1
HAS THE US LOCATION CHOICE CHANGED
SINCE THE RISE OF THE BRICS? : The
Host
country determinants that influence the US Foreign
location.
Date : August 15th, 2014
By : Nombulelo Mbokazi
Student Number : 10604421
MSc in Business Studies : International Management
First Supervisor : Carsten Gelhard
2 Table of Contents Abstract ... 3 Introduction ... 4 Theoretical Background ... 7 Location Patterns ... 10 The BRICS ... 12
Location factor determinants ... 15
Economic factors and hypothesis ... 17
Institutional factors and hypothesis ... 19
Data and Methods ... 23
Variables ... 23
Estimations ... 25
Results ... 27
Location choice analysis... 31
Location determinants analysis ... 32
Discussion ... 36
Conclusion ... 40
References ... 42
3 ABSTRACT
This paper explores if there was any change in the US Foreign Direct Outflows since the formation of the BRICS (Brazil, Russia, India, and China & South Africa).) .It further analyses the economic and institutional determinant factors which may appeal to the US foreign direct investment. The study uses a t-test and panel data for a period of ten years (1996-2012) in order to examine the change in location for the US foreign direct investments (FDI), then explore the significant determinants of FDI. The analysis have been done using a t- test, and random effects regression. This study takes into account GDP, Population,
Infrastructure, Trade openness, Natural resources as economic determinants and, Political stability, Control of corruption, Rule of Law as potential institutional determinants of foreign location. These factors are based on their relative importance from previous empirical
literature. The findings reveal that economic factors are more imperative than institutional factors for the US to invest in a country. There is statistically significant evidence of the US FDI growth since the formation of the BRICS. The findings reveal that efficiency seeking seems to be the dominant motive for the US FDI, furthermore, GDP and Population are the main host country determinants.
Analysis of empirical data also depicts that infrastructure and rule of law and are statistically significant. The paper reveals whether the US FDI attracted by the BRICS economies.
4 Introduction
The world economy has changed drastically over the past 50 years (Wilson& Purushothaman, 2003). The rapid shift in the economy impacted several money generating courses such as business expansion. In the context of this paper business expansion usually means that a firm has to leave its home country and locate into a host country either through foreign direct investment (FDI) or joint ventures, etc. Drawing back to three decades ago, literature reveals the Triad1 region used to be the main recipient of FDI. Furthermore the traditional motives were regarded to be natural resource seeking, efficiency seeking, market seeking and strategic asset seeking (Dunning, 2000). Traditional FDI location motivation trends have been analysed at country level, with FDI determinants such as economic and political stability, host government policies, market size, GDP, cultural distance, tax rates, wages, corruption, and production and transportation costs being key factors (Li, 2009). However, in the past few years there was a shift within the traditional patterns. The emerging economies are reported to have attracted more than half of global Foreign Direct Investments in the past five years (Jadhav, 2012). Regarding the reasons why firms actually opt for a country, Enright (2009) found that foreign direct investment is not uniform, but varies in terms of its motivations and observed behaviour by activity. An activity-based approach can provide distinct insights into what the firm takes into account before going into different locations. On exploring FDI patterns, focusing on the US, Sethi et al., 2003 noted the need for FDI patterns to be examined over time. They added that the change in factors could favour a multinational enterprise (MNE) and would prompt that multinational enterprise to move a new investment elsewhere.
On researching the US location choice or preferences, Flores& Aguilera (2007) discovered that multinational companies’ activities have expanded well beyond the historically preferred
5
regional locations. In addition to examining the US’ change of Multinational corporations (MNCs) in location choice Flores and Aguilera further examined the potential host country level factors that influence the MNC’s choice of location. Their findings are a bit outdated since they did their analysis two decades ago, moreover their observations were only based on two points in time. Therefore, in this paper I will examine how the US FDI location choice has evolved, underlining the BRICS2 economies. I further explore reasons why the US would prefer particular host countries as a destination. This paper therefore contributes to the literature by updating and filling in the research gaps found in Flores and Aguilera (2007), in particular the inclusion of the emerging markets. Additionally, it looks into the accuracy of some of Dunning’s (2009) predictions regarding the future prospects of foreign location motives. I additionally examine whether there is a statistically significant change in the United States outward flows since the formation of the BRICS. This analyses simultaneously challenges/scrutinizes the validity of the Dunning’s prediction regarding FDI percentage growth and the continued dominance of the Triad nations as leading MNE destinations. The study observes whether the US may be one of the main countries contributing to the BRICS’ impressive increase in foreign direct investments. I observe three different country sets at two relevant points in time in order to determine whether the US location choice has changed since the formation of the BRICS countries. To be able to differentiate the effects of the BRICS countries towards location decisions, I categorise the heterogeneous country sets into samples with all 35 countries, then I exclude the BRICS countries which makes the sample 30 countries. Lastly, I have a sample that has just the 5 BRICS countries. I observe points in time in order to determine whether there has been an increase of FDI within these sample countries, I use the years 2000 and 2012 due to the fact that they signify before and after the formation of the BRICS. In the same manner I determine the validity the of Dunning’s prediction regarding FDI
6
percentage growth and the continued dominance of the Triad nations being leading MNE destination. This study intends to reveal whether the shift in the global market, especially emerging economies has impacted the world’s largest economy with regards to FDI.
The second part of the study determines the host country factors that may appeal to the US in order for to invest in a host country. I examine some of the variables used by Flores and Aguilera (2007), however, I address Dunning’s expectation of efficiency seeking and strategic asset being reason for FDI increase. I add more variables to my estimation strategy that literature regards to impact the attraction of foreign location, these include both economic and institutional determinants (Kok and Ersoy, 2009; Blonigen, 2005; Nunnenkamp, 2002; Sethi et al., 2003; Jadhav, 2012; Ranjan, & Agrawal, 2011).
I examine different models, that is, our sample has three different country sets (complete set with 35 countries, the thirty countries and the BRICS countries). I also analyse the three country sets with just economic factors, then I analyse only the institutional factors. I choose the years between the years 1996-2012, because these years signify the period before the BRICS were formed until a just after decade after they were formed. I specifically examine if host country variables have the same effects on other countries as it has on the BRICS countries. I also get to find out whether indeed efficiency seeking and strategic assets have impacted the increase of FDI in the past 17 years.
I find that economic factors seem to explain the US location choice more than institutional factors. Findings reveal that efficiency seeking seems to be the dominant motive for the US FDI and confirm Flores and Aguilera’s findings that GDP has become a host country determinant (Flores& Aguilera, 2007). Population maintains its position as a critical factor to consider in a country before locating or forming foreign direct investment.
7
I believe that no one has specifically observed whether the BRICS formation had an impact on the US foreign direct investment or not. Therefore my findings bridge in the gap found between the US FDI and its determinants. Furthermore, I did not come across any literature that examined the consistency of Dunning’s predictions. Finally no one has studied the US’ foreign location within the time period of 1996-2012.
I structure my theoretical and empirical testing as follows; I first illustrate the theoretical derivatives and literature in sections of; location as a general topic, location patterns, the BRICS. I further categorise the host country factor determinants into economic and institutional determinants. This leads me to the development of my theoretically based hypothesis. On developing my hypothesis I explain and describe the methodology and variables. The methodology elaborates on my data samples its collection, analysis, then reveals the results found. Finally, the last section includes the discussions, limitations and conclusions.
Theoretical Background
Location choice and its determinants have witnessed drastic advancement over the years. In retrospect to literature from 38 years ago, one of the most significant models ever done on foreign location, the Uppsala Model emphasized the staged nature of firm internationalization. According to the model firms first had to enter host countries close to their home countries and with low level of commitment entry modes such as exporting. International business (IB) was argued to have to learn how to be a successful Multinational Enterprise (MNE) and thus as the managers of the firm gain experience in close host countries they venture further afield to more distant host countries and use entry modes with higher levels of commitment and risk, such as greenfield foreign direct investment (Forsgren& Hagström,2007).
Dunning (1998) illustrates location in one of the most widely adopted and highly regarded models for understanding the existence, strategic decision-making and location choice of the
8
multinational enterprise. He uses the Eclectic Paradigm or OLI in order to depict three dimensions or sub-paradigms: the ownership advantages (O) of the MNE, the location (L) advantages enjoyed or accessible to for the MNE and the concept of internalization (I).
It is the existence of these sets of advantages that either encourages or discourages a firm from undertaking foreign activities and becoming an MNE (McCann& Mudambi, 2004). The ownership specific (O) advantages refer to the firm specific advantages of the MNE which it might seek to transfer to the host countries it wishes to operate in. The O advantages build on the resource-based view and the argument is that the MNE is able to overcome the liability of foreignness in a host country because its firm specific advantages are of superiority relative to local host country competitors.
Internalization (I) is related to understanding the different ways in which the MNE can organize and exploit its O advantages (firm specific advantages). Internalization theory argues that MNEs internalize the market for an international transaction when the transaction costs associated with the transaction are so high that the transaction would not take place. By internalizing the transaction the MNE removes these costs and the value creation opportunity can be exploited. These choices are most clearly seen in the choice of entry mode on the part of the MNE, ranging from the non-equity exporting and licensing and equity joint venture or wholly owned subsidiary. These entry modes representing ever greater degrees in internalization and the associated control the MNE has over its presence in the host country (Dunning, 1998).
The final sub-paradigm of location (L) advantages refers to the attractiveness of different regions or countries in which a MNE might decide to locate their foreign direct investment (FDI) in a form of subsidiaries. This advantage emphasizes that MNEs enter host locations to access immobile, natural or advanced factor endowments of the host location, that are not
9
accessible in the cross border market place. These could be factors such as; capital, labour, inputs, skills and capabilities.
Although all these advantages rightfully are separately independent, combined they are mainly regarded as the cornerstone that holds location choice decision. However, of all these advantages, the (L) seems to have captured the attention of many scholars and managers hence locational options of MNEs or potential MNEs depend heavily on the motives for their foreign value-added activities. In particular; market, strategic asset, natural resource and efficiency seeking motives. Iammarino & McCann (2013) describe the motives as such; the market seeking motives comprise MNE investing in order to supply local demands. Resource seeking motives comprise of tangible and intangible resources such as labour, skills which are not available in the home country or are at low cost than at home country. Strategic asset seeking reasons comprise acquiring assets and knowledge of host country firms aiming at advancing long-term objectives. Efficiency seeking motives comprise expanding FDI in order to explore larger markets. Location, government regulations and endowments are important factors in efficiency seeking motives. In combination, these motives provide a fuller picture and more complete understanding of MNE activity.
As evolution emerged, the locational preferences of firms making more traditional forms of FDI have also revolved, and so have the attitudes of recipient countries to these investments (Dunning, 1998). Interviewed on his thoughts when receiving the journal of international business decade award, Dunning anticipated more efficiency-seeking and asset augmenting FDI, although, over time, he predicted that traditional delineations of the different motives for FDI would become less meaningful (Dunning, 2009).When he compared the spatial distribution of the world stock of FDI inward flows in 2007 he further foresaw attraction to several developing and transition economies - especially China and India, and some African countries. Moreover, Dunning further anticipated the greatest percentage growth over the next
10
two decades, but with the Triad nations persevering as leading destinations for all forms of MNE activity.
Location pattern
There has been drastic location shifts within the different worldwide economies. We find some evidence of this when we refer to the UNCTAD reports for the years 2000, 2006 and 2013. By the year 2000 developing countries started to get slight attention for location purposes of Foreign Direct Investments (FDI). Most of the developing countries inflows went into developing Asia (Central Asia and West Asia), a strong percentage of inflows went into just China alone. Other marginally popular destinations at that time were Latin America, the Caribbean and Brazil, which was South America’s leading destination for the fourth consecutive year. Inflows into developed countries on the other hand had just started declining from the previous years. The United States FDI outflows increased for the third consecutive year, although the rate of increase had declined. Latin America and the Caribbean were the US’ greatest developing countries’ recipients (UNCTAD, 2000).
In 2006 developing countries gained a reputation as recipients of foreign direct investment. Their total world inflows increased from an average of 20% in the years 1978-1980 to an average of 35% in years 2003-2005. However, the FDI location distribution of the countries within different regions was uneven. During this year the most competitive locations within developing countries were Taiwan, Singapore, the Republic of Korea, the United Arab Emirates and Qatar. The Russian Federation suddenly claimed dominance in FDI outflows and inflows within its region (UNCTAD, 2006). Significant location changes had taken place between the years 2006-2013. Developing economies surpassed developed economies by becoming the dominant attractive host location. The largest recipients of FDI depict changing patterns of location choice, making 45 percent of the largest recipient developing countries. However, within developing countries, Asia and Latin America are the location of choice.
11
China abruptly started competing with developed economies, from being the sixth to the third largest investor by 2012, following the United States and Japan.
The BRICS countries (Brazil, the Russian Federation, India, China and South Africa), after their formation in 2001 these countries gradually became the principal sources of FDI among emerging host countries. The attraction of transition economies declined by 9% in 2012, the similar trend took place within the FDI within South-East Europe. The Russian Federation maintained dominance in outward FDI.
Outflows from developed economies fell within the period 2007-2012, Europe and North America witnessed massive declines in their outflows, however these countries kept their position as the second leading investor country in the world. Developed countries lost their popularity as the location of choice between the years 2010-2011. Interestingly, the structurally weak, vulnerable and small economies had a rapid increase FDI. These included; small islands, developing States and the least developed countries (UNCTAD, 2013).
In retrospect to work done on the US foreign location we learn that in the last half of the 1990’s the US experienced a dramatic shift from developing countries - a drastic decline from 37% in 1996 to 21% in 2000 ( Jackson,2008). Contrary to their previous location preferences which was the UK and Germany, the US no longer regards these countries as their preferred destination, but considers the rest of the basic countries as potential locations (Filippaios & Papanastassiou, 2008).
Flores and Aguilera (2007), discovered that there has been an overall growing presence of US MNC investments around the globe and this is not evenly distributed across regions. Their analyses depict that, overall, US MNCs have geographically expanded their international capital investments, if one compares their location choices in year 2000 relative to 1980.Moreover, MNCs withdrew from certain regions such as Central America and Africa, in this period. The geographical expansion of the US MNC activities is targeted beyond eastern
12
Asia, encompassing countries rich in natural resources and dynamic emerging markets in south-eastern Asia. However, on the other hand, their findings also disclose that regions outside the TRIAD have become more attractive MNC locations in 2000.They conclude by stating that that some regions, such as South America and Africa, remain relatively less attractive host countries. Flores& Aguilera further urged scholars to assess the choices of MNC location preferences in a changing global market and for host countries seeking to attract FDI.
Though a lot has been done on location US location, as far as we know none of the studies was done specifically on a heterogeneous countries sample nor has there been any studies examining the US location choice patterns/direct investment towards the BRICS.
The BRICS
When the BRIC (Brazil, Russia, India, and China) countries were formed back in 2001, they were predicted to become the largest economies of the world in the upcoming decades. The economies combined were projected to be larger than the combined economies of the U.S., Japan, and the four largest European economies of Germany, France, Italy, and the United Kingdom (G6) in US dollar terms. (Wilson& Purushothaman, 2003)
In their report, Goldman Sachs foresaw China and India to be the leading global suppliers of manufactured goods and services while Brazil and Russia would become similarly dominant as suppliers of raw materials (Cheng et al., 2007). The term BRICs momentarily became a brand and a global financial term which would shape how a generation of investors, financiers and policymakers view the emerging markets. The BRIC states soon realized the prominence of this change in global perceptions about them and used it to structure a new group to underline their growing immensity in global politics and economics (Pant, 2007). According to Cheng et.al, (2007), there has been 30% increase of world reserves and a drastic increase of Foreign
13
Direct Investment (FDI) within the BRIC borders since the year 2000. This would capture the BRICs' role in the world economy today and into the future.
These four countries would later be joined by South Africa, the biggest economy in the Africa continent at the time. While, South Africa did not possess the characteristics of BRIC countries in terms of territorial extension and the size of the population, it possessed phenomenal characteristics such as: potential consumer market with larger middle-income group abundant supply of natural resources, well developed financial parameters, good communication and network, effective energy and transport sectors and, sound legal system and modern infrastructure supporting an efficient distribution of goods and services (Vijayakumar et al.(2010).
The BRICS countries are referred to be the largest economies outside the OECD (Tudoroiu, 2012). These countries represent around 40 percent of the world’s population and nearly a quarter of its economic output. China surpassed Japan in 2010 to become the second-largest economy in the world, and Brazil overtook the United Kingdom in 2012 to emerge as the sixth. The economic profile of the BRICS nations has continued to grow, with some suggesting that they collectively could become bigger than the United States by 2018, and by 2050 could even surpass the combined economies of the G-7 3states (Pant, 2007). According to the UNCTAD report, (2013), over the past decade, FDI inflows to BRICS more than tripled to an estimated US$263 billion in 2012. As a result, their share in world FDI flows kept rising even during the crisis, reaching 20% in 2012, up from 6% in 2000. With the rapid rise in inflows, FDI stock in BRICS countries are increasing as well, standing at 11% of global FDI stock and catching up with developed economies (UNCTAD, 2013).
14
Vijayakumar et al. (2010) reveal that the BRICS countries have been the main recipients of FDI during the last decades. Until 1984, Brazil was the major FDI recipient country among the BRICSs, overtaken by China in 1985. China became the world major recipient of FDI in the 1990s, with the intention to integrate with the world economy. Most of China FDI was intended to take advantage of its low labour costs and huge domestic market while Russia, South Africa and India constantly received an almost a small part of the world total FDI flows in the last two decades. As these countries realized the need to improve their attractiveness, so these countries began to liberalize their economies in more unconditional ways in order to receive more portfolios of FDI.
The BRICS formation also has an impact on institutions and also called come to shape geopolitical representations (Laidi, 2012). Kennel& Salmi (2008) pointed out that international integration of the BRICS countries is occurring through membership of international institutions, international trade and investment, and the movement of people. They went on to reveal that membership of the BRICS countries in the United Nations, World Bank, IMF, WTO etc increases their economic and political integration through trade, investment, and international cooperation. Hence the next decade was predicted to be characterized by increasing influence of the BRICS countries in international institutions and fierce competition for strategic resources as well as customers in both developing and traditional Western markets (Kennel& Salmi, 2008). This will in turn cause these countries to have more influence and decision power in international economic and financial institutions(Tudoroiu, 2012).The economic significance of BRICS is expected to continue to rise for the foreseeable future (Vijayakumar et al.2010). Therefore, these countries are expected to become a very important source of new global spending and FDI (Wilson& Purushothaman, 2003).
15 Location factor determinants
Most international business authors stem their firm-based location pattern studies from the foundation of Dunning’s work. Through the influence of his work, beget location or foreign direct investment arguments and conclusions. Some firms consider product differentiation to be an important determinant of location choices (Nachum & Wymbs, 2005). Country based enterprises seem to view location more or less the same as firm based companies. Makino et al. (2004) reveal that country effects have a significant impact on foreign direct investments just like industry effects. However the foreign companies may achieve fluctuating levels of performance depending on the country in which they operate.
Therefore, foreign direct investment (FDI) motives are considered to be the prerequisite of location choice since Multinational Enterprises evaluate the advantages of the anticipated destination according to specific motives (Chen & Yen, 2012). Hence ideally most countries that would receive FDI inflows were the traditionally motives “attractive” developed countries. According to Cantwell (2009) in the 1960’s market seeking was the dominant motive for FDI, he further argues that to this day market seeking still remained crucial. Sethi et al.(2003) finds that of all the motives, less developed countries attract mostly resource-seeking and efficiency-seeking foreign direct investments in product markets or labour intensive production tasks. Buckley& Ghauri, (2004) combine economic-wide factors and enterprise strategic motives, they realize a changing pattern of multinational activity. These authors find an increase in refined location decision making that is their location decision making is increasingly being based on slicing an enterprise’s activities more defined and locating those enterprise activities according to places where they are likely to optimize that specific activity. Buckley et al. (2007) share the same sentiments by concluding that managers follow coherent rules from the perspective of the company’s interest in creating types of investments to consider. According to Buckley et al. (2007) enterprises prefer nations in which they were already functioning to
16
those in which they were not. One more finding is that, companies with broad international experience demonstrated less preference for near, similar and familiar markets. Instead these enterprises would rather invest in markets that others might perceive as less viable. As the enterprises’ global familiarity rises their interests their scope.
According to Jadhav (2012) traditional economic determinants are more important than institutional determinants, moreover most of the foreign direct investments in BRICS economies are motivated by market seeking purposes. Blonigen (2005) suggest that factors that determine location choice in developed countries are different from factors in less develop countries. More interestingly, even years after the OLI triad, Dunning (2009) remains aware of the unpredictable rapid changing world, hence he observed an increase in FDI into particular regions and a decline in formerly popular destinations.
In their examination of location and global cities Goerzen et al(2013) realize that market seeking MNE are more likely to be located in global cities, whereas the other three investment motives(strategic asset seeking, resource seeking, and efficiency seeking) are might seek more efficiency outside global cities. However, I are more interested in country level factors.
Other findings rationalise the U.S. firms to invest abroad in order to serve the foreign local market, instead of producing goods to export back to the United States, while some firms do establish overseas operations to replace U.S. exports or production, or to gain host country advantages ( Jackson,2008).
Complementary, Flores& Aguilera derive their research about the shift in host country determinants of MNC foreign location choice from Dunning’s work (Flores& Aguilera, 2007). Their findings depict that neither economic factors nor institutional-cultural factors taken alone fully explain foreign location choice (Flores& Aguilera, 2007). Though there is plenty of literature which has examined the host country determinants within different economies, to my
17
knowledge there hasn’t been any studies done specifically on the US outward investment into heterogeneous economies, most of all towards the BRICS countries. Therefore, as Flores& Aguilera, I will use both economic and institutional factors in order to determine the host-country factors that may influence the location choice patterns since the rise of the BRICS. However, I will not consider the cultural factors. Furthermore I explore Dunning’s predictions regarding efficiency-seeking and asset augmenting in FDI. I also look into the validity the prediction regarding FDI percentage growth and the continued dominance of the Triad nations being leading MNE destination.
Economic factors:
Market size and growth prospects of the host country immensely affect investment location. However, in contrast with previous research within the economic perspective, both a country's GDP and its population independently explain MNCs' foreign location choice (Flores& Aguilera, 2007; UNCTAD, 2013). Although according to Flores& Aguilera, (2007) GDP has become less important in predicting the likelihood of being the recipient of US investments while population became more important in 2000 (relative to 1980). Hereof I hypothesize the following:
Hypothesis 1 a: The more impressive the GDP is, the higher the likelihood of having the US investing in a country.
Hypothesis 1 b: The larger the population, the higher the likelihood of having the US investing in a country.
Natural resources was traditionally known as one of the main FDI attractions. However the influence that these resources seems to have different outcomes when examined by different scholars. When examining the Nordic countries Bravo et al., (2002) concluded that the fortune of natural resources is inconsistent as it could be either viable or non-viable. Hence , regardless
18
of their natural resources poor countries or small countries may attract very little or no FDI, regardless of the policies the country pursues (Asiedu, 2011).On the contrary, according Jadhav (2012) resource-seeking FDI is motivated by the availability of natural resources in host countries, furthermore natural resource seeking FDI was traditionally rightfully important and remains a relevant source of FDI for various developing countries. Hence, natural resources play a vital role in overall FDI attraction or decision.
Hypothesis 2: Natural resources play a vital role in having the US invest in a host country.
In retrospect to literature, the improvements in infrastructure, has facilitated a shift in efficiency-seeking US FDI. This has contributed to a change in the FDI trend over time (Sethi et al. 2003). The availability of infrastructure is critical, it increases efficiency of investments and therefore enhances FDI investments. Emerging market countries that are best prepared to secure greater amounts of FDI. Hence advanced infrastructure in a host country in favour of foreign investment is an important consideration for investor (IMF, 2013). Flores& Aguilera (2007) and Ranjan et al. (2010) support infrastructure’s significance, which shows that these determinants are potential determinants of FDI inflow.
Hypothesis 3: The US is more likely to invest in a country with higher infrastructure.
Most of the studies find that trade openness is positively related to FDI host country attraction, however the impact of openness on FDI depends on whether the investment is market seeking or export oriented (Jadhav, 2012). Furthermore, the IMF report, (2013) supports Jadhav by reporting that investment in emerging economies is likely to be led by market-seeking investments that will focus on countries with large markets and promising growth prospects. In this regard, participation of countries in free trade agreements and regional trade integration scheme, increase regional demand and potential market size will likely increase their appeal to
19
investors. On the contrary, trade openness does not always promote FDI in some developing countries (Asiedu, 2011).Therefore I base my hypothesis on this literature.
Hypothesis 4: The higher a country is open to trade, the higher the US investing might direct their outflows to a country.
Institutions:
North (1999) defines institutions as humanly formulated constrictions which form political, economic and social interaction. These consist of both informal constraints and formal rules (North, 1991). Formal institutions refer to the hierarchy of rules (constitutional law, statute law, common law and bylaws) which determine the expected procedures and costs involved when an MNE is considering investing in a foreign country (Dikova, Rao-Sahib & van Witteloostuijn, 2010 ) . Informal institutions on the other hand refer to the norms, customs, mores, traditions, sanctions, taboos, and codes of conduct, religion etc. are located. These institutions combined with the standard constraints of economics they outline the choice set and consequently regulate transaction and production costs and henceforth the profitability and feasibility of engaging in economic activity (Williamson, 2000).
According to Blonigen, (2005) the eminence of institutions is likely a significant determinant of FDI activity, especially for less-developed countries for a several reasons. Poor legal protection of assets reduces the chance of appropriation, of a firm’s assets making investment. Inefficient quality of institutions such as corruption increases the cost of doing business and, thus, should also lessen foreign direct investment. Blonigen further argues that poor institutions could go to the extent that they lead to poor infrastructure in factors like public goods, hence due to this effect profitability declines so does FDI into a market. Cuervo-Cazurra, (2008) shares a similar view by stating that the lack of properly developed institutions limit foreign direct investment into a country. However, the effects of institutions on the performance of
20
firms differ across countries because institutions are developed and continual in path dependent and highly confined processes in a country (Makino et al.2004).Williamson (2000) also complements this by stating that rules and procedures evolve not to restrict economic activity but to simplify the processes of within different business departments, moreover to enable value-adding transactions that would otherwise not decorously have taken place. Williamson complements the findings of Makino et al (2004) by adding that institutions are nation specific, therefore the rules of the game that reduce uncertainty in economic activity contrast across national boundaries. Institutional framework of the host country, tend to play a more decisive role than they once did (Dunning, 1998). One of the essential location specific advantages are assumed to be based on the institutional structures that are specific to a country. Examples here are the tariff and tax structures, politics and law (McCann& Mudambi, 2005). The regulations for creating new businesses differ significantly across various countries. In order to incorporate and register a new business an entrepreneur has to comply with legal procedures. While some economies facilitate the process of new business entry with a straightforward and affordable process, others have lengthy, tedious and highly bureaucratic procedures that induce bribery of officials to smooth the process (Cheng et al.2007).All these factors impact foreign investments. Holmes et al. (2013) find that informal institutions influence formal institutions and that different formal institutions have distinct effects on inward FD.
While in their paper Flores and Aguilera were mainly concerned about institutional-cultural variables, I contrary fit in some different variables that are inspired by Dunning (2009) predictions. These are government and firm tackling institutional factors that might jeopardize trade and FDI (Dunning, 2009). Therefore in order to observe the impact of institutions I consider factors such as; political stability, corruption and rule of law.
21
The subject of political stability has rather evolved interestingly in the literature. Schneider and Frey (1985) found that political instability in host countries does not affect the United States’ foreign direct investment, moreover they interestingly find a further positive relationship between the investment flow and political instability in the host countries (Schneider & Frey, 1985). Kim (2014) shares interesting views about the political state of countries, that is, politically stable countries produce capital flows to invest in politically unstable countries. Other literature found inclination in the fact that more politically stable countries are expected to have a higher probability of attracting FDI inflows (Meon Sekkat, 2012; Allard et al, 2012). The IMF (2013) supports these authors by stating that a reasonably stable political environment, as well as conditions that support physical and personal security is an important benchmark that is used in judging the likelihood of adverse changes in the investment essence for foreign owned firms. Therefore, I examine the effect Political stability affects the US’ consideration into investing into host countries.
Hypothesis 5: Political stability is likely to have positive significant effects on the US investment.
The level of corruption in the host country has recently been introduced as one factor among the determinants of FDI location. Within this context corruption can be defined as paying bribes to corrupt government bureaucrats in order to get favouritisms in government produced documents such as permits, investment licenses, tax assessments, and police protection (Al-Sadig, 2009). Cuervo-Cazurra (2008) discovers that the corruption outcome is less of a negative influence on FDI in transition economies. He goes on to state that there are different effects of corruption, hence he states that corruption may have negative effects in countries that have established market institutions, whereas it could be a stimulant in transition economies that have not yet established appropriate market institutions . Cuervo-Cazurra further distinguishes types of corruption in transition economies as pervasive and arbitrary corruption. Henceforth
22
pervasive corruption is explained as a known informal cost of operating that foreign investors which accumulates additional costs. Contrary, arbitrary corruption is the insecurity that is associated with corruption, such as officials not keeping their word. Therefore, when investing into transition economies, investors would rather in host countries with arbitrary corruption over countries with pervasive corruption.
According to Jadhav (2007), investors from countries with high corruption and the lack of enforcement of anticorruption laws select similar countries when they internationalize in order to exploit their familiarity with corrupt environments and also because they face lower costs of operating as opposed to other investors. Unexpectedly, Kim (2014) concludes that host countries with higher level of corruption of governments and lower level of democracy attract more FDI inflows. I base my hypothesis on the discussed literature plus the fact that the US is a well institutionalized developed country.
Hypothesis 6: corruption is likely to negatively influence the US FDI investment in a host country.
Interestingly rule of law seems to be closely related to corruption. According to two studies that were done not so long, the effects of corruption depend on the country's rule of law and economic freedom (Houston, 2007). Hence the outcome of these studies reveal that corruption on countries with a weak rule of law have positive effects on economic growth whereas the same corruption has negative effects in countries with strong institutions. Globerman & Shapiro (2003) found that countries which have legal systems based on English common law are likely to be the US preferred FDI destination. Jadhav (2007) found the rule of law effect to be statistically significant and has positive effect on total inward investment. Therefore, based on the fact that the US has strong institutions and the fact that rule of law has proven to have a positive effect on FDI, I expect the US to engage in foreign location with strong rule of law.
23
Hypothesis 7: Strong Rule of law is likely to have a positive effect on US investment.
DATA AND METHODS
In order to examine if there has been location change since the formation of the BRICS I ran a t-test on a sample of the 35 countries selected. I paired the years 2000 and 2012, the reason for using these years was due to the fact that these years signify the periods before and after the formation of the BRICS. The paired t-test determines whether the mean difference between the US outflows from 2000 and 2012 is statistically significantly different from zero. I fundamentally looked at the differences in the values of the two years and testing if the mean of these differences is equal to zero (Zimmerman, 1997).
Therefore, for me to note if there has been indeed a difference prior or post the formation of the BRICS I ran 3 types of t-tests. One consisted of all 35 country sets when the second one consisted of 30 country sets which did not include the BRICS countries. Finally I tested for a country set which only contained the BRICS. I classified the two sets of US outflows as before the formation of the BRICS = outflows 2000 and after the BRICS formation = outflows 2012.
Variables
Dependent variable
The dependent variable, the US Outflows is a variable that indicate whether the US made direct investment or disinvestment in a certain country. The data are acquired from the OECD Foreign Direct investment statistics. My data set consist of 35 countries that the US outflows may have or not invested in between the years 1996-2012. The countries examined in this study consist of: Argentina, Australia, Austria, Canada, Chile, Colombia, Czech Republic, Denmark, Egypt, Germany, Greece, Hong Kong, Ireland, Israel, Italy, Japan, South Korea, Malaysia, Mexico,
24
Netherland, New Zealand, Philippines, Poland, Spain, Sweden, Switzerland, Thailand, Turkey, United Kingdom and Venezuela. In addition to developed economy countries, my sample also includes a number of large emerging economy countries, such as Brazil, Russia, India, China and South Africa. These 35 countries were chosen because they vary geographically and economically, therefore depict a heterogeneous economic and institutional sample (Table 1).
Table 1
Sampled Countries
Argentina Australia Austria
Canada Chile China Colombia Czech Republic Denmark Egypt
Germany Greece Hong Kong SAR China India Ireland Israel Italy Japan Korea Rep. Malaysia Mexico Netherlands New Zealand Philippines
Poland, Russian Federation
South Africa Spain Sweden Switzerland Thailand Turkey United Kingdom Venezuela
Independent variables
I examine two categories of independent variables: variables that signify the economic dynamics and one that signify institutional dynamics. The economic variables consists of GDP measure in U.S current U.S. dollars yearly, Population Total which is based on all residents regardless of legal status or citizenship except for refugees not permanently settled in the country of asylum, who are generally considered part of the population of their country of origin(WDI,2014). In order to estimate infrastructure I used different indicators: electric power consumption (kWh per capita), energy use (kg of oil equivalent per capita), and road sector energy consumption per capita (kg of oil equivalent) and telephone lines per 100 people (Vijayakumar et al 2010; Ranjan et Agrawal, 2012). In order to reduce the number of variables of interest into one solid infrastructure we used principle components (Horne, & Camp 2004).
25
Natural resources was based on a countries’ on the sum of oil rents, natural gas rents, coal rents (hard and soft), mineral rents, and forest rents. All these four variables were collected from the World Development indicator. Trade openness was derived from the World Trade Organization. In order to examine the institutional factors within the 35 countries which could potentially host the US, I assess three variables namely corruption, political stability and rule of law (Flores& Aguilera, 2010; Jadhav, 2012, IMF, 2013). These institutional data are collected for the World Governance Indicator. All eight variables (prospective determinants) are collected from 17 years, 1996-20012.
Estimation:
I constructed a panel data set that contains the presence (or the lack) of US MNCs investments for each country in our sample in each of the countries where they existed between 1996-2012.Considering that our sample includes the US FDI outflows going to 35 countries for 17 years, our panel data consists of 595 observations. Due to the limited number of observations, I opted not to delete a number of because of missing values for the variables would radically reduce the useable cases for our regression analyses which may result to weak and biased parameter estimates (Myers, 2011).
Therefore, I used the predictive mean matching in order to fill in missing values, this method would be most accurate than case deletion or list wise because it’s our data contain continuous variables (Mustillo, 2012). All measures are at the country level.
In order to evaluate the country level determinants of the foreign location choice of the US investment, our approximations are based on the following models:
OUTFLOWS it = α +β1 LGDPit + β2 LPOPit + β3 INFRSTit + β4 NTR it + β5 TRDOP it + β6 COR it + β7 POLSTA it + β8 ROL+ e it
26
OUTFLOWSit is the US FDI outflows in current US$ for country i at time t.
LGDPit presents the log of Gross Domestic Product in current US$ for country i at time
LPOPit is the Population for country i at time t
INFRSTit signifies the Infrastructure Index for country i at time t.
NTR it denotes the natural resources index for country i at time t.
TRDOPit is the Trade Openness for country i at time t and is measured by the WTO as ratio of import of Goods and Services plus Export of Goods and Services divided by GDP.
COR it is an index that estimates the level of corruption for country i at time t.
POLSTAit is an index that estimates the level of corruption for country i at time t.
ROL it is an index that estimates the level of corruption for country i at time t.
eit signifies the error term over the time t.
I first panel the above model with all variables and the entire country set, next I maintain all variables but exclude the BRICS from the 35 country set, which sums our data set into 30 countries. The third time I observe just the BRICS countries, still with the same variables. However I still was curious about the significance of the host country determinants if there were observed separately as economic factors or institutional factors. Therefore I first panel the just economic variables with all 3 different country sets (complete, other countries and the BRICS). I do the same with the institutional variables.
Instead of using cross-section and time series I use panel data methods because of their advantages of using all information available which could not have been taken into account in cross-section and time series (Baltagi, 2008). Furthermore, panel data since it accounts for individual heterogeneity, therefore it reduces the risk of obtaining biased results. Moreover,
27
panel data can comprise variables at different levels of analysis suitable for multilevel or hierarchical modelling (Torres-Reyna, 2007).Lastly, panel data increases the degrees of freedom of the dataset I have, therefore this enriches the credibility of our results. However, before running a panel data regression I had to fit the model in order to determine the appropriate method between random and fixed effects methods. The fixed effects model assumes that there could be other influences besides the model factors that could impact or bias results, therefore it controls for this. Fixed effects also assumes that time-invariant factors are unique and therefore they are not be correlated with other individual factors. Fixed effects models control for the effects of time invariant variables with time-invariant effects (Vijayakumar et al.2010). Random effects models assume that each country’s error term is not interrelated with the independent variables which allows for time-invariant variables to play a role as explanatory variables. Contrary to the fixed effects model, the random effects model’s variation across entities is assumed to be random and uncorrelated with the independent variables in the model. Furthermore, since random effects models handles differences between the countries as a random draw from a probability distribution, it is often considered to be more efficient than a fixed effects model (Clark & Linzer, 2012).
RESULTS
Descriptive analysis and Correlations:
Prior estimating panel data analysis, I implemented descriptive statistics and correlation analysis (Das,2013; Ranjan et al. 2011; Vijayakumar et al.2010). Descriptive statistics and the correlation matrix of the variables used for estimation are illustrated in table 3 and 4.1, 4.2, 4.3. All these variable have 595 observations, with FDI outflows having the highest mean and standard deviation of 1768.494 and 3013.471. The correlation matrixes reveal that GDP has moderate positive values which are significant (p<0.01 level) on both complete country set and
28
the country set with other countries but insignificant on the BRICS country set. LPOP is negatively(-.0262) correlated with the outflows and GDP from the complete country set whereas the two country sets are positively correlated with outflows but negatively correlated with GDP(-.4679 and -.8203).LPOP correlates significantly with GDP from all country sets but correlates insignificantly with the outflows from all the country sets. NTR is negatively correlated with outflows and GDP on both the full country set and the other countries (-.1021 and -.3804; -.0860 and-.4393), the BRICS country set correlates negatively with outflows and LPOP (-.0958 and-.2304). However all the NTR values are significantly correlated in all country sets, except outflows from the BRICS country set. INFRST is highly correlated with GDP from all the complete country from all country sets (.8910, .8866 and .5620). We can further see a significant correlation (p<0.01, level) in all country sets on outflows, LGDP, LPOP and NTR, except the GDP and LPOP from the BRICS country set. TRDOP correlates significantly with outflows, LGDP, LPOP, NTR and INFRST from all country sets except outflows from the other countries’ country set (.532), the same variable is highly correlated with TRDOP. There is also a high correlation with NTR from the BRICS country level. COR has a significant correlation ((p<0.01, level) with all variables from all three country sets, except outflows and LGDP from the BRICS country set. There is a negative correlation with LPOP and NTR (-.5240- and .3955; .4507and -.3432) from the complete country set and the other countries. Whereas with the BRICS country set, there is negative correlation is with outflows LGDP, NTR, INFRST, and TRDOP (-.0375 .0385,-0.2909,-.4196,-.3468 and.1622, correspondingly). I can further tell that COR is highly correlated with LGDP and INFRST from the complete country set and other countries’ set (0.6696 and 0.6311; 0.6617 and 0.6311).
On the entire country set POLSTA is highly correlated with LGDP, INFRST and COR (0.6160, .5671, .6534), the same correlation pattern can be seen with the other countries matrix. I further see negative correlations with LPOP and NTR (-.4444 and -.3629) from both matrixes. The
29
BRICS country set’s POLSTA is negatively correlated with LGDP, NTR and TRDOP (-.2426, -.3597 and -.19260). All variables from three country sets are significantly correlated (p<0.01, level), except outflows from the BRICS country set.
Lastly, ROL is significantly correlated with all variables from the complete country set and the other countries’ country set (p<0.01, level). On the BRICS country set, only outflows, GDP and LPOP are insignificantly correlated. ROL is highly correlated with LGDP, INFRST, COR and POLTA (.6508, .6391, .7605 and .6825; .6511, .6363, .7584 and .6827) from the complete country set and the other countries country set. There no high correlation on the BRICS country set, however, there is negative correlation with LGDP, LPOP, NTR, INFRST, and TRDOP (-.1731,-.0096,-.4489,-.3977and -.2797).The presence of high correlation among the independent variables will result to the problem of multi collinearity in the estimation. This should not be a problematic due to the fact that the statistical nature of panel data estimation tackles collinearity problems. Therefore I maintain all variables in the model.
However in order to determine the most suitable model between random and fixed effects we conducted a Hausman test. The Hausman Test developed null hypothesis that random effects was appropriate, alternative fixed effects was appropriate. Whenever the p value of the test is < 0.05, fixed effects should be preferred. Therefore since the Prob>chi2 was statistically insignificant 0.9291, null hypothesis could not be rejected hence random effects was preferred. In order to affirm that random effects model is the most appropriate I ran a Breusch–Pagan test, I compare it with pooled regression model and random effects is still the most appropriate model.
Table 4.1: Correlation table for complete set
outflows LGDP LPOP NTR INFRST TRDOP COR POLSTA ROL outflows 1.0000
30 LPOP -0.0262 -0.6549***1.0000 NTR -0.1021 **-0.3804 ***0.1969***1.0000 INFRST 0.2415*** 0.8910***-0.5447***-0.4160***1.0000 TRDOP 0.0729* 0.2245***-0.4198***-0.1287***0.2329***1.0000 COR 0.2473*** 0.6696*** -0.5240*** -0.3955***0.6311***0.2331***1.0000 POLSTA 0.1764 ***0.6160***-0.4444***-0.3629*** 0.5671***0.2595***0.6534***1.0000 ROL 0.2275*** 0.6508*** -0.4913***-0.4927***0.6391***0.2408***0.7605***0.6825*** 1.0000 p<0.01, *** , p<0.05, **p<0.1*
Table 4.2: Correlation table for other countries
outflows LGDP LPOP NTR INFRST TRDOP COR POLSTA ROL outflows 1.0000 LGDP 0.2940***1.0000 LPOP 0.0480 -0.4679***1.0000 NTR -0.0860**-0.4393***0.1390** 1.0000 INFRS 0.2238***0.8866***-0.4359***-0.4867***1.0000 TRDOP 0.532 0.1271***-0.3947*** -0.1121** 0.1508***1.0000 COR 0.2377***0.6617***-0.4507*** -0.3432***0.6207***0.1778***1.0000 POLSTA 0.1576***0.5944***-0.3769*** -0.3184***0.5413***0.2203***0.6273***1.0000 ROL 0.2125***0.6511***-0.4255***-0.4634***0.6363***0.1875***0.7584***0.6827*** 1.0000 p<0.01, *** , p<0.05, **p<0.1*
Table 4.3: Correlation table for the BRICS
outflows LGDP LPOP NTR INFRST TRDOP COR POLSTA ROL outflows 1.0000 LGDP 0.0466 1.0000 LPOP 0.1341 -0.8203*** 1.0000 NTR -0.0958 0.3668*** -0.2304** 1.0000 INFRST 0.0710 0.5620*** -0.1476 0.5704***1.0000 TRDOP -0.1997* 0.2254** -0.1813* 0.4329***0.1842* 1.0000 COR -0.0375 0.0385 -0.2909***-0.4196***-0.3468**-0.1622 1.0000 POLSTA 0.1058 0.2859*** -0.2426** -0.3597*** 0.0846 -0.1926* 0.4763*** 1.0000 ROL 0.0655 -0.1731 -0.0096 -0.4489***-0.3977***-0.2797***0.3134***0.2117 *** 1.0000 p<0.01, *** , p<0.05, **p<0.1*
Table 3: Descriptive Statistics of Variables for the entire country set
Variable | Obs Mean Std. Dev. Min Max Outflows| 595 1768.494 3013.471 -7512 15971 LGDP | 595 7.898 .834 6.206 8.830 LPOP | 595 19.454 1.279 17.504 21.024 NTR | 595 9.441 9.936 1.144 44.526 INFRST | 595 15.603 8.750 1.494 31.696
31
TRDOP | 595 44.881 15.571 14.12 75.39 COR | 595 -.164 .705 -1.09 2.47 POLSTA | 595 -.510 .666 -2.19 1.28 ROL | 595 -.124 .642 -1.13 1.95
Descriptive statistics and the correlation matrix of the variables used for estimation are illustrated in table 3 and 4.1, 4.2, 4.3. All these variable have 595 observations, with Outflows having the highest mean and standard deviation of 1768.494 and 3013.471.
Location choice Analysis
The analyses are illustrated in different tables, tables 2.1-2.3 depict results for probing changes with regards to the US outward investments prior and post the BRICS formation. The models presented in table 5.1 and 5.2 depict the results for host country potential location/investment determinants. Therefore, referring to tables 2.1-2.3 the mean and of the complete country set from 2000 outflows were 3532.514 and the standard deviation was 5937.555.The outflows from 2012’s mean was 6930.286, the standard deviation was12584.21.The null hypothesis of the two-sided paired t-test mean (diff) = 0 generated t statistic value of 1.9454 and the two sided p-value = 0.0600.Therefore the result is statistically significant, the null hypothesis was rejected. The country set with the other countries except the BRICS 2000 outflows resulted a mean of 3940.133 and standard deviation of 6310.874 .Their outflows from 2012 had a mean of 7738.267 and standard deviation of 13357.4. These t statistic value was t = 1.8847 and the two-sided p-value was 0.0695, which makes the result statistically significant. Therefore the null hypothesis was rejected. Finally, the BRICS country set’s 2000 outflows had a mean of 1086.8 and the standard deviation was 1481.194. The 2012 outflows yielded a mean of 2082.4 and the standard deviation was 4273.087. With the same null hypothesis of the two-sided paired t-test, the t statistic value was t = 0.5588 and the two sided p-value was 0.6061.This result is not statistically significant, therefore the null hypothesis cannot be rejected.
32
The empirical results (Table 5.2) obtained from the different Random Effects regression models show the unique R squares from each of the 9 regression models. The different squares signify the variation in US FDI investment/outflows explained by the models. The R-squared values fluctuate between the models, the complete country set model with all independent variables (model 1)’s R-square is 0.25. There is a slight change with model 2 (other countries country set + entire variables) has an R-square of 0.23. Contrary to the previous similar models, the BRICS show a drastic increased R-square of 0.98. Models 4-6 have excluded the institutional variables from the models, they are regression models of just the economic variables. Still structured as a complete data set, the other countries (30countries) and the BRICS country sets, their R-squares are (0.23, 0.22 and 0.98).There is a similar inclination as the previous models. Finally, the last 3 models are similar regression models with just the institutional variables. There’s a decrease in all models (7-9) with R-squares 0.11, 0.10 and 0.69. Therefore in this case the economic variables explain most of the variation in the US investment than the institutional variables.
Location Determinants Analysis
Models 1 -3 in Table 5.1a tests the first hypothesis, the sample covers all 35 countries and combines both economic and institutional variables. Model 1 reports that the GDP, Population and Rule of Law are statistically significant and have a positive effect on the FDI investment. This implies that our hypothesis does not find support for this smaller set of countries. Infrastructure is statistically significant but has a negative effect on the US FDI investment. Natural resources, Trade openness, have a positive effect on US FDI investment but are statistically insignificant. Contrary Political stability has a negative effect. Unexpectedly Corruption has a positive effect on US investment. Like the previous model, Model 2’s sample is made of all institutional and Economic variables but the BRICs countries have been excluded which sums it up to (N=30). The model reports a positive and significant relationship on the
33
effect of GDP and Population on the prospect of receiving US FDI investment as well as a negative effect but statistically significant for Infrastructure. Though insignificant Natural resources and Trade openness and have a positive effect but are statistically insignificant. Table 5.1a
Unexpectedly, Corruption has a positive effect and is statistically insignificant. Political stability is statistically insignificant but has a negative effect on the US FDI investment. Rule of Law has a positive and insignificant coefficient. Model 3 is identical to the previous models except that its sample has only the BRICS countries (N=5). It shows a significant and positive coefficient for GDP and Population. Contrary to the same test with the whole sample and 30 countries here the Infrastructure variable is not significant, Political stability is also insignificant and has a positive coefficient. On the contrary, the coefficients for Natural resources, Corruption and Rule of Law have positive effect on US investment decision toward the BRICS countries. Trade openness surprisingly have a negative effect on US investing into the BRICS. As hinted before, Model 4-6 in Table 5.1a eliminates the institutional variables, it only contains the 3 different country sets with only the economic variables. Model 4 is a sample of the complete 35 countries set, reports a positive and significant relationship on the effect of
34
GDP and Population on the possibility of receiving US direct investment. Infrastructure reports to have significant but negatively associated with the US investment decisions. Trade openness is positively but insignificantly associated with the US FDI investment. Natural resources on the other hand have an insignificant coefficient.
Model 4 reduces the sample to 30 countries, I analyse the effects the economic variables have on the US investment when the BRICS countries are excluded. The economic variables have the exact same effects as Model 4, however Infrastructure has is slightly different, contrary to Population from Model 4 which is significant at 5% level, and Population from Model 5 is significant at 10% level. GDP is significantly positive whereas Infrastructure is significant and has negative coefficient while the other interactions are not significant.
Model 6 examines the effects the economic variables have on the US investing on the BRICS countries. This model depicts that a positive significant effect for GDP and Population while there Natural Resources have a positive effect but are insignificant. There is a shift regarding Infrastructure, though its effect remains negative effect it’s suddenly insignificant. Trade openness has a negative insignificant coefficient.
In order to test for robustness of our arguments regarding the institutional effects on US outward investment. Models 6-9 in Table 5.1b illustrates the examination of institutional variables effects. In Model 7 our sample entails of only the institutional variables and the 35 country set. Corruption unexpectedly has a positive effect on investment, it also is insignificant. Political stability has insignificant negative effects. Rule of law on the other has a positive coefficient and is significant at 10% level. Model 8 repeats the previous analyses but for a reduced set of countries (N=30) where Corruption is insignificant and has a positive effect. Political stability still is insignificant and has negative coefficient. The Rule of Law has
35
positively significant effect. Model 8 is sampled with the same variables as the previous two models, however the countries are reduced to just the BRICS countries. Unlike the results from the previous models Corruption has a negative effect on the US investing in the BRICS
Table 5.1 b
countries. Political stability also reacted differently from the previous models, it has negative coefficients and it’s insignificant. Rule of law positively insignificant.
The consistency of our findings in Models 1-6 implies that my argument seems to be supported, meaning that hypothesis1a is supported. GDP also finds support since in it’s significant in all 6 models, therefore hypothesis 1b is also accepted. Though Infrastructure is insignificant in the two BRICS country set models, Infrastructure is significant in both the large sampled models (the complete country set and the 30 country set) .Therefore hypothesis 2 cannot be rejected. Hypothesis 3 is not accepted since in all the models are none of the Natural resources is significant. Hypothesis 4 is rejected due to the fact that it’s insignificant in all models and has
36
an inconsistent effect on the three different country sets. Corruption apparently has a positive effect on US FDI, moreover none of the models have significance in corruption. Therefore, hypothesis 5a is rejected. Political stability’s effect inconsistency, over and above it’s insignificance gives us grounds not to accept Hypothesis 5b. The Rule of law as expected has a positive effect on investment, half of the models are significant. The models with the largest and second country sets are significant, therefore there isn’t any grounds to reject Hypothesis 5c.
DISCUSSION
The results show that, overall, US outwards investments had statistically changed since the formation of the BRICS countries. There is evidence of this if I compare the US outflows in the 2012 in relation to 2000 in tables 2.1, 2.2 and 2.3. The results on the complete country set in table 2.1 reveals that the result is statistical significance, meaning that the mean is different from zero.
It is worth recapping that the data included all 35 countries’ data set. In table 2.2, though the outflows had now excluded the BRICS countries, the mean is still different from zero and it is still positively significant. Table 2.3 depicts the results for only the BRICS countries dataset, although the mean has a positive difference it’s statistically insignificant. Therefore I failed to reject the null hypothesis. When analysing the different country sets I realize that on the mean difference complete country set is less compared to the mean difference when the BRICS countries had been excluded from the data set. Through these results I gather that that the US was more likely to have been sending their outflows into the 30 countries than into the country set inclusive of the BRICS countries. The fact that the BRICS country set statistically insignificant supports this assumption. Therefore I argue that the rapid increase FDI the BRICS
