MASTER THESIS

(1)

1

MASTER THESIS

PSYCHIC DISTANCE: THE EFFECT ON NORWEGIAN AND

SWEDISH OUTWARD FOREIGN DIRECT INVESTMENT

INTERNATIONAL BUSINESS & MANAGEMENT

by

Nynke Schaap – S2581493

n.e.schaap@student.rug.nl

University of Groningen Faculty of Economics and Business

The Netherlands

Supervisor: Dr. R. Drogendijk Co-Assessor: Dr. S.N. Ponsioen

June 2015

(2)

(3)

3

PSYCHIC DISTANCE: THE EFFECT ON NORWEGIAN AND

SWEDISH OUTWARD FOREIGN DIRECT INVESTMENT

Nynke Schaap

Faculty of Economics and Business, University of Groningen, Groningen, The Netherlands

ABSTRACT

This research examines the effect of psychic distance on Norwegian and Swedish Outward Foreign Direct Investment (OFDI). Norway and Sweden differ with regard to their European Union (EU) membership status – Norway being the ‘ousider’, while Sweden is a member of the EU – but are very similar to each other in most other respects. Seven dimensions of pyschic distance have been tested, using OLS regression analysis. The empirical analysis is based on 109 host countries over the period 2002 to 2012. The results show that psychic distance cannot be ignored as an explanatory factor for either Norwegian or Swedish OFDI. However, the effect of psychic distance is even significantly stronger for Sweden than for Norway. Suggesting that Swedish firms, compared to Norwegian firms, invest less in countries with larger psychic distance. In particular the psychic distance stimuli culture, language and industrial development relate to Swedish OFDI, whereas none of the psychic distance stimuli relate to Norwegian OFDI. Therefore, the findings indicate that the effect of psychic distance significantly varies between an EU member and a non-EU member. Concluding that countries within the EU are more affected by psychic distance than countries operating outside the EU.

(4)

4

1INTRODUCTION

Markets increasingly become more global, as a rising number of organizations have progressively dedicated themselves to search for growth opportunities and collaboration beyond their country of origin. Organizations often venture into unknown territory, in order to operate on an international level (Hofstede, 1980). It is of significant importance for organizations to understand how backgrounds of people may influence managerial decisions. Different backgrounds may exist due to national cultures, as they are known to vary between countries. Subsequently, these cultural differences increase the complexity of business transactions (Hofstede, 1980; Johanson & Wiedersheim-Paul, 1975; Nordström & Vahlne, 1992). According to Boyacigiller (1990) large cultural distances between parties will increase the risk of misinterpretation and also increase the cost of interpreting information flows. Boyacigiller (1990) stated that these increases in transaction costs, both perceived and real, will in turn influence a manager’s perception of the attractiveness of doing business with a certain group of individuals. Therefore, large cultural differences among countries are predicted to influence managerial decisions such as market selection for foreign direct investment (FDI) (Davidson, 1980; Kogut & Singh, 1988). As stated by Barkema, Bell and Pennings (1996), companies entering the global game of FDI face cultural adjustment costs, especially when these companies engage in double layered acculturation. A situation where firms have to adjust both to an unknown national culture, as well as to a corporate culture (Barkema et al., 1996). Studies of performance show that greater cultural distance lowers the performance of MNCs in a host country, as foreign investments by MNCs are affected by cultural differences (Craig, Greene & Douglas, 2005; Li & Guisinger, 1992; Barkema et al., 1996; Benito, 1997). Cultural distance is therefore considered as an important source of costs and challenges for firms operating outside their home country (Beugelsdijk, Slangen, Maseland & Onrust, 2014).

(7)

7 study is on psychic distance. Researchers from the Uppsala University introduced the concept of psychic distance, studying the internationalization process of firms (Johanson & Vahlne, 1977; Johanson & Wiedersheim-Paul, 1975). Johanson and Wiedersheim-Paul (1975) view psychic distance as “factors disturbing or preventing firms learning about and understanding a foreign environment”. Psychically close countries are suggested to offer more familiar operating environments and to be more easily understood (O’Grady & Lane, 1996). Therefore, psychic distance can imply barriers for understanding information and the flow of information between firms, as difficulties and costs of interpreting and transferring the information necessary to effect outward foreign direct investment (OFDI) increase (Håkanson, 2014; Johanson & Wiedersheim-Paul, 1975; Nordström & Vahlne, 1992). Several studies have been performed to describe the role of psychic distance as an explanatory factor for the internationalization of MNCs (Barkema et al., 1996; Johanson & Vahlne, 1977). This has been studied largely due to the importance of psychic distance in internationalization processes (Conway & Swift, 2000). Traditional internationalization theories indicated that a large psychic distance between two countries makes the market entry very challenging (Bilkey & Tesar, 1977; Cavusgil, 1980; Johanson & Vahlne, 1977; Johanson & Wiedersheim-Paul, 1975). Barkema et al. (1996) examine the longevity of foreign entries and stated that psychic distance is a prominent factor in foreign entry. According to Barkema et al. (1996), further research should investigate whether the conclusions reached in their research are robust for using data on expanding firms from other home countries. In theory, it is plausible that it leads to different conclusions. Therefore, the concept of psychic distance has been identified to be a major indicator of foreign direct investment including the choice of entry mode and the market choice of companies (Schlegelmilch & Stottinger, 1998).

(8)

8 ‘outsider’, while Sweden is a member of the EU. Still, they are remarkably similar culturally as well as in terms of wealth and market size. Therefore, the main difference between Norway and Sweden is the extent to which their industrial and economic structures have converged due to a deep integration taking place within the EU (Benito, Grøgaard & Narula, 2003). Furthermore, MNC activity within the EU has largely been studied, concerning the importance of location advantages (Benito & Gripsrud, 1992; Culem, 1988; Davidson, 1980; Dunning, 1988; Jackson & Markowski, 1996; Lipsey & Kravis, 1982; Mudambi, 1995; Veugelers, 1991). According to Benito et al. (2003) there is concrete evidence that there are significant benefits for MNCs operating within the EU compared to those countries operating outside it. However, no comparative study has been performed concerning the effect of physic distance on EU members and non-EU members (Benito et al., 2003). Their different statuses– Sweden, the ‘insider, vs Norway, the ‘outsider’– makes the case of the effect of psychic distance on OFDI in these countries particularly well suited for an investigation. Consequently, I seek to analyze the impact of psychic distance on OFDI in a EU member country and non-EU member country, based on seven dimensions of psychic distance. Taken together, I expect to be able to define which psychic distance stimuli are (particularly) affecting Norwegian and Swedish OFDI. This lack of knowledge of how psychic distance influence OFDI in an EU and non-EU country will thereby fill a gap which exists in current literature.

The aim of this research is to explore the effect of psychic distance on the internationalization process of Norwegian and Swedish Multinational Corporations (MNC), and in turn influencing Norwegian and Swedish OFDI. This objective is summarized in the following overall research question:

‘What is the influence of psychic distance on Norwegian and Swedish Outward Foreign Direct Investment?’

(9)

9 monetary integration and social protection). Therefore, it is expected that the effect of psychic distance will be stronger for Sweden than for Norway.

Is there a difference between Norway and Sweden with regard to how psychic distance affects Outward Foreign Direct Investment?

Moreover, I will investigate the relevance of psychic distance for the internationalization processes of MNCs. In other words, this research helps to support management internationalization strategies. It will help to raise awareness of managers that psychic distance might have a significant influence on FDI and should be taken under consideration when making business decisions. Therefore, this research will make a significant and value added contribution to current thinking as well as offer practical tools for firms that operate with foreign partners. It will provide more insight into the field of international business by empirically testing the effect of psychic distance on Norwegian and Swedish OFDI. Moreover, the work by Dow and Karunaratna (2006) will be applied. They developed a range of psychic distance stimuli, divided into seven dimensions: culture, language, education, industrial development, political systems, religion and time zones. As a result, this research contributes by testing the aggregate effect of psychic distance, as well as the individual effects of these seven stimuli, as some stimuli might be more important for Norwegian and Swedish OFDI than other stimuli.

This research paper is organized as follows; section two outlines the relevant literature, followed by section three that provides an overview of the hypotheses. Next, the research design and methods are described; this is then followed by section five with the results of the empirical research. Section six provides the discussion based on the literature review and data collection. Section seven is the closing chapter, including the conclusion, limitations of this research and implications for future research, whereby an answer is given to the main research question.

2 THEORETICAL BACKGROUND

(10)

10

2.1 Psychic Distance and FDI

Internationalization of companies has been the research focus of many international business scholars (Brouthers & Hennart, 2007; Cavusgil & Zou, 1994; Johanson & Vahlne, 1977; Johanson & Wiedersheim-Paul, 1975; Weisfelder, 2001). FDI, defined as a “cross-border investment by a resident entity in one country with the intention of obtaining a lasting interest in a firm resident in another country”, is a key element in international integration (Al-Sadig, 2009; OECD, 2013; UNCTAD, 2014). Companies are increasingly involved in FDI and arguably need to learn to adapt to the idiosyncratic milieus of foreign markets places (Kogut & Singh, 1988). Previous academic research shows that due to the lack of knowledge about foreign countries and a propensity to avoid uncertainty, companies start investing in countries that are comparatively similar and well-known with regard to business practices (Johanson & Wiedersheim-Paul, 1975). This is in line with Davidson (1980), who stated that firm’s priority for projects is likely to increase when firms have prior experience in a host country. In addition, the relative importance of different country characteristics in determining location patterns is influenced by the experience level of the firm. Inexperienced firms exhibit greater preference for similar, near markets than firms with broader international operating experience (Davidson, 1980).

International studies (Hirvensalo & Hazley, 1998; Mayer, 1998; Ziacik, 2000) show that increasing sales and acquiring new markets have been the main motivation for making FDI. In addition to market considerations, both investment climate factors and strategic position factors have played important roles in the FDI decision making process (Hirvensalo & Hazley, 1998; Mayer, 1998; Ziacik, 2000). Borensztein, De Gregorio and Jong (1998) stated that FDI has been cited by many as an important source of economic growth and could be one of the connections between economic growth and business cycles. However, as stated above, when companies are increasingly involved in FDI, cultural differences come along. These differences are of special interest for firms who get involved in FDI, whose most characteristic feature is that they consist of units located in foreign countries (Björkman & Forsgren, 1997). Thus, these firms are located in different cultural milieus and people with different nationalities, who belong to the same company, have to get along with each other (Hofstede, 1980). Misunderstandings are therefore likely to occur when people from different cultures work together (Adler, 1986). As Hofstede (1983) points out, cultural differences may become one of the most crucial problems when managing international firms.

(11)

11 international markets based on ‘psychic proximity’ (Brewer, 2007). The Uppsala model shows the internationalization process of a firm in which the firm gradually increases its international environment. The process evolves in an interplay between an increasing commitment of resources to foreign markets on one hand and the development of knowledge about foreign operations and markets on the other. The model assumes that experiential market knowledge generates business opportunities and is therefore a driving force in the internationalization process of a firm (Johanson & Vahlne, 1990). Research has provided considerable empirical support for the Uppsala Internationalization Theory (Björkman & Forsgren, 1997). Furthermore, previous academic literature also indicates that psychic distance is a result of perceived business differences between a foreign country market and the firm’s home environment. A country will be less likely to be selected, when the perceived differences are larger. Consequently, companies initially select markets which are perceived to be similar (Stottinger & Schlegelmilch, 1998a; Petersen & Pedersen, 1996). Therefore, a significant deterrent to market selection is psychic distance (Cicic, Patterson & Shoham, 1999). However, Sullivan and Bauerschmidt (1990) found support that managers perceived no differences in cultural barriers at different stages of their firms’ internationalization. Research by Engwall and Wallenstȧl (1988) is in line with this finding, as they found no support that firms start FDI in countries culturally closer to the home country. Moreover, Benito and Gripsrud (1992) also don’t provide support for the notion that FDI is, in general, initially made in foreign markets close to the home country. Consequently, there exists a view of mixed evidence about the concept of psychic distance and its impact on OFDI. As Dow and Karunaratna (2006) stated: ‘within the realm of International Business Research, psychic distance is one of the most commonly cited (Sivakumar & Nakata, 2001), yet vaguely measured, constructs’.

2.2 European Union membership status

(12)

12 industrialization much earlier. The Swedish economy, therefore, has a disproportionately high share of large multinationals compared to the other Nordic countries (Oxelheim & Gärtner, 1994).

Table 1 Basic information about Norway and Sweden

Norway Sweden

Population (million)ª 5,1 9,7

GDP per capita (PPP) in USDª 55,400 40,900

Inward and Outward FDI stock 2013 (million)ᵇ 231 109 435 964

EU membership Not a member Member since 1995

ªCIA Fact Book (2014)

ᵇWorld Investment Report (2013)

However, when calculating the inward and outward FDI stock as a percentage of GDP per capita, the levels of FDI activity are quite similar (Table 1). This suggest, on an aggregate level at least, that both Norway and Sweden show no major differences in location advantages, and there is no clear hierarchy among these countries. Nevertheless, the one location advantage in which these Nordic countries differ is the extent to which their industrial and economic structures have converged due to a deep integration taking place within the EU (Benito et al., 2003). Concerning the EU membership status, Norway remains an ‘outsider’, whereas Sweden joined the EU in 1995 (Brommosson, 2010). Norway is associated to the EU through the European Economic Area (EEA), which is a regional integration agreement between the EU and some non-EU European countries, having rejected the EU membership twice (1972 and 1994). However, they remain unlikely to consider full membership (Benito, et al., 2003).

Sweden, on the other hand, had to harmonize their industrial structures and policies as a full member of the EU. Whereas Norway has to some extent maintained its import substituting policies, encouraging and supporting domestic industry through subsidies in several industries and non-tariff barriers (Benito et al., 2003). Norway has been obliged by the EEA to dismantle some of its subsidies, nonetheless, barriers to investment and trade are still at least double of those in most EU countries (OECD, 2000). Therefore, “Norway has thus moved from being on the periphery to being ‘on the periphery of the periphery’, by being the only non-member of the EU in Northern Europe” (Benito et al., 2003).

(13)

13 show concrete evidence that there are substantial benefits for firms operating within the EU compared to those operating outside it. Hence, MNC activity is more likely to take place within member countries of the EU. The study by Benito et al. (2003) confirms that being in the ‘core’ rather than the periphery, will derive certain benefits relative to current EU outsiders such as Switzerland and Norway. According to Benito et al. (2003), these countries are at a disadvantage relative to the core members of the EU. Though, the authors state that an ‘insider’ of the EU can experience a decline of its location advantages. This is associated with deep integration schemes, since the state must reorient its economy to the supra-regional norms established by the core. However, this is assumed to be offset by an industrial redistribution within an area based on potential access to a larger unified market and a comparative advantage (Benito et al., 2003). An ‘outsider’ country, on the other hand, will also experience a decline in their location advantages. However, as stated by Benito et al. (2003) “this is not because of industrial redistribution, but by virtue of being marginalized relative to neighboring countries that are ‘inside’ the EU”. Therefore, the different statuses in the process of regional integration, that accelerated in Europe from the 1980s onwards, makes the case of FDI in the Nordic countries particularly suitable for an investigation (Benito et al., 2003).

3 HYPOTHESES

A multidimensional instrument, developed by Dow and Karunaratna (2006), is used to measure psychic distance. They developed a range of psychic distance stimuli, divided into seven dimensions; culture, language, education, industrial development, political systems, religion and time zones. These psychic distance stimuli are all macro-level factors that are measured at national level and influence the manner in which people interpret and communicate information. The reason why these seven dimensions of psychic distance are included will be explained below and hypotheses will be given.

3.1 Culture

(14)

14 Forsgren, 1997). Large cultural differences between parties will increase the cost of interpreting information flows (Boyacigiller, 1990). These costs will in turn influence a manager’s perception of the attractiveness of doing business with a group of individuals (Johanson & Wiedersheim-Paul, 1975). Therefore, this suggests that countries with large cultural differences to Norway and Sweden are less attractive to Norwegian and Swedish OFDI. Resulting into the following hypothesis:

Hypothesis 1: Differences in culture will be negatively associated with the intensity of OFDI between Norway/Sweden and a potential host country.

3.2 Language

Challenges of doing business abroad have long been addressed by differences in languages (Johanson & Wiedersheim-Paul, 1975). Language differences have been shown to form a serious challenge to the internationalization process of firms, as differences in languages between parties will disrupt knowledge transfer and communication (Welch & Marschan-Piekkari, 2001; Welch & Welch, 2008). Language similarities present efficiencies in communication (Tushman, 1987), whereas language differences tend to increase both the risks and the costs of a transaction (Dow & Karunaratna, 2006). Welch, Welch & Marschan-Piekkari (2001) add that there exists a tendency for companies to remain within their language groups during the initial expansion as a means of containing risk. As a result, language differences influences international expansion patterns and is a key component of psychic distance (Welch et al., 2001). Summarizing, this suggest that Norwegian and Swedish firms are more likely to invest in countries where individuals speak languages that are less different from those spoken in Norway and Sweden. Formally:

Hypothesis 2: Differences in languages between Norway/Sweden and a potential host country, will be negatively associated with the intensity of OFDI between these countries.

3.3 Education

(15)

15 systems for exchanging and storing information, whereby larger parts of the population have access to these systems and knowledge (Blomkvist & Drogendijk, 2013). According to Akin and Vlad (2011), FDI level is significantly higher in countries with higher levels of education. In addition, Borensztein et al. (1998) stated that countries with a low level of education do not benefit from FDI investments. This indicates that countries with low education levels have a lower level of competitiveness on the FDI market due to lower wages for unskilled workers (Akin & Vlad, 2011). Therefore, education levels among countries have been identified as an underlying factor of psychic distance (Johanson & Vahlne, 1977; Cavusgil, 1980). It is assumed that different levels of education between countries are therefore related to lower OFDI, or formally:

Hypothesis 3: Differences in education levels between Norway/Sweden and a potential host country, will be negatively associated with the intensity of OFDI between these countries.

3.4 Industrial Development

Differences in industrial development have been found to relate to investment and trade flows between countries (Vahlne &Wiedersheim-Paul, 1977). These differences have been part of psychic distance (Vahlne & Wiedersheim-Paul, 1977) and psychic distance stimuli (Dow & Karunaratna, 2006). Business and communication norms in a developing economy are likely to be dramatically different from those of a highly industrialized economy. These differences bring in extra uncertainties and costs into transactions, and thus are likely to influence market selection decisions (Kobrin, 1976; Vahlne & Wiedersheim-Paul, 1977; Davidson & McFetridge, 1985). The importance of differences in the degree of industrialization is emphasized by Dow and Karunaratna (2006), arguing that such differences affect habits and business cultures. Consequently, these differences will affect the intensity of activities of companies in a foreign country (Dow & Karunaratna, 2006). In accordance with this argument, it is expected that differences in industrial development will negatively affect Norwegian and Swedish investments in foreign countries. Resulting into the following hypothesis:

(16)

16

3.5 Political Systems

Carlson (1974) points out that differences in political systems represent barriers to the international transfer of information. Therefore, political systems have been nominated as a potential psychic distance stimulus. These differences increase the risk that foreign companies might misjudge how other companies are likely to react in light of any potential government intervention, and how a government is likely to react in specific situations (Carlson, 1974; Child et al., 2002). Both of these phenomena are influencing market selection decisions, as they increase the risks and costs of doing business in a foreign country (Dow & Karunaratna, 2006). In the case of FDI, studies by Stein and Daude (2001) and Wei (2000) show that bad quality of formal political systems in the host country is highly detrimental to its FDI prospects. Therefore, countries that are less different, concerning political similarities, may be more attractive for Norwegian and Swedish OFDI. They would be able to benefit from experience of dealing with institutions, the government and other political actors from their home market (Cuervo-Cazurra & Genc 2008). It is therefore expected that Norwegian and Swedish firms invest less in foreign countries characterized by larger differences in political systems. Formally:

Hypothesis 5: Differences in political systems between Norway/Sweden and a potential host country, will be negatively associated with the intensity of OFDI between these countries.

3.6 Religion

(17)

17 2007; Guiso, Sapienza & Zingales, 2009). In line with this, it is expected that Norwegian and Swedish firms may prefer to involve in OFDI in countries where the main religions are similar to those religions in Norway and Sweden. Summed up and formalized:

Hypothesis 6: Differences in religions between Norway/Sweden and a potential host country, will be negatively associated with the intensity of OFDI between these countries.

3.7 Time Zone

Time zone differences are included as a final individual stimulus of psychic distance. Time zones create uncertainty about the ability for rapid communication when it is needed. The non-existent or small overlap in working hours between cities remains a problem for managers attempting to span such regions. Stein and Daude (2007) find that differences in time zones have a significant and negative effect on the location of FDI. According to these authors, time zone differences can matter even given today’s low-cost and easy communications. The transaction costs associated to the differences in time zones are important in activities that require a great deal of interaction in real time. Frequent real time communications are particularly important between foreign affiliates and headquarters, as well as between a company and it prospective foreign partner (Stein & Daude, 2007). Therefore, time zone differences have been identified as a factor of psychic distance (Dow & Karunaratna, 2006). This suggests that differences in time zones may hamper direct contact between Norwegian/Swedish firms and their representatives or units abroad. Hypothesizing:

Hypothesis 7: Differences in time zones between Norway/Sweden and a potential host country, will be negatively associated with the intensity of OFDI between these countries.

3.8 Psychic Distance

(18)

18 information necessary to effect OFDI increase (Håkanson, 2014; Johanson & Wiedersheim-Paul, 1975; Nordström & Vahlne, 1992). In the case of Norwegian and Swedish firms, it is expected that the larger the differences between Norway/Sweden and a respective host country, with regard to the psychic distance stimuli, the less OFDI by Norwegian/Swedish firms in this country will occur. Accordingly, hypothesis 8 argues:

Hypothesis 8: Differences in the aggregate psychic distance between Norway/Sweden and a potential host country, will be negatively associated with the intensity of OFDI between these countries.

3.9 Conceptual Model

(19)

19

Figure 1 Conceptual Model

4 DATA AND METHOD

In this study, the relationship between the dimensions of psychic distance and Norwegian and Swedish OFDI are analyzed. Based on the hypotheses proposed earlier, data was collected to measure the psychic distance between countries and Norwegian and Swedish OFDI.

5.1 Data Collection and Sample

The data for this study are drawn from secondary data sources: United Nations Conference on Trade and Development (UNCTAD), World Bank and the Psychic Distance Scores from Dow and Karunaratna (2006). From a review of the literature, it is recommended that longitudinal research designs take place at 10 year intervals to ensure the internal and external validity (Church, 2001; Green, Tull & Albaum, 1993). Therefore, the most appropriate time frame for collecting data about psychic distance and the effect on OFDI took place at a 10 year interval. As OFDI data is only available till 2012, the time span ranges from 2002 till 2012. Furthermore, 109 countries will be included, for which full data is available at Dow (2011). 5.2 Dependent variables: Norwegian OFDI & Swedish OFDI

(20)

20 outward flow of FDI has been used, as flow variables are subject to periodical fluctuations caused by periodic changes in the FDI environment, therefore the results might be spurious (Beugelsdijk et al., 2008; Tadess & Shukralla, 2013). Following previous studies on FDI and to obtain a more robust variable, I calculated the mean OFDI from 2002-2012 from Norway/Sweden to each country (Blomkvist & Drogendijk, 2013; Buckley, Clegg, Cross, Liu & Voss, 2007).

5.3 Independent variable: Psychic Distance

The independent variable in this study is psychic distance. For the operationalization of the psychic distance concept, the psychic distance scores by Dow and Karunaratna (2006) are used. The authors developed a multidimensional instrument for measuring psychic distance stimuli. An overview of the descriptive statistics can be found in Table 2. In order to measure cultural distance, I included the original Kogut and Singh (1980) index, based on the four original dimensions (i.e., power distance, uncertainty avoidance, masculinity and individuality) by Hofstede (1980) (Dow & Karunaratna, 2006). In total, Hofstede developed six dimensions of national culture, however full data was only available for the four original dimensions, therefore I chose not to include the other two dimensions (long term orientation and indulgence). The scores on the four dimensions by Hofstede (2001) were available for 87 countries of the 110. The scores for a further number of 14 countries were retrieved through extending the area scores of the ‘Arab world’, ‘West Africa’ and ‘East Africa’ to more countries than those included in Hofstede’s original database. Scores for another eight countries were obtained by building on earlier empirical studies that published scores of Hofstede’s cultural dimensions for these countries, which were not included in Hofstede’s study (Blomkvist & Drogendijk, 2013).1 The four cultural dimensions are tested twice: once with each of the four dimensions independently and second using Kogut and Singh’s (1988) composite index. This index is based on the deviation along each of the four cultural dimensions from the score of a given home country for each country. They are arithmetically averaged, after the deviations are corrected for differences in the variance of each dimension (Benito & Gripsrud, 1992). The Kogut-Singh index for cultural distance CDj is as follows:

1_{Scores for Kazakhstan and Uzbekistan were retrieved by earlier empirical work by Suanet and Van de Vijver}

(21)

21 Where

Iᵢᵢ = index value for cultural dimension i of country j; Vᵢ = variance of the index for dimension i;

N = home country

(22)

22 Finally, the overall operationalization of the psychic distance concept was calculated by correcting the average countries’ scores on each stimulus for the variance of the scores of all countries. Then, the formula for cultural distance proposed by Kogut and Singh (1988) was used to calculate the average distance over the seven psychic distance stimuli. This was used as an aggregate measure for psychic distance (Blomkvist & Drogendijk, 2013).

Table 2 Measurements for Psychic Distance Stimuli

Indicators Cultural distance

 Composite Hofstede score between i and j

 Absolute value of difference in IDV between i and j

 Absolute value of difference in PDI between i and j

 Absolute value of difference in MAS between i and j

 Absolute value of difference in UAI between i and j Language distance

 Distance between major languages of i and j

 Incidence of i’s major language in j

 Incidence of j’s major language in i

 Difference in ‘language’ factor Educational distance

 Difference in percentage of adult literacy between i and j

 Difference in percentage in second-level education between i and j

 Difference in percentage in third-level education between i and j

 Difference in ‘education’ factor

 Absolute value of difference in ‘education’ factor Industrialization distance

 Difference in radios per 1,000 people between i and j

 Difference in energy consumption (equiv. kilogram coal per capita) between i and j  Difference in cars per 1,000 people between i and j

 Difference in TV per 1,000 people between i and j  Difference in GDP per capita between i and j

 Difference in percentage of non-agricultural labour between i and j  Difference in percentage of urban population between i and j  Difference in newspapers per 1,000 people between i and j  Difference in phones per 1,000 people between i and j  Difference in ‘industrial development’ factor

 Absolute value of difference in ‘industrial development’ factor Political distance

 Difference in POLCON between i and j

 Difference in Modif POLITY IV between i and j

 Difference in Freedom House Civil Liberties: i and j

 Difference in ‘degree of democracy’ factor

 Absolute value of difference in ‘democracy’ factor Religion distance

 Distance between major religions of i and j

 Incidence of i’s major religion in j

 Incidence of j’s major religion in i

(23)

23  Time zone differential between i and j (hours)

 Standardized residual of the regression of Dist on Time

(Dow & Karunaratna, 2006) 5.4 Control variables

The variables controlled for are those specifically cited by scholars as important causal factors for FDI. These are (1) market size, (2) GDP growth, (3) inflation, (4)trade openness, (5) human development index and (6) corruption.

5.4.1 Market Size

Market size is controlled through the population size of each host country (measured in thousands of inhabitants), as stated in the World Bank Development Indicator (Habib & Zurawicki, 2002; Sethi, Guisinger, Phelan & Berg, 2003). Population size rather than GDP is chosen because countries such as India and China have relatively large populations compared with their GDPs (Slangen & Beugelsdijk, 2010). The enormous populations of these host countries are an important reason for MNCs to undertake FDI (Khanna, 2007), making the size of populations a better proxy for host-market size than GDP (Slangen & Beugelsdijk, 2010). Data was obtained from the World Bank Development Indicator for the years 2002-2012, using the mean of these years (Habib & Zurawicki, 2002; Sethi et al., 2003; Slangen & Beugelsdijk, 2010).

5.4.2 GDP Growth

GDP growth is generally recognized as a significant determinant of FDI flows (Al-Sadig, 2009). According to Seyoum (2011), more opportunities for generating profits will be present in rapidly growing markets. A rapid GDP growth in host countries such as China, for example, attract an increasing portion of market seeking FDI. Thus, to control for market potential and the host country’s market size, the GDP growth rate in host countries serves as a proxy for market growth (Al-Sadig, 2009; Seyoum, 2011). Therefore, it is expected that FDI flows are positively associated with GDP growth. Data was retrieved from the World Bank Development Indicator for the years 2002-2012 (Al-Sadig, 2009; Blomkvist & Drogendijk, 2013; Seyoum, 2011). To give a more robust value controlling for fluctuations, the mean was calculated of these years (Blomkvist & Drogendijk, 2013).

5.4.3 Inflation

(24)

24 may also lead to devaluation, which lowers the real value of earnings, when remitted to the home country of the investing company. Therefore, a negative relationship is expected between FDI and host country inflation, as inflation proxies for macroeconomic stability (Seyoum, 2011). Data was collected from IMFs World Economic Outlook for the years 2002-2012, using again the mean of these years (Al-Sadig, 2009; Blomkvist & Drogendijk, 2013, Seyoum, 2011).

5.4.4 Trade Openness

Trade openness (defined as imports plus exports measured as a share of gross domestic product (Blonigen & Davies, 2004)), is generally considered by economist as an important factor that promotes FDI (Asiedu, 2002; Morrisset, 2000; Noorbakhsh, Paloni & Yousseff, 2001). According to Blonigen and Davies (2004), greater openness to trade results in increased FDI activity. The ratio of exports and imports to GDP is often used to control for trade openness of a country (Al-Sadig, 2009; Gastanaga, 1998; Li,Woodard & Leatham, 2013; Sharma & Kaur, 2013). Trade openness is therefore included as a control variable to examine the impact of openness on FDI. Again data is retrieved from the World Bank Development Indicator for the years 2002-2012, calculating the mean of these years (Li et al., 2013;Sharma & Kaur, 2013).

5.4.5 Human Development Index

The Human Development Index (HDI) is widely known as a significant determinant of FDI flows. It measures the overall achievements in a country in terms of knowledge, decent standard of living and longevity (Globerman & Shapiro, 2003; Peterson, Malhotra & Wagner, 1999). HDI is derived from educational attainment (mean years of schooling and adult literacy), real GDP per capita adjusted for purchasing power and life expectancy (Seyoum, 2011). GDP per capita proxies the amount of physical infrastructure (Globerman & Shapiro, 2003). It is an indirect measure of quality of life and is considered to be an important determinant of FDI (OECD, 2001). According to Mody and Srinivasan (1998), increased levels of physical infrastructure and human capital are positively related to FDI flows. Therefore, it is assumed that higher levels of HDI will attract more FDI (positive relation between FDI and HDI). Data is obtained from the United Nations Development Programme (Seyoum, 2011) for the years 2005-20122, using the mean of these years. The index is scaled

2

(25)

25 from 0.900 and over (high HDI) to 0.349 and under (low HDI) (United Nations Development Programme, 2014).

5.4.6 Corruption

Corruption is generally considered to have a negative impact on economic growth, the level of investment (Mauro, 1995), on education services and health care (Gupta, Davoodi & Tiongson, 2000), on income inequality (Gupta, Davoodi & Alonso-Terme, 1998; Li, Xu & Zou, 2000) and on the productivity of public investment and the quality of infrastructure (Tanzi & Davoodi, 1997). According to Al-Sadig (2009) and Habib and Zurawicki (2002), all those factors are found to be important determinants for FDI activity. Therefore, it is assumed that countries with high levels of corruption will be less attractive for FDI (Al-Sadig, 2009). Data is collected from the World Bank Development Indicator for the years 2005-20123 using the CPIA database (Collier & Dollar, 2001). This database is based on the degree of corruption, transparency and accountability in a country. The index is scaled from 0 (highly clean) to 5 (highly corrupt) (WDI, 2015).

5 EMPERICAL RESULTS

This section begins with an overview of the descriptive statistics, which is followed by several robustness checks. Subsequently, a correlation matrix is provided and the results of the regression analysis are presented.

5.1 Descriptive statistics

Table 3 and 4 provide the descriptive statistics of the variables conducive to the context of this research. It outlines the number of observations, the mean, the standard deviation, as well as the maximum and minimum value of all the variables. As can be seen in Table 3 and 4, the sample size of all variables are 109 countries, selected from Europe, Africa, the Pacific, Asia, North America, Middle America and South America. An overview of all countries included in the sample can be found in appendix A. The nature of the dependent variable and the independent variables are measured on a continuous scale; whereas, the control variables are both continuous and categorical.

3

(26)

26 Table 3 Descriptive Statistics Norway

Sample Mean Std. Deviation Min. Max. Dependent Variable OFDI Norway 109 1115,78 3129,24 31,94 18480,95 Independent Variables Psychic Distance 109 2,83 1,20 0,21 5,85 Language Distance 109 0,23 0,22 0,26 0,53 Religion Distance 109 0,29 0,97 1,55 1,53

Industrial Development Distance 109 1,22 0,68 0,02 2,22

Education Distance 109 0,99 0,66 0,00 2,23

Political System Distance 109 0,31 0,30 0,00 0,83

Cultural Distance 109 3,68 1,52 0,14 8,73

Time Zone Distance 109 2,94 3,00 0,00 11,00

Control Variables

Market Size 109 56962864,59 170794967,30 183097,58 1320277083

GDP Growth 109 4,14 2,47 2,03 13,38

Inflation 109 66469,82 48601,24 35,00 256381,00

Trade Openness 109 85,61 58,84 0,12 391,67

Human Development Index 109 0,71 0,15 0,38 0,92

Corruption 109 0,72 1,24 0,00 4,00

Table 4 Descriptive Statistics Sweden

Sample Mean Std. Deviation Min. Max. Dependent Variable OFDI Sweden 109 2471,53 7085,28 0,00 43496,39 Independent Variables Psychic Distance 109 2,97 1,22 0,19 6,32 Language Distance 109 0,21 0,25 0,91 0,53 Religion Distance 109 0,12 0,85 1,29 1,53

Industrial Development Distance 109 1,25 0,68 0,02 2,26

Education Distance 109 0,77 0,53 0,02 1,88

Political System Distance 109 0,28 0,24 0,00 0,67

Cultural Distance 109 4,53 1,72 0,23 10,47

Time Zone Distance 109 2,94 3,00 0,00 11,00

Control Variables

Market Size 109 56962864,59 170794967,30 183097,58 1320277083

GDP Growth 109 4,14 2,47 2,03 13,38

Inflation 109 66469,82 48601,24 35,00 256381,00

Trade Openness 109 85,61 58,84 0,12 391,67

Human Development Index 109 0,71 0,15 0,38 0,92

Corruption 109 0,72 1,24 0,00 4,00

5.2 Robustness checks

(27)

27 interest when the assumptions are met (Field, 2009). This is important, because it is only appropriate to use linear regression if the data ‘passes’ the assumptions that are required for linear regression, which will result in a valid outcome (Berry, 1993).

(28)

28 Furthermore, as there are several explanatory variables in this research, multicollinearity is an additional concern. According to Field (2009), multicollinearity exists when there is a strong correlation between two or more predictors. To control for potential collineartiy, the variance inflation factors (VIF) are calculated for all control variables and independent variables included in the models (see Table 4). The VIF factor indicates whether a predictor has a strong linear relationship with the other predictors. A common cut-off point for VIF values are set to around 5 (Studenmund, 1992), or as high as 10 (Myers, 1990). All variables are tested on an individual level as well as together. The results showed that collinearity was present. In order to solve this, I decided to leave out the control variable Human Development Index. The results can be found in Table 4. As can be seen in Table 4, the highest VIF value in this research is 4.194. Therefore, it can be concluded that multicollinearity is not likely to be an issue within this research, as the VIF values are all below 5. All together, the sample at hand meets all the above criteria, indicating that a linear regression analysis can be performed.

Table 4 Multicollinearity statistics for OLS regression Models 2 and 3

VIF Model 2 VIF Model 3

Norway Sweden Norway Sweden

Market Size 1,157 1,151 1,254 1,214 GDP Growth 1,349 1,357 1,575 1,546 Inflation 1,399 1,398 1,604 1,606 Trade Openness 1,203 1,221 1,273 1,300 Corruption 1,349 1,369 2,460 2,461 Psychic Distance 1,300 1,366 Language Distance 1,988 2,198 Religion Distance 2,216 2,331 Industrial Development Distance 4,194 3,490 Education Distance 3,516 2,699

Political System Distance 1,260 1,180

Cultural Distance 1,191 1,246

(29)

29 5.3 Correlation Matrix

Table 5 (Norway) and 6 (Sweden) show the results of the Pearson correlation matrix between the variables, including the dependent variable and the explanatory variables. The Pearson correlation coefficient, r, is constrained to lie between -1 (negative correlation) and 1 (a positive correlation) (Field, 2009). According to Cohen (1988, 1992), a large effect exist when r is 0,50 or higher, indicating that the effect accounts for 25% of the variance. For Norway, the highest correlation is a Spearman’s r of 0,780 at a 1% level between education distance and industrial development. Also for Sweden, the highest correlation is a Spearman’s r of 0,684 at a 1% level between education distance and industrial development. It is generally understood that industrialized countries have high degrees of education (Blomkvist & Drogendijk, 2013). Moreover, as can be seen in Table 5 and 6, religion strongly correlates with language distance at a 1% level (0,559 for Norway and 0,602 for Sweden). This makes intuitive sense; languages play a fundamental role in any set of religions (Darquennes & Van den Bussche, 2011). According to Blomkvist and Drogendijk (2013), higher correlations are logically found between the psychic distance stimuli and the psychic distance index.

Table 5 Pearson Correlation: between dependent and explanatory variables

(30)

30 Table 6 Pearson Correlation: between dependent and explanatory variables

SWEDEN 1 2 3 4 5 6 7 8 9 1. OFDI 1.000 2. Psychic Distance -0,392*** 1.000 3. Language Distance -0,457*** 0,284** 1.000 4. Religion Distance -0,285** 0,348*** 0,602*** 1.000 5. Industrial Development -0,475*** 0,641*** 0,210* 0,323*** 1.000 6. Education Distance -0,268** 0,551*** 0,104 0,431*** 0,684*** 1.000 7. Political System Distance -0,039 0,208* -0,104 -0,216* -0,074 -0,266** 1.000 8. Cultural Distance -0,371*** 0,526*** 0,348*** 0,132† 0,092 -0,017 0,033 1.000 9. Time Zone Distance -0,102 0,475*** -0,012 -0,076 0,150† 0,010 0,143† 0,195* 1.000 †p < 0.10; * p < 0.05; ** p < 0,01; *** p < 0.001 5.4 Regression analysis

In order to capture the effect of psychic distance on Norwegian and Swedish OFDI, the hypotheses are tested and analyzed with an ordinary least square (OLS) regression analysis. The results for Norway are shown in Table 7 and the results for Sweden are shown in Table 8. Both tables consist of three models, including a controls-only model (Model 1), a model including the controls and the psychic distance index (Model 2) and a model including the controls and each of the seven psychic distance stimuli individually (Model 3). Both tables will be analyzed separately.

Table 7 shows the results of the regression analysis for the determinants of Norwegian OFDI. As can be seen, the adjusted explanatory value for Model 2 is 10,1%, with an F-value of 3,012 (p < 0,01). Compared to Model 1, the explanatory value increased with almost 5%. Model 3 explains 10,3% of the variance of Norwegian OFDI with an F-value of 2,028 (p < 0,05), indicating an increase of 5,3% compared to Model 1.

(31)

31 between the effect of psychic distance on Norwegian OFDI (with a marginal significant level of p < 0,10). Indicating that the larger the aggregate psychic distance between Norway and a potential host country, the less Norwegian firms will invest in this country. Therefore, hypothesis 8 is supported.

Model 3 shows the results of the seven individual psychic distance variables. The results show that the variables culture, language, education, industrial development and time zones are in the predicted direction. Suggesting that differences between countries in culture, language, education, industrial development and time zones negatively relate to Norwegian OFDI. The other two variables, religion and political systems, go against the predicted direction. However, none of the individual psychic distance variables show a significant relation. Therefore, differences in culture, language, education, industrial development, political systems, religion and time zones do not seem to explain Norwegian OFDI between 2002-2012; in other words, hypothesis 1 till 7 are not supported.

Table 7 Regression analysis: determinants of Norwegian OFDI (2002-2012)

Model 1 Model 2 Model 3

Market Size 0,048 (0,481) 0,080 (0,812) 0,075 (0,739) GDP Growth -0,177 (-1,656) -0,121 (-1,141) -0,091 (-0,793) Inflation -0,095 (-0,860) -0,072 (-0,667) 0,003 (0,025) Trade Openness 0,127 (1,236) 0,108 (1,083) 0,071 (0,689) Corruption -0,072 (-0,679) -0,001 (-0,014) 0,000 (0,001) Psychic Distance -0,272 (-2,614)† Language Distance -0,184 (-1,430) Religion Distance 0,074 (0,545)

Industrial Development Distance -0,300 (-1,609)

Education Distance -0,004 (-0,026)

Political System Distance 0,048 (0,467)

Cultural Distance -0,111 (-1,115)

Time Zone Distance -0,034 (-0,341)

F-value 2,128† 3,012** 2,028*

N 109 109 109

Adjusted R² 0,050 0,101 0,103 Note: t-values in parentheses

† p < 0.10; * p < 0.05; ** p < 0,01; *** p < 0.001

(32)

32 Model 3 explains 42,2% of the variance of Norwegian OFDI with an F-value of 7,583 (p < 0,001), indicating an increase of 28,5% compared to Model 1.

From the first model, I can conclude that the control variable that most significantly relates with Swedish OFDI is GDP growth (p < 0,01). This effect is also supported in the model including psychic distance (p < 0,05), but not supported in the model including psychic distance stimuli. The controls-only model further shows a marginal significant relation between Swedish OFDI and inflation rates, but this effect is not supported in the models including psychic distance and psychic distance stimuli. Furthermore, Model 1 does not support the control variable market size, but the models including psychic distance and psychic distance stimuli do find support for this motivation for Swedish OFDI (p < 0,05). Finally, the last control variable that shows a marginal significant relation (p < 0,10) is trade openness, but this is only found to be significant in Model 3. Model 1 and Model 2 do not support this effect. Moreover, there is no significant relationship found between the control variable corruption and Swedish OFDI. Model 2 shows the aggregate effect of psychic distance on Swedish OFDI. The results show that there is a significant relationship between the effect of psychic distance on Swedish OFDI (p < 0,01). Indicating that the larger the aggregate psychic distance between Sweden and a potential host country, the less Swedish firms will invest in this country. Therefore, hypothesis 8 is supported.

(33)

33 hypothesis predicts that differences in languages between Sweden and a potential host country will be negatively associated with the intensity of OFDI between these countries. This is also in line with the results, since Swedish firms are less likely to invest in countries that are different in language. Furthermore, the significant and negative result on the stimulus industrial development distance suggest that Swedish firms are most likely to invest in countries that show similar levels of industrial development. The final hypothesis (H5) predicts that differences in political systems between Sweden and a potential host country will be negatively associated with the intensity of OFDI between these countries. The results show indeed that Swedish firms are more likely to invest in countries at similar levels of political systems. This is consistent with hypothesis 5, however, as stated above, the relation is not significant. Therefore, hypothesis 5 is not confirmed.

Table 8 Regression analysis: determinants of Swedish OFDI (2002-2012)

Model 1 Model 2 Model 3

Market Size 0,151 (1,586) 0,183 (2,007)* 0,195 (2,421)* GDP Growth -0,290 (-2,847)** -0,217 (-2,196)* -0,117 (-1,285) Inflation -0,194 (-1,841)† -0,166 (-1,656) -0,045 (-0,489) Trade Openness -0,071 (-0,730) -0,119 (-1,265) -0,142 (-1,703)† Corruption -0,063 (-0,632) 0,033 (0,328) -0,029 (-0,250) Psychic Distance -0,344 (-3,466)** Language Distance -0,308 (-2,843)** Religion Distance 0,058 (0,516)

Industrial Development Distance -0,404 (-2,959)**

Education Distance 0,004 (0,034)

Political System Distance -0,075 (-0,946)

Cultural Distance -0,254 (-3,107)**

Time Zone Distance 0,002 (0,019)

F-value 4,424** 6,083*** 7,583***

N 109 109 109

Adjusted R² 0,137 0,220 0,422 Note: t-values in parentheses

† p < 0.10; * p < 0.05; ** p < 0,01; *** p < 0.001

(34)

34 significantly stronger for Sweden than for Norway, which is in line with the expected effect that psychic distance would have on Norwegian and Swedish OFDI.

6 DISCUSSION

This study sought to address the effect of psychic distance on Norwegian and Swedish OFDI. While extant literature provides several insights of psychic distance, the effect of psychic distance is not widely understood (Dow & Karunaratna, 2006; Sivakumar & Nakata, 2001). Thus, as a departure from prior studies, the effect of psychic distance on Norwegian and Swedish OFDI is explored.

The results show that psychic distance cannot be ignored as an explanatory factor for either Norwegian or Swedish OFDI. This supports the view that psychic distance is a significant deterrent to market selection (Cicic, Patterson & Shoham, 1999). For both countries the aggregate psychic distance significantly explains Norwegian (p < 0,10) and Swedish (p < 0,01) OFDI, showing that Norwegian and Swedish firms invest less in countries at a larger psychic distance to Norway and Sweden. This research is therefore in line with existing theories on internationalization, suggesting that psychic distance is an important explanatory factor for the internationalization of MNCs (Barkema et al., 1996; Johanson & Vahlne, 1977).

When looking into it more carefully, clear differences can be found between Norway and Sweden. Regarding the effect of psychic distance on Norwegian and Swedish OFDI, I can conclude that the effect of psychic distance is significantly stronger for Sweden (p < 0,01) than for Norway (p < 0,10). Suggesting that Swedish firms, compared to Norwegian firms, invest less in countries with larger psychic distance. It was assumed that the effect of psychic distance on OFDI is stronger for Sweden, due to the convergence of industrial and economic structures as a result of deep integration within the EU. Therefore, it was expected that Swedish firms would invest more in EU members than Norwegian firms. The results confirm this expectation.

(35)

35 psychic distance are significantly related to Swedish OFDI. The first dimension that significantly relates to Swedish OFDI is culture (p < 0,05). This is in line with current research. Previous work suggested that culture is an important explanatory factor of firms’ internationalization behavior (Boyacigiller, 1990; Child et al., 2002; Johanson & Vahlne, 1977). Countries have culturally determined ways and rules of doing business and will therefore be less likely to get involved in OFDI when large cultural differences exist (Björkman & Forsgren, 1997). Linking this with the deep interrelations between economic elements and cultural variables in EU member countries, this might explain why cultural distance is significantly related to Swedish OFDI.

The second dimension that significantly relates to Swedish OFDI is language (p < 0,05). This is also in line with current literature, as there exists a tendency that firms remain within their language groups when making foreign investments, as a means of containing risk (Welch et al., 2001). Therefore, this might explain why language significantly relates to Swedish OFDI. However, it does raise the question about the fact that differences in languages are significantly related to Sweden, but not to Norway, while both languages are quite similar.

Finally, the third dimension that significantly relates to Swedish OFDI is industrial development (p < 0,05). According to existing theories, differences in industrial development will negatively affect the intensity of FDI between countries (Dow & Karunaratna, 2006; Vahlne & Wiedersheim-Paul, 1977). As all EU member states share several common rules and institutions (Liñán & Fernandez-Serrano, 2014), it makes intuitive sense that it might explain why industrial development significantly relates to Swedish OFDI.

(36)

36 reliability of intra-EU trade, because the EU member countries can be regarded as an alliance (Gowa & Mansfield, 1993). Therefore, integration within the EU might create more certainty, indicating that countries within the EU show to be more affected by psychic distance than countries operating outside the EU.

Another explanation between these psychic distance differences might be the economic crisis that began in 2007. The EU has experienced a significant decrease in FDI inflows and outflows (see appendix C) (Ioannou, Leblond, & Niemann, 2015). To give an example, in 2010, EU OFDI decreased by 62% compared with 2009. At the same time, EU inward FDI decreased by 75% (Elsa & Radoslav, 2011). As stated by Aizenman and Noy (2006; 2009), FDI is affected by the impacts of financial and economic crises on trade relations. Consequently, this negative effect on FDI flows influences psychic distance, as an economic/financial crisis becomes visible in all components of the social system, which has a direct effect on cultural habits (Moldoveanu & Ioan-Franc, 2011). Therefore, negative FDI flows can have an important impact on the cultural environment in a country (Liñán & Fernandez-Serrano, 2014). As a result, due to the economic crisis, the effect of psychic distance on OFDI might have altered with respect to EU and non-EU members.

Furthermore, another potential explanation for differences in psychic distance between Norway and Sweden, might be the differences in dominant industries between these countries. For example, Norway’s largest MNCs are gas and oil companies (e.g. Statoil, Norsk Hydro and Telenor) (Norway Statistics, 2013), whereas Sweden’s largest MNCs are in electronic equipment, motor vehicles and machinery (e.g. Ericsson, Volvo and Preem) (Sweden Statistics, 2014). These different industries might have a significant impact on OFDI decisions. Scholars have identified four main types of determinants for making FDI; market seeking, resource seeking, efficiency seeking and strategic assets seeking (Dunning, 1993; 2000). To give an example, when Statoil in Norway would pursue a resource seeking strategy, they mainly take advantage of natural resources and will therefore not consider countries who are psychically close. Foreign investments of such firms are therefore more likely to be in distant markets (Benito & Gripsrud, 1992). Thus, the effect of psychic distance on OFDI might be influenced when motives for making FDI are taken into consideration. Therefore, differences in industries between Norway and Sweden might be another potential explanation for differences in psychic distance between these countries.

(37)

37 effect of psychic distance on OFDI significantly varies between an EU member and a non-EU member. Suggesting that countries within the EU are more affected by psychic distance than countries operating outside the EU.

7 CONCLUSION

7.1 Conclusion

The aim of this research was to empirically investigate the effect of psychic distance on the internationalization process of Norwegian and Swedish MNCs, and in turn influencing Norwegian and Swedish OFDI. Prior research has predominantly focused on the importance of location advantages of MNC activity within the EU. However, no comparative study has been performed concerning the effect of psychic distance on MNC activity between EU members and non-EU-members. As psychic distance has been identified as a major indicator of FDI (Schlegelmilch & Stottinger, 1998), this research attempted to contribute to literature by examining the effect of physic distance on an EU member (Sweden) and a non-EU member (Norway). By examining this, the main research question below can be answered:

‘What is the influence of psychic distance on Norwegian and Swedish Outward Foreign Direct Investment?’

In order to test this research question, I built on the work by Dow and Karunaratna (2006), who developed a multidimensional instrument to measure psychic distance stimuli. These psychic distance stimuli exist of culture, language, education, industrial development, political systems, religion and time zones. The individual level of these dimensions as well as the aggregate psychic distance was measured in order to test the effect of psychic distance Norwegian and Swedish OFDI.

(38)

38 differences might include the EU membership status, the economic crisis or differences in dominant industries between Norway and Sweden. Overall, it can be concluded that psychic distance is context dependent. Understanding how the context influences OFDI decisions is therefore of significant importance.

7.2 Limitations

This study has several limitations that need to be acknowledged. The first limitation concerns the databases used, which is secondary data and inherent data limitations are present. For the operationalization of the psychic distance concept, the psychic distance scores by Dow and Karunaratna (2006) are used. They have full data available for 120 countries, however due to data constraints I had to reduce my sample to 109 countries. Moreover, there was not full data available for certain countries concerning cultural distance. In order to calculate the cultural distance of these countries, scores were retrieved using the average scores of two neighboring countries. However this might influence the validity and objectivity of the research.

The second limitation concerns the timeframe (2002-2012) used, as no data was available for certain years. Due to missing data about the years 2002, 2003 and 2004 for the control variables corruption and human development index, the mean was calculated over the years 2005-2012, which might bias the results.

A third limitation is the fact that I do not test the effect whether being part of the EU is an explanatory factor why differences in psychic distance exist between Norway and Sweden. Statistical data alone are not sufficient to test the effect of psychic distance on OFDI between an EU country and a non-EU country, therefore I only tested for the presence of a correlation. The relation in psychic distance between Norway and Sweden is present, however it cannot be demonstrated that this is because of the EU membership, since statistical data alone is insufficient to prove this. Due to data and time constraints I chose to only focus on statistical data. Future studies could therefore investigate the influence of an EU membership status as an explanatory factor why differences exists between Norway and Sweden, regarding psychic distance.

MASTER THESIS

MASTER THESIS

PSYCHIC DISTANCE: THE EFFECT ON NORWEGIAN AND

SWEDISH OUTWARD FOREIGN DIRECT INVESTMENT

INTERNATIONAL BUSINESS & MANAGEMENT

Nynke Schaap – S2581493

PSYCHIC DISTANCE: THE EFFECT ON NORWEGIAN AND

SWEDISH OUTWARD FOREIGN DIRECT INVESTMENT

ABSTRACT

Table of Contents

1INTRODUCTION

2 THEORETICAL BACKGROUND

3 HYPOTHESES

4 DATA AND METHOD

5 EMPERICAL RESULTS

6 DISCUSSION

7 CONCLUSION