Host Country Characteristics and Distances on
Foreign Direct Investment Location Choice
An analysis of the development of FDI location choice from multinational
firms in the United States of America from 2000 to 2018
Wendy F. Schoonderbeek
BSc International Business & Honours College
Bachelor Thesis
University of Groningen
Faculty of Economics and Business
Supervisor: Dr. L. Em
Co-Assessor: Dr. B. Verwaeren
May 31
st, 2020
__________________________________________________________________________________ Abstract
Foreign direct investment and the location choice thereof is a widely researched topic in international business. The influence that host country characteristics, such as GDP or population, as well as distances between countries have on FDI location choice of multinational companies has been researched in the past. Flores and Aguilera (2007), for example, discussed how the influence several economic, institutional and cultural host country characteristics had on FDI location choice of US MNEs evolved from 1980 to 2000. For this thesis, I conducted a similar research, but for the more recent period 2000-2018. Additionally, I also looked at which of the distances (economic, cultural and institutional distance) mostly influenced location choice of US MNEs in this period. I found that GDP of the host country has the largest influence on FDI in all years of analysis, yet that the level of influence GDP has on FDI decreased from 2000 to 2018. This is in line with the findings from 1980-2000, as here GDP’s influence on location choice of US MNEs also decreased. Furthermore, economic distance is the only one of the three distances which has a significant influence on FDI in both 2000 and 2018. Throughout the whole 2000-2018 period, economic distance had the largest influence on FDI, followed by institutional distance and cultural distance.
Keywords: foreign direct investment, MNEs, location choice, economic distance, cultural distance, institutional
C
ONTENTSIntroduction ... 5
Research Design ... 5
Location Choice Determinants ... 5
US as a Home Country ... 6
Main Findings ... 6
Relevance ... 7
Contributions to Existing Research ... 8
Outline ... 9
Theoretical Background ... 9
Determinants of FDI – Flores and Aguilera (2007)... 9
Economic Distance ... 10
Cultural Distance ... 12
Institutional Distance ... 13
FDI Determinants from 1980-2000 to 2000-2018 ... 14
Most Influential Distance in 2000-2018 ... 15
Methodology ... 17
Data ... 17
Dependent Variable ... 17
Independent Variables ... 17
Control Variables ... 18
Preparing the Dataset ... 19
Dependent variable ... 19 Independent Variables ... 20 Control Variables ... 21 Multicollinearity ... 21 Analyses ... 21 Results ... 23
Host Country Characteristics ... 23
Interpretation of Results... 26
Distances ... 28
Interpretation of Results... 31
Conclusion... 31
Host Country Characteristics ... 32
Distances ... 32
Limitations ... 33
Avenues for Further Research ... 34
Literature ... 35
Appendices ... 40
Appendix A. Overview of Data ... 40
I
NTRODUCTION
The location choice for foreign direct investment (FDI) of multination companies (MNEs) is a widely researched topic. Flores and Aguilera (2007) found that the determinants influencing location choice of MNEs based in the United States are not stable over time. However, as they only looked at the difference in location choice determinants from 1980 to 2000, it is still unclear whether these location choice determinants change over a longer period of time and if so, how. Furthermore, distances matter: economic (Ghemawat, 2001), cultural (Tsang and Yip, 2007) and institutional (Cezar and Escobar, 2015) distances all influence FDI location choice. However, the relative influence these distances have over time is unclear, whereas this would provide useful insights adding to the current literature and challenging existing theories. My research will address both of these knowledge gaps through the question How has the importance of economic,
cultural and institutional host country characteristics and distances on FDI location choice of US-based multinationals evolved from 2000 to 2018, and how does this compare to location choice determinants over the two decades before 2000? To address this question,
I will look at how economic and institutional host country characteristics and cultural distance influenced FDI location choice for US MNEs in the period from 2000 to 2018, and compare this to Flores and Aguilera’s (2007) findings regarding the 1980-2000 period. Then, I will compute economic and institutional distances, and analyse how these – and cultural distance – evolved from 2000 to 2018.
R
ESEARCHD
ESIGN LOCATION CHOICE DETERMINANTShost country characteristics on FDI is then analysed for the years 2000, 2008 and 2018 through linear regressions, to gain insights into how the influence of each determinant on FDI changed throughout the 2000-2018 period. Furthermore, a panel analysis is used to understand the influence each of the determinants had on FDI during the whole 2000-2018 period. Consecutively, the economic and institutional host country characteristics are transformed into economic and institutional distances. Just as for the host country characteristics, the influence of each distance on US FDI location choice for 2000, 2008, 2018 and the whole 2000-2018 period is analysed.
US AS A HOME COUNTRY
This study will address the research question using the United States as home country for FDI. There are different reasons to focus on the United States as the home country. First, as Flores and Aguilera also focussed on US companies in their 2007 study, it grants an opportunity to not only provide clarity on the US companies’ current location choice host country characteristics, but to compare this to findings of research on a similar topic from 1980 onwards. Through that, the trend of US MNEs’ location choices over almost four decades can be determined. Moreover, the US is one of the leading countries when it comes to FDI outflows in the past decade: from 2011 to 2017, the US was the largest country when considering FDI outflows for four of those years (OECD, 2020). Simultaneously, US FDI outflows have plummeted from 300.39k million USD in 2017 to -90.62k million USD in 2018 (OECD, 2020). This indicates that whilst the US is one of the largest players on the worldwide FDI market, it also has a highly volatile level of FDI outflows. Furthermore, over 24% of all companies present in Fortune’s Global 500 of 2019 are US-based companies (Fortune, 2019). This makes the US the country representing the most companies worldwide with the highest revenue. These reasons make that it would be most interesting for both academia as well as businesses to better understand the methodology of US companies deciding upon FDI host countries, and to gain further insights into to what extent economic, cultural and institutional determinants matter in their location choice.
M
AINF
INDINGS2000-2018. However, the speed at which it is decreasing is shrinking. Furthermore, Flores and Aguilera (2007) found that population had a significant influence on FDI in 1980 and 2000, but that its influence decreased over the years. As I found population to not influence FDI anymore after 2000, this is in line with the decreasing trend Flores and Aguilera’s (2007) identified. For democracy, the legal system and cultural distance, Flores and Aguilera (2007) found that it was of significant influence on FDI in 1980 but not anymore in 2000, indicating a decreasing trend. The democracy score not impacting FDI significantly in the 2000-2018 period is hence in line with Flores and Aguilera’s (2007) findings. However, I found the legal system and cultural distance to have a significant impact on US FDI over the whole 2000-2018 period, which contrasts the decreasing trend Flores and Aguilera (2007) identified. Lastly, looking at the distances over the whole 2000-2018 period, all three distances showed to significantly negatively influence FDI, with economic distance influencing FDI the most, followed by institutional distance and cultural distance.
R
ELEVANCEthe influence of these distances on FDI over the years, whereas existing literature so far only focussed on overall periods.
FIGURE 1. WORLD GLOBALISATION LEVEL FROM 1980 TO 2017 (KOF GLOBALISATION INDEX, 2020).
CONTRIBUTIONS TO EXISTING RESEARCH
The results of this study will hence provide clarity on the trend of the importance of economic, cultural and institutional host country characteristics and distances on FDI location choice of US MNEs. For managers, this information is valuable when looking to invest in a foreign country, as it provides insights into most common practices from US companies conducting FDI. It helps answering questions about which of the distances have been influential in determining location choice in the past, which companies can use as a guideline for their own location choice.
O
UTLINEThe paper proceeds as follows. First an overview of existing knowledge on the determinants of FDI, including an outline of Flores and Aguilera’s (2007) main findings and an analysis of the influence of economic, cultural and institutional distances on FDI is provided. Furthermore, the hypotheses tested in this paper are developed. In the following section, the methodology is explained. This includes an outline of the dependent, independent and control variables, how the dataset is prepared and which analyses are run. Consecutively, the results are discussed and the hypotheses reviewed. Lastly, this paper concludes by giving an answer to the research question and discussing the limitations and avenues for further research.
T
HEORETICAL
B
ACKGROUND
D
ETERMINANTS OFFDI
–
F
LORES ANDA
GUILERA(2007)
choice of US MNEs from 1980 to 2000, in line with the increasing focus on host countries’ population size. The importance of whether the host country’s political institutions are similar to those in the US has decreased from 1980 to 2000. Whether the host country has a similar legal system as the US (common law) was of significant influence on US MNEs FDI location choice in 1980, but not anymore in 2000, just as cultural distance. This decreasing or disappearing influence of institutional, legal and cultural host country characteristics is by Flores and Aguilera (2007) explained using Dunning’s (1998) argument that MNEs become less susceptible to these institutional-cultural host country characteristics over time. This trend is according to Dunning (1998) driven by increasing globalisation, which is causing countries to look more homogenous and decreasing the risk perceived by investing in a country with a historically different institutional or cultural environment. 1980 Progression 1980 - 2000 Relative change 2000
Economic Host country characteristics - GDP (billion USD) - Population (millions) - Infrastructure 0.00711 0.002*** 0.005*** 0.00011 0.001 -0.001 0.001 0.001 14% -50% 20% 9091% 0.00811 0.001*** 0.006*** 0.00111*** Cultural distance -0.11*** 0.06 -55% -0.05
Institutional Host country characteristics - Political Institutions - Legal Institutions 0.48 0.31*** 0.17** 0.08 0.12 -0.04 17% 39% -24% 0.56 0.43 0.13
TABLE 1. REGRESSION COEFFICIENTS OF ECONOMIC HOST COUNTRY CHARACTERISTICS, CULTURAL DISTANCE AND INSTITUTIONAL HOST COUNTRY CHARACTERISTICS ON FDI INFLOW FROM US MNES FOR 1980, PROGRESSION FROM 1980 TO 2000 AND REGRESSION COEFFICIENTS FOR 2000 (FLORES AND AGUILERA, 2007). CONFIDENCE LEVELS: * = 10%, ** = 5%, *** = 1%.
E
CONOMICD
ISTANCEof cheap production. In this case, the economic distance between the US and the host country would be large. A large economic distance between home and host country indeed results in a lower expected risk of an investment by an MNE (Tsang and Yip, 2007). This indicates that investors expect economic distance to decrease the risk of foreign investments, and hence that economic distance has a positive influence on FDI.
On the other hand, companies could be investing in countries with low economic distance. Just as high economic distance, low economic distance could lead to an advantageous business environment. From a US perspective, it would concern FDI inflows from the US in other developed economies. These investments could for example be for research or innovation purposes. Furthermore, the potential market for US products in these developed economies is larger than the market in a developing economy would be. Ghemawat (2001) argued that companies are most likely to invest in countries with similar economic profiles, as to be able to replicate their business models and profit from their competitive advantages in the host country.
The influence of the economic host country characteristics on FDI is expected to have decreased from 2000 to 2018, however with fluctuations between the different host country characteristics. Considering Dunning’s (1998) arguments around location choice, companies will be focusing more on the size of the population and less on GDP. Hence, the number of inhabitants will have increased in importance, whereas both GDP and GDP per capita will decrease in importance. Dunning’s (1998) theories were also supported by Flores and Aguilera’s (2007) findings. Furthermore, I expect the importance of digital infrastructure to decrease in importance, as digitalization is increasing over the years (Eurostat, 2020), closing the gap between countries with and without a well-developed digital infrastructure. This would mean that the number of inhabitants is the only indicator increasing in importance, whereas all others decrease in importance. Overall, economic host country characteristics will then be of less importance on FDI location choice of US MNEs in 2018 than they were in 2000.
Hypothesis 1a: From 2000 to 2018, the four economic host country characteristics GDP, GDP per capita, population and mobile phone subscriptions are expected to on average have decreased in importance on FDI location choice of US MNEs1.
Hypothesis 1b: From 2000 to 2018, economic distance is expected to have decreased in importance on FDI location choice of US MNEs2.
C
ULTURALD
ISTANCEIn general, distances are perceived to be negative (Stahl et al., 2016), which for some MNEs also holds true for cultural distance. Having a different cultural background, norms, values and ways of doing business is expected to decrease the ease of doing business and increase the risks of investments. Beugelsdijk et al. (2018) also investigated the influence of cultural distance on FDI through reviewing 156 published papers on this subject, and concluded that cultural distance indeed has a negative impact on FDI. Hence, host countries with low cultural distance from the USA are most likely to receive US MNEs’ foreign investments (Flores and Aguilera, 2007). These results indicate that companies lay a continuous focus on mitigating risk through selecting host countries with minimal cultural distance from the home country. This is supported by the finding that cultural distance between the home and host country increases the hazard rates – hence perceived risk – of foreign direct investments (Tsang and Yip, 2007).
Although it seems clear that cultural distance has a negative impact on FDI inflows, the importance of cultural distance on FDI location choice has been decreasing between 1980 and 2000 (Flores and Aguilera, 2007). Furthermore, for high-tech companies, cultural distance could be expected to be of less importance than for example the available human capital in the host country or economic characteristics, aiding R&D and sales. Hence, although cultural distance was of importance in the past, it is expected to have decreased in importance from 2000 to 2018. This is supported by Flores and Aguilera’s (2007) findings that, from 1980 to 2000, cultural distance’s importance to determine FDI decreased by 55%. Moreover, when increasing globalisation is taken as one of the main reasons behind a decreasing focus on cultural distance, then the trend of cultural distance seen in 1980-2000 can be expected to continue similarly in 2000-2018 given the rather stable increase in globalisation (Figure 1). This brings me to hypothesis 2:
Hypothesis 2: From 2000 to 2018, cultural distance is expected to have decreased in importance on FDI location choice of US MNEs3.
I
NSTITUTIONALD
ISTANCEInstitutional distance reflects the level of difference in institutional context between two countries (Salomon and Wu, 2012). Within this institutional context, North (1991) and Scott (1995) distinguished between formal and informal institutions. Formal institutions are explicit and enforceable, such as laws and rules. Informal institutions are comprised of less tangible social constructs, such as norms and values (Salomon and Wu, 2012). As informal institutions are less measurable within society and – when expressed as institutional distance – would contain some overlap with cultural distance, I focus on informal institutions only. Institutional distance will make it more difficult to do business abroad and hence has a negative influence on investments. It namely does not only reduce the likelihood that an MNE will invest in a certain host country, but also the level of the investment. This is due to institutional distance being associated with adaptation costs: adaptation costs increase with institutional distance, which raises the threshold for MNEs to undertake FDI. It also affects the firms’ profitability, limiting the volume of FDI invested by an MNE when there has been decided to invest in a certain host country (Cezar and Escobar, 2015).
Contrarily, institutional distance could also make a certain host country more attractive for FDI. This could be the case when the home country has an unfavourable business environment, for example including export or production restrictions. In this situation, investing in a country with a more favourable business environment would mean investing in a host country with high institutional distance from the home country.
Institutional host country characteristics are expected to have increased in importance when considering their influence on US MNEs’ FDI location choice. From 1980 to 2000, institutional host country characteristics overall became 17% more important in determining location choice (Table 1) (Flores and Aguilera, 2007). This was caused by a 39% increase in the importance of political institutions and a 24% decrease in the importance of the legal system. As the political institution increased relatively more in importance compared to the legal system’s decrease, the trend of increasing importance is expected to continue for 2000-2018. Furthermore,
institutional distance is also relevant for FDI (Cezar and Escobar, 2015). As there is an increased focus on institutional factors (Flores and Aguilera, 2007), institutional distance is expected to have gained importance when it comes to US MNEs’ FDI location choice. This brings me to hypotheses 3a and 3b:
Hypothesis 3a: From 2000 to 2018, the two institutional host country characteristics legal system and political institution are expected to on average have increased in importance on FDI location choice of US MNEs.
Hypothesis 3b: From 2000 to 2018, institutional distance is expected to have increased in importance on FDI location choice of US MNEs.
FIGURE 2. CONCEPTUAL MODEL 1, HYPOTHESIS 1A, 1B, 2, 3A AND 3B: THE CHANGE IN INFLUENCE OF SEVERAL DETERMINANTS ON US MNES’ FDI LOCATION CHOICE FROM 2000 TO 2018. A DOWNWARDS ARROW (↓) DEPICTS AN EXPECTED DECREASE IN THE IMPORTANCE OF THE DETERMINANT ON US FDI LOCATION CHOICE FROM 2000 TO 2018; AN UPWARDS ARROW (↑) DEPICTS AN EXPECTED INCREASE.
FDI
D
ETERMINANTS FROM1980-2000
TO2000-2018
institutional host country characteristics are more volatile than economic host country characteristics. I expect that high volatility indicates a higher likelihood that there is a large difference in the progression from 1980-2000 and the progression from 2000-2018. The influence cultural distance has on FDI is expected to have the smallest difference in change between 1980-2000 and 1980-2000-2018, since it is expected to have decreased at a similar rate in 1980-2000-2018 as it did in 1980-2000, as explained in the section on cultural distance. This results in hypothesis 4:
Hypothesis 4: The two institutional host country characteristics will on average see the largest difference between how much they progressed from 1980 to 2000 and 2000 to 2018, followed by the four economic host country characteristics and cultural distance.
FIGURE 3. CONCEPTUAL MODEL 2, HYPOTHESIS 4: THE RELATIVE DIFFERENCE THE THREE DETERMINANTS HAVE IN PROGRESSION FROM 2000-2018 COMPARED TO THE PROGRESSION FROM 1980-2000. THE THICKER THE ARROW, THE LARGER THE DIFFERENCE IS EXPECTED TO BE.
M
OSTI
NFLUENTIALD
ISTANCE IN2000-2018
Lastly, the question is which of the three types of distance – economic, cultural and institutional – was most influential on FDI in the 2000-2018 period. Cultural distance – although declining in importance between 1980 and 2000 (Flores and Aguilera, 2007) – is expected to still have a relatively large influence on US MNEs’ location choice in 2000-2018, and a larger influence than both economic and institutional distance. In 2000, it indeed decreased much compared to 1980, yet still had a significant negative influence on FDI. In line with leading morale amongst businesses and findings from existing research (Flores and Aguilera, 2007; Tsang and Yip, 2007; Beugelsdijk et al., 2018), cultural distance is expected to have a negative influence on FDI.
favourable business environment, more than a business environment similar to that of the US. Nevertheless, institutional distance is expected to have a negative influence on FDI. I expect the influence of institutional distance on US MNEs’ FDI location choice to be smaller than that of cultural distance, as I think cultural distance is more present in the MNEs considerations for foreign strategy and almost exclusively considered as a difficulty when doing business abroad. As explained, institutional distance can however also result into a more favourable business environment. Furthermore, as this research considers only formal institutions, these are political and legal systems of which the MNE can have full clarity beforehand. This entails that the MNE can prepare a strategy to cope with institutional differences whilst being well-informed, whereas cultural distance is based on less straight-forward behavioural constructs. Henceforth, I expect the risk perceived with regards to institutional distance to be smaller than that for cultural distance.
Economic distance is expected to be least important, considering that some MNEs decide upon investing in developing countries (high economic distance with US), whereas others decide upon investing in well-developed countries (low economic distance with US). These two compensate each other, due to which the final impact is expected to be rather low. Following Tsang and Yip’s (2007) conclusions, economic distance is expected to have a positive influence on FDI. This results in hypothesis 5:
Hypothesis 5: Cultural distance will have influenced US FDI location choice most in the 2000-2018 period, followed by institutional distance and economic distance.
M
ETHODOLOGY
D
ATADEPENDENT VARIABLE
The dependent variable is the US FDI Stock Outflows (in billion USD) as reported by the US Bureau of Economic Analysis. The scope of this research is from 2000 to 2018. All countries that received FDI from the US in 2000 are included for the 2000-2018 period. This has resulted in a sample of 180 countries worldwide (Table 6, Appendix A). However, for each year, the FDI outflow from the US for several countries is unavailable. Because of this, the number of countries included in the analysis for each year ranges from 146 to 167 countries. With 19 years of observation, the total dataset would be 3,420. For FDI, there is data available for a total of 2,944 host country-year pairs.
INDEPENDENT VARIABLES
ECONOMIC HOST COUNTRY CHARACTERISTICS AND DISTANCE
As the purpose of this research is to compare to Flores and Aguilera’s (2007) findings, I use the same variables as they did in their 2007 study as much as possible. A full overview of all variables can be found in Table 7 in Appendix A. There are four economic host country characteristics, also included in economic distance: (1) affluence will be measured through the GDP, (2) magnitude through the number of inhabitants, (3) physical infrastructure through the number of mobile cell subscriptions per 100 inhabitants and (4) wage as GDP per capita. Data on these variables is available in the World Bank’s database. Due to data being unavailable for some host countries and years, the number of observations for these host country characteristics is 2,823, 2,878, 2,837 and 2,823 respectively, out of the 2,944 host country-year pairs for which US FDI inflow is known. These four variables are combined and then compared to the US to compute economic distance (N = 2,781).
definitions and the CAGE framework, including GDP and GDP per capita is validated by previous research using this measure and also fall under the ‘differences in consumer incomes’ attribute of the CAGE framework. The number of mobile cell subscriptions was used by Flores and Aguilera (2007) to express infrastructure, which is also one of the attributes that Ghemawat (2001) identified as creating economic distance. Lastly, ‘quality of human resources’ in the CAGE framework could be broadly interpreted to constitute the number of inhabitants of a host country. Also, as the number of inhabitants displays the magnitude of a certain host country, it is a logical addition to the other three variables when comparing the economic environment of two countries. Hence, considering previous research and literature on economic distance by various scholars, the four host country characteristics included to express economic distance are believed to be a good representation of the differences in economic environment between two countries.
INSTITUTIONAL HOST COUNTRY CHARACTERISTICS AND DISTANCE
Institutional distance is determined by allocating countries scores on the variables political institution and legal system. Political institution is determined by making use of a democracy score (Varieties of Democracy, 2020). The democracy score is not available for all countries and years (N = 2,576). Legal system is a dummy variable where 1 means that the country has the same legal system as the US (Common Law), 0 being another legal system. Whether the country uses common law or not is available for the full dataset (N = 2,944) via the CIA World Fact Book. These two variables are combined and compared to the US to compute institutional distance (N = 2,576).
CULTURAL DISTANCE
Lastly, cultural distance is calculated using the Kogut and Singh (1988) Index, which determines cultural distance by making use of Hofstede’s cultural dimensions (Hofstede-Insights). For most countries, scores on all six dimensions are included in the analysis: Power Distance, Individualism, Masculinity, Uncertainty Avoidance, Long-Term Orientation and Indulgence. For several countries, scores are only available for a number of cultural dimensions. There are 65 countries for which there is no data available on any of the cultural dimensions, hence no cultural distance from the US could be computed. This results in a number of observations of 2,000 for cultural distance.
CONTROL VARIABLES ENGLISH LANGUAGE
was also included in Flores and Aguilera’s study (2007). The reason for including specifically whether the official language of the host country is English is because of criticisms on the measure of cultural distance through Hofstede’s dimensions and Kogut & Singh’s index. The criticism is that the official language of a host country would also significantly influence cultural distance (Shenkar, 2012). Furthermore, investing in a country with the same official language as the US might be perceived as less risky by US MNEs than when the host country’s official language is not English. Henceforth, whether the official language of the host country is English is controlled for when assessing the influence of economic, institutional and cultural determinants on FDI location choice of US MNEs. Whether English is the official language of the host country is known for all 180 countries in the dataset throughout the whole 2000-2018 period (N = 2,944).
GEOGRAPHIC DISTANCE
Secondly, the geographic distance is measured in miles between the US and the host country. This can be found in Great Circle Distances Between Capital Cities (Eden, 2006), which measures the distance between the capital of the US (Washington D.C.) and the capital of the host country. This control variable also was included in Flores and Aguilera’s (2007) research. Reason for including geographic distance is because it is often confused with cultural distance and may also be the determinant of a host country’s attractiveness, rather than cultural distance itself (Shenkar, 2012). Moreover, I would expect that geographic distance influences US MNEs’ location choice, as investing in a country further away could be considered riskier. For example, because the perception is that a geographically distant country also differs more in terms of business environment. The geographic is available for all countries but one, resulting in a number of observations of 2,921.
P
REPARING THED
ATASET DEPENDENT VARIABLEINDEPENDENT VARIABLES ECONOMIC DISTANCE
Economic distance includes the four host country characteristics as described earlier. First, the log is taken for the values GDP and number of inhabitants, as these are relatively large values. Then, all four variables are standardized. This is necessary, as the four variables need to be expressed on a similar scale to be combined into economic distance. For each country in each year, the average of these four Z values is then taken, which represents the economic score for that country in that year. Consecutively, for each year, the economic score of each country is compared to the economic score of the USA in that same year. The absolute difference between the two is the economic distance between the host country and the USA in that year.
CULTURAL DISTANCE
Equation 1 calculates the cultural distance between the home country (in this case: the US) and a host country, based on Hofstede’s cultural dimensions (Hofstede-insights). In this formula, N is the number of cultural dimensions for which information is available for that host country. It uses the absolute difference between the score on each cultural dimension for the US and the host country and the variance throughout the whole dataset to come to a cultural distance score for each host country per dimension. Then, the average of the dimensions is taken to display the overall cultural distance between the US and the host country. To ensure that it can be compared to the other variables in the dataset, the cultural distance value is then standardized. Hofstede’s cultural dimensions scores have been calculated in the 1970s and have not been updated since. However, as cultural distance is stable over time (Beugelsdijk et al., 2015), this is not an issue for this study and the cultural distance score between the US and the host countries is the same for all years. 𝐶𝐷𝑗= ∑ {(𝐼𝑖𝑗− 𝐼𝑖𝑢) 2 ∕ 𝑉𝑖} 𝑁 𝑖=1 /𝑁
EQUATION 1. CULTURAL DISTANCE (KOGUT AND SINGH, 1988)
INSTITUTIONAL DISTANCE
year, the absolute difference between their institutional score and the USA’s score represents the institutional distance.
CONTROL VARIABLES
The two control variables are geographic distance and whether English is an official language. As the geographic distance contains relatively large values compared to the other variables, the log is used. To ensure similar measurement of the influence of these values on FDI location choice as for the other variables, these two values have also been standardized.
MULTICOLLINEARITY
As there are multiple variables expressing a similar subject – i.e. four host country characteristics to assess the economic environment and two for the institutional environment – it is important to check for multicollinearity. Table 9 in Appendix B displays the correlations between all independent- and control variables. Assuming multicollinearity from a correlation of 0.7 onwards, the variables Common Law and C1: English show signs of multicollinearity (r = 0.74). To confirm whether there is multicollinearity, the Variance Inflation Factor (VIF) is used. The rule of thumb is that VIF values of lower than 10 are acceptable (Gómez et al., 2010). However, it should be taken into account that the VIF is not resistant to outliers. Although standardized, the values that were prone to outliers in the raw data due to large differences between countries still contained outliers after standardization. The VIF is <10 for all host country characteristics as well as all distances (Table 12 and 13, Appendix B). For Common Law and C1: English, it is 2.06 and 2,05 respectively, which indicates there is no multicollinearity. As all other correlation values are lower than 0.7 and all VIF values are lower than 10, there is no multicollinearity for which measures would need to be taken.
A
NALYSESFor this study, I ran eight different analyses to address the research question. The main mode of analysis – six of the models – is linear regression for models on one year only (2000, 2008, and 2018). For the panel data (period 2000-2018), linear mixed models are used with year as fixed effect, resulting into two models. In each analysis, FDI is the dependent variable and – next to the independent variables – the two control variables are included.
5, 6, and 7 in Table 4). Although my research focuses mainly on the changes from 2000 to 2018, the year 2008 is included as mid-point to further understand the trend of the influence these host country characteristics and distances have on FDI. Tables 3 and 5 display the change in influence of the significant determinants on FDI between 2000, 2008 and 2018.
R
ESULTS
H
OSTC
OUNTRYC
HARACTERISTICSDependent variable: FDI Inflow from the US (billion GDP)
Model 1 2000 Model 2 2008 Model 3 2018 Model 4 2000-2018 GDP (billion USD) 0.890*** (0.131) 0.621*** (0.172) 0.786*** (0.136) 0.748*** (0.039) Population (million) -0.157 (0.105) 0.106 (0.134) -0.033 (0.111) -0.003 (0.031)
Mobile cell subscriptions per 100 inhabitants -0.081 (0.201) 0.231* (0.096) 0.075 (0.070) 0.118*** (0.024)
GDP per capita (USD) 0.116 (0.161) 0.172* (0.073) 0.211*** (0.062) 0.192*** (0.020) Democracy Score 0.025 (0.061) 0.007 (0.060) -0.025 (0.053) -0.004 (0.014)
Common Law (yes/no) 0.082 (0.063) 0.033 (0.070) -0.025 (0.058) 0.078*** (0.017) Cultural Distance 0.077 (0.047) 0.038 (0.051) 0.020 (0.042) 0.036*** (0.012) C1: English Language (yes/no) 0.050 (0.065) 0.145** (0.070) 0.165*** (0.060) 0.080*** (0.017) C2: Geographic Distance (miles) -0.251*** (0.076) -0.255*** (0.073) -0.208*** (0.064) -0.246*** (0.019) N 105 100 97 1,944 R2 0.761 0.784 0.825 0.726 Adjusted-R2 0.739 0.762 0.807 0.722
TABLE 2. REGRESSION COEFFICIENTS (STANDARD DEVIATION). CONFIDENCE LEVELS: * = 10%, ** = 5%, *** = 1%.
ceteris paribus and significant at the 1% level. Furthermore, the control variable geographic distance also showed to have a significant influence on FDI. Here, ceteris paribus and significant at the 1% level, if the geographic distance between the US’ capital and the host country’s capital would be 1 larger, then FDI inflow from the US into that host country decreased by 0.251 percentage points.
The influence the host country characteristics and cultural distance had on US MNEs’ FDI location choice in 2008 is displayed in Model 2. Just as in 2000, GDP and the control geographic distance had a significant influence on FDI. Additionally, the number of mobile cell subscriptions per 100 inhabitants, GDP per capita and the control whether English is an official language also were of significant influence on US FDI, in descending order of their level of influence on FDI. The exact influence of each of these determinants on FDI and their significance can be found in Table 2.
In 2018 (Model 3), the same host country characteristics as in 2008 were of significant influence on FDI, except that the number of mobile cell subscriptions per 100 inhabitants was not of significant influence anymore. GDP had the largest influence on FDI, followed by GDP per capita, geographic distance and whether English is an official language. The exact influence of each of these determinants on FDI and their significance can be found in Table 2.
When comparing Model 1 to Model 3, hence 2000 to 2018, multiple host country characteristics had a significant influence on FDI location choice of US MNEs in 2018, whereas only GDP and geographic distance were of significant influence in 2000. Model 3 can also be considered the best model as the R2 is the highest for this model. Hence, the host country
comment on Hypotheses 2 and 3a. Lastly, I expected institutional host country characteristics to have changed most in how much it progressed from 2000-2018 compared to 1980-2000 (Hypothesis 4). However, as economic characteristics is the only variable which significantly influenced FDI in 2000 and 2018, I can neither accept nor reject this hypothesis.
Model 4 analyses the impact the host country characteristics had on the whole 2000-2018 period. Here, we see that all host country characteristics except for population and the democracy score had a significant influence on FDI inflow, ceteris paribus and significant at the 1% level. GDP, mobile cell subscriptions, GDP per capita, common law, cultural distance and whether English is an official language all had a positive influence on FDI, whereas the geographic distance had a negative influence on FDI. The exact influence of each indicator on FDI for the 2000-2018 period is displayed in Table 2 above and in Figure 7 below.
2000 Change 2008 Change 2018 2000 – 2018 change GDP 1 -30.22% 1 +26.57% 1 -11.69% C2: Geographic Distance 2 -1.59% 2 -18.43% 3 -17.13% Mobile Cell Subscriptions
Not significant N/a 3 N/a Not significant N/a
GDP per Capita Not significant N/a 4 +22.67% 2 N/a
C1: English Not significant N/a 5 +13.79% 4 N/a
TABLE 3. RANKING AND CHANGE OF IMPORTANCE OF THE HOST COUNTRY CHARACTERISTICS ON US FDI LOCATION CHOICE IN 2000-2018.
FIGURE 5. ABSOLUTE INFLUENCE THE SIGNIFICANT HOST COUNTRY CHARACTERISTICS HAD ON US MNE FDI INFLOW.
0 0.2 0.4 0.6 0.8 1 2000 2008 2018 Re gre ss io n C oe ff ic ie nt Year
Absolute Influence of Host Country Characteristics on US FDI
GDP GDP per Capita Mobile Cell Subscriptions
FIGURE 6. CHANGE OF THE IMPORTANCE OF EACH SIGNIFICANT INDICATOR ON US MNE FDI.
FIGURE 7. ABSOLUTE INFLUENCE OF THE SIGNIFICANT HOST COUNTRY CHARACTERISTICS ON US FDI INFLOW IN THE 2000-2018 PERIOD.
INTERPRETATION OF RESULTS
PROGRESSION OF ECONOMIC AND INSTITUTIONAL HOST COUNTRY CHARACTERISTICS AND CULTURAL DISTANCE ON USMNES’FDI LOCATION CHOICE IN 2000-2018
GDP is the only indicator which had a significant influence on FDI in all years of analysis (2000, 2008, 2018). Overall, from 2000 to 2018, the influence GDP had on FDI location choice for US-based companies decreased by 11.69%. Although GDP per capita did not have a significant influence on FDI in 2000, it did in both 2008 and 2018 and gained 22.67% in influence on FDI from US MNEs from 2008 to 2018. Lastly, looking at the whole 2000-2018 period, I found that population and the democracy score did not influence FDI, and that the number of mobile cell subscriptions, GDP per capita, the legal system and cultural distance positively influenced FDI.
-40.00% -30.00% -20.00% -10.00% 0.00% 10.00% 20.00% 30.00% 2000-2008 2008-2018 2000-2018 P er ce ntage C ha nge Period in Years
Change of Influence of each Indicator on FDI
GDP C2: Geographic Distance GDP per Capita C1: English
0 0.2 0.4 0.6 0.8 Indicator Re gre ss io n C oe ff ic ie nt
Absolute Influence of Significant Host Country Characteristics on
FDI, 2000-2018
GDP GDP per Capita Mobile Cell Subscriptions C1: English
COMPARISON BETWEEN 2000-2018 AND 1980-2000 AS FOUND BY FLORES AND
AGUILERA (2007)
D
ISTANCESDependent variable: FDI Inflow from the US (billion GDP)
Model 5 2000 Model 6 2008 Model 7 2018 Model 8 2000-2018 Economic Distance -1.107*** (0.171) -1.008*** (0.116) -0.607*** (0.088) -0.867*** (0.028) Cultural Distance 0.029 (0.069) -0.030 (0.068) -0.046 (0.071) -0.037** (0.017) Institutional Distance -0.287* (0.147) -0.162 (0.143) -0.214 (0.155) -0.276*** (0.035) C1: English Language (yes/no) (Z) -0.010 (0.076) 0.043 (0.074) 0.059 (0.080) 0.014 (0.019) C2: Geographic Distance (miles) (Z) -0.309*** (0.108) -0.249** (0.102) -0.229** (0.111) -0.247*** (0.026) N 106 101 103 1,968 R2 0.437 0.548 0.379 0.434 Adjusted-R2 0.409 0.524 0.347 0.428
TABLE 4. REGRESSION COEFFICIENTS (STANDARD DEVIATION). CONFIDENCE LEVELS: * = 10%, ** = 5%, *** = 1%.
Table 4 contains the regression results of the influence economic, cultural and institutional distance and the control variables had on US MNEs’ FDI in 2000, 2008, 2018 and the 2000-2018 period. In 2000 (Model 5), economic distance, the control geographic distance and institutional distance had a significant influence on FDI, in descending order of influence. For economic distance, a 1 increase in this value would lead to a 1.107 percentage point decrease in FDI inflow. A 1 increase in geographic distance would have led to a 0.309 percentage points decrease in FDI, both assuming ceteris paribus and significant at the 1% level. Lastly, a 1 increase in the institutional distance between the US and a host country resulted in a 0.287 percentage points decrease in FDI from the US into that host country, ceteris paribus and significant at the 10% level.
The regression coefficients for 2018 are displayed in Model 7. In 2018, the same distances had a significant influence on FDI as in 2008, however at different levels. Economic distance still influenced FDI most, however decreased relatively much in importance compared to ten years earlier (Table 5). Furthermore, geographic distance also had a significant influence on FDI but also decreased in importance: a 1 increase resulted in a 0.229 percentage points decrease in FDI.
Economic distance and geographic distance had a significant influence on US MNEs’ FDI location choice in both 2000 and 2018. As displayed in Table 4 and 5, the influence economic distance had on FDI remained negative for both years, but almost halved in importance on US MNEs determining FDI. The same holds for geographic distance, yet this distance became around 25% less important on determining FDI. Although economic distance overall decreased most in its importance on FDI, it remained the most important determinant to influence FDI in all years and throughout the whole 2000-2018 period. The importance of each significant distance for each year and the changes in importance are displayed in Figures 8-10 below. In Hypothesis 1b, I hypothesized that economic distance will have decreased in importance on FDI, which these results have confirmed. Furthermore, I expected the influence of institutional distance on FDI to have increased over the years (Hypothesis 3b). However, as this distance only had a significant effect on US FDI in 2000, this seems to not be the case.
Lastly, Model 8 displays the impact each distance had on FDI for the whole 2000-2018 period. Here, all three distances and the control geographic distance had a significant and negative influence on FDI, ceteris paribus and significant at the 1% level. Economic distance was most important, followed by institutional distance, geographic distance and lastly cultural distance. The regression coefficients for each of these variables can be found in Table 4 above. In Hypothesis 5, I hypothesized that cultural distance influenced FDI most in the 2000-2018 period, followed by institutional and economic distance. However, economic distance was most influential on FDI in this period and cultural distance the least.
2000 Change 2008 Change 2018 2000 – 2018 change Economic Distance 1 -8.94% 1 -39.78% 1 -45.17% C2: Geographic distance 2 -19.42% 2 -8.03% 2 -25.89% Institutional Distance
3 N/a Not significant N/a Not significant N/a
FIGURE 8. ABSOLUTE INFLUENCE OF EACH SIGNIFICANT DISTANCE ON US FDI INFLOW IN THE 2000-2018 PERIOD.
FIGURE 9. CHANGE OF THE IMPORTANCE EACH SIGNIFICANT DISTANCE HAD ON US MNE FDI INFLOW.
FIGURE 10. ABSOLUTE INFLUENCE THE SIGNIFICANT DISTANCES HAD ON US MNE FDI INFLOW.
0 0.2 0.4 0.6 0.8 1 1.2 2000 2008 2018 Re gre ss io n C oe ff ic ie nt Year
Absolute Influence of Distances on US FDI
Economic Distance Institutional Distance C2: Geographic Distance
-50.00% -40.00% -30.00% -20.00% -10.00% 0.00% 2000-2008 2008-2018 2000-2018 P er ce ntage C ha nge Period in Years
Change of Influence of each Distance on FDI
Economic Distance C2: Geographic Distance
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Distance Re gre ss io n C oe ff ic ie nt
Absolute Influence of Significant Distances on FDI, 2000-2018
Economic Distance Cultural Distance Institutional Distance C2: Geographic Distance
INTERPRETATION OF RESULTS
INFLUENCE OF ECONOMIC, CULTURAL AND INSTITUTIONAL DISTANCE ON USMNES’
FDI LOCATION CHOICE IN 2000-2018
My analysis has shown that in the whole 2000-2018 period, economic, cultural and institutional distance negatively influenced FDI location choice of US MNEs. Economic distance had by far the biggest influence on FDI (r = -0.867***), followed by institutional distance (r = -0.276***). Cultural distance only had a relatively small influence on FDI in this period (r = -0.037**). This makes sense, as cultural distance also did not have a significant influence in the individual years 2000, 2008 or 2018, whereas economic distance did. Institutional distance was only significant in 2000. In the most recent year (2018) economic distance is hence the only of the three distances to have a significant impact on FDI. Considering the three years in the dataset separately –2000, 2008 and 2018 – the impact of economic distance on FDI has been continuously decreasing: while this was a relatively small decrease of -8.94% from 2000-2008, this increased to -39.78% from 2008 to 2018. The same holds for geographic distance, whereas the speed of decrease for this distance became less rapid over time (-19.42% from 2000-2008 and -8.03 from 2008-2018).
C
ONCLUSION
This research addressed the question How has the importance of economic, cultural
and institutional host country characteristics and distances on FDI location choice of US-based multinationals evolved from 2000 to 2018, and how does this compare to location choice determinants over the two decades before 2000? by looking at US MNEs’ FDI location
span the full 2000-2018 period and include 180 potential host countries. Following is an answer to the research question, based on the results and the interpretation thereof.
H
OSTC
OUNTRYC
HARACTERISTICSOverall, I found that GDP was the only host country characteristic which had a significant influence on US MNEs’ FDI location choice in both 2000 and 2018. From 2000 to 2018, the influence that GDP had on FDI decreased with 12%, compared to a 50% decrease from 1980 to 2000. It can hence be concluded that the influence of GDP on US MNEs’ location choice is decreasing over time. Furthermore, I found that GDP per capita was of no significant influence on FDI in 2000, but did become significant in 2008 and increased in importance for FDI towards 2018. Moreover, while population and democracy influenced FDI in 1980-2000, they do not anymore in 2000-2018. As they also decreased in importance from 1980-2000, this indicates that these two host country characteristics saw a continuous decrease in importance on FDI from 1980 onwards. Furthermore, the influence that the legal system has on US FDI has been fluctuating from 1980-2018.
GDP per capita increasingly influencing FDI is not in line with Dunning’s (1998) explanation of companies’ strategies shifting focus towards the size of the population rather than the income of the population. Simultaneously, population not influencing FDI in 2000-2018 directly contrasts this theory as well. As GDP also had a significant impact on FDI, it seems that – at least in the 2000-2018 period – total market affluence or purchasing power of a country is more important to US MNEs than the number of inhabitants, rejecting Dunning’s (1998) theory.
D
ISTANCESRegarding economic, cultural and institutional distance, I found that all three had a significant impact on FDI over the whole 2000-2018 period, with economic distance having the largest influence and cultural distance the smallest. Economic distance became 45% less important on determining FDI from 2000 to 2018. While institutional distance was of significant influence on FDI in 2000, it was not anymore in 2018. Cultural distance has not been significant in any of the three years separately.
With the positive influence of GDP in 2000-2018 and economic distance having a negative impact on FDI, it seems as if US companies are not as much investing in developing or low-wage countries, but more in countries with a relatively strong market, expectably to exploit the well-developed market and gain higher profits. My expectation is that the low importance of cultural distance on FDI is attributable to increasing globalization. This causes countries to look more homogenous which decreases the difficulty of doing business, even when cultural distance still exists. These results are particularly interesting as cultural distance is often considered to be one of the most present and important distances. However, my research has now shown that – for the 2000-2018 period – cultural distance is of significantly less importance than economic or institutional distance for US MNEs when they are deciding in which country to invest.
D
ISCUSSION
L
IMITATIONSAs with most research, there are several limitations to my analysis that may have negatively impacted the validity of the results. Firstly, for this study, I have used FDI outflows data for the US as a whole rather than per company. Flores and Aguilera (2007), however, used FDI outflows data of US companies present in the Fortune 500, to which I did not have access. As these companies would also make up the biggest part of the total US FDI outflows, I do not consider this to significantly impact the validity of the results. The results considering solely the extent to which each indicator or distance influenced US FDI location choice in the 2000-2018 period are not impacted by this, as these results are not directly linked to Flores and Aguilera’s (2007) research. However, this did result in a smaller sample size, as now for each year and country there was only one FDI outflow (from the US as a whole), compared to multiple outflows from different US companies in Flores and Aguilera’s (2007) research. This can be an explanation as to why certain host country characteristics or distances have shown to not have a significant impact in a certain year, whereas this goes against expectation. I attempted to investigate whether this has indeed impacted the results, yet due to data being unavailable for the 1980-2000 period I have not been able to do so.
and the fact that cultural distance is already computed through multiple variables (Hofstede’s dimensions), the variable “trust” has been excluded.
A
VENUES FORF
URTHERR
ESEARCHBased on the new insights into location choice this paper suggests, there are two avenues for further research that I would like to offer. Firstly, my research only looks at multinationals which are based in the US. Hence, the conclusions drawn based upon my results are expected to differ for MNEs from other parts of the world. For example, I would expect economic host country characteristics and economic distance to have a different impact on FDI location choice of MNEs originating from a developing country, as my research focussed only on FDI originating from a developed economy. Furthermore, whereas for example Indian MNEs are strongly considering cultural distance and political risk in their FDI location choice, Chinese MNEs do so much less (Quer et al., 2017). However, I found that the democracy score does not influence US MNEs’ location choice. This is another example of how the influence of several location choice determinants on FDI included in my research vary across countries. Hence, one avenue for further research I would like to suggest is to replicate this research, yet for MNEs from another region. As regions I suggest (a part of) Europe or Asia, as these both are regions where a number of MNEs are established, whilst being geographically distant from the US. Alternatively, it would be interesting to consider MNEs from Canada or South-America, to understand whether MNEs from areas which are geographically less distant to the US also show similar location choice patterns as US MNEs.
L
ITERATURE
Aguilera-Caracuel, J., Fedriani, E., Delgado-Márguez, D. 2014. Institutional distance among country influences and environmental performance standardization in multinational enterprises, Journal of Business Research, 67: 2385-2392.
Batsakis, G., Singh, S. 2019. Added distance, entry mode choice, and the moderating effect of experience: The case of British MNEs in emerging markets, Thunderbird International Business Review, 61: 581-594.
Beugelsdijk, S., Kostova, T., Kunst, V., Spadafora, E., van Essen, M. 2018. Cultural Distance and Firm Internationalization: A Meta-Analytical Review and Theoretical Implications,
Journal of Management, 44(1): 89-130.
Beugelsdijk, S., Maseland, R., van Hoorn, A. 2015. Are scores on Hofstede’s dimensions of national culture stable over time? A cohort analysis, Global Strategy Journal, 5(3): 223-240.
Buckley, P., Devinney, T. & Louviere, J. 2007. Do managers behave the way theory suggests? A choice-theoretic examination of foreign direct investment location decision-making,
Journal of International Business Studies, 38: 1069-1094.
Buckley, P., Ghauri, P. 2005. Globalisation, Economic Geography and the Strategy of Multinational Enterprises, Journal of International Business Studies, 34(2): 81-98.
Cantwell, J. 2009. Location and the multinational enterprise, Journal of International
Business Studies, 40: 35-41.
Cezar, R., Escobar, O. 2015. Institutional distance and foreign direct investment, Review of World Economics, 151(4): 713-733.
Choi, C. 2003. Does the Internet stimulate inward foreign direct investment?, Journal of Policy Modeling, 25(4): 319-326.
CIA World Fact Book. Field listing: Languages. Accessed online
CIA World Fact Book. Field listing: Legal system. Accessed online
https://www.cia.gov/library/publications/the-world-factbook/fields/308.html. Viewed April 21, 2020.
Dunning, J. 1998. Location and the multinational enterprise: a neglected indicator?,
Journal of International Business Studies, 29(1): 45-66.
Eden, L. 2006. Great Circle Distances Between Capital Cities. Accessed online
http://www.chemical-ecology.net/java/capitals.htm. Viewed April 24, 2020.
Eurostat, 2020. Internet use and activities. Accessed online https://appsso.eurostat.ec.europa.eu/nui/submitViewTableAction.do. Viewed April 29, 2020.
Flores, R. G. & Aguilera, R. V. 2007. Globalisation and location choice: an analysis of US multinational firms in 1980 and 2000, Journal of International Business Studies, 38: 1187-1210.
Fortune. 2019. Global 500, Accessed online
https://fortune.com/global500/2019/search/?hqcountry=U.S.. Viewed April 14, 2020. Ghemawat, P. 2001. Distance Still Matters, Harvard Business Review, 79(8): 1-11.
Gomez, S., García, G., García Pérez, J., López Martín, M. 2016. Collinearity diagnostic applied in ridge estimation through the variance inflation factor, Journal of Applied Statistics,
43(10): 1831-1849.
Hofstede, G. 1983. Cultural dimensions for project management, Journal of Project Management, 1(1): 41-48.
Hofstede, G., Bond, M. 1988. The Confucius connection: From cultural roots to economic growth, Organization Dynamics, 16(4): 5-21.
Hofstede, G., Hofstede, GJ., Minkov, M. 2010. Cultures and organizations: Software of the mind: Intercultural cooperation and its acceptance for survival. McGraw-Hill.
Hofstede-Insights. Dimensions of National Culture, Accessed online
https://hi.hofstede-insights.com/national-culture. Viewed April 22, 2020.
https://www.ilo.org/shinyapps/bulkexplorer32/?lang=en&segment=indicator&id=EAR_MRE E_NOC_GR_A. Viewed April 15, 2020.
Kogut, B., Singh, H. 1988. The Effect of National Culture on the Choice of Entry Mode,
Journal of International Business Studies, 19(3): 411-432.
Maseland, B., Dow, D., Steel, P. 2018. The Kogut and Singh national cultural distance index: Time to start using it as a springboard rather than a crutch, Journal of International Business Studies, 59: 1154-1166.
Nachum, L., Zaheer, S. & Gross, S. 2008. Does it matter where countries are? Proximity to knowledge, markets and resources, and MNE location choices, Management Science, 54(7):
1252-1265.
North, D. 1991. Institutions. Journal of Economic Perspectives, 5(1): 97-112.
OECD. 2020. Outward FDI flows by partner country (indicator), Accessed online
https://data.oecd.org/fdi/outward-fdi-flows-by-partner-country.htm. Viewed April 21, 2020.
OECD. Average annual wages. Accessed online
https://stats.oecd.org/Index.aspx?DataSetCode=AV_AN_WAGE. Viewed April 15, 2020. Palmero, A., Herrera, J., de la Fuente Sabaté, J. 2013. The role of psychic distance stimuli on the East-West FDI location structure in the EU. Evidence from Spanish MNEs, Journal of East European Management Studies, 18(1): 36-65.
Pogrebnyakov, N., Maitland, F. 2011. Institutional distance and the internationalization process: The case of mobile operators, Journal of International Management, 17(1): 68-82.
Quer, D., Claver, E., Rienda, L. 2017. Cultural distance, political risk and location decisions of emerging-market multinationals: a comparison between Chinese and Indian Firms, Journal of the Asia Pacific Economy, 22(4): 587-603.
Salomon, R., Wu., Z. 2012. Institutional distance and local isomorphism strategy, Journal of International Business Studies, 43(4): 343-367.
Shenkar, O. 2012. Cultural distance revisited: Towards a more rigorous conceptualization and measurement of cultural differences, Journal of International Business Studies, 43(1):
1-11.
Smigiera, M. 2019. Countries with highest direct investment position received from the United States in 2018, Statista. Accessed online
https://www.statista.com/statistics/188806/top-15-countries-for-united-states-direct-investments/. Viewed April 12, 2020.
Stahl, G., Tung, R., Kostova, T., Zellmer-Bruhn, M. 2016. Widening the lens: Rethinking distance, diversity, and foreignness in international business research through positive organizational scholarship, Journal of International Business Studies, 47(1):621-630.
The World Bank. GDP (current US$). Accessed online
https://data.worldbank.org/indicator/NY.GDP.MKTP.CD. Viewed April 21, 2020.
The World Bank. GDP growth (annual %). Accessed online
https://data.worldbank.org/indicator/NY.GDP.MKTP.KD.ZG?end=2018&start=1961&view =chart&year=2018. Viewed April 21, 2020.
The World Bank. GDP per capita (current US$). Accessed online
https://data.worldbank.org/indicator/ny.gdp.pcap.cd. Viewed April 21, 2020.
The World Bank. Mobile cellular subscriptions (per 100 people). Accessed online
https://data.worldbank.org/indicator/IT.CEL.SETS.P2. Viewed April 21, 2020.
The World Bank. World Development Indicators. Accessed online
https://databank.worldbank.org/source/world-development-host country
characteristics/Type/TABLE/preview/on. Viewed April 18, 2020.
Tsang, E., Yip, P. 2007. Economic Distance and the Survival of Foreign Direct Investments, Academy of Management Journal, 50(5): 1156-1168.
Varieties of Democracy. 2020. V-Dem Dataset – Version 10. Accessed online
A
PPENDICES
A
PPENDIXA.
O
VERVIEW OFD
ATAAlbania Djibouti Latvia Russia Algeria Dominica Lebanon Samoa Andorra Dominican Republic Lesotho Saudi Arabia Angola Ecuador Liberia Senegal Anguilla Egypt Libya Serbia Antigua and Barbuda El Salvador Liechtenstein Seychelles Argentina Equatorial Guinea Lithuania Sierra Leone Armenia Eritrea Luxembourg Singapore Aruba Estonia Macau Slovakia Australia Ethiopia Macedonia Slovenia
Austria Fiji Madagascar Solomon Islands Azerbaijan Finland Malawi Somalia
Bahamas France Malaysia South Africa Bahrain French Guiana Mali Spain Bangladesh Gabon Malta Sri Lanka
Barbados Georgia Marshall Islands St. Kitts and Nevis Belarus Germany Mauritania St. Lucia
Belgium Ghana Mauritius St. Vincent and the Grenadines Belize Gibraltar Mexico Sudan
Bermuda Greece Micronesia Suriname Bhutan Greenland Moldova Swaziland Bolivia Grenada Morocco Sweden Bosnia and Herzegovina Guatemala Mozambique Switzerland Botswana Guinea Namibia Syria Brazil Guyana Nauru Taiwan Brunei Haiti Nepal Tanzania Bulgaria Honduras Netherlands Thailand Burkina Faso Hong Kong Netherlands Antilles Togo Burma Hungary New Zealand Tonga
Cambodia Iceland Nicaragua Trinidad and Tobago Cameroon India Niger Tunisia
Canada Indonesia Nigeria Turkey Central African Republic Iran Norway Turkmenistan Chile Ireland Oman Uganda China Israel Pakistan Ukraine
Colombia Italy Palau United Arab Emirates Congo (Brazzaville) Jamaica Panama United Kingdom Congo (Kinshasa) Japan Papua New Guinea Uruguay
Costa Rica Jordan Paraguay Uzbekistan Cote D'Ivoire Kazakhstan Peru Vanuatu Croatia Kenya Philippines Venezuela Cuba Korea, Republic of Poland Vietnam Cyprus Kuwait Portugal Yemen Czech Republic Kyrgyzstan Qatar Zambia Denmark Laos Romania Zimbabwe
Measurement Variable Source
Dependent Variable
FDI FDI stock inflow from the US (billion USD)
US Bureau of Economic Analysis
Economic Affluence GDP (billion USD) World Bank
Economic Magnitude Number of inhabitants (millions) World Bank
Economic Physical infrastructure Mobile cell subscriptions per 100 inhabitants
World Bank
Economic Wage GDP per capita (USD) World Bank
Cultural Power Distance Hofstede country score Hofstede Insights
Cultural Individualism Hofstede country score Hofstede Insights
Cultural Masculinity Hofstede country score Hofstede Insights
Cultural Uncertainty Avoidance Hofstede country score Hofstede Insights
Cultural Long-Term Orientation Hofstede country score Hofstede Insights
Cultural Indulgence Hofstede country score Hofstede Insights
Institutional Political institution Democracy score Varieties of Democracy
Institutional Legal system Common law or not – Dummy CIA World Fact Book
Control Variable
Official Language English or not - Dummy CIA World Fact Book
Control Variable
Geographical Distance Distance from capital host country to capital US (miles)
Great Circle Distances Between Capitals (Eden, 2006)
TABLE 7. OVERVIEW OF VARIABLES INCLUDED IN THE ANALYSIS.
Variable Flores and Aguilera (2007) Variable this research
FDI outflows of US companies in Fortune 500 FDI outflows of US as a whole Average wage manufacturing workers 40h/week GDP per capita
Phonelines per 1000 inhabitants Mobile cell subscriptions per 100 inhabitants Democracy: yes/no Democracy Score
Trust Level Excluded
A
PPENDIXB.
R
ESULTS 1 2 3 4 5 6 7 8 9 10 11 1. FDI 2. GDP 0.33*** 3. Inhabitants 0.04* 0.57*** 4. Mobile cell subscriptions per 100 inhabitants 0.18*** 0.15*** -0.06** 5. GDP per Capita 0.46*** 0.20*** -0.08*** 0.46*** 6. Democracy Score 0.29*** 0.18*** -0.06*** 0.35*** 0.50*** 7. Common Law 0.12*** -0.07*** 0.02 -0.09*** -0.04* 0.05** 8. Cultural Distance -0.27** -0.04* 0.06** 0.00 -0.29*** -0.29*** -0.26*** 9. Economic Distance -0.09*** -0.31*** -0.28*** 0.34*** -0.18*** -0.18*** 0.04* 0.05* 10. Institutional Distance -0.30*** -0.14*** 0.03 -0.24*** -0.43*** -0.43*** -0.40*** 0.35*** 0.15*** 11. Control 1: English 0.11*** -0.07*** 0.01 -0.08*** -0.05** -0.05** 0.74*** -0.29*** 0.05** -0.28*** 12. Control 2: Geographic Distance -0.02 -0.02 -0.02 -0.05** -0.04* -0.04* 0.12*** 0.01 0.02 0.25*** 0.12*** TABLE 9. CORRELATIONSN Min. Max. Mean Std.
Deviation Statistic Std. Error Skewness Statistic Std. Error Kurtosis FDI 2,944 -3.81 2.01 0.00 1.00 -0.38 0.05 -0.48 0.09 GDP 2,823 -2.43 2.63 0.09 0.97 -0.01 0.05 -0.52 0.09 Inhabitants 2,878 -2.49 2.85 -0.06 0.97 0.55 0.05 0.27 0.09 Mobile Cell Subscriptions per 100 Inhabitants 2,837 -1.49 5.38 0.03 1.00 0.27 0.05 0.14 0.09 GDP per Capita 2,823 -0.67 8.17 0.03 1.03 2.73 0.05 10.45 0.09 Democracy Score 2,576 -1.61 1.78 0.06 0.99 0.06 0.05 -1.29 0.10 Common Law 2,944 -0.62 1.62 -0.03 0.98 1.08 0.05 -0.83 0.09 Economic Distance 2,781 0.00 1.75 0.76 0.41 -0.14 0.05 -1.00 0.09 Cultural Distance 2,000 -1.87 3.47 0.01 1.01 0.44 0.06 0.52 0.11 Institutional Distance 2,576 0.00 2.72 1.52 0.68 -0.27 0.05 -0.79 0.10 C1: English 2,944 -0.64 1.57 -0.02 0.99 0.99 0.05 -1.02 0.09 C2: Geographic Distance 2,921 -2.90 9.87 -0.02 1.03 4.98 0.05 49.71 0.09
