Why and How do Institutional Distance Effects differ?
A Comparison between Multinationals from Developed Countries
and Emerging Market Countries
Master Thesis
MSc International Business & Management
Evita Dominique Muller e.d.muller@student.rug.nl Student number: S3176967
University of Groningen, The Netherlands Faculty of Economics and Business
ABSTRACT
The common argument found in the literature is that institutional distance negatively affects Foreign Direct Investment (FDI). However there is a limited discussion about the heterogeneity of these effects across different types of countries. There has been restricted systematic study of the perspective of Emerging Market Multinational Enterprises (EMNEs) related to institutional distance and the determinants of FDI outflows. I argue that the effect of institutional distance on FDI differs across Developed Market Multinational Enterprises (DMNEs) and EMNEs because EMNEs may have different motivations than DMNEs to internationalize. Using a panel data set with a sample of 71 countries of origin and 191 countries of destination from 2001 to 2012, reflecting in total 40,076 FDI flows, this study shows that the negative relationship between institutional distance and FDI is not significant for emerging market countries. This implies that the effect is heterogenous across different types of countries. Specifically, EMNEs could even be attracted by high institutional distance as they seek for certainties to safeguard their business or to acquire intangible assets which are found in strong institutional environments.
Keywords:
TABLE OF CONTENTS
LIST OF TABLES AND FIGURES ... v
LIST OF ABBREVIATIONS ... vi
2. LITERATURE REVIEW ... 4
2.1 Foreign Direct Investment ... 4
2.1.1 Multinational Activity and Foreign Direct Investment ... 4
2.1.2 Determinants of FDI Flows into and from Emerging Markets ... 7
2.2 Institutional Distance ... 10
2.2.1 Institutions ... 10
2.2.2 Institutional Profile versus Institutional Distance Effects ... 11
2.2.3 Institutions and Emerging Market Multinational Enterprises ... 12
2.3 Foreign Direct Investment and Institutional Distance ... 13
3. DATA AND METHODOLOGY ... 18
3.1 Sample ... 18
3.2 Data ... 18
3.2.1 Dependent Variable ... 18
3.2.2 Independent Variable... 19
3.2.3 Control Variables ... 20
3.3 Method and Empirical Model ... 21
3.3.1 Gravity Equation and its Empirical Aspects ... 21
3.3.2 Linear Regression Model with two High Dimensional Fixed Effects ... 23
3.4 Testing the Normality Assumption ... 25
4. RESULTS ... 26
4.1 Multicollinearity and Descriptive Statistics ... 26
4.2 Regression Analysis ... 27
5. CONCLUSION ... 33
5.1 Discussion ... 33
5.2 Limitations... 35
5.3 Recommendations for Future Research ... 37
REFERENCES ... 40
APPENDICES ... 48
APPENDIX A: Funding Options for Multinationals ... 48
APPENDIX B: Emerging Markets and EMNEs ... 49
APPENDIX C: List of Countries of Origin ... 51
APPENDIX D: List of Countries of Destination ... 52
APPENDIX E: Worldwide Governance Indicators and Interpretation ... 57
APPENDIX F: Mahalanobis Index for calculating Distance, Formula and Properties ... 58
APPENDIX G: Overview Variables ... 59
APPENDIX H: Econometric Model of Robustness Check Simple Effects Model ... 60
APPENDIX I: Normality Test Ln FDI ... 61
APPENDIX J: VIF and CTL Output ... 62
APPENDIX K: Correlation Matrix ... 63
APPENDIX L: Descriptive Statistics after Sample Split ... 64
APPENDIX M: Descriptive Statistics for Four Cases ... 65
APPENDIX N: The Effect of Institutional Distance on FDI Flows, Robustness Check ... 67
LIST OF TABLES AND FIGURES
Table 1: Motivations for Foreign Direct Investment by EMNEs………. 9
Table 2: Institutional Differences Developed Countries and Emerging Market Countries……. 13
Table 3: Descriptive Statistics all Variables………. 27
Table 4: The Effect of Institutional Distance on FDI, all Markets………... 28
Table 5: The Effect of Institutional Distance on FDI, Developed Country Perspective……... 31
Table 6: The Effect of Institutional Distance on FDI, Emerging Market Country Perspective... Table A1: Funding Options for Multinationals……..………... 32 48 Table C1: List of Countries of Origin………... 51
Table D1: List of Countries of Destination………... 52
Figure F1: Formula to calculate Mahalanobis Distance………... 58
Table F1: Properties of Different Methods for Calculating Dyadic Distance………... Table G1: Overview Variables………. 58 59 Table I1: Normality Test Ln FDI………. 61
Table J1: VIF and CTL Output……… 62
Table K1: Correlation Matrix………... 63
Table L1: Descriptive Statistics Developed Countries……….………... 64
Table L2: Descriptive Statistics Emerging Market Countries..………... 64
Table M1:Descriptive Statistics Emerging Market Countries to Emerging Market Countries... 65
Table M2: Descriptive Statistics Developed Countries to Developed Countries……… 65
Table M3: Descriptive Statistics Emerging Market Countries to Developed Countries………. 65
Table M4: Descriptive Statistics Developed Countries to Emerging Market Countries………. 65
Table N1: The Effect of Institutional Distance on FDI, Developed Country Perspective…….. 67
LIST OF ABBREVIATIONS Concepts
DMNE - Developed Market Multinational Enterprise EMNE - Emerging Market Multinational Enterprise
FDI - Foreign Direct Investment
GDP - Gross Domestic Product
OLI-framework - Ownership Location Internalization framework
SOE - State Owned Enterprise
Methodology
FE - Fixed Effects
NLS - Non-linear Least Squared
OLS - Ordinary Least Squared
PML - Pseudo Maximum Likelihood
Organizations and others
CEPII - Centre d’études prospectives et d’information internationale
IMF - International Monetary Fund
OECD - Organization for Economic Cooperation and Development UNCTAD - United Nations Conference on Trade and Development
WGI - Worldwide Governance Indicators
1. INTRODUCTION
Since 1970, global FDI in- and outflows have become increasingly more complex as many poorer countries have been growing at a faster rate than richer countries. There has also been a geographical shift in the origin or destination of Foreign Direct Investment (FDI) flows (Buchanan, Le, & Rishi, 2012). In the last decade, emerging markets have assumed a prominent position in the global economy and they account for more than half of world’s GDP on the basis of purchasing power in 2017 (IMF, 2017). During the last two decades, outward FDI from emerging markets grew at a higher year average rate than those from developed economies. Their participation of emerging markets in global outward FDI flows grew from 8% in 2000, to 27% in 2016. In terms of outward FDI stocks, their participation increased from 10% on 2000 to 23% in 2016 (the World Bank, 2018). The rapid rise and development of emerging markets has attracted tremendous interest from both investors and scholars (Marquis & Raynard, 2015).
This study is interested in the determinants of FDI flows which differ across developed countries and emerging market countries and in particular the effect of institutional distance. As argued in many emerging market literature (Goldstein, 2007; Guillén, 2005; Luo & Tung, 2007), EMNEs have different determinants of outward FDI flows than their developed country counterparts. Whereas MNEs are driven by market size and cheap labor costs in emerging markets, EMNEs are driven by strategic asset seeking factors when investing abroad. In addition, emerging market countries are characterized by a weak institutional environment (London & Hart, 2004), in contrast to developed countries, characterized by a strong institutional environment (Zhou, Xie, & Wang, 2016). The idea that the institutional environment matters for doing business is widely accepted, both within and outside international business and management studies (Scott, 2008). DMNEs and EMNEs operate in multiple institutional environments simultaneously and the dissimilarity between these institutional environments creates challenges due to legitimacy problems and the MNE being unfamiliar with the institutions of the host country (Eden & Miller, 2004). The negative effect of institutional distance on FDI is widely analyzed and confirmed in previous literature (Ghemawat, 2001; Xu & Shenkar, 2002; Zaheer, Schomaker, & Nachum, 2012).
able to cope with institutional distance. Do emerging market countries have an advantage compared to developed countries in institutional different environments? Do the different FDI motivations of DMNEs and EMNEs have an influence on the effect of institutional distance on FDI? Is it easier to adapt to a weak institutional environment or to a strong institutional environment? To answer these kind of questions, this study will examine whether and why the effect of institutional distance on FDI is heterogenous across different types of countries. Therefore this study answers the following research question:
How and why does the effect of institutional distance on FDI differ between developed countries and emerging market countries?
This study aims to answer this question using a panel data set with a sample of 71 countries of origin and 191 countries of destination from 2001 to 2012. Following previous research, I will explain FDI using a gravity model with two high dimensional fixed effects extended with the measure of institutional distance. The gravity equation is the common estimation method for FDI. To give a more nuanced view on the effect of institutional
distance on FDI, the sample is split into 8 cases, namely developed-all, developed-developed, developed-emerging, developed-developing, all, developed, emerging-emerging, and emerging-developing.
This study will contribute to the existing literature in the following matters. First, I contribute to the literature by exploring the heterogenous effects of institutional distance on FDI of developed countries and emerging market countries which has not been done before. As discussed in section 2, the literature on institutions and FDI is mainly devoted to studying the impact of a strong institutional environment on inward FDI. The emerging market
perspective related to outward FDI and institutional distance stays out. The broad country coverage in the sample gives insights for studying the different impacts of institutional distance. Second, this study seeks to add to the small but growing literature of emerging markets and EMNEs in general as it will explore the institutional environments of EMNEs and the determinants for EMNEs to engage in FDI. Last, I contribute to the literature because I adopt the calls of van Hoorn and Maseland (2016) and include multiple reference countries of origin. Current literature has not adequately distinguished between the institutional
the variation in institutional profiles of the partner countries. This conflation undermines the validity of a large amount of institutional distance studies in international business research. The rapidly growing literature on emerging markets and increased data availability on countries that are institutionally relatively different from developed countries offers great opportunities for moving beyond those single reference point analyses.
2. LITERATURE REVIEW
This section will review relevant literature concerning FDI and institutional distance. Section 2.1 will introduce FDI as concept and will end with a detailed discussion about the FDI in- and outflows of emerging market countries and the FDI determinants of EMNEs. Section 2.2 presents the concept of institutional distance and will subsequently end with the discussion of this concept related to EMNEs. In the last section the relation between FDI and institutional distance will be discussed and the hypotheses for this study are presented. To prevent confusion I would like to note upfront that throughout this study EMNEs are considered to be the multinational enterprises originating from emerging market countries and DMNEs are considered to be the multinational enterprises originating from developed countries.
2.1 Foreign Direct Investment
In this study foreign direct investment will be used as a proxy for multinational activity of firms in foreign markets. To create a well-refined theoretical foundation first the basic theories of why firms engage in international business will be discussed. Also the concept of horizontal and vertical multinational activity and the difference between these and FDI will be explained. At the end of this section the different determinants for DMNEs and EMNEs to engage in FDI are analyzed.
2.1.1 Multinational Activity and Foreign Direct Investment
Firms have different motivations to engage in international business activities and to become a ‘multinational’. An important question that kept, and still keeps, researchers busy is why and how firms want to engage in economic activities in foreign countries. The two basic theoretical frameworks to answer this question are Porter’s Diamond Model and the OLI-framework of Dunning. Both OLI-frameworks try to explain multinational activity of firms and integrate firm- and nation-level factors.
sources of firm-specific advantages. The last factor is the government whose role it is to influence all the other factors.
According to the OLI-framework a firm will engage in multinational activities if Ownership, Locational, and Internalization advantages concur (Dunning, 2000). The ownership aspect includes firm specific advantages that the firm can transfer abroad to exploit and refers to the question why firms go abroad. The second aspect location includes country specific characteristics like the availability of labor or capital and refers to the question where firms go to when going abroad. The internalization aspect relates to the costs of organizing the transaction and refers to the question how firms go abroad.
Porter’s Diamond Model is home-country driven and does not take the host country conditions into account. For international business scholars this is the most common point of critique and therefore the model is often extended as for example done by Asmussen, Pedersen and Dhanaraj (2009). As opposed to Porter’s Diamond Model, the OLI paradigm does include the foreign host country and is therefore a better fit within international business studies. Guillén and García-Canal (2009) for example use the OLI-framework when arguing that multinationals exist because certain economic conditions and proprietary advantages make it desirable and possible for firms to profitably undertake production of a good or service in a foreign country. In other words, when ownership, locational, and internalization advantages of the OLI-framework concur (Dunning, 2000).
Following previous literature, the multinational firm is defined as “a firm owning and controlling value added activities in two or more countries” (Guillén & García-Canal, 2009: 25). As discussed by Buch, Kesternich, Lipponer, and Schnitzer (2009), a firm has four options when deciding to engage in business with foreign countries, namely exporting, importing, horizontal multinational activity, and vertical multinational activity. Exporting refers to producing at home and selling in the foreign country, importing refers to importing goods or services without being present in the foreign country (Guillén & García-Canal, 2009).
producing and or selling the good abroad. This producing and selling abroad are vertical and horizontal multinational activity, respectively.
As discussed by Hernandez and Guillén (2018), vertical expansion occurs when the firm locates assets or employees in a foreign country with the purpose of securing the production of either raw material, component, or input (backward vertical expansion) or securing the distribution and sale of a good or service (forward vertical expansion). There are two conditions for engaging in this expansion; first, the presence of a comparative advantage in the foreign country, like low labor or capital costs (location advantage in the OLI-framework) and second, the presence of benefits for the firm from undertaking foreign production itself rather than relying on others (internalization advantage in the OLI-framework) (Guillén & García-Canal, 2009). As argued by Guillén and García-Canal (2009), two reasons for undertaking production by the firm itself are uncertainty about the supply or distribution and high asset specificity. If uncertainty is high the firm prefers to integrate forward or backward into the foreign location to make sure that the supply chain or distribution functions smoothly. Asset specificity is high when the firm and the foreign distributor or supplier need to develop joint assets in order for the supply operation to take place. This asset specificity refers to physical asset specificity, defined as “transaction-specific capital investments that tailor processes to particular exchange partners” (Dyer & Singh, 1998: 662). When uncertainty and high asset specificity is in place, the firm prefers to expand forward or backward to avoid opportunistic behavior of the partner (Guillén & García-Canal, 2009).
Horizontal expansion occurs when the firm sets up a factory or service delivery facility in a foreign country with the goal of selling in that market. This is desirable in the presence of protectionist barriers, high transportation costs, unfavorable currency exchange rate shifts, or requirements for local adaption that make exporting unprofitable (Hernandez & Guillén, 2018). The firm should consider the advantages of licensing a local producer or establishing an alliance against those of committing to a foreign investment. The decision to engage in horizontal multinational activity is driven by the desire to protect valuable intangible firm assets like brands, technology, or know-how that make licensing too risky as the licensee might act opportunistic (Guillén & García-Canal, 2009).
Up to now, the question of why firms engage in multinational activity is theoretically underpinned and the clarification between multinational activity and FDI is made. What follows is the exploration of FDI in- and outflows related to emerging market countries and their multinational enterprises. In Appendix B background information of the characteristics of emerging market countries and EMNEs and related FDI patterns are presented.
2.1.2 Determinants of FDI Flows into and from Emerging Markets
Firms engage in FDI because they see advantages in foreign markets. The framework established by Dunning suggests three primary motivations to engage in FDI (Dunning, 1980). These motivations are natural resource-seeking, market-seeking, and efficiency-seeking (Dunning, 1980). Firms motivated by natural resource-seeking are interested in accessing and exploiting natural resources. Firms motivated by market-seeking are interested in serving other markets with attracting factors as market size or growth. Lastly, efficiency-seeking (or cost reduction) firms want to benefit from factors that enable it to compete in international markets, as they could benefit from privatization methods or institutional environments. As discussed by Hernandez and Guillén (2018), horizontal and vertical activity can be mapped onto Dunning’s classification. Market-seeking investments are horizontal investments, while both efficiency-seeking and natural resource-efficiency-seeking investments are vertical investments. Since firm specific assets have become mobile across national boundaries, Dunning (1998) proposed an additional, fourth motivation, namely strategic asset-seeking. Strategic asset-seeking is the looking for knowledge-related assets or markets necessary to protect these assets, and the institutional and other variables influencing the ease or difficulty at which such assets can be acquired (Dunning, 1998).
These motivations are general motivations and not yet related to different types of countries. Motivations for FDI into and from emerging market countries differ.
invest in emerging markets can be summarized as market size and low labor costs, which are market-seeking factors.
As this study aims to compare the effect of institutional distance on FDI between developed countries and emerging market countries, a more interesting question are the determinants of FDI outflows of EMNEs.
To my knowledge, there exist only three studies that include multiple emerging market countries in the analysis of the determinants of FDI outflows. First, Cuervo-Cazurra (2006) relies on FDI flows data, but restricts his sample to one year and to the topic of corruption. He found a negative relationship between corruption and FDI. Second, Darby, Desbordes, and Wooton (2009) constructed a panel dataset for a number of foreign affiliates of emerging and developing companies. They found that the positive impact of good public governance on FDI is moderated and even in some cases eliminated or reversed when EMNEs had prior experience of poor institutional quality at home. EMNEs with little experience are deterred much more by bad public governance conditions (Darby et al., 2009). Last, Aleksynska and Havrylchyk (2013) analyzed location choices of EMNEs, with a particular interest on institutional distance and natural resources. They found that EMNEs invest more regional than DMNEs and that large institutional distance discourages FDI inflows depending on whether investors choose markets with better or worse institutions (Aleksynska & Havrylchyk, 2013).
Given the novelty of the subject and scarcity of the data, the academic literature about FDI outflows of emerging markets is yet very limited and most existing papers are either descriptive or have a regional focus (Aleksynska & Havrylchyk, 2013).
An example of the latter is the study of Buckley, Clegg, Cross, Liu, and Zheng (2007), who examined the determinants of Chinese outward direct investment and found market size, natural resource endowments, institutional environment, and cultural and geographical proximity as the main drivers. Their paper represents one of the first large-sample empirical studies of FDI by EMNEs (Hernandez & Guillén, 2018).
counter-attack global rivals’ major foothold in their home country market, (4) bypass stringent trade barriers, (5) alleviate domestic institutional constraints, (6) secure preferential treatment offered by emerging market governments, and (7) exploit their competitive advantages in other emerging or developing market countries (Luo & Tung, 2007). Accordingly, EMNEs’ motives can be summarized as asset-seeking motives.
Another descriptive paper is the study of Guillén and García-Canal (2009) who identified similar motivations of EMNEs, shown in Table 1.
TABLE 1
Motivations for Foreign Direct Investment by EMNEs
Source: Guillén and García-Canal (2009, 29).
Additional to the various motivations of FDI between DMNEs and EMNEs, organizational features also differ. Worthwhile to highlight in the light of this study are the political capabilities. For the DMNE this capability is weak because DMNEs are used to stable political environments. For the EMNE, however, this capability is strong because EMNEs are used to dealing with unstable governments in their home countries. They have a higher change than the DMNEs to succeed in foreign markets characterized by a weak institutional environment (Cuervo-Cazurra & Gene, 2008). Furthermore, a second feature is the expansion path, whereas the DMNE follows a simple and more gradual path, from less to more distant countries, the EMNE follows a dual path, with simultaneous entries into developed countries and emerging market countries. Last, the DMNE has strong competitive advantages because they have the required resources available in-house, as opposed to the EMNE, who has weak competitive advantages as they need to upgrade their resources.
In sum, EMNEs are driven by strategic asset-seeking factors when investing abroad, while DMNEs tend to be driven by market-seeking factors, namely growing market sizes and cheap foreign labor. Lastly, EMNEs have strong political capabilities as they are used to a weak institutional environment and have therefore a comparative advantage in similar emerging market countries or developing countries than their DMNE counterparts.
2.2 Institutional Distance
This study draws on institutional theory as the leading theory. Hence, the following section will discuss the concepts of institutions and institutional distance. At the end of this section, the institutional environments in emerging market countries will be explored.
2.2.1 Institutions
the requirements of local institutions, which may differ from the institutions of their home countries (Cezar & Escobar, 2015). In an increasingly complex and integrated global economy, a significant challenge for DMNEs and EMNEs is to navigate through these divers institutionally contexts (Marquis & Raynard, 2015). The literature often distinguishes between regulative, normative, and cognitive environments, with the first term describing formal institutions and the second and third referring to culture or other informal institutions. This study focuses solely on distance with regard to the formal, regulatory institutions since it is interested in the differences between developed countries and emerging market countries, clearly visible in their institutional environments.
2.2.2 Institutional Profile versus Institutional Distance Effects
There are two arguments that prevail in the literature about how institutions matter for international business, namely institutional profile and institutional distance (Jackson & Deeg, 2008). First, institutional profile is the generic argument and refers to the idea that firms are embedded in the institutional environment of the host country and face distinct challenges and opportunities as a result of this (Meyer, Estrin, Bhaumik, & Peng, 2009). Second, a related but different argument is that institutional distance matters. This argument is more specific to international business. Institutional distance refers to the dissimilarity between the institutional environments of two or more countries in which an MNE is active (Ghemawat, 2001; Xu & Shenkar, 2002; Zaheer et al., 2012). As argued by Eden and Miller (2004), MNEs operate in multiple institutional environments simultaneously and the dissimilarity between these institutional environments creates challenges due to legitimacy problems and the MNE being unfamiliar with the institutions of the host country.
exposure to the unfavorable host-country institutional environment and institutional distance requires efforts to bridge the distance between home and host environments (Van Hoorn & Maseland, 2016).
2.2.3 Institutions and Emerging Market Multinational Enterprises
Numerous institutional differences, like political or legal factors, differentiate the business environments of developed countries from those of emerging market countries (Marquis & Raynard, 2015). Zhou et al. (2016) argue that emerging market countries often lack transparent and stable rules of the game and have a higher probability of sudden changes in their political and economic environments, such as sudden government regime changes, tax increases, or exchange rate controls. The governments are prone to external conflicts, coupes, and internal tensions which increases the risk of unstable resource exchanges and information flows (Hiatt & Sine, 2014). Due to the less developed government and regulatory infrastructures of emerging market countries, the market regulation, corporate governance, transparency, accounting standards, and intellectual property protection are not so reliable or mature as those in developed countries (Marquis & Qian, 2014; Marquis, Zhang, & Zhou, 2011). Another critical factor is the strong influence of the government and the prevalence of state-owned firms, or SOEs (State Owned Enterprises) as discussed by Musacchio and Lazzarini (2014). In response on the weak institutions, EMNEs often internalize institutions via business groups. Highly diversified business groups can be particularly well suited to the institutional context in most emerging and developing market countries (Khanna & Palepu, 1997; Ramachandran, Manikandan, & Pant, 2013).
Opposing to the weak institutional environment of EMNEs, DMNEs originate from countries with more entrepreneur-friendly regulations, better protection of intellectual property rights, less corruption, and more transparent and well-functioning capital markets, all of which makes the outcome of trading or investing in those markets easier to forecast and eventually to perform (Zhou et al., 2016). The differences in the institutional environments of developed countries and emerging market countries are compared and presented in Table 2.
In general, the institutional environments of developed countries and emerging market countries can be considered as strong and weak respectively.
TABLE 2
Institutional Differences Developed Countries and Emerging Market Countries
Source: Marquis and Raynard (2015, 303)
2.3 Foreign Direct Investment and Institutional Distance
Institutions are relevant in international business because different legal, political, and administrative systems create adaptation costs that determine the international attractiveness of a host country (Beugelsdijk, Groot, Linders, & Slangen, 2004). These costs to adapt to the institutional environment of the host country are called fixed FDI costs (Cezar & Escobar, 2015). However, as developed countries have strong institutional environments and emerging market countries have weak institutional environments, the environments matter differently for FDI.
contract enforcement are especially important (Busse & Hefeker, 2007; Naudé & Krugell, 2007). Many other studies have found a positive relationship between strong institutional quality and FDI (Buchanan et al., 2012; Gani, 2007; Jensen, 2003; Li & Resnick 2003).
In addition, numerous studies demonstrate that weak institutional environments of emerging market countries deters FDI. North (1994) argues that institutions affect economic activities because they affect transaction and production costs. When property rights are poorly protected and contracts are difficult to enforce, the risk will be high and so will be transaction costs. Inefficient institutions can raise production costs by disrupting the supply chain when for example delays happen in obtaining permits. Next to that, when corruption (Wei, 2000) or criminality (Daniele & Marani, 2011) are present in foreign countries this will also bring additional costs to FDI. Many other studies have found a negative relationship between weak institutional quality and FDI (Ali, Fiess, & MacDonald, 2010; Estrin & Meyer, 2004; Globerman & Shapiro, 2003; Peres, Ameer, & Xu, 2018).
To conclude, developed countries have strong institutions and political certainty and stability which attract FDI in contrast to emerging market countries who have weak institutions and political uncertainty and instability which deters FDI.
The above discussion refers to the institutional profile of countries which differs from institutional distance. Several recent studies propose that not only the institutional profile matters, but also the institutional distance between home and host countries (Aleksynska & Havrychyk, 2013; van Hoorn & Maseland, 2016). MNEs operate in multiple institutional environments and the dissimilarity (distance) between these institutional environments can create all sorts of misunderstanding and legitimacy problems which deters FDI (Eden & Miller, 2004).
Another important phenomenon is explained by the concept of institutional distance, as opposed to institutional profile. As discussed above, poor institutional quality (profile) is deterring FDI. However, this concept does not answer the question of why EMNEs still engage in FDI with emerging market countries, although the host country’s institutional environment is weak. This phenomenon can be explained by the notion of institutional distance. EMNEs invest in other emerging market countries because they have similar institutional environments and therefore institutional distance is low.
The studies of Habib and Zurawicki (2002) and Bénassy-Quéré, Coupet, and Mayer (2007), both argue that EMNEs familiar with weak institutions may have a comparative advantage in investing in other emerging or developing market countries that suffer from a weak institutional environment. Also Cuervo-Cazurra and Gene (2008) argue that emerging market countries have a higher change than developed countries to succeed in foreign countries characterized by a weak institutional environment.
As argued by van Hoorn and Maseland (2016) current literature has not adequately distinguished between the two effects of the institutional environment (profile) or institutional distance. The majority of studies focused on just one single reference country of origin and therefore variation in institutional distance between the reference country and the partner countries is identical to the variation in institutional profiles of the partner countries. Taking into account the calls of van Hoorn and Maseland (2016) to include a diverse group of reference countries, this study will re-evaluate the existing findings on the relationship between institutional distance and FDI, in which it is argued that greater institutional distance will have a negative effect on the amount of FDI, by including a wide range of home and host countries. This is reflected in Hypothesis 1.
Hypothesis 1: The relationship between institutional distance and FDI is negative, such that an increase in institutional distance reduces FDI flows.
developing countries, however they examined institutional quality, rather than distance. The authors found an insignificant effect of institutional quality because of the weak structure of institutions in developing countries.
As EMNEs have experience with weak institutional environments I hypothesize that the size of the effect of institutional distance on FDI will be lower as opposed to DMNEs (which do not have experience with weak institutional environments), when investing in developing countries. This is reflected in Hypothesis 2.
Hypothesis 2: The negative effect of institutional distance on FDI is lower for emerging market countries compared to developed countries when investing in developing countries.
As argued in the literature review, developed countries and emerging market countries have very different institutional environments, considered as strong and weak respectively. This study argues from the perspective of the origin country, so either the developed country or the emerging market country. Both have the options to engage in FDI with either developed, emerging, or developing countries. Transferring the general negative effect of institutional distance on FDI to the context of emerging market countries, the question arises whether this effect is equal for emerging market countries compared to developed countries.
The institutional distance index number between the Netherlands and China is the same, meaning that the institutional distance from the Netherlands to China is assumed to be the same as the distance from China to the Netherlands. Knowing that this distance is the same, should the effect on FDI be the same as well? The answer is likely to be no since the Netherlands engages in the weak institutional environment of China and China engages in the strong institutional environment of the Netherlands, both pursuing different motives to engage in FDI with such countries. As discussed previously, especially the acquisition and exploitation of firm-specific intangible assets as latest technologies and famous brands is important for EMNEs. These assets are likely to develop in institutionally friendly (strong) environments. Next to that, emerging market countries have shortages in their institutional environments. This encourages the EMNEs to internationalize in order to seek safer environments for business (Gaffney, Karst, & Clampit, 2016). In these cases, EMNEs are willing to invest in strong institutional environments, even though there is a large institutional distance.
some cases than in others. A large institutional distance should not necessarily deter EMNEs, but could even attract investors when institutions in the destination countries are better than at home. Aggregating this to a country level, even when institutional distance is the same between two countries, the effect is likely to be less strong (in absolute terms) when emerging market countries engage in FDI with developed countries compared to developed countries engaging in FDI with emerging market countries. This reasoning is reflected in Hypothesis 3.
3. DATA AND METHODOLOGY
This section will discuss the methodology of this study. Section 3.1 presents the chosen sample. Section 3.2 describes the nature and measurement of the variables used to test the hypotheses. In the last section the estimation method, namely the gravity model with high dimensional fixed effects is discussed.
3.1 Sample
This study focuses on two sets of countries, namely developed countries and emerging market countries. In this way the comparison between the two sets is possible. The choice of the time span and number of countries were contingent upon the availability of data on all variables. The dataset used in this study covers 71 origin countries sending FDI to 191 destination countries during the period 2001 to 2012, reflecting in total 40,076 FDI flows. Among the 71 countries, 26 are developed countries and 45 are emerging market countries.
3.2 Data
3.2.1 Dependent Variable
The dependent variable in this study is the flow of Foreign Direct Investment (FDI) from the home country to the host country, as a proxy for the total activity of multinationals in the home country in the respective host country in a given year.
This study follows the latest definitions of FDI and FDI flows of the United Nations Conference on Trade and Development (UNCTAD, 2017). According to UNCTAD (2017), FDI is an investment made by a resident enterprise in one economy (direct investor or parent enterprise) with the objective of establishing a lasting interest in an enterprise that is resident in an another economy (direct investment enterprise or foreign affiliate). The lasting interest implies the existence of a long-term relationship between the direct investor and the direct investment enterprise and a significant degree of influence on the management of the enterprise. The ownership of 10% or more of the voting power of a direct investment enterprise by a direct investor is evidence of such a relationship (UNCTAD, 2017).
- Acquisition or disposal of equity capital. FDI includes the initial equity transaction that meets the 10% threshold and all subsequent financial transactions and positions between the direct investor and the direct investment enterprise.
- Reinvestment of earnings which are not distributed as dividends.
- Inter-company debt transactions, i.e. transactions related to short- and long-term borrowing and lending of funds between parents and affiliates.
The FDI flow data were taken from the UNCTAD database. UNCTAD regularly collects published and unpublished national official FDI flow data directly from central banks, statistical offices, or national authorities and international organizations such as the International Monetary Fund (IMF) or the World Bank. These data constitute the main source for the reported data on FDI. In the document Methodological Notes of the UNCTAD (World Investment Report, 2016) all the sources of the FDI outflows per country are listed. I refer to this document for reviewing all these sources and methodology of the data. The FDI data are expressed in millions of dollars and reflect net acquisitions of assets. Thus, decreases of net acquisitions of assets like reverse investment or disinvestment are recorded as negative flows (World Investment Report, 2016).
Within the sample, no distinction has been made regarding the 45 emerging market countries, all countries which are labeled as emerging market according to UNCTAD are included which implies a high external validity. The 26 developed countries are the countries within the OECD convention. Six OECD countries appeared to be simultaneously emerging market countries, those countries are treated as emerging market countries (Chile, Czech Republic, Greece, Hungary, Mexico, and Poland). Appendix C provides a full list of the origin countries. With regards to the 191 destination countries, 36 are developed countries, 49 emerging market countries, and 106 developing countries. Appendix D provides a full list of the destination countries. Lastly, this study makes use of a 12 year time frame from the years 2001 to 2012 due to data availability.
3.2.2 Independent Variable
Institutional distance is measured based on the Worldwide Governance Indicators (WGI). These indicators are commonly used in the literature to measure institutional distance (Beugelsdijk et al., 2004; Buchaman et al., 2012; van Hoorn & Maseland, 2016).
Violence/Terrorism); (b) the capacity of the government to effectively formulate and implement sound policies (Government Effectiveness & Regulatory Quality) and (c) the respect of citizens and the state for the institutions that govern economics and social interactions among them (Rule of Law & Control of Corruption)” (Kaufmann, Kraay, & Mastruzz, 2010: 35). The interpretation of these indicators can be found in Appendix E.
Data for institutional distance is kindly made available upon request and received from van Hoorn and Maseland (2016). The authors calculated the average countries’ scores on the six indicators and transformed them to distances using the Mahalanobis index. This dataset is used because of time efficiency reasons and because the reliable Mahalanobis index is used, which is the commonly used index of institutional distance (Berry, Guillén, & Zhou, 2010; van Hoorn & Maseland, 2016). The formula to calculate dyadic distance by the Mahalanobis index and the differences in properties for calculating dyadic distances between the Mahalanobis and the competing Euclidean index are shown in Appendix F.
3.2.3 Control Variables
In order to obtain unbiased econometric estimates of the effect of institutional distance on FDI outflows, this study controls for a number of factors which existing literature on the gravity model have identified as important determinants of FDI (Beugelsdijk et al., 2004; Bevan, Estrin, & Meyer, 2004; Nowak-Lehmann, Herzer, Martinez-Zarzoso, & Vollmer, 2007).
As will be discussed further in section 3.3, the two basic elements of the gravity equation are the economic sizes of the countries and geographical distance between the countries. As a proxy for economic sizes, I control for the Gross Domestic Product’s (GDP’s) of the home and the host countries (Buchanan et al., 2012). Additionally, GDP per capita is used as a proxy for purchasing power of the country. The reason for using both GDP and GDP per capita is that two countries with considerably different populations may have similar GDPs but totally different levels economic development, which is more accurately reflected in GDP per capita. Data for both variables were taken from the World Bank. However, as will be discussed further below, these variables are only included in robustness checks where I estimate gravity regressions with simple host-country and home-country fixed effects as the robustness check. In my main regressions I will include home-country-year (it) and destination-country-year (jt) fixed effects. As both GDP and GDP per capita differ per country and per year, these variables are effectively captured by the fixed effects.
natural logarithm is used for GDP, GDP per capita, and geographical distance (Bano & Tabbada 2015; Bellak & Leibrecht, 2011).
Furthermore, following the extension of the gravity equation in the literature, additional explanatory variables that reflect other aspects of distance will be included (Beugelsdijk et al., 2004; Fally, 2015; Head & Mayer, 2014). Specifically, I include dummies for capturing whether country i and j share a common border, whether country i and j share a common language, whether country i and j have a colonial relationship, and whether country i and j are part of the same country (e.g. Hong Kong and China). The data for these control variables come from the CEPII database which is a commonly used source when estimating gravity equations (Mayer & Zignago, 2011).
An overview that summarizes the variables is given in Appendix G.
3.3 Method and Empirical Model
This study is quantitative of nature since its purpose is to find statistical evidence for the hypotheses. The obtained data regarding the variables were entered into STATA, which allows to statistically analyze the data. The level of analysis is on the country level. To assess the influence of the independent variable on the dependent variable I estimate gravity equations.
3.3.1 Gravity Equation and its Empirical Aspects
Tij = A x Yi x Yj / Dij
• Where Tij are the bilateral trade or FDI flows between two countries i and j,
• A is reflecting a gravitational constant,
• Yi and Yj are the economic sizes of the two countries,
• Dij is the geographical distance between the two countries i and j.
Since Tinbergen (1962) a considerable amount of research has been done on the gravity equation in an attempt to improve the empirical aspects. Although the gravity equation has always been successful in providing economically and statistically significant results, the model was criticized for its lack of theoretical foundations (Kepaptsoglou, Karlaftis, & Tsamboulas, 2010). The empirical literature reveals a wide number of publications offering methodological advancements or refinements of the gravity equation.
The first turning point was in 1979, when Anderson (1979) presented the first well-grounded theoretical foundation for the gravity model. It was based on economic theory; the constant elasticity of substitution (CES) preferences and goods that are differentiated by region of origin. During that time the usual technique for estimating the gravity model was the Ordinary Least Squared (OLS) estimation method in its log-linear form (Nowak-Lehmann et al., 2007). However, researchers indicated methodological flaws when using the OLS-method. In most empirical studies, implementation assumptions were not in line with the underlying theoretical models (Henderson & Millimet, 2008).
Another, even more important turning point was the work of Anderson and van Wincoop (2003) which refined the theoretical foundations of the gravity equation by accounting for the endogeneity of trade costs and considering institutional barriers to trade. The authors showed that costs of bilateral trade between two regions are affected by the average trade costs of each region with the rest of its trading partners (the so-called lack of multilateral resistance). The traditional gravity equation estimated with OLS was not correctly specified because it did not take into account these multilateral resistance terms that are correlated with both exporting trade potential and bilateral trade costs (Anderson & van Wincoop, 2003).
their interactions. In this respect, the use of two-way fixed effects models has been recognized as potentially useful to capture simultaneously cross-country and time interactions (Baltagi et al., 2003). Two-way fixed effects combine the fixed effects of both groups (countries) and time, in this study origin-country-year and destination-country-year respectively, which will be discussed further in section 3.3.2.
The FDI data used in this study are a classical example of multi-dimensional panel data since FDI flows are observed for each country of origin i towards multiple countries of destination j (Baltagi et al., 2017). Examining such data over more than one year t, increases the dimensionality of the data and so it becomes three-dimensional panel data.
In sum, this study will use the gravity equation to estimate the regression model. The chosen regression model is the linear regression model with two high dimensional fixed effects.
3.3.2 Linear Regression Model with two High Dimensional Fixed Effects
In this section the decision of using fixed effects rather than random effects is discussed, afterwards the concept of fixed effects is elaborated on.
The linear regression model can in principle be estimated with either fixed or random effects. According to Egger (2002), the choice between fixed and random effects models depends on the interests of the analysis, sample, data properties, and the underlying theoretical model. Random effects models should be considered if they are adequately consistent and if there is an interest in estimating time-invariant effects (Egger, 2002). In this study, the variable of interest is institutional distance which is considered as time-invariant. In this respect one could argue that a random effects model should be used to estimate the model. However, most empirical studies estimating the gravity equation indicate that fixed effects models tend to provide more robust results, especially when the dataset is large (Antonucci & Manzocchi, 2006; Baltagi et al., 2017; Glick & Rose, 2002). Fixed effects are affecting bilateral FDI flows but are unobservable for the researchers (Guimarães & Portugal, 2009).
heterogeneity of origin-year & destination-year pairs), which make it possible to account for all inter-group variability (Guimarães & Portugal, 2009).
In three-dimensional panel data, the dependent variable, in this case FDI, is observed along three indices i, j and t. Previous studies have used country pair, country specific, or time specific fixed effects. In this study, the independent variable is institutional distance which varies by country pair (ij), but not over time (t). Therefore, country pair fixed effects (ij) cannot be used since this will blur the effect of institutional distance. The baseline model that this study will use is the model with origin-country-year (it) and destination-country-year (jt) fixed effects. In this way the fixed effects will not blur the effect of institutional distance. However, there may be other pair-specific effects that are influencing FDI besides institutional distance. I will aim to control for this by including the above-mentioned control variables, namely geographical distance, common border, common language, colonial ties, and same country, which differ across country pairs (ij) but not over time (t). GDP and GDP per capita are not included in this baseline model as they are already captured by the fixed effects it and jt. More formally, the estimated model is of the form:
Ln FDIijt = β1 IDij + β2 ln DISTij + β3 Border + β4 Language + β5 SameCountry + β6 ExComColony + β7 ExColony + β8 CurColony + αit + αjt + ε,
where i denotes the country of origin or FDI-sending country, j denotes the country of destination or FDI-receiving country, and t denotes time in years. The variables are defined as follows:
• Ln FDIijt is the log of FDI-outflow from home country i to host country j at year t,
• IDij is the institutional distance between country i and country j,
• ln DISTij is the log of the geographical distance between countries i and j,
• Border is a binary variable equal to 1 if i and j share a common border,
• Language is a binary variable equal to 1 if i and j share a common official language, • SameCountry is a binary variable equal to 1 if i and j are the same country,
• ExComColony is a binary variable equal to 1 if i and j were ever colonies after 1945 with the same colonizer,
• ExColony is a binary variable equal to 1 if i ever colonized j or vice versa,
• αit is the fixed effect of country j and year t,
• ε term captures the error term.
To check the robustness of the results, a simple fixed effects model will be tested, with the simple fixed effects for i, j, and t respectively. In this model I have to control for variables that will differ across country pairs (ij) and variables that vary for each country over time (t), as these will not be captured by the fixed effects. Variables that differ across countries and over time are GDP and GDP per capita. The econometric specification of this simple fixed effects model is shown in Appendix H.
For estimating the baseline model the user-written command reg2hdfe is the Stata implementation. The approach is computationally intensive but it has the advantage of imposing minimal memory requirements. The estimation involves a transformation to remove the unobserved effects prior to actual estimation of the model.
3.4 Testing the Normality Assumption
4. RESULTS
This section presents the results of the data analysis. I first check for multicollinearity and then present the descriptive statistics in section 4.1. This is followed by a discussion of the regression results in section 4.2.
4.1 Multicollinearity and Descriptive Statistics
Multicollinearity refers to the correlation between the independent variables in a multiple regression model (Wooldridge, 2009). To detect if multicollinearity is present, both correlation matrix and variance inflation factors (VIF) are used. VIF indicates how much the estimated variance of the estimated regression coefficient is increased due to collinearity. Prior research used a threshold of 10 for the maximum VIF and a threshold of 0.2 for the maximum of collinearity tolerance level (CTL) as the guideline for multicollinearity (Asteriou & Hall, 2011; Belsley, Kuh, & Welsch, 2005). As can be seen in Appendix J, the highest calculated VIF is 1.85 with a collinearity tolerance of 0.54 for the variable ‘Ever colonial tie’. As the highest VIF is below the threshold it can be concluded that there is no multicollinearity between the variables. In addition to this, I checked for multicollinearity by reviewing the correlation matrix. The output of this matrix can be found in Appendix K. As can be seen, there is no multicollinearity as there are no correlations above 0.7, which is a commonly accepted threshold (Cooper & Schindler, 2006).
The descriptive statistics of all variables are analyzed. These are shown in Table 3. Additionally, to get a first insight of the differences between DMNEs and EMNEs, for both the descriptive statistics are separately shown in Appendix L. First, the mean of FDI for developed countries is higher than for emerging market countries, meaning that developed countries engage in a higher amount of FDI than their emerging counterparts. Furthermore, the average institutional distance is higher for developed countries compared to emerging market countries. Last, the GDP per capita of origin is lower for emerging market countries compared to developed countries, which is logical since these countries face lower standards of living and economic development. Furthermore, to develop a better understanding of the developed country perspective compared to emerging market country perspective, an overview and analysis of the descriptive statistics of the cases developed-developed,
market countries invest in developed countries. Last, the maximum of institutional distance of the case emerging-emerging is higher than for the case developed-developed. All these results are in line with theory.
TABLE 3
Descriptive Statistics all Variables
4.2 Regression Analysis
In this section the results of two models will be discussed. First, the baseline model with high dimensional effects it and jt, and second the modified model to check for robustness of the results with the simple fixed effects i, j, and t and controls for GDP and GDP per capita.
When interpreting the models two things should be looked at, both the regression coefficients and the overall fit of the model. To assess the regression coefficients three things should be interpreted. First the significance (at a 0.1%, 1%, or 5% level), second the sign (either positive or negative), and last the magnitude. The regression coefficient represents the change in the DV for a one-unit increase in the IV while other regressors are held constant. To assess the overall models, the adjusted-R2 is used. It shows the proportion of the variation in FDI that is explained by the linear combination of all the regressors. It also takes into account the number of independent variables, sample size, and degrees of freedom. A higher
coefficient reflects a better fit of the model.
The results of the main models are shown in Table 4.
Variable Observations Mean Standard
Deviation Minimum Maximum Ln FDI 19,005 3.927671 2.554972 0 11.59999 Institutional distance 19,005 1.069808 .7828336 .0007798 3.63463 Ln GDP origin 19,005 26.95531 1.759792 20.79875 30.41327 Ln GDP destination 18,847 25.84012 2.050091 18.56176 30.41327 Ln GDP per capita origin 19,005 9.939027 1.008195 5.863631 11.65929 Ln GDP per capita destination 18,847 9.217175 1.440908 4.70953 11.65929 Ln geographical distance 18,970 8.096963 1.077024 4.661588 9.885824 Common border 19,005 .0877664 .2829623 0 1 Common language 19,005 .1522231 .3592465 0 1
Ever colonial tie 19,005 .0676664 .2511792 0 1
TABLE 4
The Effect of Institutional Distance on FDI, all Markets
In the first half of the table the models 1 and 2 present the baseline models. Model 1 only includes the independent variable along with the fixed effects. Model 2 is the full model that includes also the country-pair specific control variables. In the second half of the table, models 3 and 4 show for comparison the results of the robustness check, the results are similar to the baseline models 1 and 2.
countries share a common border, speak the same language, have colonial ties, or when they are part of the same country, FDI is higher.
Furthermore, the adjusted explanatory power in model 1 is 0.4964, in model 2 it is 0.5996. For the full model 2 this indicates that 60% of the variation of FDI outflows between the countries is explained by the variables included in the model, 40% of the variation stems from other determinants, which are not included in the model. Looking at the results in the second half of the table, we see that the results are robust to using an OLS regression with country and time specific simple fixed effects and controls for GDP and GDP per capita.
To be able to test the other two hypotheses, the sample of home countries is divided into developed countries and emerging market countries. The results are reported in Tables 5 and 6. In both cases, the case of FDI into all other countries is shown in part 1 of the table, FDI into developed countries in part 2, FDI into emerging market countries in part 3, and FDI into developing countries in part 4. To check the robustness of the results, the regression is run by both estimation methods. Since the linear regression model with two high dimensional effects is the method of interest, the results of the OLS regression with simple effects are shown in Table N1 for the developed country perspective and in Table N2 for the emerging market country perspective, both in Appendix N. The results are similar using both regression methods which implies that the results are robust.
To test hypothesis 2, the cases developed-developing and emerging-developing are compared. For the case developed-developing the coefficient of institutional distance in model 1 is -1.328 with a p-value of 0.01 and in model 2 it is -0.899 with a p-value of 0.05. The corresponding adjusted R2’s are 0.6141 and 0.6845 respectively.
To test hypothesis 3, the cases developed-emerging and emerging-developed should be compared. For the case developed-emerging, the coefficient on institutional distance is
-0.381 in model 1 and -0.606 in model 2, both with a p-value of 0.01. The adjusted R2’s are 0.5939 and 0.6659 respectively.
For the case emerging-developed the coefficients are in both models insignificant. The insignificant results for emerging market countries perspective imply that institutional distance does not have a negative effect. Since it is hypothesized that the negative effect of institutional distance will be stronger for developed countries engaging in FDI in emerging market countries than for emerging market countries engaging in FDI in developed countries, which also implies that the effect in the latter case could be zero, hypothesis 3 is supported. Furthermore, coming from the same country or speaking the same language appears to be of more importance for FDI for firms originating from emerging market countries than for firms originating developed countries.
To give a sense of the economical magnitude of the effect of institutional distance on FDI an illustration of the average and the majority (90th percentile) of observations follows. The
regression coefficient is -.304 for institutional distance in the case of developed-developed, implying a decrease of 30,4% in FDI if institutional distance increases by one unit, ceteris paribus. The average institutional distance between the developed-developed countries is .5035653 as can be seen in Table M2 in Appendix M. Typical pairs of countries whose distances are close to this average are Japan-Netherlands (.502063) or Switzerland-United States (.506914). On average, Japan sent $5,515 million to the Netherlands and Switzerland sent $11,528 million to the United States per year. The decrease in FDI for both country pairs, incorporating the effect of institutional distance will subsequently be $4,673 million for Japan-Netherlands ($5,515 million – 15,3% (30,4% x 0.502063)) and $9,752 million for Switzerland-United States ($11,528 – 15,4% (30,4% x 0.506914)).
TABLE 5
TABLE 6
5. CONCLUSION
This section concludes the thesis. I first discuss the results of the empirical analysis and the related theoretical and managerial implications. In section 5.2, I discuss the limitations of this study and in section 5.3 I provide recommendations for future research. Last, section 5.4 concludes.
5.1 Discussion
I hypothesized that institutional distance negatively affects FDI flows, but discussed that this effect is heterogenous across different types of countries. Simultaneously, I hypothesized that this effect is lower for emerging market countries compared to developed countries when investing in developing countries. In addition, I hypothesized that the effect is stronger for developed countries engaging in FDI in emerging market countries than for emerging market countries engaging in FDI in developed countries. I confirm the theoretical predictions and previous findings of the large negative effect of institutional distance on FDI, however as will be discussed further, this relationship does not hold in all cases.
In general, all six indicators incorporated in the measure of institutional distance used in this study are expected to affect FDI. Voice and accountability can influence FDI because encouraging the public to give input is important for democratic development, furthermore, political stability can reduce the risk of doing business (Gani, 2007). In addition, Gani (2007) also argues that when government effectiveness is in place, governments can use discretionary power over economic affairs as well that regulatory quality promotes private sector development. Also, when rule of law is in place, agents have confidence in the government and abide by the rules, finally, when there is control of corruption in the foreign country it becomes more attractive for foreign investors (Gani, 2007). When the institutions are as described, they are supporting an efficient market, allowing for lower transaction and production costs and reduce uncertainty. However, in emerging market countries these underlying economic mechanisms are typically underdeveloped.
indeed negative effects of institutional distance created by the weak institutional environments of emerging and developing market countries.
However, I also proposed and found evidence that this relationship between institutional distance and FDI is more complex for emerging market countries. Namely, I found that EMNEs are not always deterred by a large institutional distance.
First, I showed that the effect of institutional distance on FDI is smaller in absolute terms for firms originating from emerging market countries than for firms from developed countries when considering FDI in developing countries. As highlighted by Cuervo-Cazurra and Gene (2008), EMNEs have strong political capabilities because they are used to weak institutional contexts, while DMNEs have weak political capabilities because they are used to stable institutional contexts. Here I bring in a new concept, namely the concept of physic distance. Psychic distance is defined as “the sum of individuals perceptions regarding the differences between two countries in terms of language, education, business practices, culture, and industrial development” (Johanson & Vahlne, 1977: 24). This perceived distance is a major barrier or driver to the decision of firms to enter foreign markets. Psychic proximity would reduce either perceived uncertainty or learning costs (Habib & Zurawicki, 2002). In practice this implies that when EMNEs invest in developing countries with weak institutional environments, they will be less deterred by institutional distance because they are psychologically closer to them. EMNEs face a lower size of the effect of institutional distance on FDI when investing in developing countries than DMNEs because developing countries are familiar with such environments.
Second, scholars measure institutional distance as the absolute difference in institutions between two countries, which implies that the institutional distance from e.g. the Netherlands to China is assumed to be the same as the distance from China to the Netherlands. Thus, it is assumed that expanding to countries with better (the Netherlands) or worse (China) institutions has a similar effect (because institutional distance is high). However, it is reasonable to argue that the effect of this distance on multinational activity is different depending on where the firm originates from. When a Dutch firm invests in China it invests in a weak and unstable institutional environment while the Chinese firm invests in the strong and stable institutional environment of the Netherlands. Even though institutional distance is similar, the latter may have a less negative effect on multinational activity than when the Dutch firm should adapt to the weak Chinese institutional environment.
