I suggest that such an exploratory study could be done via single or multiple case studies. In this way, researchers could go into detail of this topic and gain more information (Castellan, 2010). Fourthly, it should not be ignored the role of communication between headquarter and subsidiaries. Future studies could also pay attention to communication channel and to what extent it would impact the attention.
5.4 Reflection
Given a lack of research, the best method to conduct is multiple case studies (Yin, 2014). At least four case studies should be included: subsidiary with high financial performance and low attention, subsidiary with high financial performance and high attention, subsidiary with low financial performance and high attention, and subsidiary with low financial performance and low attention. However, this method is hard to adopt in practice for three main reasons: 1) It costs some time to find out four types of subsidiaries above; 2) Subsidiary with high financial performance and low attention is hard to access. Meanwhile, the subsidiary with low attention is another challenges, since none of the subsidiary managers are willing to claim that their subsidiaries receive less or least attention in the network. 3) Even if it is possible to find out those types of subsidiaries, spending few months in each company is necessary. The researchers have to go very deep and inside of the subsidiary and therefore get the detailed information that is needed in the study (e.g. Interview few managers in each unit).
receiving at least 50 responses among the Dutch subsidiaries with foreign headquarters. I attempted to do this research in the Netherlands at the beginning of the study. The sample of foreign-owned subsidiaries in the Netherlands was selected via ORBIS database. The sample contained 1091 companies with a foreign ultimate owner (ownership over 50.01%). After checking the parent company of each subsidiaries and trying to get personal contact through Linkedin, I sent 448 emails
with the survey as the first email from 19th April to 21st (all the emails were not sent
in a day). However, I only got 3 full responses since the most of the managers refused
to answer. In order to increase the response rate, two reminders were sent on 26th
April and 1st May. However, I only got 7 responses in total, 8 refusals and 3 returns
for promising to read (or to forward to relevant managers). The response rate did not increase even if I attempted to call in the fist week of May. Considering the time-limitation, I decided to change my strategy of data collection from regular probability method to snowball sampling.
References
Ambos, T. C. & Birkinshaw, J. 2010. Headquarters' Attention and Its Effect on Subsidiary Performance. Management International Review, 50(4): 449-469
Ambos, T.C., Andersson, U. & Birkinshaw, J. 2010. What are the consequences of initiative-taking in multinational subsidiaries? Journal of international business studies, 42: 1099-1118.
Atkinson, R., & Flint, J. 2001. Accessing hidden and hard-to-reach populations: Snowball research strategies. Social Research Update, 33.
Bartlett, C.A., Ghoshal, S., 1989. Managing Across Borders: the Transnational Solution. Harvard Business School Press: London.
Barnett, M.L. 2008. An attention-based view of real options reasoning. Academy of Management Review, 33(3): 606-628.
Baron, R. M. & Kenny, D. A. 1986. The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51: 1173-1182.
Biernacki, P., & Waldorf, D. 1981. Snowball sampling: Problems and techniques of chain referral sampling. Sociological Methods & Research, 10(2): 141-163.
Birkinshaw, J. 1997. Entrepreneurship in multinational corporations: the characteristics of subsidiary initiatives. Strategic management journal, 18(3): 207-229
Birkinshaw, J. & Hood, N. 1998. Multinational subsidiary evolution: capability and charter change in foreign-owned subsidiary companies. Academy of Management Review, 23(4): 773-795.
Birkinshaw, J., Hood, N. & Jonsson, S. 1998. Building firm-specific advantages in multinational corporations: The role of the subsidiary initiative. Strategic Management Journal, 19(3): 221-241.
Birkinshaw, J. & Ridderstråle, J. 1999. Fighting the corporate immune system: a process study of subsidiary initiatives in multinational corporations. International Business Review, 8(2): 149-180.
Birkinshaw, J. M. 2000. Entrepreneurship in the global firm. Sage Publications: London. Birkinshaw, J. & Hood, N. 2001. Unleash innovation in foreign subsidiaries. Harvard
Business Review, 79(3): 131–138.
Birkinshaw, J., Hood, N., & Young, S., 2005. Subsidiary entrepreneurship, internal and external competitive forces, and subsidiary performance. International Business Review, 14: 227–248.
Bishop, K., Simsarian Webber, S., and O'Neill, R. 2011. Preparation and prior experience in issue-selling success. Journal of Managerial Issues, 23(3): 323-340.
Bolino, M. C., & Turnley, W. H. 2003a. Counternormative impression management,
likeability, and performance ratings: The use of intimidation in an organizational setting. Journal of Organizational Behavior, 24: 237-250.
Bolino, M. C., & Turnley, W. H. 2003b. More than one way to make an impression: Exploring profiles of impression management. Journal of Management, 29: 141-160. Bolino, M. C., Varela, J. A., Bande, B., & Turnley, W. H. 2006. The impact of
impression-management tactics on supervisor ratings of organizational citizenship behavior. Journal of Organizational Behavior, 27: 281-297.
Bolino, M.C., Kacmar, K.M., Turnley, W.H. & Gilstrap, J.B. 2008. A multi-level review of impression management motives and behaviors. Journal of Management, 34 (6): 1080-1109.
Bouquet, C, & Birkinshaw, J. 2008a. Weight versus voice: How foreign subsidiaries gain attention from corporate headquarters. Academy of Management Journal, 51(1): 557-601
Bouquet, C, & Birkinshaw, J. 2008b. Managing power in the multinational corporation: how low-power actors gain influence. Journal of management, 34: 477-508
Boyd, H.H. 2002. How to Get a Respectable Response Rate. University of Wisconsin-Extension: Madison.
Björkman, I., Barner-Rasmussen, W. & Li, L. 2004. Managing knowledge transfer in MNCs: the impact of headquarters control mechanisms. Journal of International
BusinessStudies, 35, 443-455.
Castellan, C. M. 2010. Quantitative and Qualitative Research: A View for Clarity. International Journal of Education, 2: 1-14
Caves, R. E. 2007. Multinational enterprise and economic analysis. Cambridge University Press: New York.
Ciabuschi F., Dellestrand H. & Holm U. 2012. The role of headquarters in the contemporary MNC. Journal of International Management, 18, 213-223.
Cho, T. S., & Hambrick, D. C. 2006. Attention as the mediator between top management team characteristics and strategic change: The case of airline deregulation. Organization Science, 17: 453–469.
Corder, G.W. & Foreman, D.I. 2014. Nonparametric Statistics: A Step-by-Step Approach, Wiley: Manhattan.
De Jong, G, Van Dut, V., Jindra, B. & Marek, P. 2015. Does country context distance determine subsidiary decision-making autonomy? Theory and evidence from European transition economies. International Business Review, 24(5), 874-889.
Denrell, J., Arvidsson, N., & Zander, U. 2004. Managing knowledge in the dark: An
