MSc Thesis IB&M Distance in management practices: the effect on firm performance in a host country.

(1)

MSc Thesis IB&M

Distance in management practices: the effect on firm performance

in a host country.

by Djwan Ali Student number: 2295202 E-mail: d.z.ali@student.rug.nl Submission date: 18-06-2018 Supervisor: dr. M.J. Klasing Co-assessor: dr. C.H. Slager Word count: 11156

(2)

2

Abstract

This research focusses on studying the effect of cross-country distance in management practices on the performance of a foreign-owned firm in a host country. The literature review provides insights in the existing distance concepts and introduces the concept of management styles. The next section provides the causes of differing cross-country management styles and provides the formulation of a hypothesis. The methodology section describes the formation of the sample, consisting out of 207 foreign-owned manufacturing firms. Furthermore, the data of the variables is described and the results of the regression analysis are presented. Finally, the conclusion presents the discussion of the results, this research’s limitations and guidelines for future research

(3)

3

1. Introduction

Firms are known to strive for the best financial performance in order to gain the largest profit margins. However, some firms are able to gain a larger profit margin than others, while some firms even record losses from their operations. These differences in financial performance always attracted the interest of social sciences, especially in the field of international business. Numerous studies have devoted time to research the determinants of these differences. Due to the increasing internationalization activities by firms, the topic of financial performance of multinationals in the host country and the role of distance between the home and host country became the topic of interest. Ghemawat (2001) compiled existing research on various types of distances into the so-called CAGE framework (abbr.: Cultural, Administrative, Geographic, Economic), which includes the cultural- administrative- (or institutional), geographic- and economic distances. These distances act as barriers for the internationalization of firms. He argued that the greater these distances, the greater the cost of internationalization. This is naturally an important topic for firms that wish to internationalize. Therefore, this research’s focus is on the performance of multinational firms in a host country and the role of distance.

(6)

6

management practices, with the focus on practices that are associated with efficient production and delivery of components, products and/or services. From the output of their survey, it was concluded that higher management scores are strongly correlated with higher performance. This means that firms with better management practices perform better compared to firms with worse management practices. Additionally, it was found that differences in management practices explain about 30% of cross-country productivity differences (Bloom, Sadun, van Reenen & National Bureau of Economic Research, 2016). In spite of these findings, there is still not much research to be found on the effect of management practices on firm productivity and performance. Specifically on the effect of distance in management practices on firm performance. Since management practices have a strong effect on firm performance and differ across countries (Bloom & van Reenen, 2010; Bloom et al., 2016), foreign-owned firms could be affected by this distance in terms of financial performance in their host country. This research aims, therefore, to answer the following research question:

“When considering foreign owned firms operating in a host country, what is the effect of distance in management practices between their home and host country on their performance?”

(7)

7

2. Literature review

Existing literature up until now did not elaborate the distance in management practices between the home and host country of a firm. However, recent literature did provide evidence of other types of distances that influence firm performance, which are: institutional distance, geographic distance, economic distance and cultural distance. Ghemawat (2001) puts these distances together in his paper in order to create the CAGE framework, which includes the four types of distances mentioned before. Continuing, Bloom & van Reenen (2007), who assigned scores to different dimensions of good management practices, provide the most recent operationalization of management practices. However, they only focussed on management practices as a concept and did not consider the impact on a firm with differing management practices between their home and host country.

This section, therefore, will first introduce the existing conceptualization of the known distances in the CAGE framework. The next subsection provides the conceptualization of management practices, the reason why not all firms are adopting these practices and, lastly, the concept of distance in management practices is introduced.

2.1 The CAGE framework

The CAGE framework was developed by Ghemawat (2001) to help managers identify and measure the impact of distances on firm performance in a host country. These distances include, as explained before, the cultural, administrative (or institutional), geographic and economic distances. The definitions of these distances are elaborated below.

(8)

8

hierarchy vs egalitarian commitment and mastery vs harmony. Concluding, cultural distance forces firms to adapt their way of doing business in a host country. These changes take time and can be costly, which means that cultural distance between a firm’s home and host country will negatively affect that firm’s operating performance in its host country.

Continuing with institutional distance, according to North (1994), institutions are formal or informal constraints that are built to organise social relations between people and organisations. According to Kostova & Roth (2002), national institutions consist out of regulatory, cognitive and normative pillars. The regulatory pillar involves national legal institutions that impose policies and rules on people and organisations to constrain their behaviour (Kostova & Roth, 2002). This regulatory pillar is considered as a formal constraint (Estrin, Ionascu & Meyer, 2007). The normative pillar refers to a country’s moral and ethical systems that hold the roles, habits and norms about human nature and behaviour that is embedded in people’s minds (Kostova & Roth, 2002). Lastly, the cognitive pillar describes a country’s socially shared cultural systems that hold values, beliefs and assumptions about human behaviour (Kostova & Roth, 2002). The normative and cognitive pillar are considered as informal constraints (Estrin et al., 2007).

Furthermore, according to DiMaggio & Powell (1983), institutional isomorphism is a constaining process that obliges individuals or organisations in a national environment to imitate other individuals and organisations in their institutional environment. This means that firms, which operate in the same country, will start to converge in terms of behaviour due to the same national formal and informal constraints that are imposed on them. Consequently, every country will have a unique set of behaviours that are practiced by their firms. Therefore, firms that extend their operations to a country with a dissimilar institutional profile will have to adapt their business practices in order to survive in their new environment. This adaptation is a costly endeavour, which means that institutional distance, between the home and host country of a firm, will negatively affect its performance in this host country.

(9)

9

wood, cement, etc.) or fragile products. The larger the distance becomes, the higher the transportation costs become (Ghemawat, 2001). Therefore, it is safe to conclude that the greater the geographic distance from the host country to the home country of a firm, the higher are the costs a firm will experience when doing business in that host country. These costs will hamper the performance of that particular firm.

Economic distance is related to the wealth of the consumers inside a given country. According to Ghemawat (2001), it is the most important economic factor that creates differences between countries. It has an effect on the intensity of trade and the types of trading partners a country attracts. Furthermore, firms usually prefer to do business in countries with a low economic distance between them (Ghemawat, 2001). This refers to the fact that some businesses can imitate their domestic business model in the host country with a low economic distance, which helps them to exploit their competitive advantage (Ghemawat, 2001). However, if a firm goes to a country with a high economic distance, it would be a costly investment to adapt operations to the new economic environment, which will affect the profitability of their new operations (Ghemawat, 2001). Therefore, in this sense, it is safe to conclude that economic distance negatively influences firm performance in its host country. It forces firms to adapt their business models to the new economic environment, creating the risk of losing the competitive advantage the firms enjoyed domestically.

(10)

10

economic distances reduce firm performance in the host country, it is found that geographic distance increases firm performance abroad.

2.2 Good management practices

Managers tend to adopt various management practices to meet their firm’s targets and improve firm performance in order to satisfy their shareholders. There are, however, good and bad management practices. According to Bloom & van Reenen (2007), good management practices in the manufacturing sector are related to the following management practices: operations management, performance monitoring, target setting and talent management. These management practices are elaborated on in more detail in the following subsections. The scores for the four management practices can be combined into one average management score. Bloom & van Reenen (2007) found, after analyzing their survey data, that the average management score is positively associated with labor productivity. Therefore, it can be concluded that putting an emphasis on these management practices will result in increased firm performance.

2.2.1 Operations management

According to Bloom & van Reenen (2007), good operations management practices are related to applying the most recent innovations like, for example, lean manufacturing techniques such as just-in-time production and total quality management. These innovations are known to increase firm performance and efficiency, which is beneficial in the long term. Raymond (2005), who argued that advanced manufacturing techniques directly impact firm performance with increased productivity, efficiency and quality, supports this notion. Additionally, Filho, Ganga & Gunasekaran (2016) concluded that the implementation of lean manufacturing improves the firm performance of Brazilian SMEs. These findings prove that firms which continuously innovate their operations, enjoy increased firm performance.

(11)

11

2.2.2 Performance monitoring and target setting

It is well known that performance measurement (or monitoring) is important for improving productivity and, therefore, firm performance (Mathur, Dangayach, Mittal & Sharma, 2011). Zairi (1994) identified measurement as the key factor for organisational improvement, adding that organisations, which put an emphasis on measuring performance, eventually gained a leading position inside their markets. Additionally, Kaydos (1999) defined five important arguments for a firm to measure performance.

The first argument is improved control, where Kaydos (1999) argued that feedback is essential for an organisation in the sense that issues are found and resolved. This leads eventually to increased operational performance of a firm. The second argument involves clear

responsibilities and objectives (Kaydos, 1999). This means that good performance measures

can indicate who is responsible for what particular result. This clear responsibility simplifies the process of finding the cause to a problem and resolving it. The third and most important argument describes the strategic alignment of objectives (Kaydos, 1999). This involves target setting, since performance measures are a good way to communicate firms’ goals to its employees (Mathur et al. 2011). Target setting involves the setting up of key performance indicators (KPIs) in order to measure the employees and/or departments of a firm. These KPIs involve financial and non-financial measures. Financial measures include amongst other things: operating profit margin, profit margin, return on equity, return on assets. Non-financial measures include amongst other things: market share, innovation level, export level, absence rate, employee turnover rate. Hourneaux, Carneiro-da-Cunha & Correa (2017), who researched the benefits of implementing Performance Management and Management Systems (PMMS), argued that the setting of KPIs creates an institutionalized firm structure by imposing control mechanisms on the firm that is able to adapt its indicators based on the changing organisational goals. It is argued that this would increase firm performance over time.

(12)

12

can be considered as a good management practice because it provides firms with the ability to identify specific operational issues and improve those, which is beneficial for the operational performance of a firm.

2.2.3 Talent management

Talent management is related to the recognition of good and bad performing employees and/or departments within a firm. This management practice involves rewarding and/or promoting those good performers and training and/or punishing bad performers. Additionally, Bloom & van Reenen (2007) argued that promoting poor performing employees based on their employment tenure is a bad management practice leading to poor productivity and, thus, firm performance. They argued that firms should instead reward good individual performance of an employee with a promotion decision. Maslanka-Wieczorek (2014) concluded in their study on a high performance work system, that there is a relationship found between a high performance work system and effective talent management. Effective talent management involves the training of employees, providing general and individual development programs and a complete assessment (Maslanka-Wieczorek, 2014). This means that educating employees will lead to improved task efficiency and, thus, firm productivity. Therefore, it is safe to say that good talent management involves the education of employees and promotion based on good individual performance. This management practice will lead, if done correctly, to improved firm performance and can be considered as a good management practice.

2.3 Reasons of not adopting better management practices

If these practices are so beneficial for firm performance, why do not all firms adopt these? There are several arguments related to this question. The first argument is related to the cost of adopting certain management practices. Bloom & van Reenen (2007) argued that the cost of transforming firm processes and adopting better management practices can outweigh the projected benefits. Therefore, firms sometimes do not adopt management practices, which are proven to improve productivity and performance, simply because the costs surpass the projected performance in terms of sales.

(13)

13

undermines the goal of maximizing the shareholder value. Therefore, shareholders sometimes block decisions from managers that result from the implementation of good management practices if these hamper the firm’s profits.

The third and last argument involves heterogeneity. Bloom & van Reenen (2007) argued that there is no optimal level of practices due to the differing costs and benefits per market, industry and/or firms. The authors supported this with the example that investing in talent management through assessments will not bear benefits if most of the employees lack educational knowledge or practical skills.

Furthermore, Bloom & van Reenen (2007) added that internal conflict due to adoption of good management practices will hinder this process of implementation. Bloom & van Reenen (2007) name two main reasons for this. The first reason is related to the learning effects. As the diffusion of information about new management practices throughout a firm takes time, while adapting the task routine of the employees to those practices takes even more time (Bloom & van Reenen, 2007). This means that employees first need to get acquainted with the proposed changes before they are able to adopt these in their daily task routines. A second reason is that there are costs associated with adapting to better management practices, which means that this adaptation will happen gradually instead of overnight (Bloom & van Reenen, 2007). Because gradual implementation of those practices allows firms to allocate the related costs over a larger period in time.

2.4. Distance in management practices

It was theoretically explained in section 2.1 that the greater the distance, whether it is cultural, geographic, institutional or economic, the higher the costs that firms face when doing business in another country than their home country. The same can be argued to hold for distance in management practices. Think of differences in operations management (innovation is more important in some countries than in the other), performance monitoring and target setting (different markets require different KPIs) and, finally, talent management (in some countries nepotism is allowed, while others promote employees based on their performance). Bloom, Genakos, Sadun & van Reenen (2012), who found that the average management scores differed per country, confirm this notion.

(14)

14

product market competition, labour market regulation, firm ownership and human capital (Bloom & van Reenen, 2010). These causes are elaborated below.

The first cause, product market competition, involves the intensity of national competition among firms, which varies across countries. Bloom & van Reenen (2010) argued that when the competition intensity is low, less efficient firms are able to survive in marketplace. When the intensity of national competition increases, less efficient firms will be forced to apply for bankruptcy. Another reason is that firms in a country are improving their production efficiency, which causes shifting market shares among firms in a marketplace (Bloom & van Reenen, 2010). This has implications for a country’s average management score, since Bloom & van Reenen (2010) found that a higher competition intensity is strongly related with higher national management scores. Furthermore, the constraining effect of labour market regulations on managers’ capability to hire, fire, compensate or promote employees decreases the national management scores (Bloom & van Reenen, 2010). This creates differences in national management scores because labour market regulations differ per country.

The next cause of cross-country differences in management styles is the firm’s ownership structure. Bloom & van Reenen (2010) argued that some governments provide tax subsidies for family owned firms (the UK), while other countries do not (Germany, the US). These authors argued that this causes differences in management styles because family-owned businesses are usually badly managed compared to private equity firms, since family firms typically have less debt. This means that family-owned firms are able to generate positive cash flows, even if they are incurring economic losses, through the financial aid of the family owners with cheap capital (with little or no interest rates). The same holds for public companies, which are owned by the state, because these firms are also badly managed compared to private equity firms (Bloom & van Reenen, 2010). Lastly, human capital involves the education level of employees in a given country. Bloom & van Reenen (2010) claim that higher education levels of employees are associated with better management practices. Since the education level differs across countries, it can be said that human capital is another cause of cross-country differences in management styles.

(15)

15

to the learning effect, meaning that it will hamper the multinationals’ profitability due to the costs of this adaptation. These findings lead, therefore, to the following hypothesis:

H1: Distance in management practices will have a negative effect on the firm performance of

(16)

16

3. Methodology

To start off, this research is designed to measure the effect of distance in management practices on the performance of multinational firms in the host country. In order to measure the management distance, the average management country scores in the home country of a firm are compared with the average management country scores of the host country of that particular firm. These scores are derived from the World Management Survey (WMS), which was used by Bloom et al. (2014) for measuring the management scores of the studied firms in their sample. The factors and their corresponding question topics are presented in table 1 of Appendix A. Since this survey studied management practices related to operational efficiency rather than creativity and innovation (Bloom et al. 2014), this study will focus on firms with manufacturing as their main line of business.

The remainder of this section will provide a detailed description of how the data was collected and how the dataset was constructed. After this, the sampling method and the sample itself will be presented and described. Furthermore, this section will specify the variables that are used for testing the formulated hypothesis provided in the literature section. Finally, the last subsection will provide a description of the analysis that is used to test the hypothesis of this research.

3.1 Data collection, dataset formation and sampling method

For the purpose of this research, I combined information from two existing datasets. The first dataset contains the responses from the WMS performed by Bloom et al. (2014). It provides information on the management scores of 11300 manufacturing firms in 34 countries, collected between 2004 and 2014. In order to prepare this dataset for the analysis of this research, the management scores of the firms are aggregated at the country level. This results in an average management score per country.

(17)

17

foreign ownership of ≥ 10% in the dataset. It was also decided to include firms in the sample with foreign owners originating from the countries that are present in the WMS dataset.

Moreover, I exclude state-owned firms from the analysis because state owned firms set more non-economic goals compared to private owned firms, which results in inferior financial firm performance (Goldeng, Grünfield & Benito, 2008). Additionally, firms with owners originating from more than one country were excluded to prevent that the effects of management practices from multiple counties were captured in the management score of one firm. Lastly, firms with a foreign-ownership of <10% were excluded from the sample because owners with smaller ownership shares do not have the power to exercise effective control on a firms’ operations (La Porta, Lopez-De-Silanes & Shleifer, 2002).

After this, the two prepared datasets were merged into one dataset (DATASET 1). This was done by assigning to each of the home countries, of the foreign-owned firms inside the adapted BEEPS 2005 dataset, the corresponding average management country scores from the adapted WMS dataset. The foreign-owned firms that originated from a country with no reported average management country score were excluded from the sample. This resulted in a sample with foreign-owned firms originating from 13 different home countries. The list of these countries is provided in table 1 of Appendix B. Lastly, the firms with missing reported financial statements were also excluded from the sample. These adaptations of the sample resulted in a sample size of 210 firms.

Unfortunately, the management style scores from the WMS are not reported for every host country in this sample. Specifically, the WMS dataset only includes information for four of the host countries in the BEEPS data. Therefore, the distance in management practices cannot be explicitly measured for every firm in the sample. The effect of this distance is measured by controlling for host country factors, including the effect of host country management practices, through fixed effects. A more detailed explanation is provided in subsection 3.2.2.

(18)

18

reported management scores, were assigned to the management scores from one of the four countries with the smallest institutional distance. For example, the management style score of Turkey was assigned to Bulgaria because the institutional distance between these countries is smaller compared to the institutional distances between Bulgaria and Poland, Spain or Ireland. Van Hoorn & Maseland (2016) provided this research with institutional distance data used for assigning artificial management scores to the host countries. This institutional distance was measured using the Kogut & Singh distance of factor scores, which was calculated by taking the difference between home- and host country institutional factors (also provided in their dataset). The final clusters of the host countries are provided in table 1 of Appendix C. The sample size of this sample also consists out of 210 firms.

In order to see if a sample with a higher foreign ownership cut-off will have different results from the analysis compared to a foreign ownership cut-off of 10%, a third dataset (DATASET 3) is created. The difference with the main dataset is that the cut-off of foreign ownership is set to ≥ 50%. This results in a smaller sample size compared to the main dataset, namely a sample size of 184 firms.

3.2 Variables

This section will introduce the variables, which are used for the analysis of this research. It will provide a detailed description of the dependent, independent and control variables. 3.2.1 Dependent variable: Firm Performance

The dependent variable for this study is firm performance. The choice was made to use the operating profit margin of the firms in the sample as a measure of firm performance. This was decided in order to control for the fact that some larger firms have more profits due to larger revenues and production capacities. Since the operating profit margin was not included in the dataset, it had to be calculated. This was done by firstly subtracting from the firm sales (variable q57a in the BEEPS 2005 dataset) the operating costs (variable q57c in the BEEPS 2005 data file), which gave the absolute value of the operating profits of a firm. The next step involved calculating the operating profit as a margin from the total firm sales, which was done through dividing the operating profit by the firm sales. This led to the creation of the following variable: Operating_profit_rate.

3.2.2 Independent variable: Distance in Management Practices

(19)

19

host and home country of a foreign-owned firm should be considered. However, since the average management country scores of most host countries are missing, the choice was made to replace these by the so-called host country dummies. Dummy variables are usually used to express the different subgroups within a sample. In this case, those dummy variables will represent each host country. That is, the dummy variable will take on the value of 1 if the firm is located in that particular country and take the value of 0 if the firm is located in another country. These host country dummy variables will be used to take away the effects of management practices in the host countries on the firm performance of the foreign-owned firms inside these countries.

However, this method is not accurate in predicting the effect of national management practices on the performance of foreign-owned firms because these dummies incorporate other national effects such as institutional and cultural factors. Therefore, DATASET 2 with artificial host country scores is used to compare the results of the analysis of the main dataset, as described in section 3.1. By taking the difference of the national management scores between the host and home countries, the distance in management practices can be calculated. This difference is captured by the variable: MS_distance. After this, the effect of this distance in management practices can be estimated and compared to the findings of the main dataset. 3.2.3 Control variables

Control variables are factors that must be held constant in an analysis. If not, they become a threat to the internal validity of this research (Statistics How To, 2018). According to Trochim (2006), internal validity is reached when the research avoids that other factors, other than the independent variable in the proposed cause-effect relationship of this study, has an effect on the dependent variable. In order to reduce the confounding effects of other factors, this research makes use of control variables. These control variables will filter out their effect on the dependent variable, in order to isolate the causal effect of the independent variable on the dependent variable. The chosen control variables of this study with their theoretical motivations are presented in this subsection.

Firm size

(20)

20

firms, however, are more flexible and are able to adapt quicker to environmental changes and capitalise on emerged opportunities compared to larger firms. Therefore, it is safe to conclude that firm size can have a substantial impact on firm performance. This leads to the decision to isolate this impact for the sake of this research. Therefore, the control variable Large_firm (variable s4 in BEEPS 2005 data file) is created. This variable will work as a dummy variable, meaning that it will take on a value of 0 (Small/medium enterprise) or 1 (Large enterprise). The threshold between those two groups is set at 250 employees, i.e. a large enterprise is one that has equal or more than 250 employees and a small/medium enterprise is a firm with less than 250 employees.

Proportion of skilled employees

It is often argued that the presence of skilled workers improves the productivity of firms in the manufacturing sector. This point is supported by Haskel & Martin (1993), who argued that a shortage of skilled workers reduce the annual productivity growth of a firm. They added that those shortages are related to low educational achievement levels of the available workforce. Haskel & Martin (1993) concluded that by emphasizing education and training within a firm, the productivity of that particular firm can be raised. This notion is supported by Alegre, Pla-Barber, Chiva & Villar (2012) in their study about the effect of organisational learning capability on the export performance of a firm. They argued that organisational learning improves a firm’s ability to innovate. Moreover, Alegre et al. (2012) added that innovation indirectly improves the export performance of a firm, which consecutively improves the financial performance of a firm. In contrast, Sarmiento, Beale & Knowles (2007) argued that educated employees have a significant and negative relationship with job satisfaction. Consequently, a negative impact on job satisfaction for educated employees will indirectly result in a reduced level of job performance and, in turn, to a reduction of firm performance. However, Sarmiento et al. (2007) did not find a direct significant relation between the education level of employees and job performance.

(21)

21

BEEPS 2005 data file) is created that reflects the share of skilled employees (relative to the total workforce) inside a firm.

Headquarters location

It is believed that firms with their headquarters located in the country of operations will bear higher financial returns compared to firms with their headquarters located abroad. This is related to the liability of foreignness. Zaheer (1995) defined the liability of foreignness as the cost of doing business abroad, which results in a competitive disadvantage in the host country of a firm. Zaheer (1995) added that these costs can be reduced by gaining legitimacy in the host country environment. According to Chan & Makino (2007), a firm can gain legitimacy in a host country if this firm’s actions are consistent with the accepted organisational practices in this country’s institutional environment. In order to know what organisational practices are accepted in a host country’s institutional environment, a firm has to gain local knowledge. This local knowledge is more prevalent in firms that have their headquarters located inside a host country compared to firms with their headquarters located abroad. This means that firms with local headquarters will face lower costs and, hence, higher profit margins compared to firms with foreign headquarters.

Therefore, the effect of the difference in the location of a firm’s headquarters on a firm’s performance needs to be isolated. In order to do this, the variable HQ_location (variable s9 in BEEPS 2005 data file) is used. This variable can take on three values, namely 1 = headquarters located locally, 2 = headquarters located elsewhere in the country, and 3 = headquarters located abroad. Continuing, using the HQ_location variable, the control variable HQ_abroad is created. This variable works as a dummy, meaning that it can take on the values of 0 (headquarters located locally; HQ_location = 1 or 2) or 1 (headquarters located abroad; HQ_location = 3). Firm age

(22)

22

older firms were hit harder than younger firms were. They blamed this on the liability of age, which means that older firms are less flexible to adapt to changing environments like the start of a crisis. Nevertheless, it is safe to conclude that firm age has an impact on firm performance. Therefore, it is decided to use firm age as a control variable in order to separate its effect on firm performance. For this, the variable Firm_age is calculated, by subtracting 2005 (release year of the BEEPS 2005 dataset) with the variable Start_year (variable s1a in BEEPS 2005 data file).

Institutional Distance

Institutional distance refers to the difference in institutional structures between countries (Kostova & Zaheer, 1999). As explained before, institutions consist out of three pillars, which are formal (regulatory pillar) and informal (normative and cognitive pillar; North, 1994; Kostova & Roth, 2002). Kostova & Zaheer (1999) argued that the smaller the institutional distance between a firm’s home and host country, the easier it is for a firm to gain organisational legitimacy in its host country. Kostova & Zaheer (1999) continued by explaining that it is necessary for a firm to fight for its organizational legitimacy within its host country, after entering a new country. They defined organizational legitimacy as acknowledgement of an organisation by its environment. They added that organisational legitimacy is vital for a firm in order to survive and achieve success in the new host country (Kostova & Zaheer, 1999). Therefore, it can be can be concluded that institutional distance is negatively related with the firm performance of a firm in its host country. Thus, the decision was made to isolate its effect so it will not hamper with the results of the main analysis. For this, the control variable

Institutional_distance is created, which is calculated by using the Kogut & Singh factor scores.

These factor scores are derived from the dataset provided by Van Hoorn & Maseland (2016), which was used in their research about the effect of institutions on international business. Geographic distance

(23)

23

requires a high level of coordination between the spatially distant people and firms to manage for example the transportation of the necessary components needed for assembling (Ghemawat, 2001). Therefore, it is safe to conclude that geographic distance has a negative effect on firm performance, since an increased geographic distance bears higher costs. These costs lead to smaller operating profits of a company. So, the effect of geographic distance also needs to be isolated, in order to capture the effect of distance in management practices on firm performance. For this, the control variable Geographic_distance is created, which is calculated using the so-called great circle formula (Mayer & Zignago, 2011). This formula uses the latitudes and longitudes of the most important city/region (in terms of inhabitants) in each host country and home country to calculate the distance between two countries in their dataset. This variable is called Geographic_distance in the main dataset of this research.

3.3 Analysis

This subsection will present the analysis for this research. The variables in the main dataset, DATASET 1, are analysed using a linear regression analysis. This analysis is used to test the relationship between the dependent and the independent variable of this research. However, before conducting these analyses, some preliminary checks need to be performed on the data of DATASET 1. These preliminary checks are further elaborated in the next section of this research. After conducting these checks, the following four linear models are analysed in order to answer the hypothesis:

(1) Operating_profit_margin = β0 + β1Management_mean + β2Firm_size + β3Skilled_workers

+ β4HQ_locattion + β4Firm_age + Ɛ

+ β4HQ_locattion + β4Firm_age

+ β5Institutional_distance + β6Geographic_distance + Ɛ

(24)

24

+ β5Institutional_distance + β6Geographic_distance

(25)

25

4. Data description

This chapter documents the data descriptions of the variables used in the regression analyses. Additionally, other descriptive statistics (e.g. mean, range, median, minimum, maximum, etc.) are provided and displayed in table 1 of Appendix E. Furthermore, the variables that are used for this analysis were first tested for their normality. However, the requisites for the normality of the data are not met. This means that the issue of normality has to be taken into account when evaluating the significance of the tested variables. The Pearson correlation matrix of the used variables is presented in table 1 of Appendix D. The rule of thumb says that if the correlation matrix between two values is higher than 0.7, those variables cannot be included together in a model. This is not the case with the variables used in this research. Therefore, multicollinearity is not a problem here.

4.1 Descriptive statistics

This section provides the descriptive statistics of the used variables in the regression analysis of the main dataset. Firstly, section 4.1.1 will provide the description of the data of the dependent variable in relation to its distribution. The same is done in subsection 4.1.2 for the independent variable and in subsection 4.1.3 for the control variables. Only the relevant findings from the descriptive statistics for the variables are discussed in the following subsections. A summary of the frequency distribution of the variables is provided in table 1 of Appendix F. 4.1.1 Dependent variable

Firstly, the distribution of the data points of the dependent variable

Operating_profit_rate is analysed. The boxplot provided some insights in the distribution of

(26)

26

Furthermore, figure 1 and 2 of Appendix H presents the histogram and Q-Q plot of the dependent variable. This histogram shows that the data of the dependent variable is skewed to the right, which is also inferred from the fact that the mean is larger than the median (see table of Appendix E). Lastly, the dependent variable has a mean of 0.1380 with a minimum value of -0.14 and a maximum value of 0.44 (see table 1 of Appendix E). This means that the difference between the firms with the best and the worst operating profit rate in the sample is 58 percentage points.

4.1.2 The independent variable

The boxplot of the independent variable Management_mean is provided in figure 1 of Appendix H. This boxplot shows that the independent variable does not contain outliers in its data. Therefore, no cases are deleted for this variable. The values of this variable have a range of 0.64, which extends from a minimum of 2.65 to a maximum of 3.28 (Table 1 of Appenidix E). The mean value of this variable is 3.0667. The histogram and Q-Q plot of the independent variable are displayed in figure 1 and 2 of Appendix I. The histogram shows that the data of the independent variable is skewed to the left, which is also inferred from the fact that the mean is smaller than the median (see table 1 of Appendix E).

4.1.3 Control variables

This subsection provides the descriptive statistics on the data distribution of the control variables.

Large firm

This dummy coded variable depicts whether a firm can be considered as a large firm (equal to or more than 250 employees). It takes on the value of 1 (= large firm) or 0 (= medium sized or small firm). As reported in table 1 of Appendix E, the mean value of this control variable is 0.28.

Skilled employees

(27)

27

Firm headquarters

This dummy variable captures whether the headquarters of a firm in the sample is located abroad or in the same country. This variable can take on the values of 0 (local headquarters) or 1 (headquarters abroad). As reported in table 1 of Appendix E, the mean value of this variable is 0.07. Additionally, this control variable has one missing value. Therefore, the decision was made to exclude this firm from the sample.

Firm age

The control variable Firm_age signifies the age of the observed firms in the main dataset. The mean firm age is 18.09 years (table 1 of Appendix E).

Geographic distance

The control variable Geographic_distance displays the spatial distance between the home and the host country of the observed firms in the sample. As documented in table 1 of Appendix E, the mean distance is 2884.26 km with a minimum distance of 337.94 km and a maximum distance of 10240.88 km.

Institutional distance

(28)

28

5. Results

This section describes the results of the regression analysis and the corresponding robustness checks in detail with the use of tables.

5.1 Regression results

This subsection will present the results of the multiple linear regression analysis using the data from DATASET 1. These results are presented in table 1. The four models presented in this table are used to answer the hypothesis.

Regression output (foreign ownership >10%)

Model 1 Model 2 Model 3 Model 4

Large_firm 0.008 0.005 0.005 0.005 (0.014) (0.014) (0.014) (0.014) Skilled_workers 0.000 7.437E-5 0.000 0.000 (0.000) (0.000) (0.000) (0.000) HQ_abroad -0.008 0.003 -0.008 0.010 (0.023) (0.023) (0.025) (0.025) Firm_age 0.000 -9,584E-5 0.000 0.000 (0.000) (0.000) (0.000) (0.000) Management_mean 0.008 -0.008 -0.002 0.019 (0.032) 0.035 (0.037) (0.058)

Geographic_distance - -1,622E-6 - -3.062E-6

- (0.000) - (0.000)

Institutional_distance - 0.023** - -0.001

- (0.10) - (0.019)

Constant 0,11 0.139 0.155 0.166

(0.099) (0.104) (0.120) (0.162)

Host country dummies included? (YES/NO) NO NO YES YES

N 207 207 207 207

R2 _0,008 _0.034 _0.189 _0.195

adjusted R2 _-0,017 _0.000 _0.046 _0.041

F-test 0,326 1.011 1.317 1.268

Dependent variable: Operating_profit_rate * p < 0,10, ** p < 0,05, *** p < 0,01 Standard errors in parentheses

Table 1: Regression output of DATASET 1

(29)

29

the variables are significant at conventional levels. The independent variable

Management_mean is also insignificant, so no conclusions can be made regarding this predictor

in this model. The explained variance (R2) of this model is 0.8%, which is very low, even for a model with just 207 observations. The adjusted R2 is -0.017. This implies that this model contains some predictors that do not help to predict the dependent variable.

Model 2, which adds as control variables institutional and geographical distance has one significant control variable at the 0.05 significance level, namely institutional distance (b = 0.023, p < 0.05). This means that institutional distance has a significant positive effect on operating firm performance of a foreign owned firm. This goes against the expectation that increasing institutional distance would damage the performance of foreign-owned firms in the host country. The independent variable Management_distance is, again, insignificant. Therefore, no inferences can be made regarding this variable. Model 2, however, explains more variance compared to model 1. The R2 of this model isnamely 0.034 (adjusted R2 = 0.000).

Continuing, model 3 includes the same variables as model 1 plus the host country dummy variables. No variables in this model came out as significant at any level. This means that no inferences can be made regarding the effect of these variables on the dependent variable. The R2_{of this model is 0.189, which means that 18.9% of the variance of the dependent variable}

is explained by the variables in model 3. Thus, by adding host country dummy variables to the model, the explained variance increases by 18.1 percentage points. This is a large difference, but is rather logical. This is because the host country dummy variables take away the host country effects, which increases the predictability of the dependent variable. This is also demonstrated by the adjusted R2, which is 0.046. This adjusted R2 is larger than the adjusted R2 of model 1. Therefore, it can be said that the inclusion of the host country dummy variables helps the independent variable in the model to predict the dependent variable. Additionally, the adjusted R2 of this model is the highest of all four models in this analysis. This means that this model performs better compared to the other three models in terms of predictive ability.

(30)

30

predictability of the dependent variable. This is also demonstrated by the adjusted R2_{, which is}

higher than the adjusted R2_{of model 2.}

Concluding, the predictability of the models increases by adding host country dummy variables to the models. However, the control variables Institutional_distance and

Geographical_distance reduce the effective predictability of the models when the host country

dummy variables are added. Without the host country dummy variables, those controls increase the effective predictability of the models. The effect of the control variable

Institutional_distance is significant, but it has the “wrong” sign.

5.2 Robustness checks

Regression output (foreign ownership >50%)

Large_firm 0.009 0.007 0.005 0.005 (0.015) (0.015) (0.015) (0.015) Skilled_workers 0.000 0.000 0.000 0.000 (0.000) (0.000) (0.000) (0.000) HQ_abroad -0.008 0.003 0.005 0.006 (0.023) (0.024) (0.024) (0.025) Firm_age 0.000 0.000 0.000 -0.001 (0.000) (0.000) (0.000) (0.000) Management_mean 0.025 0.009 0.019 -0.010 (0.036) (0.038) (0,039) (0.061) Geographic_distance - -1.786E-6 - -1.256E-6 - (0.000) - (0.000) Institutional_distance - 0.025** - 0.019 - (0.011) - (0.020) Constant 0.053 0,083 0.163 0.223 (0.109) (0.114) (0.122) (0.170)

N 182 182 182 182

R2 0.016 0.045 0.233 0.239

adjusted R2 -0.012 0.007 0.087 0.082

F-test 0.574 1.184 1.592 1.519

Table 2: Regression output of DATASET 3

(31)

31

compared to models 1 to 4. Models 5 to 8 use the same variables as models 1 to 4, but the foreign ownership threshold is set to >50% rather than >10%. These models are presented in table 2 on the previous page. The necessity for this is that it needs to be inspected whether the results change if only firms with the majority of shares in foreign hands are included in the analysis.

Comparing model 5 to model 1, we do not observe any noticeable changes. Again, just like in model 1, none of the variables in model 5 are significant. Comparing model 2 to model 6, we also see no noticeable changes. Again, only the variable Institutional_distance is

significant (b = 0.025, p < 0.05). Likewise, also models 7 and 8 show very similar results as

models 3 and 4 with no significant effects for any of the included variables. Regression output (foreign ownership >10%)

Large_firm 0.007 0.005 0.005 0.005 (0.014) (0.014) (0.014) (0.014) Skilled_workers 0.000 7.889E-5 0.000 0.000 (0.000) (0.000) (0.000) (0.000) HQ_abroad -0.007 0.002 0.009 0.012 (0.023) (0.023) (0.025) (0.025) Firm_age 0.000 0.000 0.000 0.000 (0.000) (0.000) (0.000) (0.000) MS_distance 0.050 0.035 0.11 0.066 (0.035) (0.044) (0.041) (0.067)

Geographic_distance - -2.507E-6* - -4.161E-6

- (0.000) - (0.000)

Institutional_distance - 0.019 - -0.010

- (0.010) - (0.019)

Constant 0.119*** 0.111*** 0.146*** 0.214***

(0.017) (0,018) (0.023) (0.028)

N 207 207 207 207

R2 0.018 0.037 0.190 0.199

adjusted R2 _-0.007 _0.003 _0.046 _0.046

F-test 0.721 1.097 1.320 1.300

(32)

32

For the next robustness check, the independent variable Management_mean is replaced by the independent variable MS_distance in DATASET 2. This resulted in the creation of models 9 to 12, which are presented in table 3 (previous page). These models are compared to models 1 to 4, to see if there are any visible differences.

(33)

33

6. Conclusion

The past decade has brought growing interest in the effect of various types of distances between a firm’s home and host country, like geographic-, economic-, institutional- and economic distances, on firm performance. (North, 1994; Kostova & Zaheer, 1999; Ghemawat, 2001; Kostova & Roth, 2002) This research tried to extend this knowledge by introducing a new type of distance: the distance in management practices. This field of research did not receive as much attention compared to the aforementioned types of distances. Bloom & van Reenen (2007), however, did operationalise the concept of management practices by assigning scores to certain behaviours of management. This research tried to use these scores, in order to answer the research question of this study:

“When considering foreign owned firms operating in a host country, what is the effect of distance in management practices between their home and host country on their performance?”

Unfortunately, this research failed to answer this research question due to the insignificant results of the regression analysis. The causes of this are further elaborated in section 6.2. Additionally, existing literature already pointed out that there are indicators of differing cross-country management styles with a presumed effect on foreign-owned firms in a host country (Bloom & van Reenen, 2007). For this reason, the limitations and guidelines for future research are provided in the next subsection.

6.1 Discussion

This research aimed to test the hypothesis:

H1: Distance in management practices will have a negative effect on the firm performance of

a foreign-owned firm within a host country.

Unfortunately, the data analyses revealed insignificant results. Because of this, the effect of the independent variable Management_mean on the performance of foreign owned firms in a host country cannot be interpreted. This means that the null hypothesis can neither be rejected nor accepted. Neither increasing the threshold of foreign ownership to 50% nor replacing

Management_mean by MS_distance as the independent variable improved the significance

level of the independent variable.

(34)

34

analysis. Because of this, the distance in management styles between the home and host country of a firm could not be explicitly calculated and compared against its operating profit rate. The robustness check in section 5.2 proved that this is necessary, since the models with the artificial distances in management styles performed better (a higher adjusted R2) than the models of the main analysis.

Moreover, the models resulting from DATASET 3 (foreign ownership >50%) had the largest adjusted R2, especially for models 7 and 8. This shows us that these models performed better in terms of predictive ability when the foreign ownership threshold is increased to 50%. This is rather logical, since larger firm shares provide the foreign shareholders with more power to exercise control on the firm’s daily operations and decision making (La Porta et al., 2002). With this increased foreign control, the distance in management practices matters more in the sense that it has more influence on firm performance compared to the case when the foreign shareholders have only a 10% stake in the firm.

6.2 Limitations and future research

One of the main limitations of this research concerns the availability of data on firms. The size of the sample that was used for this research contained only 207 observed firms. This is too little to be able to successfully predict the dependent variable. The recommended sample size, with seven predictors, would be 682 or more observations on foreign owned firms (Maxwell, 2000). Therefore, future research should put more emphasis on collecting the necessary data in order to create a sample that is large enough to bear significant results. This would also improve the external validity of the findings.

Another limitation of this study is the availability of data on management practices of the host countries, where the observed firms operate. Since this data was not available, future research should either make sure to measure the management practices of firms in other countries or collect data on firms that operate in countries where the scores on management practices are already available. This would provide the opportunity to measure the distance in management practices between home and host country of a firm and test its effect on their operating performance and would improve the internal validity of this research. This is necessary, since the robustness check, which generated model 9 until 12 by using this distance as independent variable, performed better than models 1 to 4 in terms of predictive ability.

(35)

35

measurements would bear the same results, it would improve the reliability of the study. Because then the results would be consistent, no matter when the data was measured and analysed.

Moreover, instead of measuring just the distance in average management scores, the distance for each of the four management practices: operations management, performance monitoring, target setting and talent management, could be looked at separately. Then one could see which of these practices will bear significant results and provide a corresponding conclusion to those findings.

Another direction for future research is related to the measurement of management practices. The provided management scores by Bloom et al. (2014) emphasized practices that are related to efficiency. Future research should add measurements on practices that are related to other indicators of performance (e.g. innovation, leadership styles). Muller, Välikangas & Merlyn (2005) provided those measurements in their study on innovation metrics. They provided metrics which are related to the resources, capabilities and leadership of a firm. These metrics are related to the measurement of inputs and outputs regarding a firm’s innovation activities. Examples of those measurements are: percentage of capital invested in innovation activities, percentage of revenue from new products, number of available innovation tools and methodologies, number of markets entered in past year and the average time from idea submission to commercial launch (Muller et al., 2005).

Another direction for future research relates to increasing the foreign ownership threshold. As presented and discussed in section 5.2 and 6.2, increasing the foreign ownership threshold to 50% improved the predictive ability of the analysed models. This was related to the fact that shareholders holding a majority stake in firm exert more influence on a firm’s daily operations than minority shareholders (La Porta et al., 2002). This means that the effect of distance in management practices is more prevalent with foreign majority shareholders. Therefore, future research could focus their study on firms with a majority shareholder of foreign origin to study the effect of distance in management practices on firm performance.

(36)

36

References

Barbosa, N. & Louri, H. (2005). Corporate Performance: Does Ownership Matter? A

Comparison of Foreign- and Domestic-owned Firms in Greece and Portugal. Review

of Industrial Organisation, 27(1), 73-102.

Bloom, N., Lemos, R., Sadun, R., Scur, D., Van Reenen, J., & National Bureau of Economic Research. (2014). The new empirical economics of management (NBER working paper series, no. 20102). Cambridge, Mass.: National Bureau of Economic Research.

Bloom, N., Sadun, R., Van Reenen, J. & National Bureau of Economic Research. (2016).

Management as a technology? (NBER working paper series, no. 22327). Cambridge,

Mass.: National Bureau of Economic Research.

Bloom, N. & van Reenen, J. (2007). Measuring and explaining management practices across firms and countries. The Quarterly Journal of Economic, 122(4), 1351-1408.

Bloom, N. & van Reenen, J. (2010). Why Do Management Practices Differ across Firms and Countries? Journal of Economic Perspectives, 24(1), 203-224.

Business Environment and Enterprise Performance Survey (BEEPS). (2005). BEEPS 2005 [Data file]. Retrieved from: https://ebrd-beeps.com/data/2005/.

CEPII. (2011). The GeoDist database [Data file]. Retrieved from:

http://www.cepii.fr/CEPII/en/bdd_modele/presentation.asp?id=6.

Chan, C.M. & Makino, S. (2007). Legitimacy and Multi-Level Institutional Environments: Implications for Foreign Subsidiary Ownership Structure. Journal of International

Business Studies, 38(4), 621-638.

Coad, A., Holm, J.R., Krafft, J. & Quatraro, F. (2017). Firm age and performance. Journal of

(37)

37

DiMaggio, P.J. & Powell, W.W. (1983). The Iron Cage Revisited: Institutional Isomorphism and Collective Rationality in Organizational Fields. American Sociological Review,

48(2), 147-160

Estrin, S., Ionascu, D. & Meyer, K. (2007). Formal and Informal Institutional Distance, and International Entry Strategies. William Davidson Institute Working Paper No. 728.

Evans, J. & Mavondo, F.T. (2002). Psychic Distance and Organizational Performance: An Empirical Examination of International Retailing Operations. Journal of International

Filho, M.G., Ganga, G.M.D. & Gunasekaran, A. (2016). Lean manufacturing in Brazilian small and medium enterprises: implementation and effect on performance.

International Journal of Production Research, 54(21), 7523-7545.

Ghemawat, P. (2001). Distance still matters. Harvard Business Review, 79(8), 137-147.

Grazzi, M. & Moschella, D. (2017). Small, young, and exporters: New evidence on the determinants of firm growth. Journal of Evolutionary Economics, 28(1), 125-152.

Goldeng, E., Grünfield, L.A. & Benito, G.R.G. (2008). The Performance Differential between

Private and State Owned Enterprises: The Roles of Ownership, Management and

Market Structure. Journal of Management Studies 45(7), 1244-1273.

Haskel, J. & Martin, C. (1993). Do skill shortages reduce productivity? Theory and Evidence from the United Kingdom. The Economic Journal, 103(417), 386-394.

Hofstede, G. (1980). Culture’s consequences: International differences in work-related

values. Beverly Hills, CA: Sage Publications.

Hofstede, G. (2011). Dimensionalizing Cultures: The Hofstede Model in Context. Online

(38)

38

Hourneaux Junior, F., Carneiro-da-Cunha, J.A. & Correa, H.L. (2017). Performance

measurement and management systems: Different usages in Brazilian manufacturing companies. Managerial Auditing Journal, 32(2), 167-195.

Kaydos, W.J. (1999), Operational Performance Measurement: Increasing Total Productivity, Boca Raton, FL: CRC Press.

Kostova, T. & Roth, K. (2002). Adoption of an Organizational Practice by Subsidiaries of Multinational Corporations: Institutional and Relational Effects. The Academy of

Management Journal, 45(1), 215-233.

Kostova, T. & Zaheer, S. (1999). Organizational legitimacy under conditions of complexity: The case of the multinational enterprise. Academy of Management Review, 24(1),

64-81.

La Porta, R., Lopez-De-Silanes, F. & Shleifer, A. (2002). Corporate Ownership Around the World. The Journal of finance, 54(2), 471-517.

Mathur, A., Dangayach, G.S., Mittal, M.L. & Sharma, K. (2011). Performance measurement -in automated manufactur-ing. Measur-ing Bus-iness Excellence, 15(1), 77-91.

Maslanka-Wieczorek, B. (2014). Talent management and high performance work system.

Journal of international studies, 7(1), 102-108.

Maxwell, S.E. (2000). Sample Size and Multiple Regression Analysis. Psychological Methods,

5(4), 434-458.

(39)

39

Muller, A. Välikangas, L. & Merlyn, P. (2005). Metrics for innovation: guidelines for developing a customized suite of innovation metrics. Strategy & Leadership, 33(1), 37-45.

North, D. (1994). Economic performance through time. American Economic Review, 84(3), 359–368.

Sarmiento, R., Beale, J. & Knowles, G. (2007). Determinants of performance amongst shop‐floor employees: A preliminary investigation. Management Research News,

30(12), 915-927.

Schwartz, S. H. (1999). A theory of cultural values and some implications for work. Applied

Psychology: An International Review, 48(1), 12–47.

Statistics How To. (2018). Control Variable: Simple Definition. Retrieved from: http://www.statisticshowto.com/control-variable/

Trochim, W.M.K. (2006). Internal Validity. Retrieved from: https://socialresearchmethods.net/kb/intval.php.

Tsang, E.W.K. & Yip, P.S.L. (2007). Economic distance and the survival of foreign direct investments. Academy of Management Journal, 50(5), 1156-1168.

Van Hoorn, A. & Maseland, R. (2016). How institutions matter for international business: Institutional distance effects vs international profile effects. Journal of International

Van Stel, A., Millán, A., Millán, J.M. & Román, C. (2017). The relationship between start-up motive and earnings over the course of the entrepreneur’s business tenure. Journal of

(40)

40

World Management Survey. (2015). Manufacturing: 2004-2014 combined survey data (JEEA

2014) [Data file]. Retrieved from:

http://worldmanagementsurvey.org/survey-data/download-data/download-survey-data/#.

Zairi, M. (1994). Measuring Performance for Business Results, London: Chapman & Hall.

Zaheer, S. (1995). Overcoming the Liability of Foreignness. The academy of Management

Journal, 38(2), 341-363.

Zhang, M.M. (2015). Cross-national distance and insidership within networks: Japanese MNCs’ ownership strategies in their overseas subsidiaries. Asia Pacific Business

(41)

41

Appendix A

Factors Question topics

Operations management Introducing modern techniques

Rationale for introducing modern techniques

Performance monitoring

Process documentation and continuous improvement Performance tracking

Performance review Performance dialogue Consequence management

Target setting

Types and balance of targets Interconnection of targets Time horizon of targets Target stretch

Clarity and comparability of controls

Talent management

Instilling a talent mindset/ managing talent

Building a high-performance culture through incentives and appraisals

Removing poor performers/ making room for talent Developing talent and promoting high-performers Distinctive employee value proposition

Retaining talent

Table 1: The measured factors and their corresponding question topics from the WMS survey

Source: Bloom, N., Lemos, R., Sadun, R., Scur, D., Van Reenen, J., & National Bureau of

Economic Research. (2014). The new empirical economics of management (NBER working

(42)

42

Appendix B

List of the home countries of foreign-owned firms included in the sample

1. Canada 2. China 3. France 4. Germany 5. United Kingdom 6. Greece 7. Italy 8. Japan 9. Poland 10. Spain 11. Sweden 12. Turkey 13. United States

(43)

43

Appendix C

The clusters of host countries

Turkey Spain Ireland Poland

Bulgaria Estonia - Croatia

Albania Czech Republic Hungary

Belarus Slovenia Latvia

Georgia Lithuania

Tajikistan Slovak Republic

Ukraine Uzbekistan Russia Romania Kazachstan Moldova Macedonia Armenia Kyrgyz Republic Serbia & Montenegro

(44)

44

Appendix D

Table 1: Pearson correlation matrix of the variables in DATASET 1 Pearson Correlation DATASET 1

Variables 1 2 3 4 5 6 7 8 (1) Operating_profit_rate 1 (2) Large_firm 0.027 1 (3) Skilled_workers 0.056 0.049 1 (4) HQ_abroad -0.022 0.006 -0.03 1 (5) Firm_age -0.054 0.316** -0.096 -0.031 1 (6) Management_mean 0.014 0.111 0.01 0.017 0.126* 1 (7) Geographic_distance 0.012 0.022 0.005 0.053 0.015 0.365** 1 (8) Institutional_distance 0.17** 0.041 0.146* -0.140* -0,177** 0.269** 0,390** 1