The effect of institutional distances on South-South FDI flows : a cross-national comparison

(1)

UNIVERSITEIT VAN AMSTERDAM

The effect of institutional distances on

South-South FDI flows

A cross-national comparison

Master Thesis (final version)

January 2013

Lieke van der Velden 10328564

University of Amsterdam: Faculty of Economics and Business MSc Business Studies – International Management

(2)

Lieke van der Velden 2

ABSTRACT

Although South-South FDI has been a rapidly growing phenomenon little is known about the institutional determinants of their investments flows. All established research is focused on the determinant of North-South FDI flows, yet investors from the South may influenced by different institutional factors than investors from the North. This paper aims to provide a comprehensive framework in which the relationships between the institutional dimensions and South-South FDI flows are identified and compared with the established theory on Norht-South FDI flows. Findings show that South-South FDI flows are negatively influenced by geographic distance and demographic distance. Furthermore it confirms that South-South FDI flows are influenced by different institutional factors than North-South FDI flows. However, for future research it is recommended to look at the influences of institutional distances on specific southern host countries in order to get a more comprehensive view.

(3)

CONTENT

1 INTRODUCTION ... 5

2 THEORETICAL BACKGROUND ... 9

2.1 Foreign Direct Investment ... 9

2.2 South-South FDI ... 10

2.2.1 Characteristics of South FDI ... 11

2.3 FDI perspectives ... 13 2.4 Institutional Distance... 14 2.5 Hypotheses Development ... 15 2.6.1 Economic Distance ... 15 2.6.2 Financial Distance ... 16 2.6.3 Political Distance ... 17 2.6.4 Administrative Distance ... 18 2.6.6 Demographic Distance ... 20

2.6.8 Global Connectedness Distance ... 22

2.6.9 Geographic Distance ... 22

2.7 Conceptual Model ... 23

3 DATA COLLECTION AND METHOD ... 25

3.1 Data collection... 25 3.2 Sample ... 26 3.1 Dependent Variable ... 26 3.3 Independent Variables ... 27 3.4 Method ... 28 3.4.1. Gravity model ... 28 3.4.1. Model specification... 29 4 RESULTS ... 30 4.2 Regression Assumptions ... 30

4.2.3 Assumption of independence and homoscedasticity ... 32

4.2.4 Assumption of normal distribution ... 32

4.3 Regression Results ... 33

(4)

4.3.2 South-South regression results ... 34

4.3.3. Summarized results ... 35

5 DISCUSSION ... 36

5.1 Findings South-South FDI flows... 36

5.2 Findings North-South FDI flows... 39

6 MANAGERIAL IMPLICATIONS ... 41

7 LIMITATIONS AND FURTHER RESEARCH ... 42

8 CONCLUSION ... 43

References ... 44

Appendix I ... 50

Appendix II – Countries Dataset... 51

Appendix III – Descriptive Statistics ... 52

Appendix IV – Correlation matrixes ... 0

Appendix V – VIF scores ... 0

Appendix VI – Q-Q plots FDI ... 1

Appendix VII – Regression Results North-South ... 3

(5)

1 INTRODUCTION

Over the past decades multinational enterprises (MNE‟s) are getting more integrated in the global economy by engaging in foreign direct investment (FDI). Foreign direct investment has become even more important than trade of goods and services, as in 2003 the sales of foreign affiliates were twice as large as exports (Sauvant, 2005). FDI can be defined as

“cross-border investment by a resident entity in one economy with the objective of obtaining a lasting interest in an enterprise resident in another economy” (OECD Factbook, 2013) and is a key element of economic integration. In the early years only developed countries engaged in FDI, by investing in other developed countries and by investing in developing countries. This flow is referred to as “North-South FDI” flows, developed economies investing in developing economies. Recently not only investors from developed countries engage in FDI, according to the United Nations Conference on Trade and Development (UNCTAD) “a number of developing countries have emerged as significant sources of FDI in other

developing countries and their investments are now considered a new and important source of capital and production know-how, especially for host countries in developing regions”. Statistics confirm this statement as in the last 20 years FDI outflow of developing countries has doubled (UNCTAD, 2010). Most of the investment flows from developing economies go to other developing countries, which is called “South-South FDI” (Aykut and Ratha, 2004).

A key question in FDI literature remains; what are the location determinants of FDI? Most existing literature builds on the idea that foreign investments of a company are

influenced by resources and capabilities of the firm (Dunning,1980, Anand and Delios, 2002) and the need to lower transaction cost (Buckley and Casson, 1976). However, a new

generation of researchers focuses more on country-specific contextual factors, namely institutions defined as „the humanly devised constraints that structure human interaction‟ (North,1993). They claim that institutions directly influence the strategic decisions of enterprises (Meyer et al, 2009). Most of this research focuses on the quality of the

institutional environment and its effect on FDI. Yet recent evidence showed that MNEs are not only discouraged by bad institutions in host countries when investing, but are also daunted by institutional distance between the home and host country, since they prefer to invest in countries with similar institutional environment (Bénassy-Quéré et al. 2007; Habib and Zurawicki, 2002). Looking at previous studies, only Habib and Zurawicki (2002) and Bénassy-Quéré, Coupet and Mayer (2007) have studied the impact of institutional distance on bilateral FDI. Habib and Zurawicki (2002) focused on corruption and their results show

(6)

that the absolute difference of the corruption index between investor and host country has a negative impact on bilateral FDI. The latter study used several aspects of institutional differences and found that institutional difference tends to reduce bilateral FDI. Hence institutional distance can provide an explanation for MNE behavior and its location decision (Xu and Shenkar, 2002).

As mentioned earlier, South-South FDI has been a rapidly growing phenomenon and is gaining more importance in the field of international business (Aykut and Ratha, 2004). Still, there is little academic literature that addresses the topic of South-South FDI. Emerging economies are becoming global players, yet little is known about the determinants of their investments flows. Large amount of research has been done considering factors that influence the investment decision of MNEs of developed countries (Meyers et al., 2009), yet little research concentrated on factors that influence South-South investment decision. Most research is focused on the pros and cons of South–South FDI (Aykut and Goldstein, 2006). The lack of research raises questions regarding their strategies en motivations and may give implications for investors from the North. Do companies from the south behave like

developed MNEs when they invest abroad? Is their location decision influenced by the same institutional factors that influence northern MNEs, described in earlier studies? Looking at the area of institutional determinants, recent studies on the influence of institutional distance on FDI are mostly sampled from developed economies (Bénassy-Quéré et al. 2007; Habib and Zurawicki, 2002). . Institutional factors that influence South-South FDI flows are left unexplained, even though South-South FDI flows may follow different cycles and can be influenced by different factors than traditional North-South FDI (Aykut and Ratha, 2004). The few studies that do address the discussion surrounding South-South FDI have been descriptive and qualitative in nature (largely based on case studies), and most existing quantitative studies are only regionally focused (Rajan 2008). Thus there is a lack of cross-country quantitative research that addresses the institutional factors that influence South-South FDI flows.

In order to get an understanding of the role of institutional factors on South-South FDI flow the studies of Bénassy-Quéré et al. (2007) and Habib and Zurawicki, (2002) sampled from developed countries, do provide the insight of “psychic distance”, which proclaims that firm choose to enter markets perceived to be physically closer as these markets represent lower levels of uncertainty and psychic closeness facilitates learning from host countries and emphasize the importance of institutional distance. Investors from the North mostly prefer to invest in countries with similar institutions. However it can be argued that larger institutional

(7)

distance should not always harm and could even attract investors. In line with the study of Aleksynska and Havrylchyk (2011) which claims that institutional distance plays a positive role for investors from the South. Furthermore the influence of institutional distance differs across different investors as several studies imply. For instance the study of Buckley et al (2007) shows that Chinese firms prefer countries with higher political risk when choosing an investment location and thus with higher institutional distance, while German firm base their decision on the opposite. Therefore this thesis argues that investors from the South are influenced by different institutional factors than investors from the North. More specifically, institutional distance has a different influence on North-South FDI flows than on South-South FDI flows.

This leads to the main research question of this thesis:

What is the influence of institutional distance on South-South FDI flows? And does this influence of the dimensions differ between the conventional theory of North-South FDI flows and the proposed theory of South-South FDI flows?

The aim of this research is to contribute to the existing literature on South-South FDI by examining the importance of various location determinants of South-South FDI, trough combining the cross-national institutional dimensions given by the study of Berry, Guillén & Zhou (2010). Their study provides a framework with nine dimensions of distance that play a crucial role in attracting FDI. In addition, it is expected that South-South FDI flows are influenced by different factors than the established North-South FDI. In order to make the difference visible, this paper tries to capture the difference in location determinants between North-South and South-South FDI, by comparing the effect of the various dimensions of institutional distance on both South-South FDI and North-South FDI. It captures not only the influence on South-South FDI, but also the already established influences on North-South FDI in order to make a clear comparison.

The method used to perform this study is the following. All the dimensions of institutional distance given by the study of Berry Guillén & Zhou (2010) will be tested on North-South and South-South FDI flows. The dimensions used in this study are: economic distance, financial distance, political distance, administrative distance, cultural distance, demographic distance, knowledge distance, global connectedness distance and geographic distance. The influence of these dimensions on FDI will be tested by the use of the Gravity

(8)

model. Data is employed on FDI flows from 52 developing countries and 29 developed countries to seven developing „host‟ countries and allows this research to identify North-South and North-South-North-South FDI flows. The separation of flows helps to capture the influence of the dimensions on these specific FDI flows and leaves room for comparison.

Thus by applying this concept to identify the influences of cross-national institutional distances on FDI flows, this study provides more insight in the institutional determinants of South-South FDI flows and enhances previous literature by answering the question whether South-South FDI flows are influenced by different factors than FDI flows from the north. This brings new implications for policymakers from emerging countries, as these insights will provide them with additional instruments to attract investors from other emerging economies.

(9)

2 THEORETICAL BACKGROUND

This section will elaborate on the foreign direct investment and the role of institutional distance. It will go more in depth to explain the difference between South-South and North-South FDI. Lastly it will discuss the individual dimensions of institutional distance and there possible relationship with FDI flows.

2.1 Foreign Direct Investment

As mentioned earlier, more and more companies expand through foreign direct investments instead of exporting or licensing. According to the World Bank (2013) FDI refers to the net inflows of an investment that is attained in order for a MNE to acquire a lasting management interest in an enterprise operating in an economy other than that of the investor. It can be measured in stocks or flow. Stock being the cumulative investment flows up until a certain point in time and flow measures the periodical (e.g. annual) investment.

According to several researchers, firms choose between two types of FDI, vertical or horizontal production structures (Markusen and Markus, 2001; Aizenman and Marion, 2003). Vertical FDI occurs when the MNE fragments the production process, locating a stage or stages of production in different countries thus geographically split up their production process (Aizenman and Marion, 2003).When doing this, firms tend to take advantage of international factor price differences and production can be done at the least cost (Markusen and Markus, 2001). Horizontal FDI arises when the firm produces the same product or service in multiple nations. It is a substitute for exporting and a desire to place production close to customers and thereby avoid trade costs (Buckley & Casson, 1981). Markuses and Markus (2001) state that the choice between horizontal or vertical FDI depends on country characteristics and their research conclude that most investments are horizontal investments. Looking at motives for FDI, scholars have defined four kinds of FDI activity. Firstly the resource seeking motive, this motive indicates that a firm engages in FDI in order to gain access to natural resources and is supply oriented (Dunning, 2000). The market seeking motives implies that a firm exploits it firm-specific capabilities to satisfy a particular foreign market and demand (Kuemmerle, 1999). When MNE‟s are seeking more efficient divisions of labor or specialization of an existing portfolio of foreign and domestic assets, it is called an efficiency seeking motive (Dunning, 2000). Over the past years efficiency seeking FDI increased, since transport costs and most trade barriers are reduced. This motive is linked with vertical FDI investment (Beugelsdijk et al., 2008). The last motive is the strategic

(10)

asset-Lieke van der Velden 10

seeking motive, which is about protecting or expanding a firm specific advantage by the acquisition of new assets, or by a partnering arrangement with a foreign firm (Dunning, 2009). This type of FDI activity has grown rapidly over the past two decades and is linked with the resource exploration perspective

2.2 South-South FDI

Southern countries, referred to as developing or emerging markets, represent nations whose economies have grown rapidly, where industries have undergone and are continuing to undergo dramatic structural changes, and whose markets hold promise despite volatile and weak legal systems (Luo and Tung, 2007). Examples of economies that experienced structural transformations are China, Russia, India and Brazil.

In the early 1990s FDI flows to developing countries originated almost entirely in the North (Akyut and Ratha,2004). However as mentioned earlier, the share of FDI from

developing and transition economies doubled over the last 20 years. Most of this increase happened from 2004 until now. The share of FDI of developing and transition economies is reaching 16% of the total FDI outward stock (UNCTAD, 2010). As these numbers are still growing, South-South FDI is becoming an important concept to investigate for several reasons. The growing South-South FDI flows indicate that developing countries are more financially integrated with each other than believed. Furthermore South-South FDI may follow different cycles and can be influenced by different factors than North-South FDI (Aykut and Ratha, 2004).

Table 1 shows the relative shares of global FDI inflows and outflows over the past decades. It is evident that the share of FDI flows of the developed economies (EU, Japan and IS) is declining over the past years and the developing economies are upcoming (Rajan, 2008). In 2003-2005 the share of FDI inflows declined to below 60% compared to 80% in 1978-1990. However the share of the developing countries rose to almost 40%. Until the mid-1980s that the FDI outflows of the developing economies were insignificant, while it rose to 15percent of world outflows in 2005. Recently, in 2012, developing economies‟ outflows reached $426 billion, accounting for 31 per cent of the world total (UNCTAD, 2013). Hence, despite the global downturn, firms from developing countries continued their expansion abroad.

(11)

Lieke van der Velden 11 Table 1. History of the relative shares of global FDI inflows and outflows

Source: UNCTAD FDI/TNC database

The earliest developing countries sources were a small group of economies including Argentina; Brazil; Hong Kong, India, Korea, Singapore, and Taiwan (Aykut and Goldstein, 2006). Since the late 1980s a growing number of developing countries, have become substantial sources of FDI including Chile, China, Egypt, Malaysia, Mexico, Russia, South Africa, Thailand and Turkey. Among developing economies, those in Asia remain by far the largest investors. In 2012, Asian countries accounted for three quarters of the developing-country total (UNCTAD, 2013). In addition Hong Kong, Singapore, South Korea and Taiwan accounted for 52 percent of the FDI outflow from 1999 to 2004 (Aykut and Goldstein, 2006).

2.2.1 Characteristics of South FDI

According to various scholars there are several characteristics that separate southern MNEs from northern MNEs.

Comparing developed country MNEs with developing country MNEs, the latter tends to be smaller of size (Wells, 1983), possess less cutting edge technology and has resources that are less sophisticated (Barlett and Goshall, 2000). Developed country MNEs are likely to have stronger ownership advantages in branding and advertising and technology (Well, 1983; Barlett and Goshall, 2000) As well, southern MNEs may create a disadvantageous image among potential clients because of country-of-origin effects. Furthermore firms from southern countries are typically in the position of „late movers‟, as developed country MNEs are often already established and well-seasoned in particular markets (Barlett and Goshall, 2000). These factors add to the difficulties southern firms face as a result of operating in a home country characterized by a difficult institutional environment and inefficient market mechanisms (Ghemawat and Khanna, 1998). However, since developing-country MNEs are familiar with the more difficult institutional conditions of developing countries and because

(12)

of their expertise in managing such environment, they might face fewer problems than MNEs from the North when expanding into other developing countries (Cuervo-Cazurraand

Mehmet Genc, 2008).

Differences can also be found considering the investment motives of MNEs,

discussed earlier in this paper. According to Luo and Tung (2007) Southern MNEs motives to invest are mainly found in asset-seeking motives, both strategic-assets and resources. Assets sought by these firms can include technology, know-how, R&D facilities, human capital, brands, consumer bases, managerial expertise, and natural resources (Luo and Tung, 2007). This is necessary to encourage social and economic development at home countries and to compensate firm-level competitive disadvantages. MNE‟s from southern countries are less likely driven by an efficiency seeking motive as their home supply or manufacturing bases allow them to continually enjoy low cost advantages through their vertically integrated global production systems (Luo an Tung, 2007). Market seeking FDI will be undertaken for the same reason as with developed economy firm; to satisfy a particular foreign market and demand. The study of Taiwanese firms (Makino et al., 2002) showed that they choose better developed countries than Taiwan for resource exploration (asset-seeking motives) and less developed countries than Taiwan for resource exploitation motives (transfer company-specific resources over borders in order to make profit; market seeking motive). Tsang and Yip (2007) argue that northern firms invest in less developed countries e.g. the South only for resource exploitation motives.

According to several scholars, much of these southern economy MNEs tend to invest close to their home economy in the immediate neighborhood or region (Aykut and Goldstein, 2006) and in countries with similar levels of development (World Bank, 2006). Despite differences in institutional characteristics they invest close to their home country, where they have acquired a certain familiarity through trade, or ethnic and cultural ties. This in contrast to northern country firms, as they invest across the world. However various documents show that developing Asia is investing aggressively overseas (Rajan 2008), in 2004 about half of China‟s outward FDI went to natural resources projects in Latin America; Malaysia has emerged as a significant new source of FDI in South Africa (Padayachee and Valodia, 1999). Furthermore according to Stephen Gelb (2005), there has been a rapid move of Indian

investments in South Africa; in 2005 there were 35 conglomerates present in South Africa. And not only Asian developing economies, also countries in Latin-America are venturing beyond their immediate region (Aykut and Goldstein, 2006). For example Brazil has considerable investments in Angola and Nigeria (Goldstein, 2003). Thus despite the

(13)

advantages of intraregional investments multiple observations suggest that the trend of

investing close to home is declining and Southern countries are increasingly investing beyond their direct region.

2.3 FDI perspectives

Most existing literature builds on the idea that foreign investments of a company are

influenced by resources and capabilities of the firm (Dunning,1980, Anand and Delios, 2002) and the need to lower transaction cost (Buckley and Casson, 1976).

The former is associated with the resource based view (RBV) and focuses on company‟s resources which can become the sustained competitive advantage of a firm. These resources must be rare, valuable, imperfectly imitable, and non-substitutable (Barney, 1991).

According to this view FDI is seen primarily as a way by which firms can appropriate rents in foreign markets from the exploitation of their distinctive resources and capabilities

(Filatotchev et al., 2007) and emphasis the role of creating value. Dunning (1991) indicates several firm- and industry-level factors positively associated with FDI including firm size, operating experience, the possession of proprietary resources and product differentiation. The latter is linked to economic theory; it predicts that firms engage in FDI according to two principles (Buckley and Casson, 1976). Companies internalize missing or imperfect external markets until the costs of further internalization outweigh the benefits and will choose a location for their activities that minimizes the overall costs of their operations (Benito and Gipsud, 1992). Hence, it indicates that FDI is an approach to reduce the transaction costs related with coordinating activities in different markets (Filatotchev et al., 2007). According to this theory there are several important determinants of location choice for FDI namely; labor cost differentials, transportation costs, the existence of tariff and non-tariff barriers, as well as government policy (Benito and Gipsud, 1992).

The above theories consider mainly the impact of firm and industry specific factors, finding the influences of country-specific contextual factors less important (Yiu and Makinu, 2002). In other words they think of institutions as background (Peng et al, 2008). Yet, a new generation of researchers suggests that institutions are more than background conditions, they directly influence the strategic decisions of enterprises (Meyer et al, 2009). Currently

institutional theory has appeared to be a leading theoretical framework for research on enterprise strategies in emerging economies, along with transaction-cost and resource based views (Hoskisson et al., 2000). Institutions are defined as „the humanly devised constraints

(14)

that structure human interaction‟ (North,1993), further distinguished in formal institutions (rules, laws, constitutions) and informal institutions (norms of behavior, conventions and self-imposed codes of conduct). Contemporary institutional theory (Scott, 1995) indicates that organizations are influenced by what is “appropriate and meaningful behavior” (Zucker, 1983: 105) and in order to survive they must follow the fundamental rules and beliefs of the environment and enchaining the firm legitimacy (DiMaggio & Powell, 1983; Meyer & Rowan, 1977). Institutional theory emphasizes the ability of institutions to influence the firms to adapt to policies, structures and practices that are consistent with institutional preferences and thus influence the decisions of organizations (Meyer & Rowan, 1977). According to this school of thought the structure of an entity is influenced by three kinds of isomorphism; coercive isomorphism, were the firm experience pressure from other

organizations (Eden and Miller, 2004); mimetic isomorphism, the imitation of structures by other organization in response to pressures (Eden and Miller, 2004); and normative

isomorphism, regarding conformance to normative standards established by external institutions (DiMaggio & Powell, 1983).

2.4 Institutional Distance

Within the previously discussed institutional theory, Kostova (1999) developed a new concept namely institutional distance, referring to extent of dissimilarity between host and home institutions. The study of Kostova (1999) indicates that the larger the institutional distance the more difficult it is for the MNE to establish legitimacy in the host country and to transfer strategic routines to foreign subsidiaries (Kostova, 1999). This is in line with earlier ideas of cross-national distance in the international business field, looking at the eclectic paradigm of Dunning (1993). Dunning argues that countries may not only be distant in geographic sense, but also on social, cultural and political areas. These differences will make it harder for an MNE to operate across these countries and to overcome liability of foreignness (liability of foreignness lowers the profitability of foreign investors compared to their local competitors). The more the host economy differs from the context with which the MNE is familiar, the more difficult the adaptation (Ionascu et al., 2004)

Over the past two decades several studies draw on the concept of institutional

distance, yet all used different dimensions to identify distance and its influence on FDI. For instance Habib and Zurawicki (2002) focused only on corruption as a dimension; Linders et al (2005) used six dimensions of governance infrastructure quality to indicate distance between countries. Hence, the key problem with previous studies on cross-national distance

(15)

and its impact is the lack of a theoretical framework (Berry, Guillén & Zhou, 2010). Previous scholars has inclined to be one dimensional and not representing the diverse characteristics of countries. Therefore Berry, Guillen & Zhou (2010) provided us with nine dimensions of distance: economic, financial, political, administrative, cultural, demographic, knowledge, connectedness, and geographic distance. It is important to define and measure cross-national distance along multiple dimensions, as different types of distance can affect firm, managerial or individual decisions in different ways, depending on the dimension of distance under examination (Berry, Guillén & Zhou, 2010). Overall institutional distance can provide an explanation for MNE behavior and its location decision (Xu and Shenkar, 2002). 2.5 Hypotheses Development

This study uses the dimensions of Berry et al. (2010) of institutional distance in order to determine what influences FDI flows of developing countries to other developing countries. Furthermore it tries to capture the different influences of these dimensions on North-South and South-South FDI flows. Relationships between FDI flows and the different institutional dimensions may be either negative or positive, since institutional distance not only harm FDI flows but could also attract investors. The section below describes the dimensions used in this paper to identify institutional distance and explains the possible relationships between the dimensions and FDI flows.

2.6.1 Economic Distance

Economic distance is the first dimension used in this research, which is defined by differences in economic development and macroeconomic characteristics. The study of Berry et al (2010) has indicated three indicators of importance for economic distance, namely level of income (measured in GDP per capita), prevailing inflation rates and intensity of trade with the rest of the world (exports plus imports as proportion of GDP). These indicators are connected with consumer purchasing power and preferences, macro-economic stability and openness of the economy to external influences (Berry et al., 2010), and therefore important as these characteristics have been used in various studies to determine the choice of a foreign market or entry mode (Caves, 1996; Zaheer and Zeheer, 1997). According to Ghemawat (2001) income is the most important economic characteristic that creates distance between countries; moreover it has a striking effect on the levels of trade and the types of partners a country trades with. It often reflects distances in factor costs and also in technical capability (Ghemawat, 2001).

(16)

competition and sophisticated consumer demand, while a lower level of economic

development in the host country indicates a smaller market size and higher inflation rates that reduce purchasing power of the consumer (Pattnaik and Lee, 2012). According to the

conventional theory large economic distance allows developed firms to exploit there firm-specific assets and gain an advantage over local competition. In addition, as lower economic development might indicate low factor (labor) cost, efficiency driven developed firms will benefit from this condition. However, compared to northern firms, southern firms are less driven by efficiency seeking motives as they can profit of low wages in their home country (Tsang and Yip, 2007). And thus economics distance will be beneficial for southern firms in the way can be for northern firms. Moreover, we even expect that economic distance will harm South-South FDI flows. The study of Wells (1983) underlines this argument; he claims that the competitive advantage of entities from the South lies in their ability to function in similar economic environment, with similar levels of income per capita. It implies that economic distance will not favor the investment decision of southern MNEs as they find it difficult to function in an economic distant environment. In this line of reasoning we expect that economic distance will have a negative influence on South-South FDI.

H1 Economic distance has a negative influence on South-South FDI flows

2.6.2 Financial Distance

As countries differ in levels of economic development, diverse financial systems evolved over time. This can cause implications for the way in which MNE‟s and their competitors fund their operations (La Porta et al., 1998). According to Whitley there is a distinction between two financial systems; market-based and credit-based systems. In a capital market based system entails an active capital markets where individuals and institutions trade financial securities without state interference. Shareholders invest primarily to pursue financial interests and hold control over the firm by having the option to exit if the firm no longer fulfils their interests, engaging short capital commitment (Aguilera and Jackson, 2003). This makes market based systems volatile; however it makes it easier for new companies to quickly channel funds (Vitols, 2001). With credit based financial systems, capital markets are mostly underdeveloped and reinforce higher firm dependence on debt (Aguilera and Jackson, 2003). Investments credit is provided by banks (Withley, 1994), indicating that financial institutions are among leading shareholders of a firm and are the key

(17)

financial institutions. This entails close capital monitoring and contingent control over the company (Auguilera and Jackson, 2003) and results in a long-term capital commitment providing a long-term stable financial framework for companies. However with a credit based system the state can be involved in allocation of credit, especially when banks are state-owned, which can directly influence organizations (Vitols, 2001).

In this research financial distance is measured by three indicators; private credit, stock market capital and listed companies. These are important indicators used in research on cross-national financial systems (Berglof, 1988; LaPorta et al., 1998; Steinherr & Huveneers, 1994) and to examine foreign investment (Capron & Guille´n, 2009). Differences in

financial system cause barriers when doing business in host countries (Rios-Morales

and Brennan, 2007).For instance, if a firm‟s home country has a credit-based system it may experience difficulties to raise money in countries where market-based systems are common and the cost of capital is higher (Demirgüç-Kunt and Maksimovic, 2002). Furthermore MNE‟s from developed countries, with vibrant stock market and developed system, may find in difficult to raise funds in economies where financial institutions are less developed

(Pattnaik and Lee, 2012).

Overall, underdeveloped financial markets and limited access to credit markets

restricts entrepreneurial development (McKinnon, 1973) and will be less attractive for foreign direct investments. However, southern entities are more familiar with underdeveloped

financial markets thanks to previous domestic experience and will probably be less affected by large financial distance. Earlier research of Cuervo-Cazurraand Genc (2008) supports the idea that developing-country MNEs are familiar with the more difficult financial conditions and therefore face fewer problems than firms form the north. In addition southern entities might even benefit when investing in countries with underdeveloped institutions, as they would be first over northern firms to enter this particular market. In that case, it has large financial distance has positive influence on the location decision of southern firms. Hence we claim that financial distance has a positive influence on South-South FDI, complementing the study of Cuervo-Cazurraand Genc (2008).

H2 Financial distance has a positive influence on South-South FDI flows 2.6.3 Political Distance

According to various scholars, countries furthermore differ in terms of their political system with regard to differences in political stability, democracy and trade bloc membership. The study of Berry et al. (2010) characterized countries dimension along continuous political

(18)

dimensions. These indicators of this dimension are policy-making uncertainty, democratic character, size of the state, WTO member, regional trade agreement. According to several scholars ((Delios & Henisz, 2000, 2003; Garcı´a-Canal & Guille´n, 2008), the variables are correlated with the choice of which foreign market to enter and FDI flows.

Political distance is an important distance for this study as foreign investors can potentially suffer from political environment and political instability, and this can influence the FDI flow to a particular country. It possesses threats to MNEs as governments can suddenly shifts tax regulation policy or exchange rate controls (Henisz, 2000). On the other hand, changes in policies may change the relative competitiveness of ﬁrms operating within a given market, and in consequence provide opportunities for competitive advantage for

foreign investors (Spar, 2001). Yet, in countries with high state interference, firm may also suffer from expropriation (Henisz, 2000). This entails that the government transfers revenue or private property from the MNE to the government budget for public interest.

According to the existing theories, Northern firms are more reluctant to invest in a country where there are excessive regulations, or where regulations can vary unpredictably as they have well-defined political institutions in their home country. It increased the risk of doing business abroad and thus political distance has a large negative influence on their investment decision. In contrast with northern countries, most southern countries have less developed political institutions. Therefore developing-country MNEs may be better at dealing with a volatile political environment as they are used to political instability and violence (in the past) in their home countries. This makes them able to cope with different political environments linked to large political distance. The capability of coping with large political institutional distance might even give southern firms a first-mover advantage over northern firms, as they are most likely first to invest in those particular countries. Based on this argument it is expected that political distance has a positive effect on South-South FDI. This is in line with the study of Buckley et al (2007), showing that Chinese firms prefer countries with higher political risk when choosing an investment location.

H3 Political distance has a positive influence on South-South FDI flows

2.6.4 Administrative Distance

The next key dimension described in the study of Berry et al. (2010) is administrative distance and refers to differences in bureaucratic patterns due to colonial ties, language religion and the legal system (Ghemawat, 2001). It can be argued that administrative distance

(19)

is related to cultural and political distance, however „it goes beyond national political systems to include both formal and informal institutional arrangements that transcend the purely political nature of the nation-state‟ (Berry et al., 2010). Therefore it is also an important dimension to consider as it may have an impact on foreign direct investment decisions. Legal restrictions may limit the equity state that foreign investors are allow to hold and informal arrangements, such as whether bribery is acceptable, may support locally owned firms over foreign enterprises (Peng, 2003). This will not attract FDI as it increases the

transaction cost of engaging in these markets and thus has a negative influence on FDI flows. Although firms from developing countries might be familiar with informal arrangements, as they experienced it in their home country, due to the differentiating nature of these

arrangements it will increase transaction cost also for them. Legal or informal arrangements might differ from their home country, increasing the administrative distance, increasing transaction cost of engaging in FDI and negatively influence the FDI decision of southern entities. Therefore we argue that large administrative distance will have a negative influence on South-South FDI flows. This is consistent with the study of Rauch (1999), which

demonstrated that colonial ties and common language increased bilateral trade in emerging economies, both with developed and developing countries.

H4 Administrative distance has a negative influence on South-South FDI flows

2.6.5 Cultural Distance

In general, culture is the homogeneity of characteristics that separates one human group from another (Tihanyi et al., 2004). Cultural attributes determine how people interact with one another and with companies and institutions in a specific country (Ghemawat, 2001). Hofstede (1980) and other scholars (Ghemawat, 2001; Shenkar, 2001) have confirmed the importance of cultural differences across countries and their influence on foreign market entry, entry mode choice and other matters. Most theory on cultural distance argues that firms are less likely to invest in culturally distant countries (Shenkar, 2001) as it may lead to higher levels of complexity and uncertainty (Tihanyi et al., 2004).

In order to measure culture distance, different approaches are used (Luostarinen, 1980; Ronen and Shenkar, 1985), however the most common and wide spread approach is the study of Hosfstede (1980) and its cultural dimensions. He indicated that culture varies along four dimensions; uncertainty avoidance, individuality, power distance, and

(20)

masculinity-Lieke van der Velden 20

femininity. Given the popularity of Hostede‟s measures, his four dimensions are used in this study as indicators for cultural distance.

Cultural distance is an important dimension to take into consideration since several studies indicate it as a great influence on a firms investment decisions. For instance the theory of familiarity, arguing that firms were less likely to invest in culturally distant markets

(Shenkar, 2001). In contrast Dunning (1988) claimed that larger cultural distance encouraged FDI as a way of overcoming market and transactional failures, yet this statement is linked to FDI decisions of developed country MNEs. Looking at southern firms, different studies on emerging markets argue that cultural links enable firms to reduce the risks associated with FDI. Shared values are associated with low cultural distance and show great importance for emerging countries when investing. Therefore this study hypothesizes that cultural distance will have a negative effect on South-South FDI. This is in line with the study of Filatotchev et al. (2007) which claims that informal network linkages with host countries play an

important role in a developing MNEs FDI decision and are particularly strong when firms share the same cultural values.

H5 Cultural distance has a negative influence on South-South FDI flows 2.6.6 Demographic Distance

Countries differ in terms of size, age, growth and qualities of their population, and is called demographic distance (Berry et al., 2010). Indicators of the demographic dimension are life expectancy, birth rate, population under 14 and population under 65. This dimension has direct implications for the market attractiveness and growth potential of a country. Porterba (2001) and Huynh et al. (2006) have found a clear links between age structure and asset returns. Difference in demographic structure can create issues because of lack of availability of labor and similar target market segments in terms of age group (Pattnaik and Lee, 2012), this can influence an investment decision.

Demographic structures are particular important for firms that pursue market or efficiency seeking motives as it indicates the growth potential and attractiveness of markets and the availability of labor. Efficiency seeking motive is mostly related to developed country MNEs who are investing in developing economies. Previous research shows that

demographic distance has a positive influence on North-South FDI. However, demographic structures can be also linked to market seeking motives and is one of the main motives for southern firms to engage in FDI. Low demographic distance entails similar target market segments in terms of age group (Pattnaik and Lee, 2012). This similarity makes it easier to

(21)

exploit foreign markets with film specific-assets, since the firm knows how to reach the target market. Therefore, when the demographic distance increases, it will be more difficult to penetrate the market for emerging country firms and will effect FDI flows negatively. Hence, in line with the study of Pattnik and Lee (2012) we propose that demographic distance has a negative influence on South-South FDI flows.

H6 Demographic distance has negative influence on South-South FDI flows

2.6.7 Knowledge Distance

According to institutional literature, countries differ in terms of their capacity create

knowledge and to innovate. Since talent, innovation and creativity are not distributed evenly across nations, it affects the distance among countries. It may create difficulties for MNEs as knowledge developed in one country may not be transferable to another country or firm as knowledge can be tacit difficult to capture. Yet Nachum et al. (2008) argue that MNE location choices are related to the proximity of knowledge in the host country. In this study knowledge refers to the difference in numbers of patents and scientific articles per capita (Berry et al., 2010).

As mentioned earlier Southern firms lack in technology and knowledge compared to Northern firms. A common motive for southern firms to engage in FDI is strategic-asset seeking motive since they are eager to acquire know-how trough FDI to fill their resource void (Luo and Tun, 2007). When the host country has proximate and accessible knowledge, southern countries can benefit from other southern countries‟ knowledge spill-overs. Yet, to benefit from knowledge spill-overs, there has to be a gap in knowledge between country A and country B leading to high knowledge distance. Thus in order fill their resource void, high knowledge distance might have a positive effect on southern MNEs investment decision as they can benefit from host country know-how. Therefore this study proposes that knowledge distance will have a positive influence on South-South FDI in line with the earlier study of Luo and Tun (2007). This proposal is opposite to the traditional „north-south‟ theory, since northern firms driving force to invest in not a strategic asset-seeking motive.

(22)

Lieke van der Velden 22 2.6.8 Global Connectedness Distance

The dimension connectedness focuses on the connectedness of a country with the rest of the world and is identified by measures of international tourism expenditures as a percentage of GDP, international tourism receipts as a percentage of GDP and Internet users as a

percentage of the population (Berry et al., 2010). Connectedness facilitates the flow of

information between countries through interaction between people (Pattnaik and Lee, 2012). Lack of connectedness decreases openness of countries, which may lead to difficulties for

MNEs subsidiaries when investing in such economies. For instance, it makes it difficult for foreign firms to monitor spill-over effects due to poor communication channels. It implies that countries that are more globally connected are more attractive as potential host country, since it enhances spill-over effects. Overall developed economies are more globally

connected than developing economies, increasing the likelihood that developing countries will be less attractive for developed countries to invest in.

Looking at southern country, many interactions in emerging countries are not based on established communication channels, but rely on informal network linkages (Fillotev et al., 2007). Considering firms from the south, they gained experience with these informal ways of communication at home and might not be deterred by poor connectedness. Hence,

southern would be able to deal with both connectedness and the lack of it, indicating that large global connectedness will not have a negative effect on the location decisionThis is in line with the study of Hoskisson et al. (2000) that argues that developed capabilities for relationship-based management of southern firms that substituted for the lack of institutional infrastructure in their home countries and may be transferred abroad to other emerging economies where such assets would likewise be useful. Therefore we argue that global connectedness distance has a positive effect on South-South FDI.

H8 Global connectedness distance has a positive influence on South-South FDI flows 2.6.9 Geographic Distance

The above eight dimensions of distance are based on institutional differences across

countries, however the dimension of geographical distance is also included in the model as it has an recognized effect on trade and foreign investment. It is measured by the great circle distance between two countries according to the coordinates of the geographic centre of the countries (Berry et al., 2010). Geographic distance may increase transportation and

(23)

As earlier mentioned, emerging economies tend to invest first close to home. However in the earlier days, this was also the case for Northern countries. Yet, due to

globalization geographical distance became less of an issue. Moreover, developed economies are more experienced in dealing with geographic distant since most of them have numerous operations across the globe. The trend of expanding across the globe is also seen with emerging economies investments. Nevertheless these countries just recently started to invest across the world and may have less experience with increasing transportation and

communication costs which increases the cost of FDI. Furthermore Wells (1983) argues that a motive for engaging in FDI for emerging economies is the need to improve export

competitiveness and to defend the exports markets after increased competition. As

developing economies export close to home, this might be another reason for keeping FDI in the neighbourhood. Thus we propose that Southern MNEs have a greater tendency to invest close to their region and geographic distance will affect South-South FDI flows negatively. This is consistent with the observation of Aleksynska and Havrylchyk (2011) that common border has a large impact on investors from the South.

H9 Geographic distance has a negative influence on South-South FDI flows

2.7 Conceptual Model

Overall the theoretical claim of this thesis is that institutional dimensions have a different influence on South-South FDI than on the traditional North-South FDI. Administrative distance is an exception; we expect that it has a similar effect on both FDI flows. Figure 1 shows the by previous study established relationships between the institutional dimensions and North-South FDI. Figure 2 displays the new constructed conceptual model showing the expected relationships between the dimensions and South-South FDI flows, linked to the hypothesis. Not only will the constructed hypothesis be tested in this study. Firstly the

established theory is tested, North-South FDI flows; secondly it is tested for South-South FDI flows. The outcome of the different directions will be compared in order to see if the

influence of the dimension on the flows differs.

(24)

Lieke van der Velden 24 Figure 1. Established relationships between institutional dimensions and North-South FDI

Figure 2. Conceptual model. Expected influences of institutional dimensions on South-South FDI

(25)

3 DATA COLLECTION AND METHOD

This section will discuss the data collection and method used to conduct this research. Furthermore it will provide explanations of the measurements of the independent and dependent variables and the method used.

3.1 Data collection

This paper makes use of secondary data sources (UNCTAD, OECD, World Bank) in order to conduct the research. The main advantage of the use of secondary data is that it takes less time to collect data and leaves more time to interpret the results (Sorensen et al., 1996). Hakim (1982) argues that secondary data forces to think more closely about theoretical aims and substantive issues of the research. Furthermore most data collected by governments and international organizations are of high quality and reliable since they are collected by experts using rigorous methods (Ghauri and Grønhaug, 2005). Moreover, as this is a cross-country research, secondary sources make it easier to compare similar data from different countries (Ghauri and Grønhaug, 2005). However there are some drawbacks of secondary data that must be considered. Secondary data is conducted for another study or organization with different objectives and may not completely fit the questions of this particular research. Therefore secondary data has to be treated with some caution.

The focus of the study lies on emerging markets; however estimating the extent of such South-South FDI is not easy, as data are not available at the desired level of

disaggregation. There are two comprehensive databases that can be used to derive FDI flows of developing economies, namely IMF-BoP Manual and UNCTAD. The latter is far more complete and compiles data on bilateral FDI flows, unlike the IMF-BoP Manual. Therefore the UNCTAD database is the main source used to collect FDI data for this study. The main sources of the UNCTAD database are national authorities and other international

organization as IMF, World Bank and OECD.

Data on institutional distances is compiled by the study of Berry et al. (2010) and were collected from 1960 up until 2010 from different sources (Appendix I). They compiled the each distance into one variable, meaning that every indicator to determine distance been compiled into one variable. Thus every indicator used to compile distance has equal weight. However, since they received funds and more time to conduct their study, they were able to collect data on 150 countries. As the data set of this study contains 81countries, some of which data is really difficult to obtain, it is decided to use the institutional distances gathered by Berry et al. (2010).

(26)

3.2 Sample

The dataset of this study consist a total of 81countries shown in Appendix II. The South is defined as 52 developing countries for which reasonably detailed FDI data are available. The North includes 29 high-income OECD and non-OECD member countries. These

classifications are based on World Bank 2013.

Within the sample seven countries are chosen as „southern host‟ countries; Argentina, Mexico, Thailand, South Africa, Turkey, China and Romania. These are all developing countries from different regions and are among countries with substantial FDI inflows and outflows. The countries chosen are deliberately spread across the world, in order to create an extensive dataset; this excludes the bias of a more regional based dataset. In addition the cross-country diversity suggests that data pooled from these economies offer substantial variations in terms of institutions.

3.1 Dependent Variable

The dependent variable represents FDI flows from southern and northern countries. FDI flows are derived from the year 2010, as it is the most complete and recent data of FDI available (especially for developing countries) and is measured per million US$. It is collected in a bilateral form, meaning that country pair A and B has a different flow than country pair B and A. To avoid the problem of underreporting, which is typically done with FDI outflows of developing countries, this study looks at the FDI inflows to the seven host country.

As this paper separates two types of investment streams, North-South and South-South FDI flows, two datasets are compiled. Both datasets include Argentina, Mexico, Thailand, South Africa, Turkey, China and Romania as southern host countries. In the North-South dataset the dependent variable contains FDI flows from the 29 developed countries to the seven developing host countries. For example considering country pair A and B, country A is one of the developed countries and country B is a host country, the FDI flow from country A to country B is captured. In South-South dataset the dependent variables comprehends FDI flows from the 52 developing countries to the seven developing host countries. The host countries are also included in the 52 developing countries since the host countries may also invest in each other. The bilateral form of the FDI data makes this possible.

(27)

3.3 Independent Variables

Independent variables are collect from the dataset of Berry et al. (2010). Below a description is given of the used measures and how the overall dyadic distance of the dimensions is created.

Economic distance is composed of four different indicators, income, inflation and

exports and imports. Income measured as GDP per capita (2000 US$). The second indicator, inflation is measured as GDP deflator as percentage of GDP. The indicator export is

measured as exports of goods and services as % of the GDP. Imports are measured as the imports of goods and services as percentage of GDP.

Financial distance is based on three indicators; private credit, listed companies and

stock market capitalization. Firstly, the amount of private credit available, measured as domestic credit to the private sector as a percentage of the GDP. Secondly, a listed company is the number of listed companies a country holds per 1 million of the population. Lastly, stock market capitalization refers to the market capitalization of listed companies as a percentage of the GDP.

Political distance is composed out of five indicators. The first indicator is

policy-making uncertainty considering political stability and is measured by independent institutional actors with veto power. Democratic character is the second indicator and is measured by a countries democracy score. The third indicator is the size of the state compiled by the government consumption of a nation as a percentage of the GDP. Whether the country is a WTO member is the fourth indicator. The last indicator is regional trade agreement, which considers whether countries have a membership in the same trade bloc.

Administrative distance is characterized by three indicators namely;

colonizer-colonized link, common religion and legal system. Colonizer-colonizer-colonized link indicates whether countries share a colonial tie. Common religion is the percentage of the population that shares the same religion in the dyad. The indicator legal system shows whether nations share the same legal system.

Cultural distance is based on the four dimensions of Hofstede (1988) and is measured

by using World Value Survey (2004) questions. The reason why this survey is used instead of the Hofestedes rates is that culture evolves over time (Inglehart & Baker, 2000) and this survey is conducted every 3 or 4 years. This allows us to capture the evolving nature of culture. Power distance is measured by the question on obedience and respect of authority. Uncertainty avoidance is compiled by the questions on trusting people and job security. The question on independence and the role of the government in providing for its citizens are

(28)

linked to the dimension individualism. Masculinity is measured by questions on the importance of family and work.

Demographic distance focuses on the fundamental features of the population of

countries namely the difference in life expectancy rates, birth rate and the age structure of the population (Berry et al., 2010). Life expectancy is measured by the life expectancy at birth in years. Birth rate is measured per 1000 people. The age structure of a country is divided in two indicators; the population under 14 as a percentage of the total population and the population under 65, measured as the population ages of 65 and higher as a percentage of the total population.

Knowledge distance is assembled by two indicators namely; number of patent per 1

million of the population and number of scientific articles per 1 million of the population. Global connectedness distance contains of three indicators. The first indicator is the

international tourism expenditure as percentage of the GDP. Secondly is the indicator

international tourism receipts as % of the GDP. The last indicator is internet use measured as the internet users per 1000 people.

Geographic distance is measured by the great circle distance between two countries

according to the coordinates of the geographic center of the countries (Berry et al., 2010).

3.4 Method

3.4.1. Gravity model

In order to conduct this research a gravity model is used, as it is applied repeatedly over the years to analyze trade between countries (Tinbergen, 1962). The traditional gravity model predicts a gravity relationship for trade flows similar with Newton‟s law of Gravitation. A mass of goods or labor or other factors of production supplied at origin i, ,is attracted to a mass of demand for goods or labor at destination j, , but the potential flow is reduced by the distance between them, . The equation gives the predicted movement of goods or labor between i and j, :

(29)

Lately, this model is also used to explain FDI investments (Bergstrand and Egger, 2007) and portfolio investments (Martin and Rey, 2004). To discover the relationship between the FDI flows and the institutional determinants, our study builds on this model.

3.4.1. Model specification

The gravity model estimates are acquired using hierarchical linear regression. A regression analysis is a statistical tool that estimates the relationships among variables in a set model. It identifies the impact of various simultaneous influences upon a dependent variable. Hence it shows how the value of the dependent variable changes when one of the independent

variables is varied.

Combining the regression equation with the gravity model, for this study, the equation is written as follows:

( ) ( ) ( )

The dependent variable in this equation is , which represent the FDI flow from county i to country j. In order to answers the hypotheses the equation is done twice, the first time represent the North-South FDI flow, the second time it represents the South-South FDI flow. The independent variables are displayed as follows; represent economic distance

between country i and country j, represent the financial distance between country i and country j, is the political distance between countries i and j, represents

administrative distance between countries i and j, the cultural distance between country i and

j is represented by , embodies the demographic distance between country i and country j, represent knowledge distance between country i and j, is the global connectedness distance between country i and country j, and represents geographic distance between country i and j. The term is the intercept of the equations and can be seen as the expected mean value of FDI flows when all independent variables are zero. The amount of change in the dependent variable due to the independent variables is represented by the term etc, also known as the regression coefficient (Hair et al., 2009). The is termed the residual, which is the difference between the actual and the predicted values of the dependent variable (Hair et al., 2009).

(30)

4 RESULTS

4.1 Descriptive statistics

This section will analyze the descriptive statistics in order to get a better understanding of the collected data

The North-South dataset consist of 29 northern countries, with a total of 114

observations in the six host countries. Looking at the descriptive statistics (appendix III), the largest investment made from the northern countries is 8919 million dollars and represents the flow from the Netherlands to Mexico; the minimum FDI flow is 1 million dollars and represent the flow from Portugal to Turkey (FDI North-South, M = 511,776; SD = 1188,330). The South-South dataset contains 54 southern countries, with a total of 124 observations in the seven host countries. The minimum FDI flow is 0,1 million dollars, from Laos to

Thailand; the maximum FDI flow is 1678 million dollars and represents the flow from Brazil to Argentina (FDI South-South, M = 80,319; SD = 228,760). Furthermore the variable

Knowledge distance shows that, although values range from 0,0001 to 7,390, most values are

situated at the low end. The same holds for Economic distance, Financial distance and

Global Connectedness, indicating that most distances between the countries in the

South-South dataset are not extremely high. Another noteworthy descriptive is the fact that the variable Cultural distance has in both datasets a small amount of observations (N=53; N=60). Interpreting this variable should be done with care.

4.2 Regression Assumptions

Before performing the regressions analysis, certain assumptions have to be checked. There are four assumptions that have to be considered: the assumption of linearity, independence, homoscedasticity and normality. Several statistical tests were performed in order to asses these assumptions.

4.2.3 Assumption of linearity

To perform a linear regression analysis, there should be a linear relationship between the DV and the IVs. A correlation analysis provides information about the linear relationship

between two variables. This analysis shows both the strength of the relationship and its direction. However, among the IVs there should not be a correlation because when one IV highly correlated with another, then, when one variable changes its value so does the other. This is called multicollinearity and can cause problems for the validity of the analysis. In

(31)

order to check for this whether the variables correlate the Pearson correlation coefficient is used. Furthermore by observing VIF scores, it can be determined if the correlations among IVs are severe enough to harm the regression analysis. The phenomenon of multicollinearity is present when the scores are above 10.

The correlation matrixes of the North-South and South-South datasets are presented in Appendix IV. As the matrix of North-South shows, the DV FDI correlates with the IVs

Economic distance (r= 0.297, n= 114, p< 0.01), Financial distance (r= 0.243, n= 112, p<

0.01), Administrative distance (r= -0.223, n= 114, p< 0.05), Knowledge distance (r= 0.412, n= 113, p< 0.01) and Global Connectedness (r= 0.358, n= 114, p< 0.01). The IVs Political

distance and Geographic distance display a large negative correlation (r= -0.513, n= 113, p<

0.05). This implies that larger geographical distance will influence political distance

negatively; when geographical distance increases, political distance decreases and vice versa.

Economic distance positively correlates with Knowledge distance (r= 0.329, n= 113, p< 0.01)

and Global Connectedness (r= 0.282, n= 114, p< 0.01), which is not surprising. A high level of economic welfare is often paired with higher levels of knowledge and connectedness to the world, as more is spend on education or internet. Global Connectedness also shows a positive correlation with Knowledge distance (r= 0.422, n= 113, p< 0.01). Another remarkable

correlation is the negative relationship between Financial distance and Demographic

distance (r= -0.367, n= 112, p< 0.01), meaning that when countries are less demographically

distant, financial distance between the countries increases.

Looking at the correlation matrix of the South-South dataset, other correlations occur. The DV only has negative correlations with Financial distance (r= -0.213, n= 111, p< 0.05),

Demographic distance (r= -0.191, n= 123, p< 0.05) and Geographic distance (r= -0.243, n=

124, p< 0.01). However correlation does not imply causation Noteworthy is the high

correlation of Cultural distance with Political distance (r= 0.447, n= 57, p< 0.01), despite the low amount of observations of the variable Cultural distance. This implies that an increase in cultural distance also increases political distance. Another remarkable observation is that

Financial distance and Demographic distance are positively correlated in this dataset (r=

0.457, n= 111, p< 0.01), which is the opposite of the previous discussed data set.

Although some independent variables correlate, this does not necessarily cause

multicollinearity The VIF scores in Appendix V show whether multicollinearity is a problem. Luckily all variables of the two datasets have scores around 1 and 2. Hence all VIF scores are below 10 meaning that there is no problem with multicollinearity.

(32)

Lieke van der Velden 32 4.2.3 Assumption of independence and homoscedasticity

The next assumption is the assumption of independence, meaning that residuals are not correlated from one observation to the next. This is measured by the Durbin-Watson test, which value can range from 0 to 4. A value close to 0 suggests strong positive correlation; a value of 4 indicates a strong negative correlation. Therefore, for the residuals to be

uncorrelated, the value should be around 2. Looking at the North-South model, the Durbin-Watson statistic is 2,177. The Durbin-Durbin-Watson statistic for the South-South model is 1,842. Thus both values are near 2, indicating there is no correlation and the assumption of independence is valid.

The assumption of homoscedasticity is accessed by looking at the plots of residuals

versus time and residuals versus predicted value. A random pattern is shown for the

dependent variables, indicating no hetreoscedasticity.

4.2.4 Assumption of normal distribution

An important assumption to check is if the dependent variable has a normal distribution. This can be done by using statistical and graphical methods. The dependent variables of both datasets are plotted in a Q-Q plot (Appendix VI). These plots show that the dependent variables are not normally distributed. In order to check this more accurately, statistics test are used. The Skweness and Kurtosis should be around zero for the dependent variable to be normally distributed. For the DV of both datasets it is far above zero. The last test to check for normality is the Shapiro-Wilk test, if the test is not significant the data is normal

distributed. Thus every significance value above 0,05 shows a normal distribution. It is evident that both DVs are significant (p<0,05), meaning that they are not normally distributed. Al these observations indicate that the DV does not meet the normality

assumption and therefore has to be transformed. After using a logarithmic transformation the values for both DVs are within the boundaries (table 2 and 3). Skewness and Kurtosis are closer to zero, Shapiro-Wilk tests are insignificant and Q-Q plots show a normal distribution.