Master’s Thesis
The influence of Corruption on the relation between Business Regulations and FDI: a quantitative analysis of Emerging Markets.
Author: Bart Bakker Schut
Student number: 3798925
Supervisor: dr. Hammad Haq
Co-assessor: dr. Sathyajit Gubbi
Program: MSc International Business & Management
Date: 13-06-2020
Word count: 11,352
ABSTRACT
The objective of this study is to determine to which extent the efficiency of business regulations influences the relative inflow of FDI specified on emerging markets, and to explore to which degree this relation is influenced by corruption. This study follows IMF’s (2019) view on emerging markets and is limited to the time period of 2010-2018. The results indicate a positive, insignificant relation between the efficiency of business regulations and the net inflows of FDI relative to the economy size. Further, the results reveal a negative insignificant moderating role of corruption within emerging markets. Three additional checks confirm the robustness of these findings. Limitations and directions for further research are provided eventually.
Key words: Foreign direct investment; FDI; Business regulations; Ease of doing business;
Corruption; Corruption Perception Index; CPI; Emerging markets; Emerging market economies; Emerging economies; Transition economies.
Acknowledgements
First, I would like to thank my supervisor dr. Hammad Haq and co-assessor dr. Sathyajit Gubbi for the time and effort they spend on providing valuable feedback. Without this supervision, this thesis would not have been on the level as it is today. Writing this thesis was a challenging task for me. Therefore, I would like to make use of the opportunity to thank my family for their unconditional commitment and encouraging words during this process.
TABLE OF CONTENTS
1. INTRODUCTION ... 5
2. LITERATURE REVIEW ... 8
Theory of Economic Regulation ... 8
Foreign Direct Investment ... 8
Business Regulations ... 9
Corruption ... 10
Emerging Markets ... 11
Hypotheses development ... 13
Conceptual model ... 15
3. METHODOLOGY ... 16
Research design ... 16
Sample selection ... 16
Dependent variable ... 17
Independent variable ... 18
Interaction term ... 18
Control variables ... 19
Data analysis ... 22
4. RESULTS ... 24
Descriptive Statistics, Correlations and Assumption testing ... 24
Regression statistics ... 27
Robustness check: Excluding the country year observations of 2010 ... 29
Robustness check: Employing WGI’s Control of Corruption ... 30
Robustness check: Employing the Doing Business Rankings ... 31
5. DISCUSSION AND CONCLUSIONS ... 32
Theoretical implications ... 32
Practical implications ... 34
Limitations ... 35
Further research ... 35
REFERENCES ... 37
Appendix 1: Overview of the case-study ... 48
Appendix 2: Initial analysis - Graphs and tables ... 49
Appendix 3: Robustness check 1 - Graphs and tables ... 50
Appendix 4: Robustness check 1 – FE GLS regression results ... 51
Appendix 5: Robustness check 2 - Graphs and tables ... 52
Appendix 6: Robustness check 2 – FE GLS regression results ... 53
Appendix 7: Robustness check 3 - Graphs and tables ... 54
Appendix 8: Robustness check 3 – FE GLS regression results ... 55
Appendix 9: Overview of the regression results ... 56
1. INTRODUCTION
The macro-economic relationship between business regulations and foreign direct investments has often been studied (Contractor, Dangol, Nuruzzaman, & Raghunath, 2020; Corcoran &
Gillanders, 2015; Malik & Jyoti, 2018; Olival, 2012; Singh, 2014; Zhang, 2012). In particular, Bayraktar’s study (2013) demonstrates the association between business regulations and the relative inflow of foreign direct investments as percentage of GDP, concentrated on the BRIC countries (i.e. Brazil, Russia, India and China) over a time period from 2004 to 2010. Even though the characteristics of emerging markets are described extensively (Arnold & Quelch, 1998; Haley & Haley, 2006; Hitt, Dacin, Levitas, Arregle, & Borza, 2000; Hoskisson, Eden, Lau, & Wright, 2000; Kvint, 2010; Meyer & Tran, 2006), there is some degree of variation within definitions. As a consequence, the literature provides no fixed selection of emerging market economies. Previous studies concentrate on the relation between business regulations and foreign direct investment within certain subgroups, for instance the combined samples of
“emerging” markets from the middle East and Asia (Aziz, 2018; Shahadan, Sarmidi, & Jan Faizi, 2014; Sjöholm & Lipsey, 2012) and the Gulf countries (Mina, 2007). Further research on this relation is devoted to ex-socialist countries (Jovanovic & Jovanovic, 2018) and the ASEAN countries (Vogiatzoglu, 2016), referring to the Association of Southeast Asian Nations. Other scholars concentrate on the relation between business regulations and foreign direct investments within “emerging” African nations (Akame, Ekwelle, & Njei, 2016; Morris
& Aziz, 2011; Nangpiire, Rodrigues, & Adam, 2018). The term “emerging” implies alteration over time and might, therefore, diminish the relevance of prior literature. The academic literature does not indicate whether the relationship between business regulations and foreign direct investments holds within a contemporary group of emerging markets.
Relative to developed economies, emerging markets are characterized by less adequate regulatory discipline, lower standards with respect to intellectual property and enforcement, complicated structures regarding loyalty and authority, opaque connections between business and politics, and governmental tendency to adjust regulations unpredictably and frequently (Arnold et al., 1998; Hitt et al., 2000). Bekaert (1995) contributes by describing the relatively weak credit rating agencies. Their rapid economic advancements are not constantly followed by proportional developments on social, legal and political ground which mitigates emerging markets’ governance systems to adequately address corruption (Kaymak & Bektas, 2014).
Certain scholars find support for a negative relationship between corruption and foreign direct investments (Habib & Zurawicki, 2002; Mauro, 1995; Méon & Sekkat, 2005; Mudambi,
Navarra, & Delios, 2013), suggesting that corruption is considered as unethical activities, which induce operational inefficiencies and costs. Contradictory to these findings, particular studies indicate a positive relationship between corruption and foreign direct investments (Barassi &
Zhou, 2012; Dreher & Gassebner, 2013; Egger & Winner, 2005; Leff, 1964; Méon & Weill, 2010; Sklar & Huntington, 1969). Their empirical results declare the ‘greasing’ role of corruption to be beneficial for foreign investors, arguing that corruption could be favourable in order to bypass administrative or regulatory inefficiencies. Even though the academic literature on the relationship between corruption and foreign direct investments is extensive, studies demonstrate rather contradictory findings.
As demonstrated by Dreher & Gassebner (2013), this study concentrates on the interaction between the efficiency of business regulations and the prevalence of corruption in particular, which seems to influence the relative attractiveness of a business environment. In assessing this attractiveness, Dreher et al. (2013) consider nascent entrepreneurship as their dependent variable and employ the first sub-index of the Doing Business measures, which consists of the procedures, time, cost, and minimum capital to open a new business. Corruption positively influences the relation between the “costs of starting a business” and the dependent variable within 43 developed economies over the 2003-2005 period. Duvanova’s (2014) case-study on transition economies suggests that corruption might flourish in moderately as well as excessively regulated business environments, and “clean” business environments are not necessarily lightly regulated. From here, combined figures from Transparency International (2019a) and World Bank (2019a, 2019c) indicate that a business environment with presence of relatively low corruption and efficient business regulations, such as Sweden, attracts a relatively high and stable annual inflow of foreign direct investments. The Russian business environment, with relatively high levels of corruption and efficient business regulations, inconsistently attracts foreign direct investments. On the other side, business environments with inefficient regulations combined with little corruption indicate this volatility too, as the Uruguayan economy. A business environment such as the Venezuelan, with relatively high degrees of corruption and inefficient business regulations, seems the least attractive for organizations engaging in foreign direct investments. The academic literature does not clarify if, and to which extent, this interaction term influences emerging economies’ relative attractiveness for foreign direct investments.
The main contribution of this research to the present literature is as following: it determines the direction and strength of the moderating effect of corruption on the relation between business
regulations and foreign direct investments, based on a profound, contemporary group of emerging markets. This study adopts the following focal research question accordingly:
What is the influence of Corruption on the relation between Business Regulations and FDI in Emerging Markets?
The following section deliberates the theoretical background, which results in the development of the first and second hypotheses. The research design, sample, variables and analytical strategy are discussed in the methodology section. The results of the performed analyses are presented, in addition to the theoretical and practical implications of this study. Concluding remarks, limitations of this study and interesting directions for further research are discussed eventually.
2. LITERATURE REVIEW
This literature review firstly explains the theory of Economic Regulation, then it discusses the concepts of foreign direct investment, business regulations, corruption, and emerging markets.
The hypothesis are developed afterwards.
Theory of Economic Regulation
As its main theory, this paper employs the theory of Economic Regulation which suggests that variation in economic regulation influences the relative attractiveness of a business environment (Peltzman, 1976; Posner, 1974; Stigler, 1971). Economic regulation is defined as
“both direct legislation and administrative regulation of prices and entry into specific industries or markets” (Joskow & Rose, 1987, p. 1450). The regulations within a particular business environment could restrain or facilitate cross-border investments, and market regulations influence the profitability of firm’s operations (Contractor et al., 2020). When the local government is defined by relatively high degrees of bureaucracy and corruption, the result of potential investments within this business environment will become more uncertain (Dollar, Hallward‐Driemeier, & Mengistae, 2005). Regulations that facilitate liberalization, privatization and deregulation vigorously lower the circumstances for corruption (Djankov, La Porta, Lopez-de-Silanes, & Shleifer, 2002), which suggests that extensively regulated economies offer more opportunity for corrupt practices. However, certain scholars question this direct effect and investigated the interaction between these variables by means of case-studies (Dreher & Gassebner, 2013; Duvanova, 2014). This study assumes that firms adopt a wholistic view on the regulations that might affect the attractiveness of a particular business environment before engaging in foreign direct investments (Contractor et al., 2020).
Foreign Direct Investment
Foreign direct investment (FDI; used interchangeably) is a type of cross-border investments in which an investor acquires “a lasting management interest, 10 percent or more of voting stock, in an enterprise operating in an economy other than that of the investor” (World Bank, 2013).
The global amount of foreign direct investment between the years of 2009 and 2015 raised from 1.17 trillion to 2.06 trillion USD, after which it reduced to 1.29 trillion USD in 2018 (United Nations, 2019). Most of this variance is explained by the total inflow of foreign direct investments of the developed economies, which rapidly decreased from 1.25 trillion USD towards 592 billion USD in the year of 2018. More relevant for this study, UN (2019) indicates that the total amount of inward foreign direct investments in developing economies is growing,
from 518 billion USD in 2007 towards 718 billion in 2018, currently surpassing the inflow of the developed economies. Hence, this trend suggest that it becomes increasingly interesting for firms to invest in developing rather than developed markets. FDI occurs in the form of either green- or brownfield investments (Calderón, Loayza, & Servén, 2004); greenfield investments primarily involve the creation of new assets under foreign firm’s control, while brownfield investments basically involve the transfers of existing assets to a foreign firm. Calderón et al.
(2004) additionally indicate the distinction between cross-border mergers and acquisitions. In an acquisition, the control of assets and operations is transferred towards the acquiring firm, while the assets and operation of two firms are coupled to organise a new entity in a merger.
From a country-level perspective, the rationale for attracting higher amounts of foreign direct investment comes from the positive perception of its side-effects (Alfaro, Chanda, Kalemli- Ozcan, & Sayek, 2004). Foreign direct investment is oftentimes considered as an accelerator for economic growth, since it enhances the amount of total capital, labour, and profits from exports and it facilitates the transfer of technology (Akame et al., 2016). Meyer (2015, p. 2) states that local firms might benefit from foreign direct investment in various manners, such as
“learning from example, labour mobility, export market access, improved supply bases, or direct relations as suppliers or customers.” Alfaro et al. (2004, p. 90) suggest that the
“productivity gains, technology transfers, the introduction of new processes, managerial skills, and know-how in the domestic market, employee training, international production networks, and access to markets” would be beneficial for host business environments. Meyer (2015) proclaims that the advantages of foreign direct investment depend on local stakeholders’
abilities to benefit and learn from investor’s practices, which is commonly described in the literature as “absorptive capacity” (Durham, 2004; Girma, 2005; Kinoshita, 2000). This is in line with the perspective of Alfaro et al. (2004) who imply that country's ability to benefit from these externalities is restricted to local conditions. In order to attract foreign direct investments, Akame et al. (2016) argue that economies with limited natural resource endowments initiate more political, legal, institutional and socio-economic reorganizations in order to attract FDI.
Business Regulations
The “humanly devised constraints that shape human interaction” or more informal, the rules of the game within a society, structure economic, social and political exchange within societies (North, 1992, p. 5). North (1992) suggests that if institutions are the rules of the game, the organizations who are operating in this environment can be considers as its players. By its ability to influence transaction and production costs, regulations influence business
environments. Furthermore, North (1992) describes enforcement to be performed by three entities; the first entity is featured by self-imposed codes of conduct, the second by punishment or retaliation, and the third by the state’s power to carry out coercive enforcement or societal sanctions. Even though regulations are one of the most disputed topics in politics and law, the terminology seems to be misconceived in modern legal discussions (Orbach, 2012). This study adopts the definition of “regulation” as “the governmental intervention in the private domain or a legal rule that implements such intervention” (Orbach, 2012, p.6), in which the implemented rule is considered as a binding legal norm intended to outline individuals’ or firm’s conduct. Orbach (2012, p. 6) further proclaims that the state, or regulator, might be “any legislative, executive, administrative, or judicial body that has the legal power to create a binding legal norm.”
In order to benefit from cross-border investments, host countries need a robust business environment in the form of adequate regulations (Busse & Groizard, 2008; Munemo, 2014).
The burden of excessive regulations on firms has oftentimes been a topic for political debate (Mudambi et al., 2013), leading to the distinction between the public interest and public choice perspectives. The public interest view considers regulations as solution for market failure (Pigou, 2013), in which benevolent interference mitigates inefficiencies emanated from market power asymmetries. On the contrary, the public choice perspective questions the existence of benevolent interference, arguing that regulations should be considered as “redistributive processes” imposed by self-interested entities in order to attain particular advantages (Becker, 1976; Stigler, 1971). In particular, the authors argue that most disproportional official costs of entry are beneficial to the bureaucrats and politicians in question.
Corruption
Following the conventional view of Friedman (1970, p. 6), it is the firm’s responsibility to “use its resources and engage in activities designed to increase its profits so long as it stays within the rules of the game, which is to say, engages in open and free competition without deception or fraud”. The latter is related to the adopted definition of corruption; fraudulent activities initiated to obtain advantage for private gain, mostly attempted by individuals or organizations entrusted with a position of authority (Morris, 1991; Shleifer & Vishny, 1993). In other words, corruption involves intentions to undermine the rules of the game (Breen & Gillanders, 2012).
Senior (2006, p.27) defines corruption by five conditions which must be met simultaneously, a situation in which “a corruptor covertly gives a favour to a corruptee or to a nominee to influence actions that benefit the corruptor or a nominee, and for which the corruptee has
authority”. Even though corrupt activities are by definition illegal, there is an opaque zone in which some entities operate. In certain regions, corruption is thoroughly embedded in social activities that it could be considered as social norms that co-exists besides formal institutions (Breen & Gillanders, 2012). It involves a wide range of activities, such as embezzlement or bribery, which varies in acceptance and occurrence (Kaufmann & Vicente, 2005).
Heidenheimer and Johnston (2002) distinguish three kinds of corruption: black, grey and white.
An action which would be condemned by both the elite and the public to be penalized on the bases of principle is classified as ‘black’ corruption. The ‘grey’ corruption practises are mainly condemned by the elites, while the public’s opinion remains divided and ambiguous. When a punishment of a fraudulent action is not convincingly backed by both the public and elite opinion, or tolerated it is regarded as white corruption. More specific, Johnson & Sharma (2004, p. 2) divide corruption into thirteen categories:
- Bribery and graft e.g. extortion;
- Patronage e.g. material favours in exchange for support;
- Kleptocracy e.g. privatization of public funds;
- Misappropriation e.g. embezzlement;
- Acceptance of disproportional or improper presents e.g. “speed” money;
- Non-performance of duties e.g. cronyism;
- Influence-peddling e.g. conflicting interests;
- Protection of maladministration e.g. perjury;
- Power abuse e.g. intimidation;
- Rent-seeking e.g. illegitimately charging for artificial services;
- Manipulation of legislation e.g. favouritism;
- Electoral malpractice e.g. vote buying;
- and illegitimate campaign donation e.g. unregulated gifts to influence policies (Johnson
& Sharma, 2004, p.2).
Emerging Markets
The concept of emerging markets was announced in the early 1980s, when the initial term of
‘emerging market economy’ was coined by International Finance Corporation’s economist van Agtmael, describing economies with relatively low incomes per capita, which experienced significant governmental reorganisations intended to stimulate economic development (Šević, 2005). Before this introduction, the term “less developed countries” was used to describe the economies which did not meet the degree of economic development of dominant, developed
countries such as US, Japan and Western Europe (O’Sullivan & Sheffrin, 2003). At the beginning of this century, the concept of the BRIC countries was introduced by Goldman Sachs’
economist O’Neill (2001) after which various concepts were introduced by different economics, e.g. BEM, CIVETS, EAGLE, MINT, NEST, Next Eleven, Pacific Pumas, Tiger, and VISTA. Besides these groupings, prominently evolving economies such as Mexico and Turkey are included in some selections. Even though the perspectives on emerging markets share many features, it seems hard to establish one generally agreed upon definition of emerging markets (Šević, 2005). The International Monetary Fund (2019), defines the following 18 countries as relevant emerging markets, in alphabetical order: Argentina, Brazil, Chile, China, Colombia, Hungary, India, Indonesia, Malaysia, Mexico, Pakistan, Peru, Philippines, Poland, Russia, Thailand, Turkey and South Africa. The rationale for adopting IMF’s view is provided in the Methodology section of this paper. As the prevalent characteristics of emerging markets, Daniela-Neonila and Roxana-Manuela (2014) mention a large population with a growing middle class, the initiation of policies aimed at accelerating economic growth, the expansion of international trade and investment, a developing living standard of residents, and growing social stability and tolerance. Another prominent description comes from Vladimir Kvint (2010, p.
25), who defines an emerging market as “a society transitioning from a dictatorship to a free- market-oriented economy, with increasing economic freedom, gradual integration within the global marketplace, an expanding middle class, improving living standards, social stability and tolerance, as well as an increase in cooperation with multilateral institutions.” Compared to high-income countries, Meyer & Tran (2006, p. 179) define emerging markets as “economies with high growth or growth potential, but without the sophistication of the institutional framework seen in western Europe and North America.” This sort of comparison is also made by Haley & Haley (2006), who explicitly mention the relatively low living standards and insufficient access to goods and services, relative to high-income countries. Hoskisson et al.
(2000) discuss two criteria, which are the accelerated economic development and the governmental initiatives towards free-market development and liberalization. Kaymak &
Bektas (2014) assert that this accelerated economic development was not consistently followed by proportional improvements in legal, social and political institutions. With respect to emerging markets, the authors state that economic growth, economic intensity, openness to trade, infrastructure, reduced country risk, and economic freedom are negatively related to corruption. Relatively high regulatory costs diminish economic growth and enhance the volume of the informal sector (Loayza, Oviedo, & Servén, 2004). According to Arnold and Quelch (1998), MNEs are challenged by operating within these emerging market environments. The
authors mention government’s tendency to adjust business regulations unpredictably and frequently, in combination with inadequate regulatory discipline. When operating in emerging markets, MNEs might be confronted with inadequate market data, product diversion and imitation, undeveloped distribution systems, and insufficient communication channels (Arnold
& Quelch, 1998). On social grounds, Arnold and co-author (1998) discuss the presence of complicated structures regarding loyalty and authority, within opaque relations between business and politics. Bekaert (1995) contributes by declaring relatively weak credit ratings and high inflation.
Hypotheses development
Efficient business regulations decrease firm’s investment costs and diminish the uncertainties of investing cross-border, which enhances an economy’s relative attractiveness for foreign direct investments (Zhang, 2007). This general, positive association between foreign direct investments and business regulations, measured by the Doing Business indices, has frequently been studied and confirmed (Anderson & Gonzalez, 2012; Contractor et al., 2020; Davis &
Williamson, 2018; Malik & Jyoti, 2018; Olival, 2012; Piwonski, 2010; Singh, 2014; Zhang, 2012). Corcoran & Gillanders (2015) find strong evidence for a relation between the Doing Business rankings and FDI stock, regarding the investment flows from developed towards developing economies. More relevant for this study, Bayraktar (2013) demonstrates the association between business regulations and the relative inflow of foreign direct investments as percentage of its GDP. Other research demonstrate a positive relation between efficient, low- cost regulatory frameworks and foreign direct investment, concentrated on developing geographic nations in Africa (Akame et al., 2016; Morris & Aziz, 2011; Nangpiire et al., 2018) and combinations of the middle East and Asia (Aziz, 2018; Shahadan et al., 2014; Sjöholm &
Lipsey, 2012; Vogiatzoglu, 2016). This study adopts the first hypothesis accordingly:
H1: Efficient Business Regulations positively influence Cross-border investments in Emerging Markets.
With respect to the general, direct influence of corruption on foreign direct investments, the literature demonstrates contradictory findings. Particular studies argue for a negative relationship suggest that foreign investors have strong preferences to refrain from corrupt practises, as it is considered as wrong and it might generate operational inefficiencies (Habib
& Zurawicki, 2002; Mauro, 1995; Méon & Sekkat, 2005; Mudambi, Navarra, & Delios, 2013).
Even though some corrupt activities seem to the diminish the regulative burdens at first, Guriev (2004, p. 503) proclaims that “the equilibrium level of red tape is above the social optimum”
and therefore, does not enhance the relative attractiveness of any particular business environment. Besides confirming the negative impact between corruption and FDI, Cuervo- Cazurra (2006) contributes to this discussion by asserting that investors which have been exposed to bribery in their home business environments might not be discouraged by corruption in their hosts’. On the contrary, certain studies indicate a positive relationship between corruption and foreign direct investments (e.g. Barassi & Zhou, 2012; Dreher & Gassebner, 2013; Egger & Winner, 2005). These studies argue for the “greasing the wheels” role of corruption to be beneficial for foreign investors, suggesting that corruption would be helpful to circumvent administrative and regulatory restrictions. Rather than sabotaging economic exchange, corruption might enable entities to bypass inefficient regulations, e.g. accelerating the procedures of starting a business or registering property through bribing (Breen &
Gillanders, 2012).
More relevant for this study, the interaction between business regulations and corruption seems to influence the attractiveness of a business environment following Dreher & Gassebner (2013).
Their study suggests that corruption positively influences the relation between the “costs of starting a business” and nascent entrepreneurship within 43 developed economies over the 2003-2005 period. The subsequent case-study advocates for the hypothesized interaction term to influence the relative inflows of FDI (Transparency International, 2019a; World Bank, 2019c, 2019a). A systematic overview of this case-study is provided in the first appendix. In the first situation, the business environment is characterized by low corruption and efficient business regulations. Examples of states that fit this profile are New Zealand, Sweden and Denmark. In absolute terms, the inflow of foreign direct investments is relatively high, while the relative inflow as percentage of GDP remains stable over time, approximately between .5 and 1.5 percent. The second situation considers a business environment with higher levels of corruption and efficient business regulations. Countries as Macedonia, Mexico, and Russia show these characteristics clearly. Referring to the potential “greasing” effects of corruption (Breen & Gillanders, 2012; Dreher & Gassebner, 2013), these host business environments might be attractive for cross-border investments. Their inflows of foreign direct investments are volatile in absolute terms, e.g. Russia’s massive 69.2 billion in the year of 2013 compared to 8.8 billion during 2018. From 2012, Mexico seems to attract more foreign direct investments both in absolute and relative terms, growing from 17.7 billion USD toward 37.5 billion in 2018.
Compared to the first situation, their range of relative inflow fluctuates from approximately .5
to 5.5 percent over time. Within the third situation, there are business environments with inefficient regulations, combined with low levels of corruption. The country that fits this profile best is Uruguay; while its business regulations are not entirely exhaustive, the presence of perceived corruption is considered as low. Although the Uruguayan regulatory framework might impose inefficiencies in terms of unnecessary costs to businesses, it might benefit from the absence of corruption (Aidt, 2009; Mauro, 1995; Méon & Sekkat, 2005). This combination of inefficient regulations and low corruption is, however, rather uncommon. Similar to countries of the second situation, both the absolute and relative Uruguayan foreign direct investments are volatile; from an enormous amount of 6.0 billion USD, representing 11.8% of country’s GDP in 2012, to a current inflow 2.0 percent. The fourth situation describes a business environment with high degrees of corruption and inefficient business regulations, which seems the least attractive for foreign investment. Nations as Libya around, Venezuela and Iraq have these characteristics in common. The inflow of foreign direct investments remains low over time in absolute terms, and the relative inflow varies roughly between 1.5 and negative percentages. Libya, for instance, coped with a negative foreign direct investment of 1.1 billion in 2009 towards 956 million USD in the year of 2018, representing currently 0.2% of its GDP.
Therefore, this study hypothesizes that the interaction between business regulations and corruption positively influences the amount of FDI emerging markets receive. This study formulates the following hypothesis accordingly:
H2: Corruption positively influences the relationship between efficient Business Regulations and Foreign Direct Investment in Emerging Markets.
Conceptual model
The conceptual model indicates the hypothesized positive relationship between efficient business regulations and the inflow of foreign direct investments in emerging markets, which is positively influenced by the presence of corruption.
Fig. 1: The conceptual model. Source: author’s own.
3. METHODOLOGY
This section considers the methodology of this quantitative research, deliberating the research design, sample, variables and measurements, data analysis and limitations.
Research design
In answering the focal research question, this study relies on secondary data and the variables are of quantitative nature. The justification for secondary data is twofold. First, it provides a high standard regarding validity which would most probably not be met if I would personally collect this data, given the scope and time frame of this study. For example, it seems unfeasible to determine the efficiency of business regulations, or the amount of foreign direct investments within a particular country on a specific moment in time. Second, it is unattainable, if not impossible, to conduct a questionnaire which adequately grasps historical circumstances, e.g.
the level of perceived corruption in India during the year of 2012. Accordingly, this study relies on authorities’ databases such as the International Monetary Fund, Transparency International and World Bank. The rationale for performing quantitative research is that it facilitates studying a greater amount of observations in order to make meaningful generalizations towards emerging markets (Swanson & Holton, 2005).
Sample selection
As stated in the literature review, and supported by the International Monetary Fund (hereafter;
IMF), the subsequent 18 economies are considered as emerging markets, in alphabetical order:
Argentina, Brazil, Chile, China, Colombia, Hungary, Indonesia, India, Malaysia, Mexico, Pakistan, Peru, Philippines, Poland, Russia, South Africa, Thailand and Turkey (IMF, 2019a).
Established in 1945, the IMF organizes 189 countries and addresses “fostering global monetary cooperation, secure financial stability, facilitate international trade, promote high employment and sustainable economic growth, and reduce poverty around the world” (IMF, 2019b). Stock market analysts, e.g. Dow Jones & Company Incorporate, FTSE International Limited, J.P.
Morgan, MSCI Incorporate, Russel Investments, or S&P, predominantly base their classifications on the development of the stock markets. Therefore, these analysts might develop a limited perspective on the concept of emerging markets relative to IMF. In their World Economic Outlook (WEO), IMF employed three criteria in their classifications of emerging markets, respectively the country’s income level per capita, its export diversification, and the extent of integration within the global financial system (IMF, 2019c). Moreover, IMF’s
perspective consists of a comprehensive, broad set of markets compared to prevalent groups as the BEM, CIVETS, EAGLE, MINT, NEST, Next Eleven, Pacific Pumas, Tiger, and VISTA.
Since the terminology of “emerging” implies alteration over time, the relevance of prior work might diminish at a certain moment. In order to make relevant contributions, while coping with data availability, this study is limited to the time period of 2010-2018.
As a consequence of missing data in World Bank’s Doing Business scores for a few countries between the years 2010 and 2013, this study is based on a unbalanced panel dataset of 126 country years observations. Both the length of the selected time period and the number of observations is comparable to other studies within this field (e.g. Abdella, Naghavi, & Fah, 2018; Corcoran & Gillanders, 2015; Davis & Williamson, 2018; Morris & Aziz, 2011).
Further, the residuals should meet the requirements regarding normality, skewness, kurtosis and homoskedasticity. In order to meet these requirements, the first analyses involves 126 country year observations as a consequence of elimination of 7 country year observations. In contrast with prior literature (Al-Sadig, 2009; Zhang, 2012), this study does not exclude negative FDI inflows. These values can be negative when the actual inflow of FDI is lower than the divestments within a particular year. Excluding these values would bias the findings. Given 18 countries over a 9-years’ time period, the total population consists of 162 country years respectively. This study attempts to make inferences regarding this set of emerging market economies within the predefined time period. Because the sample represents almost the total population and it is not in this study’s interest to make inferences about a wider set of countries, a larger number of country year observations seems not needed (Vogiatzoglu, 2016, p.362). By employing a confidence level of 95 percent, the number of observations gives no reasons for concern (Israel, 1992).
Dependent variable
The inflow of foreign direct investments as percentage of GDP is considered as the dependent variable, which enables comparison of the relative inflows (Addison & Heshmati, 2003;
Adeoye, 2009; Arbatli, 2011; Bayraktar, 2013; Globerman, Shapiro, & Tang, 2006; Munemo, 2014; Popescu, 2014). These data represent the (net) inflows of investment to acquire “a lasting management interest, 10 percent or more of voting stock, in an enterprise operating in an economy other than that of the investor” (World Bank, 2013). It represents the aggregated value of equity, short- and long-term capital, and reinvestment of profits as indicated in firm’s balance of payments, divided by the respective economy’s GDP. This information will be retrieved from World Bank’s WDI database, within the defined time period.
Independent variable
In line with previous research (Contractor et al., 2020; Divanbeigi & Ramalho, 2015; Djankov, Georgieva, & Ramalho, 2018) World Bank’s Doing Business Index scores are consulted to measure the efficiency of business regulations. As a benchmark study of regulation, the index measures various aspects of 190 business regulations worldwide which directly influence firm’s activities. The following sub-indices are considered, arranged by theme: Starting a business, Dealing with construction permits, Getting electricity, Registering property, Getting credit, Protecting minority investors, Paying taxes, Trading across borders, Enforcing contracts and Resolving insolvency (World Bank, 2019a). All information mentioned in this thesis with respect to the indices is retrieved from the Methodology for Doing Business section (World Bank, 2019d). Assisted by scholars, World Bank compiled its questionnaire, concentrated on simplified and standardized business scenarios to assure comparison across nations. Over 12,000 experts with presence in either one of the respective countries fill in the questionnaire.
These individuals are requested to support their reasoning with the applicable law, regulations and tariffs. After several verification and consistency tests, the scores of ten sub-indices are aggregated on country-level. The individual indices are weighted equally, irrespective of intercorrelation and their explanation of the over-all variance. Prior to publication, the results are validated with the respective governments. This process results in a range of values from 1 to 100, in which the score of 1 indicates the presence of heavily constraining business regulations, while a score of 100 stands for strongly enhancing business regulations. World Bank considers the nature of the measurement in assigning scores to the respective countries.
Simplified business regulations, e.g. the costs related to starting a business, mostly yield relatively higher scores. This does not hold consistently over all indices, however. For example, further reaching property rights might be beneficial for businesses. Therefore, the efficiency of business regulations is adopted as the independent variable of this study.
Interaction term
The moderating effect of corruption on the main relationship is considered as interaction variable. Aligned with previously published academic papers on this subject (Jain, 2001; Mo, 2001; Svensson, 2005; Tanzi & Davoodi, 1997), the level of perceived corruption is retrieved from the Corruption Perceptions Index (CPI; used interchangeably). Following the study of Wilhelm (2002), the perceived corruption index could be considered as a reliable measure of
the concept. This index is compiled by Transparency International (2019a), which allocates scores to countries based on the following eleven respective measures:
- the occurrence of bribery;
- the diversion of public financial means
- the prevalence of functionaries employing public office for private gain;
- the governmental ability to restrain corruption;
- the presence of “red tape” and disproportionate bureaucratic constrains;
- the extent of meritocratic and nepotistic arrangements within civil service;
- the degree of effectivity in prosecution of corrupt officials;
- the presence of decent laws on prevention of conflict of interests and financial disclosure for officials;
- the juridical protection of journalists, investigators and whistle blowers;
- the occurrence of state capture by closely settled interests;
- and the societal access to information about public affairs (Transparency International, 2019a).
CPI allocates scores to the degree of corruptness of a country’s public sector, as perceived by local business managers and experts. Since corruption by definition involves illegal practices, oftentimes intentionally hidden, combined with the absence of an objective indicator to measure the concept of corruption directly, Transparency International relies on perceptions. The data is drawn from 13 sources originated from 12 independent institutions specialized in business environment and governance studies. The scores of the measures are aggregated on a scale from 1 to 100, in which a score of 1 indicates very corrupt, and 100 very clean institutions (Transparency International, 2019b). In order to prevent interpretation issues, the index scores are reversed so that high numerical values indicate high perceived corruption levels, vice versa, which is in line with prior literature (e.g. Dreher & Gassebner, 2013). Data from the time period of 2010-2018 are utilized.
Control variables
This study controls for the following variables, which are GDP total, GDP growth, trade, infrastructure, population total, population growth, and democracy.
GDP total
The gross domestic product, or size of the host market, is positively associated with foreign direct investments (Bénassy-Quéré, Coupet, & Mayer, 2007; Bevan & Estrin, 2000; Demirhan
& Masca, 2008; Dunning, 2004; Moosa & Cardak, 2006; Singh & Jun, 1999; Wang & Swain, 1995). In addition, there seems to be a negative relation between the GDP of a country and the prevalence of corruption (Dreher & Herzfeld, 2005; Tanzi & Davoodi, 1997; Welsch, 2008).
Therefore, this study considers the control variable of GDP total. The data is retrieved from World Bank’s WDI database.
GDP growth
Economic growth influences the attractiveness of a particular business environment from a market-seeking perspective (Dunning, 1988) and corruption seems to be lower in economically developed countries (Serra, 2006). Emerging markets are particularly defined by their rapid economic developments in (Arnold & Quelch, 1998; Hitt et al., 2000; Hoskisson et al., 2000;
Meyer & Tran, 2006). Accordingly, and aligned with prior research (Al-Sadig, 2009; Anderson
& Gonzalez, 2012; Calderón, Loayza, & Servén, 2004; Contractor et al., 2020; Vogiatzoglu, 2016), the control variable of GDP growth is employed. This variable indicates the percentage of annual decrease or increase in country’s gross domestic product. This information is retrieved from World Bank’s WDI database.
Trade openness
Another characteristic of the emerging markets is the tendency towards economic market reforms (Hitt et al., 2000; Hoskisson et al., 2000; Kvint, 2010). Trade openness concedes the extent of trade liberalization (Olival, 2012), and is therefore a strong determinant of foreign direct investments within emerging markets (Ang, 2008; Blonigen & Piger, 2014; Jadhav, 2012;
Kinoshita & Campos, 2003; Moosa & Cardak, 2006; Ranjan & Agrawal, 2011; Sabir, Rafique,
& Abbas, 2019). This measure is an aggregate of the total exports and imports, divided by the country’s GDP. The data is retrieved from World Bank’s WDI database.
Infrastructure
Adequate infrastructure expands resource availability and enhances the productivity of capital (Biswas, 2002; Blonigen & Piger, 2014; Globerman et al., 2006; Groh & Wich, 2012;
Sethi, Guisinger, Phelan, & Berg, 2013; Singh & Jun, 1999). More specific, the literature suggests that infrastructure is positively associated with FDI inflows (Bellak, Leibrecht, &
Damijan, 2009; Kumar, 2006; Rehman, Ilyas, Mobeen Alam, & Akram, 2011; Shah, 2014).
This study adopts the second pillar of WEF’s Global Competitiveness Index, Infrastructure, to measure the eponymous concept. It considers both transport and the electricity and telephony infrastructure across 137 economies. The transport infrastructure involves the quality of overall
infrastructure, roads, railroads, ports, air transport and the amount of available airline seat kilometres. The second aspect measures the quality of electricity supply, mobile telephone subscriptions and fixed telephone lines (WEF, 2018). Data is retrieved from WEF’s official database.
Population total
Prior research point out that economies with a large population are more likely to attract foreign direct investments, relative to low populated nations (Bevan, Estrin, & Meyer, 2004; Blonigen
& Piger, 2014; Kolstad & Villanger, 2004; Tsai, 1994). In addition to GDP, the total population indicates the magnitude of the market. Large population might enhance the attractiveness of a business environment in terms of market size and availability of labour force, referring to the resource and market-seeking motives (Dunning, 1988). This information is retrieved from World Bank’s WDI database, and indicates the total population per country and year.
Population growth
Following prior published papers related to foreign direct investment, business regulations or corruption (e.g. Alfaro, Chanda, Kalemli-Ozcan, & Sayek, 2000; Busse & Groizard, 2008;
Mauro, 1995; Mo, 2001) this study controls for population growth. As one of the prevalent characteristics of emerging markets, prior literature (Arnold & Quelch, 1998; Kvint, 2010;
Meyer & Tran, 2006) mentions a growing population. In addition, Barros & Cabral (2000) argue that larger countries attract more foreign direct investments. Therefore, this research controls for population growth accordingly. This information indicates the percentage of annual increase or decrease of the total population, and is retrieved from World Bank’s WDI database too.
Democracy
Corruption is relatively low in democratic regimes (Serra, 2006), and foreign direct investments tend to be higher in democratic countries (Busse, 2003). Hence, this research controls for democracy. Aligned with prior research (Nalla & Mamayek, 2013; Rahman, 2014), this study utilizes data from the Democracy Index as a proxy for democracy. This index, published by The Economist Intelligence Unit (hereafter; EIU), is based on 60 indices. EIU allocates scores to individual countries, after which they are ranked and categorized. EIU considers the authoritarian regimes, with scores between respectively 0 and 3, hybrid regimes scoring 4 to 6, flawed democracies scoring 6 to 8, and full democracies with scores from 8 to 10 (EIU, 2020).
This study involves, however, the annual absolute country scores of the time period from 2010 to 2018.
Data analysis
Aligned with the hypothesis, this research attempts to declare the moderating influence on the relationship between the dependent and independent variable by performing descriptive, correlation and regression analysis. The program Stata is used to perform the statistical analysis as a consequence of its comprehensiveness. The descriptive statistics provides an overview of the mean and standard deviation of the variables. Based on the country names, panel IDs are created. The study additionally tests normality, skewness, kurtosis, and heteroskedasticity within the data. The correlation analysis will determine if, and to which extent, the variables are interrelated. Correlation statistics are performed by the paired Pearson tests. If the correlation values exceed the threshold of .8 (Berry & Feldman, 1985), or the VIF value exceed the value of 4 (Hair, Black, & Babin, 2010; Miles & Shevlin, 2001), the corresponding variable suffers from collinearity and is therefore excluded from further analyses.
The regression analysis demonstrates the relationship between the dependent, independent and moderating variable, while considering the set of control variables. The regression statistics involve four models. The first model solely demonstrates the relation between the dependent and control variables. Within the second model, the dependent, control, and independent variable of the Doing Business score is included. The third model extends this set of variables with CPI, and the fourth model includes the interaction term. Since these are panel data, the difference in fixed and random effects within the GLS regression must be considered. The fixed effects regression analysis is performed, after which this procedure is replicated for the random effects. The Sargan-Hansen specification test compares the model of fixed with random effects, if the corresponding p-value is exceeds .05, the random-effect model is more reasonable or preferred. The regression analysis will gradually be extend by the number of variables with corresponding models.
Three robustness checks are performed for the similar regressions. If there is no significant change compared to the initial models, the coefficients are considered as robust, which indicates that the regression coefficients can be accepted for their structural validity (Lu & White, 2014).
The first robustness check excludes all country year observations of the year 2010, simply to determine whether the hypothesized mechanisms change over time in terms of direction or significance. The second robustness check employs Control of Corruption (hereafter; CC)
instead of the CPI measure. The CC measure is one of the six indices of the World Governance Indicators, which captures “perceptions of the extent to which public power is exercised for private gain” and involves both petty and grand forms of corruption (Kaufmann, Kraay, &
Mastruzzi, 2010, p.4). In relation to foreign direct investments, CC is widely adopted in the literature as an adequate proxy for corruption within a particular business environment (Anokhin & Schulze, 2009; Cuervo-Cazurra, 2006; Gani, 2007; Jadhav, 2012). Data is retrieved from World Bank’s database. In line with the CPI measures, the data has been transformed from -2.5 to 2.5 scaling, towards 1 to 100. The data is reversed afterwards, aligned with the CPI measure. The third robustness check implements the relative Doing Business rankings instead of the absolute country scores, following prior studies (e.g. Bayraktar, 2013; Corcoran &
Gillanders, 2015; Jayasuriya, 2011). Even though the same concept is measured by the same methodology, the marginal differences between positions is ignored by adopting the rankings as dependent variable. The outcomes might, therefore, be different. The data is obtained from World Bank’s publications.
4. RESULTS
Descriptive Statistics, Correlations and Assumption testing
The descriptive and Pearson paired correlation statistics of are provided in table 1. This study relies on 126 country year observations. To prevent issues regarding multicollinearity, mean- centred values of the dependent, independent and interaction term are employed (Afshartous &
Preston, 2011). Both GDP and population are expressed in millions to enhance comparison and usability.
The paired correlation among the variables GPD total and Population total appears to be high (.8011) and significant at the .01 percent level, as well as the paired correlation between Infrastructure and the Doing Business scores (.7200). According to Berry & Feldman (1985), either one of the correlated variables should be excluded from the analysis if the correlation exceeds the threshold of .8. The expected influence of economy size on the inflow of FDI intuitively seems more relevant than population size, therefore the latter has been excluded from further analyses. The variable of Infrastructure is not excluded since it does not exceed the threshold as proposed by Berry et al. (1985). In addition, the VIF values indicate no concern for multicollinearity. Through the four models, the corresponding rates for this variable are 1.87, 3.07, 3.26 and 3.39 respectively which does not exceed the threshold of 4 as argued by Hair, Black & Babin (2010).
Before the (paired) correlation and regression statistics are performed, this study investigates to which extent the residuals are normally distributed, and whether the dataset suffers from skewness, kurtosis and heteroskedasticity (Osborne & Waters, 2003). In order to meet these requirements, based on residuals, I gradually excluded the observations which were considered as outliers until the Shapiro-Wilk (1965) and Jarque-Bera (1981) indicate that the assumptions were met. This process resulted in the exclusion of 8 outliers. The Shapiro-Wilk test for normality of the residuals indicates a W-value of .9900 which corresponds to a probability of .5027. Given the common threshold of .05, the normality is objectively proven. Further, the scores of residuals exceed the threshold with respect to skewness (.3070) and kurtosis (.1198).
The corresponding joint adjusted R-squared value indicates a probability of .1707 which demonstrates no concern for either skewness or kurtosis. However, the assumption of homoskedasticity is not met. The Breusch-Pagan test (1979) demonstrates the following corresponding chi2-values: 37.72, 30.09, 41.04 and 41.90. These values are all significant at the .01 level, hence the hypothesis of constant variance is rejected. Fortunately, the statistical program of Stata offers the vce(robust) option which is formally known as the Huber-White estimator. By performing this test, the regression statistics are robust for heteroskedastic effects
and the standard errors are adjusted (White, 1980). The Kernel density graph relative to the normal distribution and standardized normal probability plot (PP) demonstrate the distribution of residuals, and could be in the second appendix.
The dependent variable, FDI as percentage of GDP, is positively correlated with the measures of business regulations, corruption, and democracy. Hence, emerging economies with relatively high inflows of FDI are more likely to provide efficient business regulations, cope with low levels of corruption and are rather democratic. The dependent variable of Business regulations is negatively associated with the prevalence of corruption and population growth. Infrastructure and trade are positively significant associated with the dependent variable. This indicates that emerging economies with more efficient regulatory frameworks related to business encounter lower degrees of corruption and population growth, dispose of more adequate infrastructure, and implement more liberal policies regarding international trade. In addition, the prevalence of Corruption is negatively associated with infrastructure, trade, population growth and democracy. The paired correlation matrix further reveals a positive correlation between emerging economies’ total GDP with corresponding growth and quality of infrastructure.
However, it is remarkable that this control variable is negatively associated with openness to trade, population growth and democracy. In addition, the correlations demonstrate that GDP growth is positively related to population growth within this set of emerging economies. Better infrastructure seems to be affiliated with more openness to international trade, and less population growth. The latter variables are correlated in a negative manner too, and population growth tends to be positively associated with democracy.
Table 1: Pearson Paired correlation statistics
Mean S.D. FDI DBS CPI DBS*CPI GDPt GDPg INF TRA POPg DEM
FDI 2.7110 1.9812 1.0000
DBS -.1476 7.4297 .2891***
(.0010) 1.0000
CPI .4472 11.1604 -.4632***
(.0000) -.4293***
(0.0000) 1.0000 DBS*CPI -35.3844 71.5417 -.2067**
(.0202)
.1979**
(.0263)
.4787***
(.0000)
1.0000 GDPt 1085.8740 2244.8190 -.1332
(.1371)
-.1657 (.0637)
.1047 (0.2431)
-.1238 .1671
1.0000
GDPg 4.0305 2.7424 .0815
(.3643)
-.0565 (.5295)
.0100 (.9118)
-.1403 (.1172)
.1735*
(.0520)
1.0000
INF 4.0944 .6435 .1554
(.1043) .7200***
(.0000) -.4108***
(.0000) .1681*
(.0599) .2000**
(.0247) -.0814
(.3648) 1.0000
TRA 64.8074 35.4153 .0533
(.5531) .6923***
(.0000) -.2960***
(.0008) .0115
(.8987) -.2175**
(.0144) .1113
(.2147) .5554***
(.0000) 1.0000
POPg 1.0640 .5581 -.0525
(.5591) -.3203***
(.0003) -.1869**
(.0361) -.1690*
(.0586) -.2427***
(.0062) .1986**
(.0258) -.2712***
(.0021) -.2764***
(.0017) 1.0000
DEM 63.2436 12.6240 .3215***
(.0002)
-.0025 (.9780)
-.4370***
(.0000)
.2752***
(.0018)
-.5157***
(.0000)
-.1155 (.1977)
-.1206 (.1786)
.0708 (.4310)
.1501*
(.0934)
1.0000
The first figure indicates the paired correlation coefficients, the figure in the parentheses indicate the corresponding p-values. *** Indicates P<0.01, ** indicates P<0.05, and *Indicates P<0.1.
