“Codes of good governance and FDI ”.

(1)

“Codes of Good Governance and FDI ”

Thesis MSc. in Business Studies - International Management University of Amsterdam

Iris Bakker – 10454063 First supervisor: dr. Haxhi, I. Second supervisor: Erik Dirksen Date: January 29th 2014

(2)

4.3. Regression ……….. .. 30 4.3.1. Developing countries ………. 32 4.3.2. Developed countries ……….. 33 4.4. Logistic regression ……… 34 5. Discussion ……….... 37 5.1. Hypothesis 1……….. 37 5.2. Hypothesis 2……….. 38 5.3. Hypothesis 3……….. 38 5.4. Hypothesis 4……….. 39

5.5. Hypotheses 5a. and 5b. ………. 40

5.6. Limitations and further research……… 41

6. Conclusion ………... 42

(3)

Abstract

Purpose: The purpose of this study is to examine the relationship between codes of good

governance and FDI in general and in emerging markets in particular. Moreover, it tries to find an answer to whether a symbolic nature on codes of good governance in emerging markets exists. Design/methodology/approach: For a sample of 84 countries who issued codes of good governance a distinction is made between developed and developing countries. For the study several variables (number of codes and year of first issuance) have been compared to find relations between FDI inflow and codes of good governance and other variables (enforcement, early

starter/late adopter and type of issuer) are used to search for a symbolic nature on codes of good governance in emerging markets.

Findings: The analysis shows that a positive relationship between the adoption of the first code of good governance and the amount of FDI inflow exists. When a country adopts a code of good governance the amount of FDI inflow increases. Moreover, a positive relationship between the number of codes of good governance and FDI inflow exists. The more codes of good governance a country issues, the more FDI is attracted. There are some indications for a symbolic nature on codes of good governance in emerging markets however, no hard evidence is found. This should be examined in other studies.

Contribution: This study contributes to previous literature since it is one of the earliest studies to examine the relationship between codes of good governance and FDI in general, and emerging markets in particular. Moreover, no research has been done on the possible symbolic nature on codes of good governance in emerging markets. From a practical perspective this research can be an eye-opener for managers who are planning to invest in developing countries. While there are some indications for a symbolic perspective on codes of good governance in emerging markets,

companies who are planning to invest abroad should take into account that codes of good governance do not automatically indicate a strong institutional environment and lower risk.

(4)

1. Introduction

Increasing economic integration makes international investment more interesting than ever before (Lynn et.al., 2011). Emerging markets provide a lot of opportunity for investment. The growth potential in these economies is higher than in the more developed markets. Emerging markets have been outperforming mature markets over the last seven years and this trend will continue. The United Nations Conference on Trade & Development (UNCTAD) in their 2013 World Investment Report states that developing countries take the lead in foreign direct investment (FDI). In 2012 developing economies received a higher inflow of FDI than developed countries. Also, Aizenmann (2003) sees an increase in FDI in emerging economies. Thus, based on the increase of FDI it seems like companies are more interested in investing in emerging markets.

However investing in developing economies is not without risk. Ghemawat & Khanna (1998) argue that doing business in emerging economies is challenging due to information asymmetry and uncertainties. Political and economic risks are also significant risks. Codes of good governance can be helpful by creating a more stable environment and accordingly reduce risks for multinationals associated with investing in developing countries. Codes of good governance are known as a set of ‘best practice’ recommendations regarding the behavior and structure of the board of directors of a firm. They are designed to create more transparency in the corporate governance system (Aguilera and Cuervo-Cazurra, 2004). As mentioned before, codes of good governance can contribute to a more stable institutional environment and therefore reduce risks for multinationals whom are willing to invest in a particular country. Moreover, companies benefit from implementing codes of good governance. For example, Machuga and Teitel (2007) show that the quality of earnings improves after implementation of governance codes. Also Fernandez-Rodriguez et. al. (2004) and Alves and Mendes (2004) uncover a positive relationship between the adoption of codes and firms’ equity returns. Even multilateral organizations see positive relationships between the issuance of codes and growth and development. For example, The World Bank and the Organization for

(5)

and corporate governance in particular helps countries grow and develop economically. Good corporate governance helps developing and transition countries to upgrade to international business standards (Aguilera & Cuervo-Cazurra, 2009). Issuing codes has advantages for both countries as well as firms. According to Cuomo & Zattoni (2008) and Aguilera & Cuervo-Cazurra (2004) the trigger for countries to issue codes of good governance are legitimacy and efficiency reasons. In addition, Aguilera & Cuervo-Cazurra (2009) describe two reasons for countries as well as firms to make their corporate governance practices more effective; (1) to prevent future corporate scandals (2) to attract investors. Evidence is found that investors use corporate governance practices to gauge the quality of firms in uncertain domestic markets (Sanders & Boivie, 2004). Aguilera & Cuervo-Cazurra (2009) add to this proposition that codes of good governance are a signal to foreign investors to increase FDI. This idea is supported by Filatotchev et. al. (2007) and Mengistu & Adhikary, (2011): a good corporate governance system is an important factor in entry-mode decisions of investors. Aguilera & Cuervo-Cazurra (2004) argue that codes develop faster in countries with more FDI exposure. Accordingly a link might exist between codes of good governance and FDI.

Codes of good governance are voluntary in nature and follow the comply-or-explain principle. Because of this comply or explain principle a debate exists on whether governance codes are an effective tool, or whether the following of governance rules should be more mandatory. This debate is especially heated regarding countries that have weak institutions and underdeveloped governance systems such as emerging markets (Aguilera & Cuervo-Cazurra, 2009).

Evidence on non-compliance of governance codes exists. For instance, MacNeil and Li (2006) find significant evidence of non-compliance and show that compliance is not properly monitored. Research suggests that compliance of codes in developing countries is scarce. For example, research on codes in Cyprus by Krambia-Kapardis and Psaros (2006) shows a minimum level of compliance of governance codes. The institutional environment and the development of the stock market is an important factor of the way code compliance is monitored, even if it is just for informal

(6)

and legitimation purposes (Aguilera & Cuervo-Cazurra, 2009).

The main reasons for countries to issue codes without monitoring the compliance thereof could be the attraction of FDI, thus codes could be used symbolically instead of practically. Westphal and Zajac (1998) prove the existence of a symbolic perspective on corporate governance. In addition, Cuomo and Zattoni (2008) show that symbolic governance practices exist: “when practices become institutionalized, their adoption brings legitimation to the adopting organizations or social systems, even if sometimes these practices fulfill symbolic rather than task-related requirements.” Therefore, a symbolic nature of codes of good governance is not unrealistic. While several authors uncover a positive relationship between the adoption of codes of good governance and equity returns for companies, other authors argue a positive relationship between codes of good governance and FDI inflow. For example Aguilera and Cuervo-Cazurra (2004) state that the main reason for adopting codes of good governance is making the country more attractive for investors. Moreover

Filatotchev et.al. (2007) state that good governance affects FDI and entry-mode decisions of multinationals. The World Bank, La Porta et.al (1998) and Levine (1999) also see a relationship between codes of good governance and FDI. They state that countries with an effective corporate governance system become attractive locations for foreign companies to prosper and invest.

Therefore, a relationship between codes of good governance and FDI seems to exist. This study will shed more light on this relationship and the possible symbolic nature on codes of good governance in emerging markets. No specific research has been done before to investigate the relationship between codes of good governance and FDI, and the possible symbolic nature on codes of good governance in developing countries in particular. The research question will be:

Does the adoption of codes of good governance affect FDI attraction in general, and more specifically in emerging markets?

(7)

emerging markets in particular. Moreover, this research will investigate if a symbolic nature on codes of good governance in emerging markets exists.

There are several arguments why a relationship between codes of good governance and FDI may be expected. As mentioned before, Aguilera and Cuervo-Cazurra (2004) see a relationship between codes of good governance and the attraction of investors. Filatotchev et. al. (2007) do also see a relationship between codes of good governance and multinational FDI- and entry-mode decisions, while the World Bank, La Porta et.al. (1998) and Levine (1999) see a positive relationship between a good governance system and prosper locations to invest from foreign companies. Arguments for a symbolic nature on codes of good governance in emerging markets lie in the fact that codes of good governance are voluntary in nature. A debate is heated on whether codes of good governance should be more mandatory, especially in emerging markets (Aguilera and Cuervo-Cazurra, 2009). Evidence on non-compliance of governance codes in emerging markets already exists (Mac Neil and Li, 2006; Krambia-Kapardis and Psaros, 2006). This study therefore investigates the

effectiveness of codes of good governance in emerging markets.

This research is relevant for several reasons. First, from a practical perspective it can be an eye-opener for multinationals that are willing to invest in developing economies. If indeed a symbolic nature of codes of good governance exists, multinationals should take into account that codes of good governance do not automatically mean strong institutions and lower risk. From an academic perspective this research complements research done on codes of good governance. Since codes of good governance have increased in relevance and continue spreading throughout the world Aguilera & Cuervo-Cazurra (2009) point out: there is need for additional research on codes of good

governance. Research on codes is increasing since 1997, however there is little systematic analysis on codes of good governance. According to Aguilera & Cuervo-Cazurra more research is needed on the effectiveness of codes since there is some conflicting evidence (Aguilera & Cuervo-Cazurra, 2009). This research will fill a gap in the research on codes of good governance and the relation with FDI. Moreover, it will investigate the effectiveness of codes of good governance in developing

(8)

economies.

Earlier research on codes of good governance focused on the development of codes of good

governance and underlying reasons (Aguilera & Cuervo-Cazurra, 2004). Other researchers focused on type of issuers and the relationship between cultural differences and the number of codes issued (Haxhi & Van Ees, 2010; Haxhi et.al., 2013). No research has been done on the relationship between codes of good governance and FDI in general and emerging markets in particular. This research contributes to literature in several ways. First, no research has been done to

investigate the relationship of FDI inflow and codes of good governance. Second, no research has been done to investigate if a symbolic nature on codes of good governance in emerging markets exists. From a managerial perspective there are also contributions from this study, if a symbolic nature of the use of codes of good governance exists, this can be of major influence on managerial and strategic decisions in gaining FDI in emerging markets. If proof is found for a symbolic nature on codes of good governance managers should take into account that codes of good governance do not automatically mean strong institutions and lower risk and managers therefore may change their strategic decisions.

For the empirical research 84 countries that issued codes of good governance are used. Data is collected for the dependent variable: FDI inflow, and the independent variable: the number of codes (issuance). In this study there is controlled for GNI/Capita and Country Size (population).

Enforcement and early starters/late adopters are used as a moderator. Also, information on the type of issuer is collected and used for the empirical research. Three models are tested. First, to

investigate the relationship between codes of good governance and FDI, the group of countries is tested twice. Once before the first code of good governance was issued (time 1) and once after the code was issued (time 2). For the second model the independent variable (issuance) and dependent variable FDI inflow are tested to investigate the relationship between the number of codes of good governance and FDI. To examine a symbolic nature on codes of good governance in emerging markets enforcement and early starters/late adopters are used as a moderator. These moderating

(9)

variables are integrated in the second model. When enforcement or early starters/late adopters positively moderate the relationship of the number of codes and FDI a more symbolic nature on codes of good governance may be expected. For the last model the type of issuer is studied to examine a possible symbolic nature on codes of good governance in emerging markets. Two tests are conducted to test the likelihood for a particular group (normative or coercive) to be the type of issuer of the first code of good governance. When the type of issuer in emerging markets is more likely to be in the normative group, a more symbolic nature on codes of good governance may be expected.

Then next section reviews previous literature and the hypothesis will be stated. It starts with elaborating on emerging markets, second, it will elaborate on FDI in general, thereafter codes of good governance in general and the relationship between codes of good governance and FDI and issuance. Subsequently it will focus on the symbolic nature on codes of good governance. The third chapter will present the data and methods of this study, followed by the results, discussion and conclusion.

(10)

2. Literature review

2.1. Emerging economies

Emerging economies are known as low-income, rapid-growth countries using economic

liberalization as their primary engine of growth. Research of Crittenden and Crittenden (2012) and Hoskisson et. al. (2000) suggests that doing business in emerging economies is different than doing business in the home market. Rules, regulations and marketplace expectations are different in emerging economies. Transitional economies are evolving in nature, which makes their

environment uncertain and complex. Entry into these markets tends to be easy however; it can be more difficult because of a weaker legal structure and fewer market players than in developed countries (Olson, 1996; Lynn et al., 2011). Governance in emerging markets also differs from developed economies according to Crittenden and Crittenden (2012). They conclude that the critical driver of governance in emerging markets is economic turbulence. In less developed economies institutions may be weak because rules are absent, suboptimal or poorly enforced (Aron, 2000). Weak institutions as a result of political instability can also weaken economic growth (Gani, 2011). Accordingly, it is important to improve the quality of governance. Globerman and Shapiro (2002) noted the importance to improve governance to attract investment especially for developing economies.

2.2. Foreign Direct Investment

Foreign direct investment (FDI) is known as investment in, controlling and managing value-added activities in other countries. The United Nations defines FDI as an equity stake of 10% or more in a foreign-based enterprise. FDI flow is used to describe the amount of FDI moving in a given period and can move in two ways: FDI inflow, which refers to FDI moving into a country in a year, and FDI outflow, which means FDI moving out of a country in a year (Peng, 2012). Governments often promote FDI inflow. Countries benefit from FDI inflow for several reasons: FDI can provide in employment and capital inflow and technology spillovers for domestic companies (Aitken &

(11)

Harrison, 1999). From a company perspective, there are also benefits. FDI requires a significant equity ownership position. The benefits of direct ownership lie in the combination of equity ownership rights and management control rights (Peng, 2012).

Investing in foreign countries is not without risk. When firms conduct FDI they take part in

institutional processes in both home- and host countries (Rosenzweig & Sing, 1991;Xu & Shenkar, 2002). In this way, firms are exposed to isomorphic pressures form home- and host-country

institutional environments (Cui & Jiang, 2012). Governments play an important role in providing a stable environment for FDI. Those responsibilities lay in a stable political and economic

environment, the rule of law and infrastructure. Moreover, an educated and technically skilled work force, low wages an open economy and stable currency are also essential according to UNCTAD (1997). MNEs take all these factors in consideration when conducting FDI. Especially for

developing countries it is difficult to measure up to the same standards as developed countries. Reasons for MNEs to invest in developing countries are cost-reduction strategies (Sethi et.al, 2003). FDI inflow in emerging economies increased in the last three years. In the year 2012 the amount of FDI inflow in emerging economies transcended the amount of FDI inflow in developed economies. Asian countries remained the largest sources of FDI, whereas the BRICS countries (Brazil, the Russian Federation, India, China and South Africa) continued to be leader of FDI among emerging investor countries (UNCTAD.org).

2.3. Codes of good governance

A rise in the adoption of codes of good governance in industrialized countries is seen. Adoption of codes of good governance results in more firm transparency and accountability (Aguilera & Cuervo-Cazurra, 2004). Codes of good governance are known as a set of best practice

recommendations regarding the behavior and structure of the board of directors for companies. The goal of codes of good governance is improving the quality and accountability of companies. This improvement results in shareholder protection. Codes of good governance can be a major tool when

(12)

other mechanisms, such as takeover markets and the legal environment fail in protection of

shareholder rights (Aguilera & Cuervo-Cazurra, 2004). Therefore, codes of good governance can be a way to reduce institutional risks.

Codes of good governance follow the comply- or- explain principle, which means that compliance with the codes is voluntary however, when a firm is not able, or does not want to comply to the codes, the firm should explain why (MacNeil & Li, 2006). What this comply- or explain principle entails varies across countries (Haxhi et. al., 2013). Despite the comply-or-explain principle, research shows that publicly traded companies tend to respond to the main code recommendations (Conyon and Mallin, 1997; Gregory and Simmelkjaer, 2002).The first code of good governance was developed in the US in 1978, triggered by the conglomerate merger movement, managers’ hostile takeovers and the shareholder rights movement in the US (Aguilera & Cuervo-Cazurra, 2004; 2009).

After the issuance of the first code of good governance in 1978 several other countries followed by issuing codes. However, after 1992 the issuing of codes increased more rapidly. This was due to two factors: (1) the Cadbury committee by that time issued the Cadbury Report (Aguilera & Cuervo-Cazurra, 2009; Haxhi et. al., 2013), and (2) simultaneously with the Cadbury Report other international entities began to show the importance of codes of good governance for institutions and good corporate governance. They showed that codes of good governance could stimulate growth and development in countries (Aguilera & Cuervo-Cazurra, 2009), which eventually resulted in 89 countries that have issued codes at this moment in time.

There are two differences between the development of codes of good governance in countries. First, countries vary in the year that the first code was issued and secondly, the number of codes issued differs per country (Aguilera & Cuervo-Cazurra, 2009). This difference in the year of first issuing can be explained by the legal system of countries. Early starters like the US, Hong Kong, Ireland, the United Kingdom and Canada share in common a common-law or English-based legal system. This legal system, in contrast to a civil-law system, is more flexible and it is therefore easier to

(13)

implement new laws. The first civil-law country that issued a code was Sweden in 1994. During that year the first developing country also issued a code. It was not until 1997 that other developing countries followed (Aguilera & Cuervo-Cazurra, 2009). The difference in the number of codes issued by countries can be explained by two factors: (1) research of Haxhi & Van Ees (2010) shows that the difference in the number of codes issued per country depends on cultural differences. They state that individualistic countries have a stronger tendency to develop codes. While (2) Aguilera & Cuervo-Cazurra (2009) point out that a relationship seems to exist between the development of the capital market and the number of codes issued. Countries with larger and deeper capital markets have more codes of good governance(Aguilera & Cuervo-Cazurra, 2009). Research shows that the adoption of (parts of) the practices of the codes has advantages for companies. In a sample of 200 British firms Weir and Laing (2000) demonstrate that adopting some of the practices is directly related to higher firm performance. They show that market returns where higher when companies adopted the practices.

A distinction can be made between endogenous and exogenous forces for triggering the issuing of codes (Aguilera & Cuervo-Cazurra, 2004). Endogenous forces means that pressure for the

development of codes comes from a force within the country, while exogenous forces come from outside a country. Aguilera and Cuervo-Cazurra (2004) identify six types of code issuers: (1) stock exchange, (2) government, (3) directors’ association, (4) managers’ association, (5) professional association, and (6) investors’ association. Extending on the research of Aguilera and Cuervo-Cazurra, Haxhi and van Ees (2010) use these issuers to distinguish three groups: coercive (stock exchange and investors’ groups); normative (government and directors’ and professional

associations) and mimetic (managers’ associations). Research of Haxhi and van Ees (2010) shows a relationship between the type of issuer and cultural differences. In cultures with high power

distance the type of issuer is more likely to be in the normative group, whereas in countries with low power distance the type of issuer is more likely to be in the coercive group.

(14)

2.4. Codes of good governance and FDI

As stated before the goal of codes of good governance is improving the quality and accountability of companies (Aguilera & Cuervo-Cazurra 2004). Reasons for countries to adopt codes of good governance vary. According to Cuomo & Zattoni (2008) the most often used reasons for issuing codes are legitimacy and efficiency. While Aguilera & Cuervo-Cazurra (2009) point out that the main reason for issuing codes in a country is making the country more attractive for investors. The development and adoption of codes signals a country’s commitment to improve its corporate government system (Aguilera & Cuervo-Cazurra, 2004) and therefore gains trust from investors. Attracting investors results in more FDI inflow, and therefore a relationship between codes of good governance and FDI inflow may be seen. A relationship between codes of good governance and FDI inflow is also seen by Filatotchev et.al. (2007). They state that good governance affects FDI and entry-mode decisions of multinational enterprises. Moreover, The World Bank states that countries with effective corporate governance systems become not only attractive locations for domestic companies to prosper and invest, but also for foreign investors, and thus promote

economic growth (World Bank 2000, La Porta et.al, 1998 and Levine, 1999). Dunning (1977) sees good corporate governance systems as a location advantage for countries. Thus, codes of good governance results in better governance systems, while a good governance system is related to more attraction of investors (FDI). Therefore, a relationship between codes of good governance and FDI inflow seems to exist. Codes of good governance may be seen as an improvement of the governance system and in this way reduces institutional risks.

To summarize, the adoption of codes of good governance can have a positive influence on FDI inflow in a particular country. First, the adoption of codes of good governance provides a signal to investors. It shows that the country wants to improve its corporate governance system. Secondly, the World Bank states that countries with an effective corporate governance system are attractive locations for companies to invest. This idea is supported by Dunning (1977), he states that good corporate governance systems are location advantages for countries. Moreover, Filatotchev et. al.

(15)

(2007) see a connection between FDI and a good governance system. Taken into account that codes of good governance are related to improvement of governance practices, lower institutional risks and attractive investing locations, the amount of FDI should have increased after the

implementation of the first code of good governance.

Research on institutional factors that influence FDI flow has already been done. For example, Globerman and Shapiro (2002) research the effect of governance indicators on FDI flows. While Gani and Al-Abri (2013) have investigated institutional variables in relation to FDI. Since adoption of codes of good governance may be seen as institutional and governance practices it is interesting to investigate the relationship between the adoption of the first code of good governance and FDI inflow, therefore:

H1: Adoption of the first code of good governance positively affects FDI inflow.

The relationship is shown in the following model:

(16)

2.5. Issuance

The number of codes of good governance issued per country varies. This number ranges from 1 (for most countries) to 44 (US). Research has been done to investigate the relationship between the number of codes of good governance in a particular country and Hofstede’s cultural dimensions. Haxhi and van Ees (2010) show that countries with high power distance are more likely to issue more codes of good governance, while countries with a low perception of power distance issue lesser codes. In other research on the number of codes of good governance by Aguilera and Cuervo-Cazurra (2004 a relationship is found between the number of codes adopted and the presence of foreign institutional investors. They explain this by stating that investors see codes of good

governance as a form to protect their investments, and accordingly are willing to pay more for good governance (McKinsey and Company, 2000). As mentioned before, adapting codes of good

governance signals a country’s commitment to improve its corporate government system (Aguilera & Cuervo-Cazurra, 2004). The more codes a country issued, the more commitment a country shows to improve its corporate governance system. Aguilera & Cuervo-Cazurra (2004) state: “the number of national codes of good governance created in a country is taken as an indicator of the country’s commitment to improve its corporate governance system”, it therefore may be expected that companies are more willing to invest in countries with more codes of good governance as they see improvement in the corporate governance system.

There are therefore several arguments to relate the number of codes of good governance to FDI inflow. First, Aguilera and Cuervo-Cazurra relate the number of codes of good governance to the presence of foreign institutional investors. Moreover, they see a relationship between the number of codes of good governance and the commitment of a country to improve its governance practices. This good governance in turn is related to the attraction of investors. McKinsey and Company (2000) state that investors see codes of good governance as a form to protect their investments and accordingly are willing to pay more for good governance. The previous statements indicate a relationship between the number of codes of good governance and FDI inflow.

(17)

As mentioned before, several researchers have already investigated the relationship of governance practices to FDI flow (Globerman and Shapiro, 2002; Gani and Al-Abri, 2013). Since the number of codes of good governance is seen as improvement of government practices it is interesting to investigate the relationship of issuance on FDI flow. Therefore (see also figure 2.):

H2: The more codes of good governance a country issued, the more FDI is attracted.

Figure 2. Conceptual model 2.

2.6. Emerging markets and symbolic codes of good governance

The next section will elaborate on the possible symbolic nature of codes of good governance in emerging markets using enforcement, early starters/late adopters and type of issuer as indicators for symbolic practices. First symbolic governance in general will be explained.

2.6.1. Symbolic governance

To investigate the possible symbolic nature of codes of good governance it is important to look in more depth into symbolic practices. Evidence of symbolic business practices exists. Westphal and Zajac (1998) state that symbolic actions can provide significant positive stakeholder reactions. In

(18)

their article of 1994 they also point out that CEO’s incorporate long-term incentives plans and not actually use these to suggest a potential separation of substance and symbol in CEO compensation contracts. Westphal and Zajac (1998; 2001) also find prove for adaption of symbolic governance under pressure. In developing economies this can be seen as economic pressure for growth and development for instance. Moreover, Cuomo and Zattoni (2008) point out the existence of symbolic governance practices: “when practices become institutionalized, their adoption brings legitimation to the adopting organizations or social systems, even if sometimes these practices fulfill symbolic rather than task-related requirements.” Okhmatovsky and David (2012) define ceremonial corporate practices. They consider corporate codes ceremonial when they do not provide specific guidelines and targets to produce changes in organizational behaviors. They argue that some firms develop vague corporate codes with minimal details, whereas others develop corporate codes with more specific requirements. How this works for corporate governance codes has not yet been

investigated. Non-compliance of codes of good governance may be seen as adopting codes symbolically. Evidence is found of non-compliance on codes in developing economies (Krambia-Karpardis and Psaros, 2006). Non-compliance of codes of good governance can be related to enforcement. When enforcement is low, more risk of non-compliance on codes exists.

2.6.2. Enforcement

To determine whether codes are used symbolic rather than practical it is important to investigate the enforcement of countries. Enforcement is seen as law effectiveness and it is proven that there is a large difference in enforcement between developed economies and transition economies (Pistor, Raiser and Gelfer, 2000). When addressing the topic enforcement, a distinction can be made between private ordering, private enforcement of the law or public enforcement of the law

(Djankov, et. al., 2003). Wherein private ordering means transactions that take place in the absence of laws or other enforcement institutions, private enforcement of the law means private agents serve the state to enforce the final judgment and public enforcement may be seen as the enforcement and

(19)

prosecution of the law by the government. Private enforcement however, is the main mechanism in most markets, whereas public enforcement plays a bigger role in weak institutional environments (Berglöf and Claessens, 2006). Differences in enforcement technologies are dependent on the national and local authorities, political competition at various levels of government and the strength of civil society. These differences vary across countries but also across regions within a country (Slinko, Yakovlev and Zhuravskaya, 2002). The historical origin of the legal (common law or civil law) and institutional system (English, French, German or Scandinavian) also matters for

enforcement (Glaeser and Schleifer, 2002). Moreover, the social- and cultural factors are of major influence on enforcement (Djankov et.al, 2003). Societies who are socially and culturally

heterogeneous tend to have weaker enforcement. A distinction can be made between developed economies, in which political institutions are part of the general enforcement function and

developing economies where business and politics tend to merge. When the rich influence justice (as is seen in many developing economies) litigation does not work (Glaeser, Sheinkman and Schleifer, 2003). When rich influence justice this is seen as corruption and corruption lowers the effectiveness of enforcement (Acemoglu and Verdier, 2000). Enforcement is generally agreed to have a positive relationship to economic growth (Berglöf and Claessens, 2006). North (1991) argues that: “how effectively agreements are enforced is the single most important determinant of economic performance.” Enforcement can be of influence on entry mode-decisions for firms. Berglöf and Claessens (2006) state that a weak enforcement environment makes it more difficult for firms to make commitments to a particular country, and therefore could be of influence on FDI decisions for firms. Financial contracts ask for commitment from a company and a weak

enforcement environment results in more risks and therefore makes it more difficult to make those commitments to a country. The problem arising with weak enforcement is that mechanisms to commit may be missing or incomplete.A response from governments to weak enforcement can be the development of codes of good governance. Especially in developing countries such adjustments are made to decline the impact of a weak enforcement environment (Berglöf and Claessens, 2006).

(20)

To summarize, a difference in enforcement between developed and developing countries exists. Developing countries tend to have lower enforcement than developed countries (Glaeser,

Sheinkman and Schleifer, 2003). Enforcement may be related to codes of good governance and FDI in three ways: (1) Governments sometimes respond to weak enforcement by the development of codes of good governance (Berglöf and Claessens, 2006), (2) enforcement is generally related to economic growth (Berglöf and Claessens, 2006), and (3) enforcement can be of influence on entry mode decisions (FDI) of firms. Weak enforcement makes it more difficult for firms to make commitments to a particular country and therefore could be of influence on FDI decisions of firms. It thus may be stated that a relationship between codes of good governance, FDI and enforcement exists.

Gani and Al-Abri (2013) already investigated the relationship between business environment indicators and FDI flow. In this respect one of the determinants of the business environment is enforcement mechanisms. Their research suggests that the presence of enforcement mechanisms encourages more investment and attracts FDI. It is therefore interesting to investigate the

relationship of enforcement on codes of good governance and FDI. This leads us to the following hypothesis (see also figure 2.):

H3. Enforcement positively moderates the relationship between codes of good governance and FDI.

Enforcement can be of importance on the relationship between codes of good governance and FDI regarding a symbolic nature of codes of good governance. If codes of good governance are issued which are not enforced it may be presumed that those codes are more symbolic than practical.

2.6.3. Early starters/ late adopters

Another factor that can be of influence on the symbolic nature of codes of good governance is early starters versus late adopters. In this study early starters are seen as countries that issued their first

(21)

code of good governance before 2006. Early starters is a term known from marketing and

innovation theory. An early starter is a minority group that is first to try new ideas, processes, goods and services. Early starters rely on their intuition and vision. Earlier adopters have a greater ability to cope with uncertainty and risk and a more favorable attitude toward change. Early adopters face advantages in terms of being able to have early, unique access to new products or technology, however they also serve as a trial, which can have negative effects as well (Rogers, 1962). In the case of the adoption of codes of good governance early adopters are countries that are first in issuing codes, before other countries have issued codes. In this way the early adopters face the advantageous and risks of being an early adopter. The adoption of codes of good governance can be seen as a change in the institutional environment of a country. Institutions are defined as the “rules of the game” and include both formal (regulatory) and informal (normative and socio-cognitive) categories (North, 1990; Scott, 1995).Three mechanisms of institutional change are identified by DiMaggio and Powell (1983): coercive, normative and mimetic. Coercive mechanisms are seen as resulting from political influence and problems of legitimacy. The pressures may be seen as force, persuasion or as invitations to join collusion. Normative processes are associated with

professionalization, whereas mimetic processes may be seen as a reaction to uncertainty. Countries model themselves on other countries. This modeling can be intentionally or unintentionally.

Institutional theory suggests that the main reason for adopting practices is because they are taken for granted rather than the improvement of efficiency; adoption is socially expected (Meyer & Rowan, 1997; DiMaggio & Powell, 1983). This idea is supported by Aguilera and Cuervo-Cazurra (2004). Following the idea of institutional theory it may be stated that early starters most likely develop codes to improve their governance practices, while late adopters follow early starters because this is socially expected. Bush (1987) refers to this as cultural borrowing. There are second mover advantages for late adopter countries in the code issuing process. These advantages are for instance free-riding effects. Westphal and Zajac (1994) show in their research on long term

(22)

benefits for late adopters are derived from adoption alone. Therefore, it might be possible that late adopters in the code issuing process only adopt codes of good governance instead of actual

implement them. When this is the case in a country a symbolic nature of the codes of good

governance may be expected. While early starters most likely develop codes of good governance to improve their governance practices, it might be possible that late adopters follow early starters as they see the amount of FDI inflow increase when countries adopt codes. This process may then be seen as normative pressures to institutional change (Di Maggio and Powell, 1983).

To summarize, adopting codes of good governance is a form of institutional change. In terms of late adopters, the adoption of codes of good governance may be seen as normative pressures to

institutional change (Di Maggio and Powell, 1983). This can be related to FDI inflow, while late adopters follow early starters as they see the amount of FDI inflow increase after adoption of a code. The relationship between early starters and FDI flow has already been investigated by Chiang (2008), he investigated the effect of early starters in the IT business in the US, Japan, Taiwan and South Korea on FDI flow in general. It would be interesting to investigate the relationship between codes of good governance, early starters and FDI flow. Therefore (see also figure 2.):

H4: Early starters/ late adopters positively moderates the relationship between codes of good governance and FDI inflow.

If a positive moderating effect of early starters/late adopters exists, this can be of influence on a symbolic nature of codes of good governance. While early starters most likely develop codes to improve their governance practices, it is to be expected that late adopters follow early starters as they see the amount of FDI inflow increase when countries adopt codes. As stated before, the research of Westphal and Zajac (1994) shows symbolic adoption of long term incentive plans of CEO’s. Benefits derived from adoption alone, implementation became unnecessary. They state: “ the later the date of LTIP adoption, the lower the likelihood of its subsequent use.” This can also be

(23)

related to codes of good governance. Following the principle of DiMaggio and Powell (1983) the code issuing process can be mimetic, to reduce uncertainty.

2.6.4. Type of issuer

When investigating codes of good governance it is also important to know who the first issuers of the codes are. This can shed more light on whether codes are used subsequently and how strongly they are enforced. Research of Aguilera and Cuervo-Cazurra (2004) shows that there are 6 types of issuers: (1) stock exchange, (2) government, (3) directors’ association, (4) managers’ association, (5) professional associations, and (6) investors’ associations.

As seen before, DiMaggio and Powell (1983) present three mechanisms for institutional change; normative, coercive and mimetic. In the case of code issuers, Haxhi and van Ees (2010) relate codes issued by the stock exchange and investors to coercive isomorphism. This coercive isomorphism arises from the fact that these new practices most of the times become part of the requirements for publicly traded firms. Codes issued by directors’ associations, professional associations and the government are more influenced by normative isomorphism that is consistent with legitimate values and norms, while codes issued by manager associations and the government are more influenced by normative may appear when for instance other successful peer players have adopted codes

(Aguilera & Cuervo-Cazurra, 2004). Therefore, Haxhi and van Ees (2010) use three groups to explain the type of issuer of codes of good governance. These groups are: coercive (stock exchange and investors), normative (directors’ associations, professional associations and the government) and mimetic (managers’ associations). Haxhi and van Ees (2010) in their research relate the type of issuer to Hofstedes’ dimensions. They state that for countries with a low perception to power distance the type of issuer is more likely to be in the coercive group. They argue that in countries with a low receptivity to an unequal distribution of power authority is based on the field and therefore financial institutions have a high reputation and status. In contrast, in countries with high power distance a hierarchical state model exists, in which the government is in charge. This respect

(24)

of hierarchy makes it easier for governments to succeed in issuing codes of good governance. And accordingly, Haxhi and van Ees (2010) see a relationship between high power distance and the type of issuer being in the normative group. When we compare Hofstedes’ scores for developing (for example India and Brazil) against developed countries (Europe for example) we see that developed countries have a lower perception to power distance compared to developing countries (34 against 87 and 78 respectively). Therefore, a relationship might exist between the normative group and developing countries and in contrast, the coercive group and developed countries.

As stated before governments often promote FDI inflow because of the benefits the country has from FDI. These benefits are for instance capital inflow and technology spillovers for domestic countries (Aitken & Harrison, 1999). Other studies show benefits in terms of multinationals

providing information about export markets to local producers and thus making it easier for them to access markets abroad (Harrison, 1994). Governments want to encourage foreign investors to enable markets with high export potential and little know-how about foreign markets to go abroad. Another advantage for governments to attract FDI is that foreign-owned firms generally pay higher taxes than domestic firms, which leads to an increase in overall wages and increase in the standard of living in the host country (Harrison, 1994). If indeed countries adopt codes of good governance to attract FDI, (and thus a symbolic nature may be expected) it is more likely that the type of issuer is in the normative group. While governments encourage FDI and benefit from FDI inflow.

To summarize, a relationship between the type of issuer being in the coercive group and high power distance countries exists. Moreover, a relationship between low power distance countries and the type of issuer being in the normative group exists (Haxhi and van Ees, 2010). Since developing countries are having lower power distancein generaland developed countries are having a higher perception to power distance a relationship might exist between developing countries and the type of issuer being in the normative group. While Haxhi and van Ees (2010) investigated the

relationship between the normative group and low power distance, it would be interesting to investigate the relationship between the normative group and developing countries. Therefore (see

(25)

also figure 3.):

H5a: In developing countries the type of issuer is more likely to be in the normative group.

In contrast, Haxhi and van Ees (2010) relate the coercive group to higher power distance. For this study it would be interesting to investigate the relationship between developed countries and the coercive group. Therefore (see also figure 3.):

H5b: In developed countries the type of issuer is more likely to be in the coercive group.

The mimetic group in this respect is not tested due to the limited number (3 in total).

(26)

3. Data and methods

3.1. Data Collection

Data is collected on the number of codes of good governance for 84 countries in total. The source of information was the ECGI database (2013) from www.ecgi.org. Information about the year of the first code of good governance issued per country was also available on this database. Information about the identity of the issuers came from a database from dr. I. Haxhi and was extracted from the codes itself, available on www.ecgi.org. Data on the amount of FDI inflow per country was

provided from the database of the World Bank, available at

http://data.worldbank.org/about/country-classification. I used data from FDI inflow 3 years before the first code was issued and 3 years after the first code was issued and calculated the average amount by hand. This resulted in two new variables known as: average 3 years before (A3YB) and average 3 years after (A3YA). Information about whether codes were enforced or not was also extracted from the codes itself available at www.ecgi.org. I used two control variables (GNI/Capita and Country Size). GNI/Capita is known as GNI/Capita based on purchasing power parity (PPP) and information was available at the World Bank database. I took the GNI/Capita on time of the first code issuing. With Country Size the size of the population is meant and information on this was also available on the World Bank database. I again used the data on Country Size at the time the first code was issued. Early starters in this study are known as countries that issued codes before the year 2006, while late adopters are countries that issued codes after the year 2006. Information on this topic could also be extracted from the ECGI database. Whether a country was a developed or developing country was found in the World Bank database. The World Bank classifies

developing countries as middle-income or low-income countries, whereas developed countries are classified as high-income countries by the World Bank.

(27)

3.2. Variables

For the dependent variable FDI inflow was used, the independent variables are type of issuer, number of codes and the year of code issuing. To distinguish between developed and developing countries I made use of a dummy variable (0/1). Enforcement and early starters/late adopters were used as a moderator. I also made dummy variables for these groups. The control variables that were used are GNI/Capita and Country Size (population).

Table 1. Variables and sources

Variables Definition Source

Type of issuer Following the principles of Haxhi & van Ees (2010) there are three type of issuers: coercive (stock exchange and investors’ groups); normative (government and directors’ and professional associations) and mimetic (managers’ associations).

The database of dr. Haxhi and the codes itself extracted from www.ecgi.org

Number of codes Number of codes per country The ECGI database www.ecgi.org Year of code issuance The year of the first issuance of a code The ECGI database

www.ecgi.org Developing country N = 44 Developing countries who issued codes. The world

bank classifying them as middle-income or low-income countries

The database of the World Bank http://data.worldbank.org/about/co untry-classifications

Developed country N = 40 Developed countries who issued codes. The world bank classifying them as high income countries

The database of the World Bank http://data.worldbank.org/about/co untry-classifications

GNI/Capita GNI/Capita based on purchasing power parity (PPP). PPP GNI is gross national income (GNI) converted to international dollars using purchasing power parity rates. For this research the GNI/Capita at the time the first code was issued is used.

The database of the World Bank. Extracted

from:data.worldbank.org/indicator/ NY.GNP.PCAP.PP.CD

Country Size (population) Refers to the total population. In this respect to the

population at the time of the first code was issued. The database of the world bank. Extracted from: data.worldbank.org/indicator/SP.P OP.TOTL

FDI The average amount of Foreign Direct Investment 3 years before and 3 years after the first issuance of a code

(FDI inflow year 1 + FDI inflow year 2 + FDI inflow year 3)/ 3 Code Issued (FDI inflow year 4 + FDI inflow year 5+ FDI inflow year 6)/ 3

Enforcement The enforcement of a country’s codes of good

governance The database of dr. Haxhi and the codes itself via www.ecgi.org Early starters Countries who issued codes before the year 2000 The ECGI database www.ecgi.org Late adopters Countries who issued codes after the year 2000 The ECGI database www.ecgi.org

(28)

4. Results

To check whether the data was useful for the research at first a normality check was conducted. Out of several tests (Histogram, Shapiro, skewness and kurtosis) it appeared that the data was not normally distributed and accordingly I have tried to transform the data and check whether this was effective enough to start the analysis. After data transformation the data did not transform to a normal distribution and therefore the non-transformed data was used for the analysis. I made use of the Central Limit Theorem (when N > 30 a normal distribution may be assumed). See table 2 for descriptive statistics.

Table 2. Descriptive statistics

4.1. Correlation

The relationship between FDI inflow (as measured by Average 3 years after) and the number of codes (issuance) was investigated using Pearson product-moment correlation coefficient. Preliminary analyses were performed to ensure no violation of the assumptions of normality, multicollinearity and homoscedasticity. There was a strong, positive correlation between the two variables, r =.566, n = 85, p <.000, with high levels of FDI inflow associated with higher number of codes (see also table 3).

Min. Max. Mean SD

Average 3 years before -17030000 63483333333 5295174504 9778665960

Average 3 years after -1143076712 263249000000 11637815320 31031403435

Number of codes 1 44 4.56 6.16

GNI/Capita 310 74510 12797 15157

(29)

Table 3. Correlation table

** correlation is significant at the 0.01 level (2-tailed) * correlation is significant at the 0.05 level (2-tailed)

4.2. Paired samples t-test

A paired samples t-test was conducted to evaluate the impact of the intervention of codes of good governance on FDI-inflow. There was a statistically significant increase in FDI-inflow scores from Time 1 (M= 5295174504, SD= 9778665960) to Time 2 (M= 11637815320, SD = 31031403435), t (85) = -2,379, p <.05 (two-tailed). The mean increase in FDI- inflow scores was 6342640815 with a 95% confidence interval ranging from -11644110229 to -1041171401. The eta squared statistic (0.063) indicated a medium effect size. For a sample of 85 countries these results support

hypothesis 1 which predicted that after issuing of the first code of good governance the amount of FDI inflow would have increased.

To see if a difference between developed and developing countries exists two separate analysis were conducted. Two paired samples t-test were conducted to evaluate the impact of the

intervention of codes of good governance on FDI-inflow, for developed and developing countries. For developed countries there was an increase in FDI-inflow scores from Time 1 (M= 7346772936, SD= 11583946640) to Time 2 (M= 18397311766, SD = 43281448432), however this was not significant: t (40) = -2.002, p > 0.05. The mean increase was 11050538827. The eta squared statistic (0.0093) indicated a medium to large effect size. For the developing countries the mean increase of 181986163 was significant: t (44) = -2.18, P <0.05, for Time 1 (M= 3214861264, SD =

(1) (2) (3) (4) (5) (6)

1. Average 3 years after 1.00

2. Issuance .566** 1.00

3. Enforcement .178 .343** 1.00

4. Early starter / late adopter .331** .535** .129 1.00

5. GNI/Capita .244* .315** .413** .143 1.00

(30)

7393412757) to Time 2 (M= 5034722902, SD= 9247530963). The eta-squared statistic (0.10) indicated a medium to large effect size. For details see table 4.

Table 4. Results Paired samples t-test

Mean Std. Deviation 95% conficence Lower Upper 1. General 6342640815** 24578545384 -11644110229 -1041171401 2. Developed countries 11050538827 43281448432 -22215232065 114154406 3. Developing countries 181986163** 9247530963 -3501114823 -138608451

** significant at the 0.05 level 4.3. Regression

Hierarchical multiple regression was used to assess the ability of the control measure (Number of Codes) to predict FDI-inflow, after controlling for the influence of Country Size and GNI/Capita. Preliminary analyses were conducted to ensure no violation of the assumptions of normality,

linearity, multicollinearity and homoscedasticity. Country Size and GNI/Capita were entered at Step 1, explaining 14% of the variance in FDI-Inflow.

After entry of the Number of Codes at Step 2 the total variance explained by the model as a whole was 37.7%, F(3, 77) = 15,541, p < .000. The control measure explained an additional 23,7% of the variance in FDI-Inflow, after controlling for Country Size and GNI/Capita, R squared change = .237, F change (1, 77) = 29.264, p < .000. In this model the control measure was statistically significant with the Number of Codes (beta = .515, p < .001). These results support hypothesis 2 which stated that a positive relationship between the number of codes of good governance and FDI inflow exists.

After entry of the moderating effect of Enforcement at Step 3 the total variance explained by the model as a whole was 38,3%, F (4, 76) = 11.804, p < .000. In this model the moderating measure

(31)

was statistically not significant with FDI-Inflow (beta = .269, p > .390).

In the final model the moderating measure of Early Starters/Late Adapters were entered. The total variance explained by the model was 35,4%, F(5, 75) = 9.760, p <.000. However, this moderating effect showed no significance related to FDI-Inflow (beta = .315, p > .247). Therefore, based on the results of model 3 and 4, hypothesis 3 and 4 are not supported. Hypothesis 3 stated that enforcement moderates the relationship between the number of codes of good governance and FDI inflow. Based on the results there are no indications for this moderating effect. Hypothesis 4 stated that early starters/late adopters would have a positive moderating effect on the relationship between the number of codes of good governance and FDI inflow. There are no indications for this relationship based on the results of the hierarchical regression analysis (see also table 5).

Table 5. Standardized Coefficients (Beta) general

1 2 3 4

GNI/Capita .289 .119 .113 .144

Country size .288 .236 .248 .242

Number of Codes .515* .259 -.103

Enforcement .269 .330

Early Starters/late adopters .315

R Square .140 .377 .383 .394

R Square Change .140** .237* .006 .011

* significant at the 0.01 level ** significant at the 0.05 level

(32)

4.3.1. Developing countries

Two separate analysis were used to see if a difference between developed and developing countries exists. For the developing countries hierarchical multiple regression was used to assess the ability of the control measure (Number of Codes) to predict FDI-inflow, after controlling for the influence of Country Size and GNI/Capita. Country size and GNI/Capita were entered at Step 1, explaining 74% of the variance in FDI-Inflow.

After entry of the Number of Codes at step 2 the total variance explained by the model was 74.2% F(3,34) = 32.599, p < .000. The control measure explained an additional 2% of the variance in FDI-Inflow after controlling for Country Size and GNI/Capita, R squared change = .002, F change (1,34)= .256., p >.616. In this model only the control measure Country Size was statistically significant with Number of Codes (beta = .75, p <.001).

After entry of the moderating effect of Enforcement at step 3 the total variance explained by the model as a whole was 75.8%, F (1, 33) = 25.822, p < .001. In this model the moderating measure was statistically not significant with FDI-Inflow (beta = .429, p > .151).

In the final model the moderating measure of Early Starters/Late Adapters were entered. The total variance explained by the model was 76.5%, F(5, 32) = 20.736, p <.001. However, this moderating effect showed no significance related to FDI-Inflow (beta = .271, p > .337). For details see table 6.

Table 6. Standardized Coefficients (Beta) Developing countries

1 2 3 4

GNI/Capita .077 .076 .069 .053

Country size .852* .819* .835* .865*

Number of Codes .055 -.364 -.068

(33)

Early Starters/late adopters -.271

R Square .740 .742 .758 .765

R Square Change .740* .002 .016 .007

* Significant at the 0.01 level

4.3.2. Developed countries

For the developed countries again hierarchical multiple regression was used to assess the ability of the control measure (Number of Codes) to predict FDI-inflow, after controlling for the influence of Country Size and GNI/Capita. Country size and GNI/Capita were entered at Step 1, explaining 51.5% of the variance in FDI-Inflow.

After entry of the Number of Codes at step 2 the total variance explained by the model was 53.1% F(3,38) = 14.353, p < .001. The control measure explained an additional 16% of the variance in FDI-Inflow after controlling for Country Size and GNI/Capita, R squared change = .016, F change (1,38)= 1.285, p >.264. In this model only the control measure Country Size was statistically significant with Number of Codes (beta = .819, p <.001).

After entry of the moderating effect of Enforcement at step 3 the total variance explained by the model as a whole was 53.9%, F (4, 37) = 10.818, p < .001. In this model the moderating measure was statistically not significant with FDI-Inflow (beta = .120, p > .432).

In the final model the moderating measure of Early Starters/Late Adapters were entered. The total variance explained by the model was 55.6%, F(5, 36) = 9.012, p <.001. However, this moderating effect showed no significance related to FDI-Inflow (beta = .159, p > .251). For details see table 7.

(34)

Table 7. Standardized Coefficients (Beta) Developed countries 1 2 3 4 GNI/Capita .290 .278 .280 .296 Country size .750* .776* .802* .839* Number of Codes -.130 -.213 -.152 Enforcement .120 .147

Early Starters/late adopters -.159

R Square .515 .531 .539 .556

R Square Change .515* .016 .008 .017

* significant at the 0.01 level

4.4. Logistic regression

Direct logistic regression was performed to assess the impact of a number of factors on the likelihood that a type of issuer of a developing country is normative. The model contained three independent variables (coercive, normative and mimetic). The full model containing all predictors was statistically significant, Chi-square (2, N= 83) = 8.877, p <.05, indicating that the model was able to distinguish between developed and developing countries. The model as a whole explained between 10.1% (Cox and Snell R square) and 13.5% (Nagelkerke R squared) of the variance type of country (Developing or Developed), and correctly classified 65.1 % of cases. As shown in Table 4, only one of the type of issuers made a unique statistically significant contribution to the model (Normative). This predictor of the type of issuer recorded an odds ratio of 4.167. This indicated that the odds of the type of issuer being in the Normative group is 4.167 times higher for developing countries than for developed countries, all other factors being equal. Based on the results of the logistic regression support for hypothesis 5a is found. Hypothesis 5a stated that the type of issuer

(35)

for developing countries is more likely to be in the normative group (compared to the coercive and mimetic group) (see also table 8).

In contrast, another direct logistic regression was performed to assess the likelihood that a type of issuer of a developed country is coercive. The model again contained the three independent variables coercive, normative and mimetic. The model which contained all predictors was

statistically significant, Chi-square (2, N=83) = 8,812, p <.05, indicating that the model was able to distinguish between developed and developing countries. The model as a whole explained between 10.1% (Cox and Snell R square) and 13.4% (Nagelkerke R squared) of the variance type of country (developed or developing), and correctly classified 65.1% of cases. As shown in Table 5, only one of the type of issuers made a unique statistically significant contribution to the model (coercive). This predictor of the type of issuer recorded an odds ratio of 3.922. This indicates that the odds of the type of issuer being in the Coercive group is 3.922 times higher for developed countries than for developing countries, all other factors being equal. The odds ratio of 4.444 as seen for the mimetic group indicates a higher likelihood for developed countries to be in the mimetic group than in the coercive group however, this outcome is not statistically significant. The results confirm hypothesis 5b, which stated that for developed countries the type of issuer is more likely to be in the coercive group (see also table 9).

Table 8. Logistic Regression Predicting Likelihood of TOI for Developing Country

B S.E. Wald df p Odds

Ratio 95.0% C.I. for Odds Ratio Lower Upper Coercive 8.218 2 .016 Normative 1.427 .505 7.978 1 .005 4.167 1.548 11.217 Mimetic .223 .743 .090 1 .764 1.250 .291 5.388 Constant -.446 .320 1.943 1 .163 .640

(36)

Table 9. Logistic Regression Predicting Likelihood of TOI for Developed Country

B S.E. Wald df p Odds

Ratio 95.0% C.I. for Odds Ratio Lower Upper Normative 8.160 2 .017 Coercive 1.366 .502 7.422 1 .006 3.922 1.467 10.481 Mimetic 1.492 .828 3.243 1 .072 4.444 .876 22.536 Constant -.981 .391 6.297 1 .012 .375

(37)

5. Discussion

The purpose of this research has been to investigate if a relationship between codes of good governance and FDI inflow exists. The relationship of codes of good governance and FDI inflow has been investigated in general, and a distinction between developing- and developed countries has been made. Furthermore, this research has been searching for indications of symbolic codes of good governance in emerging markets. These indications were sought in enforcement and early

starters/late adopters as moderating variables for the relationship between codes of good

governance and FDI inflow. Moreover, the type of issuer in relation to developed or developing countries has been investigated.

5.1. Hypothesis 1

To investigate the relationship between codes of good governance and FDI inflow the average amount of FDI inflow 3 years before- and 3 years after the first code of good governance has been studied. The results of the data analysis indicate that after adoption of the first code of good governance the amount of FDI inflow increased significantly. These results confirm hypothesis 1 and are consistent with the ideas of Filatotchev et.al. (2007), the World Bank (2000), La Porta et.al (1998) and Levine (1999). While Aguilera and Cuervo-Cazurra (2004) point out that adopting codes of good governance results in better governance systems, the above mentioned state that effective corporate governance systems result in the attraction of foreign investors. When making a distinction between developed – and developing countries, this relationship is only significant for developing countries. The difference can be explained by the fact that the choice of investment location for MNE’s depends on institutional security (Jensen, 2006). Blonigen and Wang (2005) argue that different patterns in emerging markets and developed countries exist. While in developed countries the institutional environment is already more stable, the adoption of a code has less effect in terms of creating more stability than for developing countries.

(38)

The results also indicate a relationship between the number of codes of good governance and the amount of FDI inflow in general, and therefore hypothesis 2 is supported. When we compare the developing with the developed countries on this topic a small difference is seen. For the developing countries the relationship is stronger than for the developed countries. However, in general it may be stated that the more codes of good governance a country issued, the more FDI is attracted. These results are in line with the ideas of Aguilera and Cuervo-Cazurra (2004). They see a relationship between the number of codes of good governance and the presence of foreign institutional investors in a particular country. Moreover, they point out that the more codes of good governance are issued, the more commitment a country shows to improve its corporate governance system, and therefore a country is more attractive for investment.

There are no indications for a positive moderating effect of enforcement on the relationship

between codes of good governance and FDI in general, or for emerging markets in particular. This can be explained by research of Ahlquist and Prakash (2008). They describe a different relationship

(39)

between FDI and host countries. According to them the presences of MNE’s in a particular host country affects FDI inflow. They state that MNE’s can influence institutions in for instance developing countries. MNE’s can force for more enforcement using ‘voice.’ This explains why enforcement does not have an effect on the relationship between codes of good governance and FDI. MNE’s do not always look at whether rules or laws are enforced since they have power in host countries. The presence of other MNE’s therefore could give them more security for gaining in FDI than the fact that rules and regulations are enforced.

Also no moderating relationship for early starters/late adopters and the number of codes of good governance and FDI in all the models is seen. This can be explained by the idea of Aguilera and Cuervo-Cazurra (2004). They state that adopting codes of good governance shows commitment to improve the corporate governance system and in this way the country becomes more attractive for investors. In this respect it does not matter whether a country is an early starter or late adopter since adopting a code of good governance is an improvement in the institutional environment in general and therefore gains trust from investors.

“Codes of good governance and FDI ”.