How are foreign direct investments affecting the rate of corruption of a country : an analysis of the relation between foreign direct investments and corruption

(1)

VLADIMIR HILCA – UNIVERSITEIT VAN AMSTERDAM

How are Foreign Direct Investments Affecting the

Rate of Corruption of a Country

An Analysis of the Relation Between Foreign Direct Investments And Corruption

Vladimir Hilca 31/08/2017

Candidate Number: 11137258

Candidate E-mai: Vladimir.hilca@yahoo.com

Programme: Political Science – International Relations Module: The Political Elites and the Economy

Department: Graduate School of Social Sciences Supervisor: Dr. Gijs Schumacher

(2)

2

discussion covers various forms of corruption that exist in society and how they each affect various aspects of public life. Determinants of corruption are discussed, as well as the perception of corruption which has been linked to some incidences of corrupt practices and affects public, commercial, and political views of corruption.

The paper discusses the FDI and corruption on an extensive list of countries and continues by investigating the relation between FDI and corruption through a Granger causality relation and times-series cross-sectional regression. This analysis reaches an overall conclusion that

corruption affects foreign direct investments.

The final part of the analysis is represented by discussions on the manner foreign investors are affected by corruption in Romania with a case study on the Microsoft file in which the

(5)

5

Introduction

Corruption is recognized as a problem affecting most aspects of economic and social life, particularly in small and developing countries. Corruption can be defined as the abuse of public power to achieve a private benefit (Wright & Craigwell 2012). In developing countries,

corruption affects sustainable development, as well as the achievement of long-term economic growth. At a global level, an estimated one trillion US dollars are reportedly lost annually as a result of nepotism, bribery, patronage, embezzlement of public funds, and public officials’ illegal sales of government property (OECD, 2014). Funds lost through corruption account for 5% of the total world gross domestic product (GDP) (OECD, 2014). The World Bank identifies corruption as the primary obstacle in eradicating poverty and boosting prosperity for the 40% poorest people (World Bank, 2017)

Historically, corruption has been among the issues that most concern public opinion. Corruption is ranked among the top ten gravest social dangers, according to public opinion (Vásárhelyi 1999). It has also gained traction as a topic in daily political discourse, whether the conversation involves the recent Chinese economic miracle, American politics, or the post-communist transitions. Corruption has been a subject of scholarly research for more than half a century, as it has transformed from a simple activity to a globally recognized vice (Tverdova 2011). To have a broader perspective on corruption, understanding the history of the idea is essential. Historically, various forms of government existed for different purposes in society, and each propagated a different form of corruption (Buchan, Hill, 2014). The overall governing body would be viewed as corrupt if the state primarily pursued divisive interests such as personal gain instead of the community’s common shared interests (Morrison, 2015).

The general negative opinion on corruption is still influential in the modern economic setting (Rose-Ackerman, 2008), even if the issues have been present in the public eye for more than a century. For instance, the emergence of the modern forms of representatives in the eighteenth century was a concern in those times, particularly given the probability of parties propagating partisan-inspired interests (Rose-Ackerman, 2008). The quest for answers is not just a recent endeavor; even the Greek philosophers believed that democracy exponentially increases

(6)

6

the likelihood of political corruption, which is the corruption of the government by the people (Hindess 2001).These philosophers believed that a democratic government faces the danger of being dominated by both the poorly educated and the have-nots, who are likely to reflect the ambitions of corrupt demagogues, as well as their prejudices and ignorance.

The empirical literature on corruption measures is largely based on expert options, which are subject to ideological bias. The corruption rankings of countries are also heavily biased, as they are based on general perceptions of past and current politico-economic performance. The literature review is crucial in developing the study of the relationship between corruption and FDI while identifying the determinants and the general effect of corruption on the economy in general.

Foreign Direct Investments (FDI) are long term investments of a foreign entity into a host country. The FDI implies the presence of the two economic actors – the investor and the country receiving the investment. These investments mostly have a positive effect on the recipient country such as creating new jobs and developing the economy, but negative aspects such as the dependence on foreign capital and policies potentially harming the local businesses, are

highlighted as well (Meyer, Sinani, 2009). The overall views on FDI are optimistic, with these being considered economic growth generators for the host countries rather than elements of threat (Fetscherin, 2010).

An in-depth analysis of the literature on both corruption and foreign direct investments may lead to the conclusion that very few authors speak about the direct relationship between the two of them. This gap in the literature is the main reason for this particular analysis, especially in Romania, where the opinion shifts from perceiving foreign investors as saviors of the economy to seeing any foreign company as a mythological vampire, draining local resources.

This paper aims to analyze the relationship between corruption, being represented by two indicators: Corruption Perception Index (CPI) and Control of Corruption Index (CCI), and FDI. To do so, a Granger causality analysis will be applied to a panel of 174 countries, and it will use the FDI inflows and the CPI and CCI. For more precision, ten more control variables, which will be detailed further in the paper, will be introduced for the countries where a strong relationship is present, and for which a times-series cross-sectional analysis will be conducted. The second stage of the analysis focuses on Romania and how the rate of corruption influences foreign

(7)

7

investors. This case will be analyzed by looking at the evolution of FDI in the country linked to corruption, the efforts undergone for fighting the corruption, and a discussion about the

Microsoft file – the most known case where a foreign investor’s name was dragged in a corruption scandal in Romania.

Corruption

Determinants of Corruption

Some aspects of a government’s activities cultivate an environment that facilitates corruption (Tanzi, 1998). The sources who determine corruption are vast and can vary from regulations, to taxation, from spending decisions to the provision of goods and services (Tanzi, 1998; Sanyal, 2005; Dreher & Schneider, 2010)

Permits, regulations, authorizations, and licenses are required to engage in activities in many countries. The state performs its role through rules and regulations, easily affected by

bureaucratic hindrances, especially in developing countries. Several government offices authorize activities by offering permits, licenses, and other required authorization documents. The existence of authorizations and regulations provides monopoly power to the government official in charge of inspecting or authorizing the activities. Government officials sometimes use their public authority to acquire bribes from citizens and investors. In some countries, individuals assume the roles of intermediaries in facilitating the acquisition of permits and thus ”grease the system” (Dreher and Gassebner, 2013). Surveys from various transitioning and developing countries indicate that managers of enterprises are often subjected to a long and tedious bureaucratic process that can be expedited through bribery (Tanzi 1998).

Another way to facilitate government corruption is through taxation and other fiscal policies. Even though taxes are based on laws that require no contact between tax inspectors and

taxpayers, thus hedging towards a minimal occurrence of acts of corruption, some situations increase the risk of corrupt practices in the administration of customs and tax. Tax laws can be interpreted in various ways, resulting in a situation in which taxpayers seek the assistance of tax administrators to comply with these statutes (Tanzi 1998). Acts of corruption by tax

(8)

8

administrators often go undiscovered. If they are discovered, tax administrators are only lightly penalized.

Corruption can affect public expenditure and influence the overall spending decisions of the national and local administration. High-level investment projects are frequently associated with increased levels of corruption as a result of civil servants’ discretion over decisions involving public investment projects. Political groups and individuals have the opportunity to receive a commission from companies chosen to execute public projects. Procurement spending, which includes the purchase of goods and services on the government’s part; money received from aid; and money acquired from the sale of a country’s natural resources are also prone to corruption, due to a lack of transparency (Sanyal 2005).

The government is closely involved in the provision of resources, goods, and services below the market price. Sometimes these resources, goods, and services are characterized by limited supply, rationing, and queuing, as well as excess demand surpassing supply levels (Dreher & Schneider 2010). Public officials decide how to apportion the limited supply. People interested in the goods provided are likely to offer bribes to access the goods.

Other determinants that propagate acts of corruption include the financing of parties, the level of public sector wages, and the quality of bureaucracy (Tanzi, 1998).

Empirical evidence seems to suggest that countries with lower incomes per capita are more corrupt than richer countries. Democratic institutions only exert control over acts of corruption if they have a long history and established credentials. Countries that are politically unstable experience high levels of corruption, and colonial heritage reflects the current levels of corruption – a former colony tends to be more corrupt (Serra 2006).

Corruption Perception

Today, corruption is primarily understood in economic terms, that is, regarding its content and effects (Sandholt & Koetzle 2000). Corruption involves money being offered in exchange for favors, which affects economic growth and development (Tanzi, 1998). Other forms of corruption include favoritism, extortion, and the involvement of public officials in the

(9)

9

manipulation of information for personal gain (Wang, 2013). Corrupt acts committed by high-ranking political decision-makers are referred to as “grand” or “political” corruption (Swaleheen 2011). Grand or political corruption involves relatively larger bribes, which an interested party pays to influence policy formulation or to acquire a contract (Swaleheen 2011). Corruption involving bureaucrats and judicial officials is referred to as “bureaucratic” and “judicial” corruption, respectively. This form of corruption aims at altering the implementation of already existing policies (Swaleheen 2011). The phenomenon of grand corruption has become such a significant problem that Transparency International (2016) has developed a legal definition for it. According to this definition, grand corruption appears when a person or a body deprives a

substantial part of the population of a fundamental right or causes a loss greater than 100 times the annual minimum subsistence income (Transparency International, 2016).

In past decades, corruption was understood from a general perspective, whilst nowadays is linked to democracy. Emphasis has been placed on corruption as a political issue since it is a situation in which businesses, public servants, and politicians abuse their privileged positions to pursue economic gain (Hindess 2001) while preventing a government from functioning properly (Hindess 2001).

Corruption and the corruption perception are usually considered to be cultural

phenomena, mostly because they primarily depend on how a society interprets and understands the rules and what constitutes a deviation from this regulation (Melgar, Rossi & Smith 2010). They also rely on the moral views and personal values of the people: while some people refrain from paying bribes, others might be inclined to do so and view the action as justifiable (Melgar, Rossi & Smith 2010). The disposition to pay a bribe is positively correlated with an individual’s perception of corruption. High levels of corruption perception have also been associated with the tradition of gift-giving, which is likely to raise levels of corruption (Egunjobi, 2013).

Elevated levels of corruption perception have a more debilitating effect than corruption itself, considering that they create a culture of distrust towards certain public or private

institutions. A significant discrepancy exists between corruption perception and the current level of corruption, although corruption perception influences corruption (Tverova, 2011). Moreover, a high degree of corruption perception is a factor likely to affect the economy (Rontos,

(10)

10

Business ethical behaviorbegan to be studied as a result of the Russian market reforms and the vagueness surrounding what constitutes ethical behavior in a given business setting (Tsalikis & Seaton 2008). The Business Ethics Index (BEI) is utilized by Russia, Romania, Poland, and Bulgaria, all ex-communist countries, to measure consumer perceptions of ethical business behavior. Of these four ex-communist countries, Romania has the highest BEI score (Tsalikis & Seaton 2008).

Most members of the public base their opinions on world affairs on their beliefs regarding how much their nation trusts other countries. International confidence is affected by age, partisanship, political trust, and social trust (Brewer et al. 2004). This fact is of crucial relevance for citizens when they decide between embracing isolation or internationalism, as demonstrated clearly in the example of citizens favoring military action against countries in the Middle East because they view these countries as threatening and unfriendly. International trust is important in shaping public opinions across a broad range of topics, including corruption (Brewer et al. 2004).

To increase public awareness of corruption, organizations such as Transparency International and the World Bank Institute publish annual measures, such as Transparency International’s Corruption Perceptions Index (CPI). The ranking of countries by their perceived corruption has been used as an indicator of development and governance. According to

Transparency International, this ranking is primarily designed to spotlight corrupt countries as a way of encouraging good governance and transparency. It is also an incentive for reform in the highly ranked countries.

However, this ranking practice is likely to create a perverse effect, as the perceptual indices result in the loss of much-needed investment (Warren & Laufer 2009). Perceived integrity is crucial for businesses and interested investors (Stevens 2013). If an organization is perceived to have unethical investors, then the organization will have trouble attracting foreign investors (Stevens 2013). Thus, in a way, the perceptual indices may contribute to higher corruption rates in poorly ranked countries (Warren & Laufer 2009). Increasing need for information on governance standards, commitment to upholding the rule of law, and perceived corruption by multinational companies (MNC) led to seeking new consumer markets (Warren & Laufer 2009).

(11)

11

Another effect of public opinion regarding political corruption is that it may destabilize developing economies, as increasing misconduct is likely to alter public sentiments regarding democratic politics. This evolution affects citizens’ support for specific institution or

administrations, especially in Latin America (Canache & Allison 2005). Citizens’ views on political corruption inform their appraisal of the existing agencies and authorities of democracy. Political corruption affects many democratic nations, considering that it is regarded as the abuse of public power that distorts government output and democratic procedures. However, when citizens correctly perceive corruption, then they are more likely to adjust opinions (Canache & Allison 2005).

Anti-corruption

There are studies stating that every year, billions of dollars are allocated for education and health sector assistance and end up enriching a few corrupt individuals (de Jesus Soares 2015).Corruption continues to be a global issue, with increasing numbers of countries

experiencing major corruption scandals (Transparency International, 2015). International bodies such as the Organization for Economic Cooperation and Development (OECD), the World Trade Organization (WTO), and the International Monetary Fund (IMF) have been at the forefront of combating acts of corruption internationally. Corruption has created problems for different political systems, as it directly challenges fundamental principles of democratic governance such as equality and openness (Anderson and Tverdova, 2003; Chang and Chu, 2006).

The government plays a major role in propagating corruption through privatization of public services and government utilities. Reducing the regulations imposed on economic

activities is one means of reducing acts of corruption. Economies may also be less susceptible to corruption if the majority of politicians and high-ranking government officials are more educated and professional (Treisman 2007).

Policy reforms addressing self-regulation, voluntary involvement, and corporate principles can encourage cooperation and commitment in fighting against corruption (Hess 2009). Such an approach can be effective in encouraging MNCs to implement compliance and ethics programs. Amnesty programs for corporations that report their corrupt payments are one

(12)

12

method of encouraging the reporting of instances of corruption. The national legislation can also expand the definition of criminal laws involving corrupt practices and utilize corporate monitors as a way of enforcing anti-corruption laws (Hess 2009). Unfortunately, corporations cannot effectively and efficiently fight corruption. Joint efforts between corporations, civil society, and the government are essential in ensuring that each party lives up to their commitment and assigned obligations, as well as in assisting each party so that these goals are achieved.

To conclude, the literature on corrupt practices and the response of companies and governments is vast but rather despondent to experts’ opinions and cases. There are ambiguities and overlaps in understandings of terms. The articles and studies focus on a single piece of the puzzle, and more often than not, the entire picture is rather incomprehensible, and the causes of a certain development in terms of corruption in a country are obscure.

Foreign Direct Investments

Nowadays, the premises of a country’s economic development lie in the abundance of capital. Public investments sources are often limited by various national interests that tend

become priorities: salaries and pensions, controlling the budget deficit, etc., and in the absence of investments, the economy loses. When the public resources are limited then investments should be encouraged by private funding (Bonciu, 2007).

Foreign Direct Investments (FDI) is a healthy growth source on which the economy can be based on even during time periods when economic stability is shaking, and the growth is under pressure (Şerbu, 2007). Firstly, one such positive aspect is represented by the FDI being complementary to public financing sources and can provide the necessary capital for economic development (Şerbu, 2007). Secondly, FDI creates new jobs not only for the company investing but also for the partner companies in the recipient country (Şerbu, 2007). Thirdly, they do not represent just a flow of capital but also of technology, knowledge, organizational practices, all which stimulate and generate economic growth. Foreign investors impose their working practices inside the company they are developing and introducing new technologies which increase the

(13)

13

employees’ efficiency and the company’s competitiveness. These effects spread through the entire chain which is involved in making a product or delivering a service (Şerbu, 2007). Fourthly, FDI is characterized by long term stability, what defines FDI is the investor’s sustainable interest in the company where he/she invests (Şerbu, 2007).

FDI: Concepts, determinants, and the impacts on the recipient

Concepts

Accurately determining the value of FDI and their components is essential for developing economic policies and capitalizing the benefits they bring to host economies.

The FDI implies the existence of at least two operators – the issuer of the investment and the receiving agent – located in two different countries (Dumiter, Todor, 2014). The

Organization for Economic Co-operation and Development (OCDE) issued a document for defining FDI called ‘The Benchmark Definition of Foreign Direct Investment’ which defines FDI According to this definition, FDI is a “category of investment that reflects the objective of establishing a lasting interest by a resident enterprise in one economy (direct investor) in an enterprise (direct investment enterprise) that is resident in an economy other than that of the direct investor. 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 company. The direct or indirect ownership of 10% or more of the voting power of a company resident in one economy by an investor resident in another economy is evidence of such a relationship” (OECD, 2008).

Determinants of FDI

The developing activities of the multinational firms and the creation of production networks have stimulated the interest of the academic environment in the direction of understanding and explaining the decision-making process of the company that leads to offer FDI. One of the best-known explanations of this process belongs to John Dunning being called the Eclectic Paradigm,

(14)

14

and it proposes four conditions to be reached by an abroad company to achieve added value (Dunning, 1995; Dunning, 1988).

The first condition is represented by the extent to which a company has ownership

advantages over other foreign enterprises in the markets it is trying to penetrate (Dunning, 1995; Dunning, 1988). The second one is the extent to which a company perceives to be in its interest to invest value in the previously mentioned ownership advantages, rather than selling them to the competition (Dunning, 1995; Dunning, 1988). The third action is the extent to which the

interests’ of a company are satisfied by creating ownership advantages abroad (Dunning, 1995; Dunning, 1988). The last condition talks about the extent to which a company considers the offshore production to be long-term sustainable (Dunning, 1995; Dunning, 1988).

The economic and social transformations to which all countries are subjected as a result of the globalization process is constantly generating new determinants for FDI whose effects are not fully highlighted in the empirical literature (Blonigen, 2005). Most likely the nature of these determinants is in a permanent dynamic provoked both by the general changes of the business field and its implementation abroad and by the economic policies and the attitude of the FDI receiving countries. At the same time the FDI impact on the host country and its negative or positive effects generated by foreign investors depends on the promotion strategies of the FDI, and thus a multitude of determinants can be named as influencers (Blonigen, 2005).

The socio-economic environment is one of the determinants. A comprehensive analysis on this topic is provided by Blonigen in 2005, who defines the socioeconomic environment as represented by the host country’s macro economic environment, the stability of prices, the political stability, and the degree of economic freedom (Blonigen, 2005). The quality of the public institutions measured by the levels of corruption and governmental efficiency is another factor for determining FDI inflows (Blonigen, 2005), as is legislation translated into the clarity and implementation of regulations (Blonigen, 2005). The degree of economic development which is measured by the evolution of GDP, the extent of innovation, industrialization, and the investments in research and development can be a key element for receiving foreign investments (Blonigen, 2005). The level of economic openness determinant refers to the FDI legislation in the recipient country (Blonigen, 2005). The extent of infrastructure explained through

(15)

15

The labor market defined through its size, level of education, price, and productivity is another determinant (Blonigen, 2005). The final example of determinants is the scale of the market, which takes into account the size of the population, salaries, and the unemployment rate (Blonigen, 2005).

The impact of FDI on the receiving countries

One of the most common and debated issues relates to the effects the FDI have on the overall economic and social welfare of a country. Moreover, if these effects are positive, the question arises to what extent they can be boosted by the governments of the countries receiving foreign investments (Gorg, Strobl, 2001). At the same time, there are voices that claim that the larger presence of multinational companies (MNC) in a country creates too much dependence on the foreign capital and policies, which can end up being harmful to the local businesses who cannot compete with the MNCs (Meyer, Sinani, 2009). It is worth mentioning that over the past

decades, political opinion on FDI changed more to these being viewed as economic growth generators for the host countries, and less regarding their adverse effects on the economy

(Fetscherin, 2010). One of the most visible results is the increase in the international competition for foreign investments and the emergence of success stories such as Ireland, Poland, or China (Fetscherin, 2010). Despite this, the effects produced by the foreign investments in a country depend on various factors like the characteristics of the sector in which MNCs invest, the time period in which the impact of a foreign investment is analyzed, the nature of the implementation of a foreign company (e.g. buying a local company, mixing with a local company etc), the strategies of the investors, and the strategies of the host governments in regards to FDI (Welch et al., 2007)

The direct impact of FDI begins to manifest at a macroeconomic level, starting from the financial capital provided by the MNC (be it from direct equity participation, through reinvested profits, or loans to the local companies) and passing through the transfer of technology and innovation - boosted by the R&D (research and development) spending – by exposing local businesses to a set of management, marketing, and entrepreneurial skill set that were initially missing or were insufficiently developed, but also by the development of human resources

(16)

16

(which not only mean raising wages or incomes, but also training programs for new jobs and skills) (Hanousek et al., 2011). At this level, the FDI effects can be observed through the increased performances of the local companies in which the MNCs invest their degree of competitiveness on the market, and also in the business practices promoted by these. From a macroeconomic level, the effects of the FDI presence can propagate at sectoral level by

modifying the structure of the markets and the degree of focus on the end-users and suppliers. At the macroeconomic level, the effects can be seen through higher contributions to the budget and the modifications brought to the exports structure (Hanousek et al., 2011).

The indirect impact of FDI on host economies arises in the form of teaching the local companies as a result of mimic behavior. These effects include technology transfer, human resources development, or increasing the performance of domestic firms competing with multinationals’ subsidiaries (Hanousek et al., 2011).

Although, as mentioned earlier FDI is considered to be a source of economic growth and competitiveness, there are situations where their impact on the recipient countries is negative – for example the presence of foreign firms may lead to a deterioration in market position for local companies, to unfair competition (given the resources that foreign companies have), and the bankruptcy of local producers. From this perspective, the clear role of a host country’s strategy is to direct the foreign investments towards regions where the benefits would be optimally

exploited, and the negative ones to be reduced as much as possible (Bonciu, 2009).

To conclude the analysis of the literature regarding corruption and FDI, there are few synoptic ideas that need underlining. The purpose of the previous sections was to examine the concepts of corruption and foreign direct investments to offer an in depth understanding of them and to set the aim of the paper by identifying the lack of analysis of the two concepts with relation to each other.

Firstly, an overview of corruption was provided followed up by five potential governmental sources which can determinate corruption: authorization and regulations, taxation, spending decisions, provision of goods and at below market prices, and other discretionary decisions (Tanzi 1998; Sanyal 2005; Dreher & Schneider, 2010). The notion of perception of corruption

(17)

17

has been tackled and showed to belong to a cultural phenomenon as it is influenced by how the members of society view it (Melgar, Rossi & Smith 2010). This section also discusses the efforts of Transparency International to raise awareness about the corruption perception and how, by doing so, it creates a perverse effect as the perceptual indices result in the loss of investments (Warren and Laufer 2009). This study concluded with a subchapter about anti-corruption and the importance governments, and international actors have in decreasing this phenomenon (Treisman 2007).

Secondly, the concept of foreign direct investments was investigated and stated their

importance as an economic growth source even during time periods of instability (Şerbu 2007). Further on, the key determinants for foreign investors to invest in a country were discussed through the prism of Dunning’s (1998; 1995) Eclectic Paradigm and were later backed up by Blonigen’s (2005) views upon the determinant factors to receive FDI. The macro level benefits of FDI such as higher contribution to the local budget and micro-level benefits like modifying the structure of the markets and the degree of focus on the end-users and suppliers (Hanousek et al. 2011) were engaged. The section concluded with some negative aspects of FDI like the possible outcome of unfair competition and bankruptcy of local producers as viewed by Bonciu (2009).

Taking into account the previously mentioned arguments, the literature shows potential in establishing a relation between corruption and FDI, a relation which shall be further investigated in the following chapter.

Research Hypothesis

As the paper focuses on the relation between FDI and corrupt practices and presents a case study on the corruptibility of the foreign investors in Romania, one hypothesis was

formulated firstly at a macro level and tested by one method, and secondly at a micro level tested via a different method: H1: Corruption Affects Foreign Direct Investments. H2: Foreign

(18)

18

H1: Corruption Affects Foreign Direct Investments

This hypothesis, tested at a macro level, refers to the macroeconomic link between corruption and FDI and is to be verified with a Granger causality test. The purpose of this particular analysis is to determine if there is a significant link between corruption and FDI, whether FDI is affected by the level of corruption or the standard of corruption is influenced by FDI and what is the lag of the causality. The main purpose of the analysis is to prove that Granger causality is encountered at a global level between corruption and FDI, by the literature. Furthermore, a times-series cross sectional regression introducing GDP and inflation will be presented with the purpose of showing whether or not there are variables which may influence the relation more.

Theoretically, corruption is viewed as either a helping or a grabbing hand for inward FDI. The analogy of the helping hand views corruption as a way of stimulating commerce, especially in economies that are characterized by pre-existing government failure (Lui 1985; Shleifer and Vishny 1993; Kaufmann and Wei 1999; Egger and Winner 2006). In other words, foreign investors may view the corruption of a country as a catalyzer for their business. In contrast, the grabbing hand views corruption as an unnecessary vice that increases a firm’s operating costs, and its benefits are outweighed by general equilibrium wage (Murphy, Shleifer, and Vishny 1991; Shleifer and Vishny 1993; Boycko, Shleifer, and Vishny 1995; Egger and Winner 2006). These two analogies are utilized in the analysis of the overall effect of corruption on FDI and in determining whether the effects are negative or positive.

There are two types of FDI, which are market-seeking and resource-seeking. Market-seeking FDI involves directly investing in a host country to provide the market with locally produced goods and services, instead of serving the market through export. Resource-seeking FDI, on the other hand, involves directly investing in a host country to achieve cost minimization by acquiring resources that are either unavailable or too costly in the home country (Brouthers, Gao & McNicol 2008).

Corruption in host countries has increasingly gained prominence as a contributing factor to declining FDI, particularly since less corrupt countries transact with more corrupt nations. The

(19)

19

corruption gap between the host and country of origin strongly affects investor behavior. As the difference between two countries increases, the likelihood of the two countries to deal decreases (Habib & Zurawicki 2002).

Economically, corruption has a detrimental effect on aggregate FDI flows, as it is an important barrier to investment. The main reason corruption deters FDI is the increase in

insecurity, especially when it comes to cost. Corruption also acts as a form of tax on investments. Overall, corruption decreases the probability of firms investing in a host country and gives rise to joint venture ownership structures involving local partners rather than wholly-owned investments (Habib & Zurawicki 2002).

Corruption raises costs, heightens uncertainty, and produces bottlenecks, which greatly affects FDI (Habib & Zurawicki 2002). Corruption propagates market distortions by affording some foreign investors preferential access to certain profitable markets. This limits equal and open market access to all competitors, which aggravates concern for foreign investors.

Corruption is viewed as unethical, morally wrong behavior, and foreign investors are known to avoid countries that are identified as having high levels of corruption. Foreign investors may also shun investing in corrupt countries as it is costly, risky, and hard to manage.

H2: Foreign Investors are Subject to Contamination by the Corrupt Practices in Romania

This hypothesis tested at a micro level will investigate whether or not the foreign investor in Romania indulged in the corrupt practices of the country. Corruption is more of an

institutional problem than a market failure. Malfunctioning bureaucracy makes firms resort to corruption. The probability of businesses operating on a global scale encountering bribery has increased greatly. Approximately 40% of international businesses paid bribes to facilitate their operations (Kwok & Tadesse 2006). In some regions, corruption and illegal business practices are the norms, considering the widespread practices. The existence and significant growth of corruption in Romania may limit international trade and create a barrier for future international markets.

(20)

20

Developing nations like Romania are suffering from the effects of corruption as government officials benefit from the resources flowing into the country. Multinational

enterprises (MNE) in emerging economies and developing countries are pressured into regularly engaging in corrupt practices (Collins, Uhlenbruck & Rodriguez 2009). Foreign investors make a critical strategic choice to be involved in corruption, given that the situation in today’s world recognizes corruption as a norm and not an exception. In countries where the standard of

corruption is locally institutionalized, MNEs’ performance, reputation, and operations are greatly affected (Spencer & Gomez 2011).

Most efforts against corruption, in particular by the OECD, address the supply side of corruption, which represents the payer of a bribe. Emphasizing anti-corruption from the supply side shields foreign investors from becoming subject to contamination by corrupt practices (Goldsmith 1999). However, this model has limitations, considering that a focus on one side of global corruption is a less efficient way of detecting and preventing corruption. As mentioned above, government officials are motivated by economics, poverty, government shortcomings, culture, geography, and climate to demand or accept bribes. These factors also motivate foreign investors to offer bribes.

Currently, MNCs are particular in their choice of host countries for foreign subsidiaries. Concerns are raised about increases in operational costs and risks, which are exacerbated by pervasive corruption. Toleration of corruption facilitates and encourages the participation in bribery of more government officials and foreign investors over time (Baughn et al. 2010). High levels of corruption undermine a government’s legitimacy. The public institutional setting is a fundamental element in shaping investor behavior as well as the behavior of an MNC, as some of the corruption’s characteristics sometimes become institutionalized, such that they end up

becoming fundamental components of the institutional environment. However, the pressure that a foreign investor faces to engage in bribery varies according to how institutionalized corruption is in their home and host countries.

(21)

21

Methodology

The research project refers to one hypothesis begin analyzed at two different levels and tested via different methods. The scope of analysis is two pronged: starting from the assumption that there is an effect of the corruption standards of the FDI a country attracts, and also that a strong FDI presence may cause the adoption of best practices and a rejection of corruption, the analysis investigates the wide, macro view, with an econometric approach for all countries where a relation exists, and the micro view shall be tested with the help of a qualitative research approach for Romania.

Primary data sources

For testing the macro aspect of the hypothesis, three sets of data were extracted from three different sources: FDI, Corruption Perception Index and Control of Corruption from World Bank and Transparency International.

The first variable was extracted from World Bank’s database, representing 174 countries FDI net inflows for the 1996-2015 period. The second variable was extracted from Transparency International’s online database, which is an international NGO fighting against corruption and consequently measuring the progress it makes through an index created by them. This index, the Corruption Perception Index (CPI) was measured starting from 1995. However, only

progressively, the number of countries for which the index was calculated increased. Lastly, Control of Corruption index (CCI), used to show how public power was used for private benefits, is provided by the World Bank’s online database as part of a wider analysis of

worldwide governance indicators. The indicator(s) was calculated starting from 1996 and, until 2002 when the index was updated bi-annually only. There are differences between the corruption perception index and the control of corruption, as can be noticed from the gaps in Azerbaijan, Barbados, Dominica, Malawi, and others. These difference, visible in the data, also represent a reason for running the analysis for both types of indices, instead of a single measurement of corruption. Figure 1 provides an overview of the FDI levels in 2015. The graph proves the

(22)

22

significant discrepancies between developed countries, which are investment hubs, receiving and investing, as compared to developing countries in a quest for gaining FDI flows.

(23)

23

Preliminary data treatment

It becomes evident that the analysis was greatly limited by data’s span of time. While common years for all three data sets exist only from 1996 onwards, there are still other issues which had to be tackled.

In the first place, some countries’ corruption level was measured discontinuously, meaning that certain years have no associated value. Moreover, due to geopolitical changes, some countries’ borders modified during the time interval under analysis. This is especially the case for ex-Yugoslav countries, some of which had to be dismissed from the analysis on the base of discontinuity. No inferences can be made when talking about different political entities.

Another important issue refers to the change in calculation methodology. Two such changes are to be mentioned.

The first one, the CPI’s methodology of calculation changed during the 2011-2012 period. Until 2011, the index was assigned a value between 0 and 10, with “10” meaning no corruption at all and “0” meaning maximum corruption level. In 2012, a new index was designed on a scale from 0 to 100. Probably such a decision was taken for achieving better accuracy and comparison power. Transparency International’s website specifically mentions that the two sets, separated by methodology, are not comparable. This specific difference in the indicator would have meant that this precious data set would have become unusable for the analysis’ purpose. Therefore, it was decided in this research to divide all CPI indicators from 2012 to 2015 by 10 for obtaining a common measurement scale. While this could be seen as an arbitrary decision, it had to be taken to keep the data series feasible.

The second change, the CCI was firstly calculated bi-annually. This was the case until 2002 when they began to update in on an annual basis. Thus, for the years of 1997, 1999 and 2001 there is no CCI assigned for any country whatsoever. I resolved this issue by artificially calculating CCI for these missing years by using the average of the two years of proximity. So, for example, CCI’s value for 1997 was with the help of an arithmetic mean between CCI’s 1996

(24)

24

and CCI’s 1998 values. CCI has assigned values between -2.5 to +2.5 where ‘-2.5’ is the weakest governance performance and ‘+2.5’ is the strongest.

The pivotal time series consists of the FDI. Therefore, both CPI and CCI, which are seen as proxies for countries’ corruption levels, depend on FDI’s existence which is available for all countries and the 1996-2015 years. In other words, there is no use if there are CPI and CCI series for a given country, but FDI series is not available. On the other hand, CPI and CCI were seen as representing the same values of interest since there is a strong correlation (r=.89) between the two (Van Dijk, Van Mierlo 2010). Thus, when one of them is missing the other one can take its place for the necessary calculation (VAR and Granger causality especially) along FDI.

These being said, the countries were divided into two categories from the start. The division is purely observational and is determined by the access to information. The reason behind this is the fluency it adds to the analysis. The first category is represented by the countries for which there are FDI data series and one of the two proxy data series for corruption (i.e., CPI or CCI). These were formally named “series of two” countries. The second category is represented by the countries for which there are FDI data series and both data series proxies for corruption (i.e., CPI and CCI). These were formally named “series of three” countries.

For the first hypothesis, the detailed methodology had to undergo a series of steps, six to be specific. Firstly, the series was made to be stationary with the help of first order differentiation. In the case of GDP series, the values were first transformed to natural logarithms and only then differentiation was made. Secondly, the stationarity of each series was checked with the help of ADF test (Augmented Dickey-Fuller). Thirdly, the cointegration of each pair of series corresponding to each country was checked with the aid of Johansen cointegration test in order to avoid spurious regression (Johansen 1995). Fourthly, Wald Lag-exclusion test was performed for each pair of series corresponding to each country was undertaken for finding out the best lag for each VAR model (Harrell 2001). Fifthly, the VAR model was calculated for each pair of series corresponding to each country, taking in consideration whether each pair is or not cointegrated and the most efficient lag number for inclusion (Hatemi-J, 2004). Sixthly, the Granger-causality test was undertaken for each VAR model with the purpose of revealing the possibility of forecasting the dependent time series with the help of the independent time series (Granger 1969).

(25)

25 Caveats

The above ADF test can be seen as a filter. Those series which did not pass the test were not taken into consideration. Due to fewer observations per series, a second differentiation would be irrelevant.

For some countries’ series, the information provided by the primary sources were incomplete, i.e., some of the years were missing. These were excluded from the analysis. Consequently, only those series remaining with 19 observations after differentiation were taken into consideration. Hence, each set spans between 1996 and 2015.

As result of the previous two points, some of the countries have both CPI values (Corruption Perception Index provided by Transparency International) and CCI values (Control of Corruption values provided by World Bank), while others have only one of the two. However, all countries have FDI data (net foreign direct investments) which are considered sine qua non for each country under analysis.

Reasoning and Analysis

1. Dickey-Fuller test

In this case, the analysis used differentiation of first order. The resulting series was accepted into analysis at a 10% significance level. The significance level of acceptance is higher than usual (5%) because there is no possibility to do a second order differentiation because of the number of data points each series has. A second order differentiation would further decrease the reliability of Granger analysis which is the final goal. In synopsis, a series of conclusions can be drawn.

From the set of countries having pairs of two data series, the following countries were rejected: Algeria, Bahrain, Croatia, Eritrea, Estonia, Iran, Laos, and Mongolia. From the set of countries having pairs of two data series, the following countries were rejected: Russia, Sweden.

(26)

26

These countries were rejected because they are not stationary series and failed to pass the Dickey-Fuller.

In this stage of the analysis, all the other countries are accepted in the subsequent analysis and are proper to be used in the following analyses. These accepted time series can be seen to have reliable parameters which can be utilized for the analysis. They are considered to be detrended and seasonally adjusted.

2. Johansen cointegration test

Without checking if the series under study have a long-term equilibrium relation, spurious regression can result, meaning that the determination found between variables is caused by a hidden trend. If cointegration is found to exist, the aspect has to be included into VAR modeling. If not, unrestricted VAR can be applied (Johansen 1995).

Johansen cointegration test has two forms: trace test and maximum eigenvalue test. I used the first form of the test, under which the null hypothesis states there is no cointegration relationship. If the null is rejected, I can conclude that there exists at least one cointegration relationship between variables. All the results in the table in the Appendix reveal the countries have data sets with at most one cointegrating relationship. All pairs which did not fulfill the 5% significant level are not shown. These latter will be analyzed through unrestricted VAR.

3. Wald lag-exclusion Test, VAR and Granger Analysis

Before applying VAR and subsequently Granger-causality test within VAR framework, Wald lag-exclusion test was used for obtaining the best number of lags which are to be included in the model. Because of the limited number of data points within each series, a maximum number of three lags are allowed.

Wald lag-exclusion test results are not showed here, for reasons of space, but they were used implicitly in VAR construction for each pair of series. Thus, one can easily realize the

(27)

27

results of Wald lag-exclusion test by checking the interpretation of Granger. Therefore, I will proceed with Granger directly.

What Granger test reveals is the possibility of forecasting one time series (i.e., the dependent time series) with the help of another time series (i.e., the independent time series). The null hypothesis states that there is no possibility of forecasting between the two series. Consequently, if the opportunity to forecast exists, it means that one series determines the other. Therefore, Granger-causality test result can also be viewed to reflect the relation strength of two-time variables. The result is accepted at 5% significance level. If 5% significance level is not met, the null hypothesis is rejected.

Further, there are two types of VAR used: VEC (Vector Error Correction Estimates) which is used when the series are cointegrated and UV (Unrestricted VAR) which is used when the series are not cointegrated.

The conclusions of the analysis are as follows: Pairs of two data series:

Type of Relation Country

FDI influences CCI and CCI influences FDI Angola, Benin, Trinidad, and Tobago

FDI influences CCI, but CCI does not influence FDI Cameroon, Dominica, Dominican Republic, Ecuador, Gabon, Gambia. Grenada, Guyana, Kenya, North Korea, Lesotho, Myanmar, Pakistan, Paraguay, Romania, Sri Lanka, UAE

CCI influences FDI but FDI does not influence CCI Chad, Costa Rica, Egypt, Mozambique, Panama, Saudi Arabia, Somalia, Uzbekistan, Zambia

Pairs of three data series:

FDI influences CPI and vice-versa Indonesia, Israel, Thailand FDI influences CCI and vice-versa Singapore

FDI influences CPI but CPI does not influence FDI Argentina, Austria, Japan, Nigeria, Spain, Turkey CPI influences FDI, but FDI does not influence CPI Colombia, Hungary, Ireland, South Korea, Mexico FDI influences CCI but CCI does not influence FDI Indonesia,

(28)

28

For all the other countries, the null hypothesis was not rejected, meaning there are no possibilities to forecast one series based on the other. A few analytical observations from the global Granger causality analysis may be made.

In most cases, the span of influence is one year, regardless of the direction of the influence. This confirms a certain myopia of investors, as well as public administration systems, which are reluctant to change and any lesson learned, has a short time span to make a difference.

Exceptions to the one-year influence span are Nigeria, Kenya, Cameroon, Japan and the Czech Republic.

Out of 49 countries in which influences have been found, in 36 cases (73.5%) the influence is from FDI to the corruption indicator (may that be Cost of Corruption or Corruption Perception Index). Out of these 36 cases, seven also have a bidirectional influence (Benin, Angola, Trinidad and Tobago, Indonesia, Israel, Singapore, Thailand). In 25 cases (51%), including the bi-directional influence, the corruption indicator affects the FDI series.

4. Times-Series Cross Section Analysis

To provide more clarity to the analysis, I run a time-series cross sectional analysis. This has been done with a view to check what other factors are influencing FDI besides corruption (CCI was used as a proxy). Similarly, I wanted to find what other factors, besides FDI, are influencing corruption.

In the following, I aim to find an overall relationship within a sample of 49 countries where Granger causality was previously found to exist. For doing so, I will use time-series cross-sectional analysis (i.e., panel-data analysis). This general methodology is allowing the analysis to account for the differences in both time-series variation within an entity and cross-sectional variation among entities (i.e., countries in our case). All variables involved were first transformed to natural logarithms. In this section, a series of quantitative control variables will be used to perform the analysis, variables which will be described in the following paragraphs.

The first variable is inflation which is expected to have an adverse impact on FDI and little to no impact on corruption. In general, investors are discouraged by the high volatility of currency

(29)

29

(Udomkerdmongkol, Görg, Morrissey 2006). The second variable is debt service as a percentage of GDP and is expected that a high debt service to increase a country’s risk of default and thus discourage FDI (Wamboye, 2012). The third one is the official exchange rate as local currency unit (LCU) per 1$. High exchange rates are expected to encourage investments by making local currencies cheaper for investors (Udomkerdmongkol, Görg, Morrissey 2006). The fourth variable is a risk premium on Lending as lending rate minus treasury bill rate measured in percentages. It is expected for the high-risk premium to discourage investments by increasing a country’s risk of default (Páez, 2011). The fifth one is a domestic credit to private sector as a percentage of GDP. It is expected for high domestic credit to positively encourage investments, as aggregate demand increases (Bongini, Iwanicz-Drozdowska, Smaga, Witowski 2017). The sixth variable is population growth in percentage. It is expected that high population growth to positively encourage investments by decreasing the cost of labor (Aziz, Makkawi 2012). The seventh variable is the research expenditure as a percentage of GDP. It is expected for a high R&D spending to encourage investments by making the local workforce more productive and qualified (Erdal, Göçer 2015). The eighth variable is unemployment as a percentage. It is expected for high unemployment to encourage FDI by decreasing the cost of the labor force (Haddad 2016). The ninth one is energy imports as a percentage of total energy used. It is expected for high imports to deter FDI by making a country more vulnerable to international price movements (Slimane, Huchet-Bourdon, Zitouna 2015). The tenth and final variable is GDP per capita at current dollars. It is expected for a high GDP per capita to encourage FDI because high GDP per capita means dynamic consuming market (GuechHeang, Moolio 2013). The effects these control variables would have on corruption can be conjectured through FDI variable. Thus, if FDI and corruption are negatively correlated, a control variable positively/negatively affecting FDI would decrease/increase corruption level correspondingly.

Finally, the two variables of interest are FDI and corruption (as measured by World Bank’s Control of Corruption Index). Each of them will take the roles of both exogenous (i.e., as regressor) and endogenous (i.e., dependent) variables by turn.

The analysis is done by following four steps. Firstly, the data series possessing unit root was firstly made stationary through first differentiation. Secondly, FDI data was transformed through natural logarithm for making the variation more significant. Therefore, the FDI data represents

(30)

30

the exponents of E number. Thirdly, I estimate fixed and random effects models, firstly considering FDI as the dependent variable and secondly considering CCI as the dependent variable. Fourthly, I identify which of the two models is most suitable to use by applying a Hausman specification test which assesses the consistency of two estimators out of which one is less potent (Greene, 2012).

In the variable column the intercept (C) is present as well. This is the constant term and it is generated by the regression (Frost 2013).

Results and Interpretation

Table: Fixed and Random effects coefficients results for all 49 countries during 1996-2015, considering CCI as dependent variable

Variable Fixed Random

Coefficient t-Statistic Prob. Coefficient t-Statistic Prob. C -0.20827 -4.60192 0 -0.30244 0.075004 0.0001

FDI 0.003734 4.981819 0 0.003773 0.000749 0

Debt Service -8.36E-05 -0.07777 0.938 0.000212 0.001054 0.8407 Domestic Credit to Private Sector 0.001416 3.303198 0.001 0.001793 0.00042 0

Energy Imports 0.000147 1.575839 0.1154 0.000207 9.15E-05 0.0238 Exchange Rate 2.10E-05 1.389286 0.1651 5.24E-06 1.42E-05 0.7115 GDP -7.58E-06 -5.39061 0 -3.16E-06 1.34E-06 0.0191 Inflation -7.95E-06 -0.16698 0.8674 -9.39E-06 4.75E-05 0.8434 Population Growth 0.013616 1.495791 0.1351 0.007968 0.00894 0.373 Research Expenditure -0.00569 -0.27438 0.7839 0.044903 0.020033 0.0252 Risk Premium on Lending 0.000809 0.384755 0.7005 0.000778 0.00207 0.7071 Unemployment -0.00178 -0.5532 0.5803 -0.00142 0.003049 0.6405

(31)

31

Table: Hausman Test for above table

Test Summary Chi-Sq. Statistic

Chi-Sq.

d.f. Prob. Cross-section

random 192.4892 11 0

Table: Fixed and Random effects coefficients results for all 49 countries during 1996-2015, considering FDI as the dependent variable

Variable Fixed Random

Coefficient t-Statistic Prob. Coefficient t-Statistic Prob.

C 20.75828 10.66144 0 19.03246 14.50671 0

CCI 7.649965 4.981819 0 2.263966 2.933338 0.0034 Debt Service 0.036766 0.755942 0.4499 0.087919 2.587401 0.0098 Domestic Credit to Private Sector 0.005071 0.25978 0.7951 0.004694 0.335315 0.7375

Energy Imports 0.031461 7.688065 0 0.023116 8.103972 0 Exchange Rate -0.00063 -0.9163 0.3598 -0.00034 -1.13805 0.2554

GDPPERCAP 6.09E-05 0.941265 0.3468 -7.48E-06 -0.18614 0.8524 Inflation 0.002667 1.239222 0.2156 0.003266 1.547097 0.1222 Population Growth 0.0469 0.113679 0.9095 0.356602 1.161341 0.2458 Research Expenditure -0.75057 -0.7993 0.4243 -0.95846 -1.5836 0.1136 Risk Premium on Lending -0.19745 -2.07956 0.0379 -0.09572 -1.41587 0.1572 Unemployment 0.008449 0.058085 0.9537 0.043142 0.606274 0.5445

Table: Hausman Test for above table

Test Summary Chi-Sq. Statistic Chi-Sq. d.f. Prob. Cross-section random 33.80719 11 0.0004

(32)

32

As seen from Hausman Test, the null hypothesis can be rejected for both tables meaning that cross-sectional fixed effect model is the most proper for the time-series cross-sectional (i.e., panel) data given. Therefore, the variables influencing CCI are FDI, domestic credit to private sector and GDP per capita. On the other hand, the variables influencing FDI are CCI, amount of energy imported from total energy consumed and the risk premium on lending. All these are significant at <5% significance level.

By introducing control variables, I managed to show that I needed to account for other variation as well to detect the actual relationship existent between FDI and corruption. I can firstly conclude that an increase in FDI will increase corruption, unlike GDP per capita and the amount of credit to the private sector which have an opposite effect on corruption. Secondly, an increase in corruption will increase FDI. At the same time, the amount of energy imported influences FDI in the same manner as corruption, while enhancing risk premium will negatively affect FDI. In summary, the hypothesis that corruption affects FDI at a macro level is accepted.

Corruption and Foreign Direct Investments in Romania: The Corruptibility of Foreign Investors

Romania is a particularly interesting case study in which concerns corruption and FDI. The former has been seen as a solution to competitiveness after the dissolution of the centralized communist system; it was a major issue in the process of joining the European Union. The latter was either the savior of the Romanian economy (the myth of the foreign investor) or a leech draining the resources of the country. Hence, an objective analysis starting from the results of the global view of the puzzle is necessary.

The conditionality to get corruption under control was one of the most important issues on the agenda of the process of adhering to the EU. Romania, alongside Bulgaria, had to have a special mechanism developed to fit the European standards, namely the CVM (Cooperation and Verification Mechanism).

(33)

33

The CVM is a safety measure created by the European Commission in order to help Romania and Bulgaria implement their commitments to the EU to fight corruption and organized crime (European Commission 2007). This was needed as a result of the European Commission assessing that these two countries still needed help in surpassing corruption matters (European Commission 2007). The CVM is still in force at the time of this thesis, proving that the

anticorruption measures have produced insignificant results.

The rationale of the second hypothesis and choosing Romania to test it is thus simple: a country with a European mechanism in place to enforce the reduction of corruption, with a defined strategy in the late 90s, early 2000s to bring in more foreign investors, ends up a decade later convicting a major foreign company for corruption and even later, has the Prime Minister making clear statements against foreign investors. The freeze frame moment is almost evident. In 2001, six years before joining the European Union, the PM at that time, Adrian Nastase was making a clear case against corruption as a deterrent for foreign investors in a discourse at Birmingham, stating:

“The government will not allow bureaucracy to shake our national potential or create opportunities for corruption. The fight, both against corruption and bureaucracy, is our key priorities. The administrative procedures for those who want to do business in our country have been simplified. We are aiming for a more efficient administration, which, with British support and from the European Commission, is already starting to prefigure. With this in mind, we have created a special department of foreign investor relations, subordinated to me, to ensure that significant foreign investors receive the assistance they need and easier access to the decision-making process. We strengthen the law enforcement system and adopt the Anti-Corruption Program, based on World Bank analysis. We are extremely serious about eliminating this cancer, corruption, on the body of our country. In the first months of government, 195 people, including senior officials, were charged with corruption.” (Romanian Government, 2001). The same politician was convicted in 2014 for corruption, in a statement of the judiciary system against like practices.

The question about what happened is however not the hypothesis of this paper. Still, it is interesting to see how Microsoft managed to get involved in a corruption scandal and whether the intensely affected environment of a host country ends up affecting the investor. Basically, this section tests the rotten apple theory in terms of investors. A rotten apple in a bag of sane

(34)

34

apples will make them rot, while a sane apple in a bag of rotten apples will only rot itself (Chomsky 1986).

It is easily noticeable from the Granger analysis that Romania is part of the group of countries with pairs of two data series. Most of the other EU member countries are, however, part of the second cluster: the pairs of three data series.

The conclusion of the analysis for Romania is that FDI influences Control of Corruption, with a 1-year lag (influence span). The hypothesis is thus accepted; the presence of foreign direct investments in Romania, combined with the political requirements brought by the international presence, supranational controls (as is the case with EU investors) brought a certain reduction in corruption.

In case of Romania, the evolution of FDI and CPI is noticeable in Figure 2. The Figure contains adjusted series in order to reveal in a snapshot the peaks and valleys of the evolution and to pinpoint key events which may confirm or infirm the second hypothesis. Thus, there are several years of importance: 2004 (medium high), 2006 (medium high), 2007 (medium low), 2008 (maximum high) and 2011 (maximum low).

Figure 2: FDI and CPI evolution in Romania

0 2 4 6 8 10 12 14 16 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Ad ju sted Va lu e o f I n d icato r Year

Romania's Evolution between 1996 and 2015 in terms of FDI and

CPI- adjusted series

CPI

(35)

35

In which follows, an analysis of these years shall be made, underlining the investments as well as the corruption prevention efforts of the state and the improvement in the governance, and concluding with the Microsoft corruption scandal in Romania.

The evolution of FDI in Romania

FDI has contributed heavily to the modernization of the Romanian economy and its integration into the European economy and the international chains of production. The foreign companies hire one-third of the workforce from the public field with more than 1.2 million employees (Foreign Investors Council, 2017). Judging by the technological development, higher exports, and the import of know-how, the growth of the foreign direct investments in Romania in the past years is undisputed. The country’s openness towards FDI begins in the early 2000s - the years of large privatizations which further encouraged FDI with Romania’s growing economy combined with the process of joining the EU. The time period of 2003-2008 sees a massive growth of FDI with a volume increase of five times (Figure 2). Following 2009, the country loses part of its attractiveness due to the effects of the economic crisis and increase investments up to three times in regards to the previous year (Foreign Investors Council, 2010). This was followed up by the economic and political turmoil of president’s Basescu impeachment scandal starting 2011 which made FDI stay below the level of 2004. This reflects in the FDI stock: Romania has the lowest volume of FDI stock per citizen reported to the GDP at the end of 2015 when

compared to Bulgaria, Czech Republic, Poland, and Hungary (Council of Foreign Investors, 2015). In other words, the country has the lowest rate of FDI attraction despite having a

consistent set of positive factors (geographical position, a large market, and cheap work force). During 2010-2015 foreign companies continued recruiting work force despite their fiscal value remaining constant. This suggests that most companies seek long term investments in Romania and are not interested in short term gains. The work of these businesses represents an average of 70% of the country’s exports and 60% of its imports (Council of Foreign Investors, 2016).

The most prominent foreign investors are the Netherlands, Austria, and Germany which own over 50% of the FDI stock (National Bank of Romania, 2015). Romania caught the eye of

How are foreign direct investments affecting the rate of corruption of a country : an analysis of the relation between foreign direct investments and corruption