MASTER THESIS IB&M
Culture and (Informal) Institutions and their Consequences
The Effect of Corporate Social Responsibility Pressures on the Location Choice of Multinational Enterprises:
Evidence from Corruption Charters and FDI Flows
Student name Jorrit J. Schotman
Student number S2386607
Address Molenweg 12A
7055 AX Heelweg
Phone number +316 21 464 124
Email address J.J.Schotman@student.rug.nl
University University of Groningen
Faculty Business and Economics
Program International Business and Management
Supervisor Dr. van Hoorn
Co-Assessor Dr. Dong
Date 20th June 2014
Subject Master Thesis IB&M
TABLE OF CONTENT
1. INTRODUCTION ... 5
2. BACKGROUND ... 7
2.1. The History of Corporate Social Responsibility ... 7
2.2. The Business Case for Corporate Social Responsibility ... 8
2.3. Corruption as an Element of Corporate Social Responsibility ... 9
2.4. The Problem of Corruption ... 10
2.5. Concluding Remarks ... 11
3. THEORY AND HYPOTHESES ... 12
3.1. Corruption as a Determinant of MNEs’ Location Choice ... 12
3.2. Corporate Social Responsibility Pressures from the Home-Country ... 14
4. DATA AND METHOD ... 17
4.1. Data ... 17
4.1.1. Dependent variable: FDI flows ... 17
4.1.2. Key independent variable: the host-country corruption rate ... 18
4.1.3. Other variables ... 19
4.2. Method ... 24
5. RESULTS ... 26
5.1. The Effect of Corruption on the Location Choice of MNEs ... 26
5.1.1. Descriptive results ... 26
5.1.2. Regression results ... 27
5.2. The Effect of Corruption over Time ... 28
5.2.1. Descriptive results ... 28
5.2.2. Regression results ... 28
5.3. The Influence of Home-Country Pressures ... 29
5.4. Robustness Check ... 31
6. DISCUSSION ... 34
6.1. The Effect of Corruption on the Location Choice of MNEs ... 34
6.2. The Effect of Corruption over Time ... 35
The Effect of Corporate Social Responsibility Pressures on the Location Choice of Multinational Enterprises: Evidence from Corruption Charters and FDI Flows.
JORRIT J. SCHOTMAN
University of Groningen, the Netherlands
ABSTRACT
Prior research has indicated that corruption has become an integral part of corporate social responsibility (CSR). Yet despite much research, the effect of this development on the relation between corruption and the attractiveness of host-countries has not become clear. The effect of corruption on the location choice of multinational enterprises (MNEs) has therefore been examined in this thesis, while taking into account home-country pressures. The research outcomes have shown that the prevalence of corruption has a negative effect on the attractiveness of host-countries. In addition, this negative effect has increased over time and is stronger for MNEs which are under more pressure not to engage in corruption. The OECD Convention on Combating Bribery appears to be a good indicator of home-country pressures and seems a first important step towards a corruption free world. Increasing the number of anti-corruption initiatives with a global reach to fight corruption is therefore recommended. However, most importantly more corrupt countries and MNEs which engage in corrupt activities should realize that their license to operate internationally is at stake, in case they decide to continue with business as usual.
Keywords: International Location Choice, Corruption, Multinational Enterprises, Development over Time, Home-Country Pressures, OECD Convention
1. INTRODUCTION
The development of corporate social responsibility (CSR) has been a lengthy and highly debated process, in which proponents and opponents continue to disagree (Brueckner, 2010: 25; Murphy and Schlegelmilch, 2013; Windsor, 2001). Nowadays it is however widely accepted that businesses do indeed have responsibilities beyond simply making a profit (Crane and Matten, 2010: 51). Multinationals enterprises (MNEs) are increasingly being pressured to act socially responsible and are being watched by an ever-broadening group of stakeholders (Carroll, 1991; Crane and Matten, 2010: 51; Porter and Kramer, 2006; The Economist, 2008). Nike for example experienced public outrage and high damages when its products appeared to be produced partly with child labor (Wolf, 2014).
According to Branco and Delgado (2012), the rejection of corruption should be seen as an integral part of CSR and is deemed incompatible with sustainable development in view of the social, economic and environmental damages caused by it. In addition, more and more scientific articles have stressed out the costs of corruption for societies and for MNEs (Hills, Fiske and Mahmud. 2009; Hindess, 2005; León, Araña and de León, 2013; Nichols, 2012). This has resulted in a significant change in the attitude towards corruption, namely one in which engaging in corruption [e.g. bribing government officials] is no longer an acceptable business practice (Stansbury, 2009). MNEs have therefore started to improve their social behavior, in order to promote their own self-interest and avoid being linked to CSR scandals, such as engaging in corruption (Branco and Delgado, 2012; Crane and Matten, 2010: 51; Hart, 2010; Islam and Deegan, 2010). Nevertheless, corruption among MNEs has remained disconcertingly high (Dutton, 2008) and many countries seem to be unable to improve their institutions and curb corruption (Mauro, 2004).
risks and costs of corruption (Doh, Rodriguez, Uhlenbruck, Collins and Eden, 2003). Corruption seems therefore to have become a negative indicator of the attractiveness of host-countries (Bellos and Subasat, 2012). However whether or not the increasing pressure not to engage in corruption, has increased the negative relation between corruption and the attractiveness of host-countries as well, has remained unanswered. In addition, the effects of new and stronger home-country pressures have not yet been researched academically. This thesis tries to fill in this literature gap by examining the effect of corruption over time, while taking into account home-country pressures.
2. BACKGROUND
2.1. The History of Corporate Social Responsibility
The development of CSR is a lengthy and continuous process of which Howard Bowen is credited as being “the father” (Brueckner, 2010: 25; Murphy and Schlegelmilch, 2013; Windsor, 2001). In his book Social Responsibilities of the Businessmen, which was published in 1953, Bowen posed the question to which extend the interest of business in the long run, would merge with the interest of society. In addition, he defined the social responsibilities of business, at that time, as the obligation of businessmen to pursue those policies, to make those decisions and follow those lines of action which are desirable in terms of the objectives and values of society. At first, CSR was considered a “radical doctrine”, as calls for CSR were vehemently opposed by those believing in the freedom of the business enterprise. (Brueckner, 2010: 25). One of those believing in the freedom of the business enterprise was Milton Friedman who published his book called Capitalism and Freedom in 1962. In his book he stated that “in economic freedom there is one and only one social responsibility of business – to use its resources and engage in activities designed to increase its profits so long as it stays in the rules of the game” (Friedman, 1962: 133). Later in 1970, Friedman emphasized that “if there are social responsibilities, they are the social responsibilities of individuals not of businesses” (Brueckner, 2010: 25). Another scientist who rejected CSR was Drucker, who wrote a book in 1973 called Management: tasks, responsibilities and practices. Drucker (1973: 315) clarified that it was the success of the business system which led to new and in many cases exaggerated expectations and overconfidence in managers and management.
(Carroll, 1991), which is still seen as one of the most important CSR models available in the literature (Brueckner, 2010: 26).
2.2. The Business Case for Corporate Social Responsibility
Today it is widely accepted that businesses do indeed have responsibilities beyond simply making a profit (Crane and Matten, 2010: 51). More and more firms are emphasizing their social responsibilities and “doing well by doing good” has become popular among the business community (Lee, 2011; The Economist, 2008; Walton, 2010). The business case for CSR, which states that corporations can benefit from voluntarily being socially responsible, has been promoted (Hart, 2010). Multinational enterprises have therefore started to take on social responsibilities to promote their own self-interest (Crane and Matten, 2010: 51), which includes maintaining and improving their reputation, government relations, customer loyalty and employee recruitment and retention (Walton, 2010). Yet problems arise where the interests of businesses do not merge with the interests of society, in which CSR is at risk of being reduced to camouflaging business as usual with a social twist (Brueckner, 2010: 28).
2008). MNEs are experiencing the costs of being linked to CSR scandals of their own firm and supply chain partners and exert pressure on suppliers to conform to the expectations that have been placed upon them by society (Islam and Deegan, 2010). Even though CSR might not always be pursued for the good of society, MNEs need to comply with stakeholder pressures in order to maintain a good reputation and keep their license to operate (Amaeshi and Adi, 2007; Lee, 2011).
2.3. Corruption as an Element of Corporate Social Responsibility
Corruption can be defined as the misuse of public office for private gain (Sandholtz and Gray, 2003). However, more general definitions substitute the word authority for public office, to include corruption that arises strictly between private parties (Rodriguez, Siegel, Hillman and Eden, 2006). The United Nations Development Program (2008) classified these definitions as limited, as they portray corruption as a one-way process driven by the greed of corrupt officials. In fact, almost all corrupt transactions have two players – the person who is receiving the bribe and the corporation who is offering it (UNDP, 2008). Firms often pay bribes in hopes of obtaining a business advantage, such as lower costs, greater efficiency, or access to relationships or markets (Nichols, 2012). However, the balance of power is not necessarily on the side of the corrupt person with entrusted power (UNDP, 2008). One might argue that when many economic actors effectively “buy” public officials, it becomes a necessity for all businesses to do so (Crane and Matten, 2010: 509). In some situations corruption may therefore be considered as a competitive requirement and not a lack in morality and honesty (Linder and Linder, 2008). These arguments however lead us directly into the debate about ethical absolutism and relativism and the perspective of Western democracies, in which this situation would be regarded as beyond ethically acceptable given the destructive effects of corruption on societies (Crane and Matten, 2010: 511).
et al. (2009), corruption has unique characteristics that distinguish it from other social problems, and while it is happening everywhere, it is rarely seen. Unlike children in sweatshops or toxics being dumped in rivers, the clandestine exchange of cash or property for a favorable decision from someone in power is rarely photographed or measured (Hills et al., 2009).
Corruption imposes significant costs on those societies with a high corruption rate namely: (1) reducing governmental service, (2) constraining economic growth, (3) decreasing trust in government and (4) reducing legitimacy of the market economy and democracy (Hills et al., 2009). In the article of León et al., (2013), it was found that corruption is one of the most important handicaps that a society can face, since it is capable of generating important economic and social impacts that limit the rise in social welfare in the medium and long term. This finding was also confirmed in the article of Hindess (2005), in which a range of damaging economic and social effects were linked to corruption, such as a raise in costs of goods and services, a reduction in productivity and a threat to the viability of democratic institutions. The totality of these costs and liabilities strongly suggests that the consequences for any given firm of paying a bribe would burden rather than benefit the firm. In addition, no nation can miss the clear and highly publicized conclusion that corruption is economically devastating (Nichols, 2012).
2.4. The Problem of Corruption
international businessperson faced with a decision regarding a bribe, and yet businesses continue to pay bribes (Nichols, 2012).
One of the biggest problems appears to be that corruption remains pervasive, despite a proliferation of anti-corruption policies and regulations (Dutton, 2008). Many countries seem to be unable to improve their institutions and curb corruption. One reason why rooting out corruption is so difficult, might be that when corruption is widespread, it just does not make sense for individuals not to engage in corruption, even if everybody would be better off if corruption were eliminated. The general observation is therefore that some countries seem to be stuck in a bad equilibrium characterized by both widespread corruption and slow economic growth. (Mauro, 2004). Corruption is not a localized crime but rather a crime without borders and efforts to control it need to come from multiple fronts and should be done over a longer period of time (Salifu, 2008; Tanzi, 1998).
2.5. Concluding Remarks
3. THEORY AND HYPOTHESES
3.1. Corruption as a Determinant of MNEs’ Location Choice
For long executives have been concerned about negative public relations from corruption, but they increasingly become aware of the additional risks and costs which they face including (1) operational costs [corruption adds additional expenses], (2) legal risks [the chance of prosecution], and (3) competitive risks [it can be a competitive disadvantage if they refuse to pay bribes]. (Hills et al., 2009). A possible solution to reduce the costs and risks of corruption would be by avoiding more corrupt countries (Doh et al., 2003). The location choice of MNEs seems therefore to be influenced by many complex factors including the host-country corruption rate.
When looking at corruption as a determinant of the location choice, two dominant views can be distinguished namely (1) “sand the wheels” and (2) “grease the wheels”. The first view “sand the wheels”, suggests that corruption deters investments, because corruption is supposed to be sign that the government is malfunctioning and so adds additional costs. (Bellos and Subasat, 2012; Helmy, 2013). Cuervo-Cazurra (2006) agrees with this view and states that “corruption creates uncertainty regarding the cost of operating in a country and acts as an irregular tax on businesses which increases costs and distorts incentives to invest”. The second view “grease the wheels” states that if corruption on the other hand substitutes for poor governance and overcomes the obstacles that an inefficient bureaucracy tends to create, then corruption can attract more MNEs (Bellos and Subasat, 2012; Helmy, 2013). In 2008, Cuervo-Cazurra tried to find an answer to the question why MNEs were still choosing some more corrupt transition economies as a suitable location. By distinguishing between pervasive corruption [corruption that is certain and widespread] and arbitrary corruption [corruption which is uncertain] an answer could be given. Pervasive corruption is a strong deterrent to the location choice, because it creates an additional known cost to MNEs. In contrast, arbitrary corruption does not act as a deterrent, because it merely creates higher uncertainty in the investment, uncertainty that is already prevalent in transition economies, since transition economies often have unclear rules to govern business operations. (Cuervo-Cazurra, 2008).
2012; Cuervo-Cazurra, 2008; Helmy, 2013), stakeholders do not accept this behavior anymore and demand that MNEs stop engaging in corrupt activities (Stansbury, 2009). In addition, publicity about corruption scandals have damaged MNEs severely during the past two decades (Hills et al., 2009) and have made them more concerned about being active in countries with higher corruption rates (Doh et al., 2003). As a result, more corrupt countries are expected to be a less attractive host-country (Al-Sadig, 2009; Brada, Drabek and Perez, 2012; Mudambi, Navarra and Delios, 2013). In this thesis it is therefore argued that countries with a higher corruption rate are relatively less attractive in comparison with host-countries with a lower corruption rate. In addition, it is argued that the increasing CSR pressure with corruption as an integral part has increased the negative effect of corruption on the location choice of MNEs over time.
Hypothesis 1: The prevalence of corruption has a negative effect on the attractiveness
of a host-country for foreign investors.
Hypothesis 2: The negative effect of corruption on the attractiveness of a host-country
for foreign investors has increased over time.
3.2. Corporate Social Responsibility Pressures from the Home-Country
Prior to the pronounced changes that accompanied the globalization of business in the 1980s and 1990s, research on corruption and its relationship to firms’ activities were almost non-existent. As foreign firms entered and new firms were born within developing and transition economies, managers and scholars grew more aware of the magnitude of corruption and the need to understand and address it. (Rodriguez et al., 2006). By becoming into contact with overseas supplier and competitors, corporate managers were often confronted with very different ways of thinking about corruption and evaluating business ethics (Bierstaker, 2009; Crane and Matten, 2010, 412). Several authors tried to link these different national corruption levels with culture and economic development (Husted, 1999; Watson, 2003). Watson (2003) for example explained that what constitutes corruption in one culture, might be considered legitimate or at least a quasi-legitimate business or social practice in another culture. According to Rodriguez et al. (2006), corporate leaders are mindful of the fact that business norms and standards, regulatory frameworks and political systems, corruption, and stakeholder demands for CSR can vary substantially across nations, regions and lines of business.
corruption on the attractiveness of host-countries. MNEs which face stronger pressures not to engage in corruption are therefore expected to locate relatively less often in countries which are more corrupt.
Hypothesis 3: The negative effect of corruption on the attractiveness of host-countries
is stronger for MNEs that face stronger home-country pressures not to engage in corruption.
4. DATA AND METHOD
In this chapter the research data and research method are described, in order to enable others to redo or complement this thesis. First of all, it is important to note that two distinct datasets have been used. The first dataset which consists of unilateral data has been used to find significant support for hypotheses 1 and 2. After adding all independent variables and removing all incomplete data entries of dataset 1, a total sample size of 165 countries remained over the period 1996 to 2012. Appendix A presents three frequency tables of dataset 1, in which a total of 2,805 data rows have been displayed. The second dataset which consists of bilateral data has been used to find significant support for hypothesis 3. After adding all independent variables and removing all incomplete data entries of dataset 2, a total of 193 countries and 178 partner countries remained over the period 2009 to 2011. Appendix B presents four frequency tables of this second dataset, in which a total of 15,609 data rows are displayed.
4.1. Data
As explained in Chapter 3, the inflow and outflow of FDI have been used to objectively measure the location choice of MNEs. The FDI flows have been used as the dependent variables of this research and are displayed in million US Dollars. In addition to the dependent variables, thirteen independent variables and four interaction variables have been used to find significant support for the formed theory and hypothesis. The dependent variable and the most important independent variable [the host-country corruption rate], have been discussed in a separate subparagraph below. The remaining independent variables have been clarified together in subparagraph 4.1.3. Table 1 and Table 2, which are displayed on page 22 and 23, present a description of the variables of both datasets. In addition, a statistical summary has been given in which the mean and standard deviation have been displayed.
4.1.1. Dependent variable: FDI flows
obtained from UNCTAD. UNCTAD which is governed by 194 Member States, is the United Nations body responsible for dealing with development issues, particularly international trade. (UNCTAD, 2014; UNCTAD, 2012). The dataset which has been obtained from UNCTAD concerning the inflow of FDI is focused on 237 countries over the period 1970 to 2012.
Dataset 2 consists of bilateral data concerning the outflow of FDI. The advantage of this dataset is that it can examine the influence which home-country pressures have on the relation between corruption and the attractiveness of host-countries. The bilateral FDI outflow data has been used as the dependent variable of the second dataset and has been obtained from the International Monetary Fund (IMF) (IMF, 2011). The IMF is an organization of 188 countries, working together to foster global monetary cooperation, secure financial stability, facilitate international trade, promote high employment and sustainable growth, and reduce poverty around the world (IMF, 2014). All available 246 countries over the maximum time span of 2009 to 2011 have been selected, in order to increase the sample size.
4.1.2. Key independent variable: the host-country corruption rate
4.1.3. Other variables
In this subparagraph all other independent variables are discussed, which are expected to have an effect on the location choice of MNEs. The effect of these variables has been taken into account, as they might distort the effect which corruption has on the location choice of MNEs. This distortion could result in wrong interpretations and conclusions. In addition to the independent variables described below, an error term should be taken into account. This error term stands for all missing variables which affect the location choice of MNEs, but which are not taken into account during this research.
Period Two Dummy Variable. A dummy variable has been used to indicate whether a data
row is from the second period. The first period includes the years 1996 to 2004 [eight years], whereas the second period includes the years 2005 to 2012 [seven years]. This variable has only been created for dataset 1 and is only relevant for hypothesis 2.
Home-Country Convention Membership Dummy Variable. The home-country pressures
which MNEs face have been operationalized via a membership dummy variable of the OECD Convention on Combating Bribery of Foreign Officials in International Business Transactions. This dummy variable indicates whether a MNE’s home-country is a member of the convention. MNEs from home-countries which are a member are expected to face relatively more pressure not to engage in corruption, which will decrease the attractiveness of more corrupt host-countries for them. A country ratification status of the OECD Convention has been obtained from the OECD (OECD, 2014). This dummy variable has only been created for dataset 2 and is only relevant for hypothesis 3.
Year Variable. This variable has been created for dataset 1 and consists of all years from 1996
to 2012. The variable has been used for the first robustness check of the results.
Less Corrupt Home-Country Dummy Variable. In dataset 2 a dummy variable has been used
GDP in Million US Dollars. According to Bevan and Estrin (2004), the gross domestic
product (GDP) should be seen as an important determinant of the location choice of MNEs. The GDP is the total sum of gross value added by all resident producers in a country plus any product taxes and minus any subsidies not included in the value of the product. The data has been collected from the Worldbank and is displayed in current million US Dollars (Worldbank, 2012).
GDP Growth Rate. The GDP growth rate is an indicator of the country’s growth rate in gross
domestic product. The data has been collected from the Worldbank and is displayed in annual growth percentages. By using this variable a relative comparison can be made between the economic development of countries. The GDP growth rate is expected to be positively related to the location choice of MNEs, because economic development is in general positively related to investments (Bevan and Estrin, 2004; Worldbank, 2012).
Distance. According to Bevan and Estrin (2004) one of the most important determinants of a
MNEs’ location choice are gravity factors. However, in order to use distance as an independent variable, a bilateral dataset should be used. This independent variable has therefore only been used for dataset 2. The distance between the home-countries and host-countries has been collected from CEPII, which is a French research centre in international economics. CEPII produces studies, researches, databases and analyses on the world economy and its evolution. (CEPII, 2012; CEPII, 2014).
Population Proximity. Instead of the distance variable, a new variable has been developed for
dataset 1, in order to control for gravity factors. This new variable consists of a combination of population data and distance data. The population data of all host-countries has been obtained from the Worldbank, whereas the distance data has been obtained from CEPII. The variable was created by dividing a host-country’s population by the total distance between this country and all other host-countries of the sample size. The variable therefore displays the distance of a country’s population from all host-countries of the sample size. (CEPII, 2012; Worldbank, 2012).
Population Density. The population density is a calculation of a country’s population divided
Worldbank, whereas the data concerning the country area sizes has been obtained from CEPII. This variable has been used in both datasets, in order to control for local gravity factors. (CEPII, 2012; Worldbank, 2012).
Landlocked Dummy Variable. This dummy variable indicates whether a host-country is
landlocked. In this thesis it is expected that landlocked countries are less attractive host-countries, because landlocked countries are more difficult to reach, e.g. no harbours. The data for this variable has been obtained from CEPII (CEPII, 2012).
Colonial History. In this thesis it is expected that investing in former colonies is more
attractive for MNEs from previous ruling home-countries. A dummy variable indicating a colonial history could only be used for dataset 2, because this dataset contains bilateral data. A new variable has been created for dataset 1, in order to measure the effect of a colonial history in this dataset as well. The variable was calculated by multiplying the network of colonies owned by three, multiplying the network of a long colonial history by two and adding this all together with the network of a short colonial history. A colonial network consists of all countries which were ruled by the same country and so share the same colonial history. The data for both variables has been obtained from CEPII (CEPII, 2012).
Language Internationality. In this thesis it is expected that a higher internationality of the
TABLE 1
TABLE 2
4.2. Method
This paragraph clarifies the research method which has been used to find significant support for accepting or rejecting the hypotheses. The first dataset has been used for hypotheses 1 and 2, whereas the second dataset has been used for hypothesis 3.
The first hypothesis can be seen as the starting point of the developed theory, as it researches the negative effect which the prevalence of corruption is expected to have on the attractiveness of host-countries. The attractiveness of host-countries has been operationalized by looking at the location choice of MNEs and has been measured via the FDI flows. The FDI inflow of more and less corrupt host-countries has therefore been compared. Several regression analyses have been executed, in order to find out whether possible differences in the FDI inflow of more and less corrupt countries can be prescribed to the effect of corruption. The more corrupt country group consists of all positive corruption rate numbers, whereas the less corrupt country group consists of all negative corruption rate numbers.
According to the developed theory, the negative effect of corruption on the attractiveness of host-countries has increased over time. A two period distinction has therefore been made from 1996 to 2004 and from 2005 to 2012, in order to execute a straightforward comparison between the effects of corruption in both periods. In addition, an interaction term between the host-country corruption rate and a period two dummy variable has been used to research whether a possible found negative effect of corruption on the attractiveness of host-countries has indeed increased over time. Formula 1 shows the most extensive model of dataset 1, in which also this interaction term can be seen. Chapter 5 will discuss the different models which have been used to find significant support.
(1) FDI Inflowit = 0 + 1(Host-Country Corruption Rateit) + 2(Period Two Dummy
Variableit)+3(Host-Country Corruption Rate X Period Two Dummy Variableit) + 4(Other Variablesit) +
tcorruption rate and a home-country convention membership dummy variable has been used in order to research the influence of home-country pressure on the possible found negative effect of corruption on the attractiveness of host-countries. Formula 2 shows the most extensive model of dataset 2, in which also this interaction term can be seen. Chapter 5 will discuss the different models which have been used to find significant support.
(2) FDI Outflowit = 0 + 1(Host-Country Corruption Rateit) + 2(Home-Country Convention Membership Dummy Variableit)+3(Host-Country Corruption Rate X Home-Country Convention Membership Dummy Variableit) + 4(Other Variablesit) +
t5. RESULTS
The research results which are discussed below indicate whether or not there is significant support for accepting the formed theory and hypotheses described in Chapter 3. The baseline results of each hypothesis have been described in a separated paragraph. In the fourth and final paragraph the robustness of the results will be discussed.
5.1. The Effect of Corruption on the Location Choice of MNEs
The first dataset has been used in order to find significant support for the first hypothesis, which states that the prevalence of corruption has a negative effect on the attractiveness of a host-country. The descriptive results below compare the attractiveness of more and less corrupt host-countries. The more corrupt country group was formed by all host-countries which have a corruption rate above zero, whereas the less corrupt country group was formed by all host-countries which have a corruption rate below zero. Two regression analyses have been carried out, in order to research whether the differences found in the attractiveness of both groups has been caused by differences in the prevalence of corruption.
5.1.1. Descriptive results
5.1.2. Regression results
The results of the regression analyses of dataset 1 are presented in Table 3. In this paragraph the research outcomes of the first two models will be discussed, as these two models have been executed to find significant support for the first hypothesis. Model 1, which analysis the relation between the host-country corruption rate and the inflow of FDI, gives significant support to conclude that the FDI inflow is negatively influenced by the host-country corruption rate [P < 0.001]. Model 2, which controls for the other independent variables, gives significant support to conclude that the FDI inflow is negatively influenced by the host-country corruption rate as well [P < 0.001]. It can therefore be concluded that there is significant support for accepting hypothesis 1.
TABLE 3
5.2. The Effect of Corruption over Time
The first dataset has been used in order to find significant support for the second hypothesis as well. The second hypothesis states that the negative effect of corruption on the attractiveness of a host-country has increased over time. As before, the descriptive results will be clarified first, after which the regression results will be discussed. The dataset has been split into two periods from 1996 to 2004 and from 2005 to 2012, because this gives the most straightforward evidence on whether the effect of corruption on the attractiveness of a host-country has changed over time.
5.2.1. Descriptive results
By splitting the data in two periods it becomes clear that the FDI inflow in general has increased over time with 87%. Even though the difference in the total FDI inflow between more and less corrupt countries has decreased over time from 62% to 42%, it continued to exist. It appears therefore that the less corrupt countries have remained more attractive over time. The results display that the mean difference between both groups increased [P < 0.001], whereas the median difference decreased over time [P < 0.001]. This difference could be caused by a large increase of FDI towards a small group of less corrupt countries. Both results have been found via a comparison of the Independent Sample T-Tests and the Mann-Whitney U-Tests which are presented in Appendix E. Several regression analyses have been executed, in order to research whether the negative effect of corruption on the FDI inflow has increased over time.
5.2.2. Regression results
other independent variables, supports this conclusion [P < 0.01]. It can therefore be said that there is significant support for accepting hypothesis 2. According to model 5 the results can be interpreted with a P < 0.05 as follows:
FDI inflow in millions US Dollars = -1,082.08 + (-2,056.21 x host-country corruption rate) + (1,649.89 x period two dummy variable) + (-1,388.43 x host-country corruption rate x period two dummy variable) + (0.01 x GDP in million US Dollars) + (103.47 x GDP growth rate) + (0.56 x population density) + (14.88 x colonial history).
5.3. The Influence of Home-Country Pressures
The second dataset has been used in order to find significant support for hypothesis 3, which states that the effect of corruption on the location choice of MNEs is stronger in home countries where MNEs face stronger pressure not to engage in corruption. In dataset 2 which is a bilateral dataset, all home-countries have been marked with a dummy variable indicating whether or not a home-country is a member of the OECD Convention on Combating Bribery. Membership of the OECD Convention is in this thesis expected to be associated with stronger home-country pressures not to engage in corruption.
be concluded that there is significant support for accepting hypothesis 3. According to model 9 the results can be interpreted with a P < 0.05 as follows:
FDI inflow in millions US Dollars = (6,191.77 x home-country convention membership dummy variable) + (-5,760.52 x host-country corruption rate x home-country convention membership dummy variable) + (0.42 x Distance) + (1,658.24 x landlocked dummy variable) + (7,127.15 x colonial history) x (5,092.91 x language internationality).
TABLE 4
5.4. Robustness Check
In this paragraph the robustness check of the research results will be discussed. Two different interaction terms have been used, in order to check the robustness of accepting hypotheses 2 and 3. The robustness of accepting hypothesis 1 has already been tested by using two different FDI flow datasets, which have both given significant support for accepting hypothesis 1.
A first potential threat to the initial findings is the period two dummy variable, which has been used in order to find significant support for hypothesis 2. The period two dummy variable has been used to find the most straightforward answer to hypothesis 2, but resulted in the loss of the continuous year variable. This loss in information might have biased the results. The robustness of accepting hypothesis 2 has therefore been tested again by using years as a continuous variable instead of the period two dummy variable. The year range of 1996 to 2012 has been mean centered and used as an interaction term with the host-country corruption rate.
TABLE 5
Table 5 displays model 10, 11 and 12 which check the robustness of accepting hypothesis 2. No problems have been found with the multicollinearity of the independent variables. Model 10, which researches the effect of the year variable, displays that time indeed, has had a positive effect on the FDI inflow [P < 0.001]. The rise in the total FDI inflow over time appears therefore to be robust. Model 11, which analysis the interaction term between the host-country corruption rate and the year variable, shows that the negative effect of corruption on the FDI inflow has increased over time as well [P < 0.01]. Model 12, which also controls for the other independent variables, support this conclusion even though it shows a much weaker increase of the negative effect of corruption over time [P < 0.15]. The initial results of accepting hypothesis 2 appear therefore to be robust.
TABLE 6
Robustness Check of Hypothesis 3
6. DISCUSSION
The discussion of the research results has been divided in three different paragraphs in which the hypotheses have been separately discussed. The discussion clarifies how the results should be interpreted. In addition, several recommendations have been given.
6.1. The Effect of Corruption on the Location Choice of MNEs
The research outcomes of hypothesis 1 provide significant support to conclude that the prevalence of corruption indeed has a negative effect on the attractiveness of host-countries, as the hypothesis appears to be robust in both datasets. The results therefore support the developed theory and earlier academic results which have indicated that more corrupt countries are expected to be less attractive host-countries (Al-Sadig, 2009; Brada et al., 2012; Bellos and Subasat, 2012; Cuervo-Cazurra, 2008; Helmy, 2013; Mudambi et al., 2013).
6.2. The Effect of Corruption over Time
During the past decade, some academic authors have started to argue that corruption has become an integral part of CSR (Branco and Delgado, 2012; Crane and Matten, 2010). This development is supposed to have pressured MNEs not to engage in corruption anymore, and is expected to have made more corrupt countries less attractive host-countries over time. The research outcomes clearly support this theory.
The results show an increase in the total FDI sum and FDI mean over time. Even though the total inflow towards more corrupt countries has increased faster, the FDI mean difference between the more and less corrupt country group increased. The FDI median difference between both groups decreased. A reason for this development could be a large increase of the FDI inflow towards a small number of less corrupt countries. Outliers have namely much effect on the mean, but not on the median. The regression analyses models 4 and 5 display a strong significant support to conclude that the effect of corruption on the attractiveness of host-countries has been more negative in the period 2005 to 2012 and has so become more negative over time. The robustness check of model 11 and 12 display a similar although less strong support to conclude that the effect of corruption has become more negative over time as well. The GDP, the GDP growth rate, the population proximity, the population density and the colonial history show again a significant positive effect on the attractiveness of host-countries.
6.3. The Influence of Home-Country Pressures
Local business standards including the prevalence of corruption vary widely and are mainly influenced by isomorphic pressures which force MNEs towards similar local behavior (Spencer and Gomez, 2011). These pressures increase the difficulty of creating a corruption free world, even though corruption is already universally disapproved (Hess and Dunfee, 2000). The OECD Convention on Combating Bribery of Foreign Officials in International Business Transactions appears to have created a new international isomorphic pressure against corruption, by using the influence of home-country pressures including civil and criminal penalties. The results indicate that the OECD Convention has strengthened home-country pressures not to engage in corruption and has so made more corrupt countries less attractive for MNEs from membership countries.
The result shows that the total FDI outflow of members is much higher in comparison with the non-members. The FDI mean and median difference between more and less corrupt countries is also much larger for member countries. The regression analyses models 8 and 9 shows that this difference is indeed partly caused by a stronger negative effect of corruption for MNEs from OECD Convention membership countries. In addition, the robustness checks of model 14 and 15 display that the OECD Convention indeed has had a stronger influence on the effect of corruption in comparison with general isomorphic pressures.
7. CONCLUSION
7.1. Final Remarks
The research outcomes have shown that the prevalence of corruption has a negative effect on the attractiveness of host-countries. In addition, this negative effect has increased over time and is stronger for MNEs from OECD Convention membership countries. This increasing negative effect of corruption on the attractiveness of host-countries seems to have been caused by the increasing CSR pressure with corruption as an integral part. MNEs are experiencing more pressure not to engage in corruption, which has increased the costs and risks of corruption and has forced MNEs to stop business as usual with a social twist. More corrupt regions are so more often avoided in order to circumvent reputation damage caused by corrupt activities. The OECD Convention has strengthened this effect by creating criminal and civil penalties on corrupt activities and seems to be a step towards a corruption free world. An increasing number of anti-corruption initiatives with global reach to fight corruption in cross-border business deals such as the OECD Convention are therefore recommended. Most importantly more corrupt countries and MNEs which continue to engage in corrupt activities should realize that their license to operate internationally is at stake.
7.2. Reflection
REFERENCES
Al-Sadig, A. 2009. The effects of corruption on FDI inflows. CATA Journal, 29 (2): 267-294.
Amaeshi, K.M. and Adi, B. 2007. Reconstructing the corporate social responsibility construct in Utlish.
Business Ethics: A European Review, 16 (1): 3-18.
Bellos, S. and Subasat, T. 2012. Corruption and foreign direct investment: A panel gravity model approach.
Bulletin of Economic Research, 64 (4): 565-574.
Bevan, A.A. and Estrin, S. 2004. The determinants of foreign direct investment into Europe a transition economies. Journal of Comparative Economics, 32 (4): 775-787.
Bierstaker, J.L. 2009. Differences in attitudes about fraud and corruption across cultures: Theory, examples and recommendations. Cross Cultural Management, 16 (3): 241- 250.
Brada, J.C., Drabek, Z. and Perez, M.F. 2012. The effect of home-country and host-country corruption on foreign direct investment. Review of Development Economics, 16 (4): 640-663.
Branco, M.C. and Delgado, C. 2012. Business, Social responsibility and corruption. Journal of Public Affairs, 12 (4): 357-365.
Brueckner, M. 2010. The business with the environment. Melbourne: Cengage Learning Australia Pty Limited.
Carroll, A.B. 1979. A three-dimensional conceptual model of corporate performance. Academy of Management
Review, 4 (4): 497-505.
Carroll, A.B. 1991. The pyramid of corporate social responsibility: Toward the moral management of organizational stakeholders. Business Horizons, 34 (4): 39-48.
CEPII. 2012. CEPII Statistics, Accessed February – June 2014; http://www.cepii.fr/CEPII/en/bdd_modele/
bdd.asp
CEPII. 2014. About CEPII, Accessed February – June 2014; http://www.cepii.fr/CEPII/en/cepii/cepii.asp
Crane, A. and Matten, D. 2010. Business ethics: Managing corporate citizenship and sustainability in the age
of globalization. New York: Oxford University Press.
Cuervo-Cazurra, A. 2008. Better the devil don’t know: Types of corruption and FDI in transition economies.
Journal of International Management, 14 (1): 12-27.
Dimaggio, P.J. and Powell, W.W. 1983. The iron cage revisited: institutional isomorphism and collective rationality in organizational fields. American Sociological Review, 48 (2): 147-160.
Doh, J.P., Rodriguez, P., Uhlenbruck, K., Collins, J. and Eden, L. 2003. Coping with corruption in foreign markets. Academy of Management Executive, 17 (3): 114-127.
Drucker, P.F. 1973. Management: tasks, responsibilities and practices. New York: Harper & Row Publishers.
D’Souza, A. 2012. The OECD anti-bribery convention: Changing the currents of trade. Journal of Development
Economics, 97 (1): 73-87.
Dutton, G. 2008. Do strong ethics hurt U.S. global competitiveness? World Trade, 21 (3): 36-41.
Friedman, M. 1962. Capitalism and freedom. Chicago: University of Chicago Press.
Hart, S. 2010. Self-regulation, corporate social responsibility and the business case: Do they work in achieving workplace equality and safety? Journal of Business Ethics, 92 (4): 585-600.
Helmy, H.E. 2013. The impact of corruption of FDI: Is MENA an exception? International Review of Applied
Economics, 27 (4): 491-514.
Hess, D. and Dunfee, T.W. 2000. Fighting corruption: A principle approach: The c principle (combating corruption). International Law Journal, 33 (3): 593-626.
Hills, G., Fiske, L. and Mahmud, A. 2009. Anti-corruption as strategic CSR: A call to action for corporations. Boston: FSG Social Impact Advisors.
Hindess, B. 2005. Investigating international anti-corruption. Third World Quarterly, 26 (8): 1389-1398.
Husted, B.W. 1999. Wealth, culture and corruption. Journal of International Business Studies, 30 (2): 339-359.
IMF. 2011. Trade and Investment Query Builder. Accessed February – June 2014; http://elibrary-data.imf.org/QueryBuilder.aspx?s=321&key=1445290&f=1&ts=1&ys=2002&ye=2011&ms=1&me=12 &ds=1&de=31&did=323&id=197&qb=24f4807a6c3f9caea0de18839c7c48dc&dl=1
Islam, M.A. and Deegan, C. 2010. Media pressure and corporate disclosure of social responsibility performance information: A study of two global clothing and sport retail companies. Accounting & Business
Research, 40 (2): 131-148.
Kostova, T. and Zaheer, S. 1999. Organizational legitimacy under conditions of complexity: the case of the multinational enterprise. Academy of Management Review, 24 (1): 64-81.
Kristol, I. 1978. Two cheers for capitalism. New York: Basic Books.
Lee, M. 2011. Configuration of external influences: The combined effects of institutions and stakeholders on corporate social responsibility strategies. Journal of Business Ethics, 102 (2): 281-298.
León, C.J., Araña, J.E. and de León, J. 2013. Valuing the social cost of corruption using subjective well being data and the technique of vignettes. Applied Economics, 45 (27): 3863-3870.
Linder, C. and Linder, R. 2008. Does business corruption have to be considered in all cases as moral misconduct? Two statements. Journal of Business Systems, Governance and Ethics, 3 (2): 1-15.
Mauro, P. 2004. The persistence of corruption and slow economic growth. IMF Staff Papers, 51 (1): 1-18.
Mudambi, R., Navarra, P. and Delios, A. 2013. Government regulation, corruption and FDI. Asia Pacific Journal of Management, 30 (2): 487-511.
Murphy, P.E. and Schlegelmilch, B.B. 2013. Corporate social responsibility and corporate social irresponsibility: Introduction to a special topic section. Journal of Business Research, 66 (10): 1807-1813.
Nichols, P.M. 2012. The business case for complying with bribery laws. American Business Law
Journal, 49 (2): 325-368.
OECD. 2014. Ratification Status as of 8 April 2014. Accessed February – June 2014; http://www.oecd.org/
daf/anti-bribery/WGBRatificationStatus_May2014.pdf
Porter, M.E. and Kramer, M.R. 2006. Strategy & society: The link between competitive advantage and corporate social responsibility. Harvard Business Review, 84 (12): 78-93.
Rodriguez, P., Siegel, D.S., Hillman, A. and Eden, L. 2006. Three lenses on the multinational enterprise: politics, corruption and corporate social responsibility. Journal of International Business
Salify, A. 2008. Can corruption and economic crime be controlled in developing economies – and if so, is the cost worth it? Journal of Money Laundering Control, 11 (3): 273-283.
Sandholtz, W., and Gray, M.M. 2003. International integration and national corruption. International
Organization, 57 (4): 761-800.
Spencer, J. and Gomez, C. 2011. MNEs and corruption: The impact of national institutions and subsidiary strategy. Strategic Management Journal, 32 (3): 280-300.
Stansbury, C. 2009. The global infrastructure anticorruption centre. Leadership & Management in
Engineering, 9 (3): 119-122.
Surroca, J., Tribó, J.A. and Zahra, S.A. 2013. Stakeholder pressure on MNEs and the transfer of socially irresponsible practices to subsidiaries. Academy of Management Journal, 56 (2): 549-572.
Tanzi, V. 1998. Corruption around the world. IMF Staff Papers, 45 (4): 559-595.
The Economist. 2008. A special report on corporate social responsibility. The Economist, 383 (8563): 1-24.
UNCTAD. 2012. UNCTAD Statistics. Accessed February – June 2014; http://unctad.org/en/Pages/
Statistics.aspx
UNCTAD. 2014. About UNCTAD. Accessed February – June 2014; http://unctad.org/en/Pages/AboutUs.aspx
UNDP. 2008. Tackling corruption, transforming lives: Accelerating human development in Asia and the
Pacific. New Delhi: Macmillan India
Walton, S.B. 2010. Do the right thing: Measuring the effectiveness of corporate social responsibility. Public
Relations Tactics, 17 (7): 10-11.
Watson, D.M. 2003. Cultural dynamics of corporate fraud. Cross Cultural Management, 10 (1): 40-54.
Weyzig, F. 2009. Political and economic arguments for corporate social responsibility: Analysis and a proposition regarding the CSR agenda. Journal of Business Ethics, 86 (4): 417-428.
Windsor, D. 2001. The future of corporate social responsibility. International Journal of Organizational
Analysis, 9 (3): 225-256.
Worldbank. 2012. Data Indicators. Accessed February – June 2014; http://data.worldbank.org/indicator#topic-9
Worldbank. 2014. About the Worldbank. Accessed February – June 2014; http://www.worldbank.org/en/about/
what-we-do
BIOGRAPHICAL SKETCH
APPENDIX C
APPENDIX D
APPENDIX E
