A cross-industry analysis of the relationship
between market commitment and host country
corruption.
Master’s thesis offered to achieve the degree of MSc in
Business Administration – Strategy track
Name: Brecht Lauwers
Student ID: 11391170
Page numbers: 45
Supervisor: Hesam Fasaei
2
Statement of originality
This document is written by student Brecht Lauwers who declares to take full responsibility for the contents of this document. I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
3
Table of Contents
1. Introduction ... 4
2. Theory development ... 7
2.1. Institutional theory, corruption and the firm ... 7
2.2. Resource dependency theory, corruption and market commitment ... 10
2.3. Corruption across industries ... 16
3. Methodology ... 18 3.1. Sampling strategy ... 19 3.2. Dependent variables ... 20 3.3. Explanatory variable ... 21 3.4. Control variables ... 24 4. Results ... 26
4.1. Descriptive statistics and correlation analysis ... 26
4.2. Regression analysis ... 29
5. Discussion ... 32
5.1. Major findings ... 32
5.2. Contributions of the study ... 35
5.3. Limitations and future research ... 36
6. Conclusion ... 38
4
1. Introduction
We can hardly argue that globalization is one of the most hotly contested phenomena of our time. It is defined as “a shorthand expression for a variety of processes encompassing worldwide integration of financial systems, trade liberalisation, deregulation and market opening, as well as pressures towards cultural, economic, and social homogeneity” (Mathews, 2006). However we define it, we cannot contest that it has significantly altered the way in which firms conduct business (Agarwal & Ramaswami, 1992; Barkema & Drogendijk, 2007; Brouthers & Hennart, 2007; Nielsen, 2003; Pan & Tse, 2000; Zaheer, 1995). Firms nowadays need to manage a wide range of competitive pressures on product-, technological- and market issues (F. J Contractor & Lorange, 2002). This results in an increase in firm internationalisation- and governance issues in order to efficiently manage the complexity of operating in a globalised world (Barkema & Drogendijk, 2007; Root & Visudtibhan, 1992; Zaheer, 1995). Following this, the choice of how to enter a foreign market has become an important issue and research topic, which has drastic implications for a firm’s competitive advantage (F. J Contractor & Lorange, 2002; Nielsen, 2003).
It’s important to realise that not only firms have been affected by globalisation; its effects have also affected governments. This is important because governments tend to have a large influence on the economic activity within the countries they govern (Auerbach & Gorodnichenko, 2012; Ilzetzki, Mendoza, & Végh, 2013; Knack, 2003; Rodriguez, Uhlenbruck, & Eden, 2005; Wade, 1990). As discussed by Wade (1990), efficient and corruption-free public sectors are necessary for a society to develop and prosper; they did not however state that all governments
5 could provide these. Previous research on the quality of government and corruption has shown that corruption often leads to undesirable and costly outcomes because it can negatively affect investment and growth (Brunetti, Kisunko, & Weder, 1998; Lee & Oh, 2007; Rodriguez, Siegel, Hillman, & Eden, 2006). These costs of corruption, as they are called, can differ according to firm-specific factors, but they still tend to have an impact on the way firms operate in markets with certain levels of corruption (Rodriguez et al., 2005). This especially is the case for a firm’s market commitment decision, or the decision of how to enter a foreign market (Agarwal & Ramaswami, 1992; Klapper, Laeven, & Rajan, 2006; Pan & Tse, 2000; Zhou & Guillén, 2015). However, research also showed, that in some cases corruption could actually be considered as the “grease” that makes commerce possible and profitable, especially in regions with a lower quality of public goods and governance (Dreher & Gassebner, 2013; Egger & Winner, 2005; Lee & Oh, 2007; Rodriguez et al., 2005).
A recent study of (Petrou & Thanos, 2014) found truth in both of these arguments and suggest that the relationship between the mode of entry of a firm in a host country, and the level of corruption in that particular host country, does not necessarily has to be linear. They prefer to see the connection between corruption and market commitment as U-shaped, where corruption will have a negative effect on market commitment “up until a threshold point” where it can become a facilitator of business. The main weakness of their work is that it only considers a single, highly regulated industry, namely the banking industry. This limitation makes it hard to generalize their findings and possibly limit them as an industry specific effect. For firms expanding abroad, it is essential to know what kind of costs this decision would bring; these costs include the costs of corruption. It is therefore imperative to overcome the main limitation of Petrou and Thanos’s (2014) analysis so that firms
6 can better consider the possible negative or positive effects of corruption. Building further on their research, the goal of this study is to further substantiate this relationship by subjecting it to a cross-industry analysis and come to a deeper understanding of both market commitment risks and corruption costs.
The hypothesises will be tested using a sample including one hundred chosen companies, registered in the United States, representing in total 1231 subsidiaries. All of these subsidiaries are located within a set of 30 preselected countries. The dependent variable in this study will be the level of market commitment of American firms. This will be measured by the firms’ share of equity in its subsidiary that is operating within one of these 30 countries that have been selected for this study. The explanatory variable in this study will be the level of host country corruption, measured by a dummy variable constructed using the corruption perception index.
This study makes a number of contributions to the debate on how host country corruption influences FDI. The results cause us to conclude that, depending on the corruption levels in the host country, MNEs have to make different choices when managing corruption in a multitude of countries. Secondly, this study contributes to the internationalisation literature by including the institutional theory and the resource dependency perspectives to explain how a firm’s market commitment decision is influenced by host country corruption. Finally, this study contributes to the cross-industry analysis literature of the impact that corruption has on firms’ operations abroad. This study provides a glance on how firms manage different levels of corruption across industries by managing their capital commitment decision. The results suggest that it is more straightforward for a firm to calculate and minimise its entry risks when corruption levels are high, making it more feasible for them to commit more resources.
7
2. Theory development
2.1. Institutional theory, corruption and the firm
Recently, gross domestic product (GDP) per capita growth in developed economies has seen a more than proportionate growth in government expenditure (Gemmell, Kneller, & Sanz, 2016; Peacock, 2004).This is important because it is a sign of the evolving structure of advanced economies (Auerbach & Gorodnichenko, 2012). Particularly because the level of government expenditure as percentage of GDP per capita “affects the structure of employment, the direction of private enterprise effort in conforming to government policies and in supplying government needs, as well as the composition of personal incomes, which are increasingly dependent on transfers and taxes to finance them” (Peacock, 2004). The importance of government expenditure can be illustrated by the following: Finland’s government spending as a percentage of GDP for example was 49.5 in 2008, and in the Netherlands this level was 45.9 in the same year (Brady & Lee, 2014). As such, it’s hard to argue that governments play a vital role in modern economies (Gemmell et al., 2016; Hoskisson, Wright, Filatotchev, & Peng, 2013; Knack, 2003). The security of property they provide, or the enforceability of contractual rights for example, are essential public goods necessary for an economy to thrive (Chan, Isobe, & Makino, 2008; Hoorn & Maseland, 2016; Knack, 2003; Wade, 1990). Few would dispute that this should be their main purpose, to protect property rights and ensure the supply of public goods like clean air, public safety, healthcare, etc. While at the same time withholding from imposing a directional thrust in the market (Wade, 1990). In
8 general, the importance of the public goods they provide can simply not be overestimated. More specifically, “these public goods should be able to create and sustain (a) efficient, rent-free markets, (b) efficient, corruption-free public sectors able to supervise the delivery of a narrow set of inherently public services, and (c) decentralised arrangements of participatory democracy” (Wade, 1990). The higher the level of efficiency with which governments meet these conditions and create policies that will ensure that they’ll be met in posterity, the more development and prosperity will follow (Knack, 2003; Wade, 1990).
As discussed, governments tend to dominate economic transactions, creating institutions and laws in order to facilitate these transactions to benefit the general public (Goel & Nelson, 2010; Hoskisson et al., 2013; Knack, 2003; Lee & Oh, 2007; Peng, Wang, & Jiang, 2008). This motivation however can sometimes be of a rather ambiguous nature. There are studies for example that find that the size and scope of government can significantly affect the quality of public goods and the level of corruption within a country (Goel & Nelson, 2010). Bad government and corruption have always been a factor to be taken into account when and where corporations and political institutions interact (Lee & Oh, 2007). As such, we cannot describe corruption as just a current phenomenon; it has existed throughout the ages and as such, has always been a factor that had to be taken into account when a firm had the intention to enter foreign markets (Goel & Nelson, 2010). Research furthermore pointed out that the credibility of rules is positively correlated to economic growth (Brunetti et al., 1998). This may come as no surprise when imagining a situation where a firm must operate in an environment characterized by unclear rules on property rights for example, or constant policy surprises, or where it is hard or even impossible to enforce a contract. Evidently, an uncertain environment makes it more
9 risky and costly for entrepreneurs to commit resources, especially for projects characterized by large sunk costs (Brunetti et al., 1998; Zaheer, 1995). This riskiness will eventually reduce aggregate private investment and overall economic growth (Brunetti et al., 1998; Rodriguez et al., 2006). Part of this credibility of rules issue and its connected costs for firms is caused by corruption (Brunetti et al., 1998). Thus, efficient and corruption-free public sectors are seen as a necessary condition for a society to develop and prosper, but as we saw, not all states are in a position to provide these (Wade, 1990).
Research showed that “government regulation of entry into new markets, whether by foreign or domestic firms, is associated with higher levels of bribery and corrupt exchanges” (Rodriguez et al., 2006). This notion is important considering that MNE’s subsidiaries tend to encounter pressures to conform to host country conventions and expectations, including corruption (Spencer & Gomez, 2011). These findings built further on the institutional theories of the firm. These theories characterise organisations as to be “influenced by normative pressures, sometimes arising from external sources such as the state, other times arising from within the organization itself” (Zucker, 1987). This is important because formal and informal institutions determine the “rules of the game” and significantly guide the strategy of firms operating within a certain market (Boddewyn & Brewer, 1994; Hoskisson et al., 2013; Spencer & Gomez, 2011). Examples of these formal and informal institutions are standard operating procedures, professional certification and state requirements, … The reason why firms adopt local formal and informal institutions can be found in the hope that adopting these will lead to isomorphism with the institutional environment, which in turn increases the probability of survival in the host country market (Hillman & Wan, 2005; Spencer & Gomez, 2011; Zucker, 1987). Research on
10 the topic has been increasingly focused on emerging economies whose institutions differ significantly from developed economies, but they also found that a high level of variance can be found even within the group of “developed” economies (Hoskisson et al., 2013; Peng et al., 2008).
When considering that corrupt exchanges, like bribes, do not tend to take the same form everywhere, they will usually be a factor that create additional costs for MNEs (Lee & Oh, 2007). These costs mostly flow out of a subsidiary’s efforts to understand and adapt to all forms of corruption that it may encounter in local markets (Zaheer, 1995). It is therefore vital to understand and describe the nature of corruption itself. This can be done using a framework, recently developed by (Rodriguez et al., 2005). It suggests that corruption should be examined from two different dimensions: pervasiveness and arbitrariness. Pervasiveness is defined as a firm’s average likelihood of encountering corruption when it interacts with government officials in a market; it’s independent of the nature of the transactions between these officials and the firm. Arbitrariness on the other hand is defined as the degree to which ambiguity, associated with these corrupt transactions, exists. This is the most pronounced of the two effects of corruption, it comes mainly from that uncertainty in evaluating the probability of gaining the preferential treatments that were promised at the transaction of the briberies (Lee & Oh, 2007; Rodriguez et al., 2005). It impacts a firm’s operations in a market and will influence its market commitment decision (Agarwal & Ramaswami, 1992; Klapper et al., 2006; Pan & Tse, 2000; Zhou & Guillén, 2015).
11 Most firms enter foreign markets via four common modes of entry (exporting, licensing, joint venture, and solo venture), all having different levels of resource commitments (Agarwal & Ramaswami, 1992; Pan & Tse, 2000; Zhou & Guillén, 2015). The decision to enter a market, and the associated decision to what degree committing its resources to that certain market, are important strategic decisions because changing the latter is hard to manage without considerable loss of time and money (Agarwal & Ramaswami, 1992). In general, firms, especially those origination from transparent countries, will tend to have a preference towards establishing wholly owned subsidiaries (WOS) because this gives them a higher level of control over their business and the way its operations are organized and conducted (Duanmu, 2011). But as we have previously discussed, this way of operating can give rise to extra costs due to operating in a corruption environment (Zaheer, 1995). Thus, the question we must focus on now is how going do countries with a high level of corruption influence a firm’s market commitment decision or entry strategy. The answer to the question has already been partly given: firms adopt local formal and informal institutions in the hope that adopting these will lead to isomorphism with the institutional environment, which in turn increases the probability of survival in the host country market (Hillman & Wan, 2005; Spencer & Gomez, 2011; Zucker, 1987). As discussed by numbers of institutional theorists’ firms will “conform to their institutional context so as to achieve (external) legitimacy, which, in turn, renders their existence and actions desirable and appropriate in the view of customers, suppliers, and the government” (Rodriguez et al., 2005). Empirical research on this topic indeed showed that MNE’s subsidiaries encounter pressures to conform to host country conventions and expectations; this includes corruption (Javorcik & Wei, 2009; Spencer & Gomez, 2011). This however, may not necessarily be a bad thing. Reportedly, paying bribes in
12 developing countries in Asia can cause companies to be four times more likely to get deals there, than those firms that do not bribe (Lee & Oh, 2007). It is the legitimacy issue, as discussed by (Rodriguez et al., 2005), flowing out of a reluctance to conform to the institutional context of the host country, that can give or prevent MNEs from having access to valuable resources that will be vital for having a profitable future in the host country market (Duanmu, 2011).
As such, the choice of the firm’s market environment is one of the most fundamental decisions a MNE must make when deciding to enter a foreign market. This because, for subsidiaries of MNEs, operating abroad has always been accompanied by extra costs incurred due to the liability of foreignness, or “all of the additional costs that a firm operating in a market overseas incurs compared to a local firm” (Zaheer, 1995; Zhou & Guillén, 2015). To overcome this liability, firms tend to enter markets where they will be able to deploy their firm’s specific assets, which will allow them to exploit local opportunities (Agarwal & Ramaswami, 1992; Barkema & Drogendijk, 2007; Root & Visudtibhan, 1992; Zaheer, 1995). To this end, a MNEs choice of ownership structure will have a fundamental effect on how they can allocate and control resources, coordinate with other parts of the corporation, and on the prosperity of their operations in the host country (Duanmu, 2011). It is important to realise here that even though a firms’ economic choices may be constrained by the institutional context in which it operates, there will still be a significant role for a firms’ choice in the adaptation to this context (Barkema & Drogendijk, 2007; Oliver, 1997; Zaheer, 1995). More specifically, firms coming from different institutional backgrounds may have developed differed ranges of skills and competences to be able to prosper in different kinds of environments (Duanmu, 2011). An illustration of this is the study of Cuervo-Cazurra (2006) which finds that “corruption results in
13 relatively lower FDI from countries that have signed the Organization for Economic Cooperation and Development Convention on Combating Bribery of Foreign Public Officials in International Business Transactions”, but also that “corruption results in relatively higher FDI from countries with high levels of corruption” (Cuervo-Cazurra, 2006). An explanation of these results, based on the resource based theory is that, “for a relatively corrupt host country, foreign investors from less corrupt countries may have the motivation to seek local partnerships in order to obtain the necessary ‘skills’ and ‘networks’ to navigate local environments, whereas foreign investors from more corrupt countries may have possessed such skills and therefore are ‘competent’ enough to operate alone” (Pfeffer & Salancik, 2003). This is explanation is also suggested by Cuervo-Cazurra and (2006), stating that laws against corruption in the home country may effectively deter MNEs against engaging in corruption, but when MNEs are exposed to home country corruption, they may not be deterred by corruption abroad and may be even be actively seeking it. More generally, a firm’s ability to manage the institutional context in the countries it operates in can even be described as a firm specific competitive advantage (Boddewyn & Brewer, 1994; Hillman & Wan, 2005; Oliver, 1997). An example of this can be found in a study from McWilliams, Van Fleet, and Cory (2002), they found that firms tend to use political strategies with the goal of raising rivals costs by “blocking the use of substitute resources that could give them the opportunity to capitalize on resources that are valuable, rare and costly to imitate”.
Now it’s important to focus on how a MNEs choice of ownership structure, when venturing abroad, is influenced by host country corruption. This can be done by using the previously described framework developed by (Rodriguez et al., 2005), suggesting that corruption should be examined from two dimensions: pervasiveness
14 and arbitrariness. Following their work, the two facets of corruption will both have different effects on how foreign firms enter markets. A high level of arbitrariness increases the likelihood that a MNE chooses a local partner when entering the host country market via foreign direct investment (FDI). They suggest this can be explained by the fact that choosing a local partner will increase the MNE’s subsidiary’s external legitimacy, reducing the overall costs caused by corruption. A high level of pervasiveness, they found, will increase the likelihood of entry via a WOS. They found that this decision also increased the likelihood to engage in corruption. The research of Spencer and Gomez (2011) indeed confirms that MNE subsidiaries face pressures to conform to host country expectations and conventions. This however, can lead to internal legitimacy problems if the MNE has established strong ethical or anti-corruption norms (Rodriguez et al., 2005). As such, we can assume that arbitrariness tends to be the most pronounced of the two effects of corruption. This flows out of its tendency to increase uncertainty in evaluating the probability of gaining the preferential treatments that were promised at the transaction of the briberies (Lee & Oh, 2007; Rodriguez et al., 2005). The uncertainty impacts a firm’s operations in a market and will therefore influence its market commitment decision (Agarwal & Ramaswami, 1992; Klapper et al., 2006; Pan & Tse, 2000; Zhou & Guillén, 2015).
The common point made by resource-based view theorists is that the more important local resources are for certain industries the higher local equity participation tends to be, because access to resources is vital for sustained profitability, but this access often tends to be controlled by local incumbents (Duanmu, 2011; McWilliams et al., 2002; Zaheer, 1995). As we have already discussed, this market commitment decision factors in a lot of different kinds of costs
15 that will impact a firms’ operations (Agarwal & Ramaswami, 1992; Klapper et al., 2006; Pan & Tse, 2000; Rodriguez et al., 2005; Zhou & Guillén, 2015). We further discussed that firms tend to mitigate these costs by entering markets where they will be able to deploy their firm’s specific assets, but that this ability can be influenced by host country corruption (Agarwal & Ramaswami, 1992; Barkema & Drogendijk, 2007; Dreher & Gassebner, 2013; Egger & Winner, 2005; Zaheer, 1995). Most of the previous research however presumed that corruption directly enters the cost function of MNEs, which causes these studies to suggest a negative relationship between FDI and corruption. We have already discussed that this is not always the case, that another stream of research proposes an opposing view, stating that corruption in some cases can actually be beneficial for firms (Duanmu, 2011; Lee & Oh, 2007; Rodriguez et al., 2005). A recently proposed third view may be the unifying solution for this problem. Petrou and Thanos (2014) found truth in both of these opposing arguments and suggest that the relationship between the mode of entry of a firm in a host country, and the level of corruption in that particular host country, does not necessarily have to be linear. They prefer to see the connection between corruption and market commitment as U-shaped, where corruption will have a negative effect on market commitment “up until a threshold point” where it can become a facilitator of business. They indeed found a negative corruption effect on market commitment at low- to moderate levels of corruption and an opposite effect at higher levels of corruption, which explains the concept of the “helping hand”. In this study, we will build further on their findings and investigate the following hypothesises;
Hypothesis 1: “low corruption levels will be positively associated with a MNE’s host country capital commitment decision”.
16
Hypothesis 2: “moderate corruption levels will be negatively associated with a MNE’s host country capital commitment decision”.
Hypothesis 3: “high corruption levels will be positively associated with a MNE’s host country capital commitment decision”.
These findings would establish that there is a U-shaped relationship between host country corruption and a firms’ market commitment. Meaning that firms will commit less capital when host country corruption levels are moderate and will commit more when host country corruption levels are low or high”.
2.3. Corruption across industries
Further substantiating Petrou and Thanos's (2014) research is particularly important because of two reasons. Firstly, because it is essential for firms that are expanding abroad to know what kinds of costs this decision would bring; these costs include the costs of corruption. And secondly because their work has a significant flaw; their study solely considers a single, highly regulated industry, namely the banking industry. This limitation makes it hard to generalise their findings and possibly limit them as an industry specific effect. It is therefore imperative to overcome this limitation of Petrou and Thanos’s (2014) analysis by subjecting it to a cross-industry analysis and come to a deeper understanding of both market commitment risks and corruption costs.
We already discussed that governments tend to dominate economic transactions, facilitating-, or hindering them (Knack, 2003; Rodriguez et al., 2005; Wade, 1990). We also saw that the regulations, institutions and laws they install do
17 not always benefit the general public and that the quality of government and levels of corruption vary greatly across boarders (Brunetti et al., 1998; Goel & Nelson, 2010; Lee & Oh, 2007; Peng et al., 2008). Also, corruption costs mainly flow out of the arbitrary nature of corruption itself, which tends to increase uncertainty in evaluating the probability of gaining the preferential treatments that were promised at the transaction of the briberies (Lee & Oh, 2007; Rodriguez et al., 2005). It’s this uncertainty that mostly impacts a firm’s operations in a market and will therefore influence its market commitment decision (Agarwal & Ramaswami, 1992; Klapper et al., 2006; Pan & Tse, 2000; Zhou & Guillén, 2015). The study of Svensson and Reinikka (2002) also found that “firms typically had to pay bribes when dealing with public officials whose actions directly affect the firms’ business operations”. They found that this was especially the case when firms had difficulties avoiding government interference, for example, when importing, exporting, or when doing business requires the use of public infrastructure services. What’s interesting about their work is that they used firm-level data and found a certain degree of consistency cross-country, but also across industries. This led them to conclude that “variation in policies/regulations can explain the incidence of corruption across firms”, but they not only meant cross country variation in policies, but also cross-industry variations in policies (Svensson & Reinikka, 2002). This is why it’s now important to discuss the fact that corruption levels not only vary across countries, but also across industries (Duanmu, 2011; Klapper et al., 2006; Svensson & Reinikka, 2002).
Two recent studies of Fredriksson, Vollebergh and Dijkgraaf (2004) and Damania, Fredriksson and List (2003) focused on the impact that corruption can have on the environmental policy implemented within a country. They found with cross-industry consistency that “greater corruptibility reduces the stringency of energy
18 policy by shifting the government's relative weight away from welfare towards bribes, making it cheaper to purchase government influence” (Fredriksson et al., 2004). Another study on this topic, conducted by Cole (2007), further concluded that when “the degree of corruptibility is sufficiently high (low), FDI leads to less (more) stringent environmental policy, and FDI thus contributes to (mitigates) the creation of a pollution haven”, further substantiating the ‘grease the wheels’ hypothesis on corruption. Furthermore, Duanmu's (2011) empirical research found that within China, corruption influences the entry mode of firms in such a way that “resource intensity industry, the size of the project, and capital intensity of the project increase the likelihood of JV being chosen over WOS, whereas projects with large foreign capital injection are more likely to be WOS”. They furthermore concluded that this decision was not only influenced by country specific levels of corruption, but also by the particular industry corruption levels. Another cross-industry analysis conducted by Klapper et al. (2006) on market entry rigidity concluded that regulatory barriers do tend to adversely affect firms when entering in less corrupt countries, but they do not adversely affect firms when entering in corrupt countries. This conclusion also substantiates the ‘grease the wheels’ hypothesis (Klapper et al., 2006). As such, the following hypothesis will be tested here;
Hypothesis 4: “The proposed U-shaped corruption-capital commitment relationship will be consistent across industries”.
3. Methodology
In this chapter, the focus will be on the approach of the research design of the used models, used for testing the validity of the proposed hypothesizes. We start out
19 by describing the sampling strategy and more specifically explaining which database we use and how we select observations. Secondly, we will discuss how the dependent, explanatory and control variables will be operationalized and tested.
3.1. Sampling strategy
Considering that we will substantiate the findings of Petrou and Thanos (2014) by conducting a cross-industry analysis of the “grabbing/helping hand” thesis, the sample used here will consist of a number of preselected companies, representing multiple industries and having foreign subsidiaries in a set of preselected countries. These countries will be selected based on their ranking within the corruption perception index (CPI) (Transparency International, 2016). The main difference with their study however will be that secondary data, obtained from an online database named ORBIS, will be used instead of primary sources.
ORBIS itself is a database containing comprehensive information on both listed and unlisted companies located worldwide. The database contains information on more than 120 million companies. This data includes information on key company financials, all in standardized format; financial strength indicators, contacts and directors, ratings, stock data, original filings/images, private equity data and portfolios, patents, detailed corporate and ownership structures, industry research, business and company-related news. From this data, we select the 100 biggest companies headquartered in the United States of America, based on their operating revenue in the year 2016. We exclude companies for which no recent financial data exists. We further collect data from their subsidiaries, more specifically, their location
20 and the percentage of the subsidiarity which is owned by the American “mother” company. The obtained data contains information on 25485 subsidiaries, all (partly) owned by the preselected one hundred American companies. We exclude data of subsidiaries where no clear information exists on how much of its equity is owned by one of the American companies. We further select 30 countries, based on their ranking within the CPI, which will be illustrated later, and exclude all data from subsidiaries not located within these selected countries. The final sample includes the one hundred chosen companies, registered in the United States, representing in total 1231 subsidiaries, all located within the following preselected 30 countries: Argentina, Belgium, Canada, China, Colombia, Germany, Denmark, Egypt, Spain, United Kingdom, Hungary, Indonesia, Japan, South Korea, Lithuania, Malaysia, Nigeria, Netherlands, New Zealand, Peru, Philippines, Poland, Portugal, Qatar, Russia, Sweden, Singapore, Slovenia, Taiwan and Venezuela.
3.2. Dependent variables
The dependent variable in this study will be the level of market commitment of American firms, operating in a selected list of countries. We already discussed that in general, the market commitment decision, or the extent of equity ownership of a subsidiary in a foreign country, is important because it will have implications on resource commitment, risk, returns and control (Chari & Chang, 2009).
For the measurement of foreign market commitment, the focus will be here on the firm’s capital investment in the host country, measured by the percentage of the share of equity owned by the firm in the foreign venture. The reason why we choose
21 share of equity to measure capital commitment is that the latter is a complex decision taking into account a lot of variables like “valuation costs due to information asymmetry between foreign and local firms, problems in integrating local firm managers in culturally distant countries, the cost of separating desired assets of the local firm from non-desired assets, and the cost of resource commitment under exogenous uncertainty” (Chari & Chang, 2009). The former tends to be influenced by the same factors, making it appropriate to measure host country market commitment (Chari & Chang, 2009).
3.3. Explanatory variable
The point of this study is to investigate how host country corruption influences the market commitment decision of a MNE when entering that market. As such, we will include it as the explanatory variable, trying to exactly determine what its effects are. Corruption can be defined in many different ways, but here we will use the concept as is defined by Transparency International (2016) as a “form of abuse of public power for private gain”. The measurement of corruption can be quite tricky because of its subjective and elusive nature, which can make it hard to quantify (Wei, 2000). In general, there are three types of measures available to measure it. Firstly, measures based on ratings given by in-house experts. Secondly, measures based on surveys of executives within the business environment of a specific country. And lastly , which is also the most used method in corruption studies, is a an average of existing measures, an example of which is the CPI (Habib & Zurawicki, 2002; Wei, 2000).
22 This index will also be used in this study. The index itself is a collection and combination of surveys, conducted by 12 different organisations. Its values range from 0 to 100 with 0 indicating the highest degree of corruption and 100 the lowest degree. In our analysis, we will transform this scale in a dummy variable with 3 categories, high-, moderate- and low levels of corruption with the cut-off values of 45 and 70. The 30 selected host countries for the subsidiaries are all allocated in one of these groups, resulting in exactly 10 countries in each “class of corruption”. Their classification is shown in Table 1;
23
TABLE 1. CLASSIFICATION OF SELECTED COUNTRIES
COUNTRY CORRUPTION LEVEL
10 low corruption levels
DENMARK (DK) (90) NEW ZEALAND (NZ) (90) SWEDEN (SE) (88) SINGAPORE (SG) (84) NETHERLANDS (NL) (83) GERMANY (DE) (82) CANADA (CA) (81) UNITED KINGDOM (GB) (81) BELGIUM (BE) (77) JAPAN (JP) (72)
10 moderate corruption levels
PORTUGAL (PT) (62) POLAND (PL) (62) QUATAR (QA) (61) SLOVENIA (SI) (61) TAIWAN (TW) (61) LITUANIA (LT) (59) SPAIN (ES) (58) SOUTH KOREA (KR) (53) HUNGARY (HU) (48) MALAYSIA (MY) (49)
10 high corruption levels
CHINA (CN) (40) COLOMBIA (CO) (37) INDONESIA (ID) (37) ARGENTINA (AR) (36) PERU (PE) (35) PHILIPPINES (PH) (35) EGYPT (EG) (34) RUSSIA (RU) (29) NIGERIA (NG) (28) VENEZUELA (VE) (17)
BASED ON THE 2016 EDITION OF THE CPI AS PUBLISHED BY TRANSPARENCY INTERNATIONAL.
24
3.4. Control variables
To get a more realistic view on how the industry in which firms operate affects the relationship between market commitment and corruption, it is necessary to control for a number of firm- and institutional specific variables that can have an influence on this relationship.
Firm performance. It is first of all important to control for firm performance.
Mainly because studies have shown that there is a positive correlation between MNE internationalization and their overall performance (Bausch & Krist, 2007; Herrmann & Datta, 2005). Research further showed that this relationship works both ways, suggesting that firms with highly profitable operations are more inclined towards a choice of high capital commitment (Bausch & Krist, 2007). This may be because high profitability levels can create a higher willingness to take more risks when expanding abroad (Herrmann & Datta, 2005). These risks include corruption costs when expanding to a host country with a certain level of corruption (Zaheer, 1995). The reason why high profitability can facilitate this choice in this particular situation is that firstly, a firm’s ability to manage the institutional context in which it operates can be the reason for its positive performance. Previous research already suggested that the ability to manage an institutional context can be seen as a firm specific asset and a source of a distinct competitive advantage (Boddewyn & Brewer, 1994). This in turn renders a firm more competitive compared to other firms, contributing to its high performance (Boddewyn & Brewer, 1994; Hillman & Wan, 2005). Secondly, a high profitability level can render a firm more willing to take on more risks, facilitating a capital commitment choice (Bausch & Krist, 2007; Boddewyn & Brewer, 1994; Hillman & Wan, 2005; Oliver, 1997). In this study, a firm’s performance will be
25 measured by the return on equity which is the amount of net income returned as a percentage of shareholders equity. It is expressed as a percentage and measures a MNE’s profitability by exhibiting how much of a firm’s profits are generated by the money that its shareholders have invested in it.
Firm experience. It has already been suggested that there is a positive
relationship between industry experiences and the number and variety of opportunities identified in new firm creation (Gruber, MacMillan, & Thompson, 2012). Furthermore, the research of Zhou and Guillén (2015) and Barkema and Drogendijk (2007) suggests that the more diverse the prior foreign experience of a firm is, the better it can cope with, and overcome the liability of foreignness. They suggest that the reason for this is that firms not only transfer knowledge from their country of origin to the host country, but that firms acquire and combine knowledge gained through experiences in the host country itself. As such, it is not a farfetched assumption that MNEs operating in multiple markets, with some of them characterised by a certain degree of corruption, actually learn how to deal with these practices and are better prepared when faced with corruption than firms who don’t have any experience regarding this matter. Because of this, it’s important to control for firm experience, measured here by subtracting the current year, 2017, by the year in which the MNE was established, when analysing the industry effect.
Firm size. As larger firms have more resources and can spread risks in a more
efficient way than smaller firms, they are more likely to engage in FDI (Lim & McCann, 2013). Moreover larger firms are more likely to take higher risks and thus choose full ownership more often (Javorcik & Wei, 2009). We will measure firm size by taking the firm’s total number of employees. We furthermore log this number because this will better reflect the effect of firm size on the capital commitment
26 decision. Especially because when size will increase by a certain percentage, the effect on share of equity will be of the same magnitude, regardless of firm size (Audia & Greve, 2006).
4. Results
This section will deal with the reporting of the results of this research. We will firstly discuss the descriptive statistics of the variables used in this study. Secondly, multiple regression analysis’s will be conducted to test the hypothesises proposed in the theory development part of this study.
4.1. Descriptive statistics and correlation analysis
The descriptive statistics and bivariate correlation analysis of all variables used in the models of this study are provided in Table 1. The correlations are expressed using the Pearson product-moment correlation coefficient (r). To start out, the correlations between the dependent variable and the control variables will be described. Secondly, correlations between the dependent variables and the independent or the explanatory variables will be described. And lastly, finishing off with a brief glance on the correlations between the explanatory variables and the control variables.
When considering the correlations between the dependent variable (share of equity) and the control variables (return on equity, firm size and firm experience), only return on equity was positively correlated with the dependent variable (r = 0,08). Firm size tends to have a negative effect on the share of equity (r = -0.12) and firm experience also had a relatively weak negative effect on share of equity (r = -0.10).
27 All of these correlations where statistically significant at the 0.01 level. When looking at the control variables themselves, only between firm experience and firm size (r = 0.19), and between firm experience and return on equity (r = 0.05) where significantly positive correlations measured.
Considering the correlations between the dependent variable and the explanatory variables, we can conclude that all the correlations measured were statistically relevant. Low corruption levels had a negative effect on a firm’s capital commitment decision, measured by share of equity (r = -0.19). Furthermore, moderate corruption levels had a relatively weak positive effect on the share of equity (r = 0.06). Also, high corruption levels had a positive effect on the share of equity, but compared with moderate levels, this correlation was a little stronger (r = 0.17).
Finally, when considering the correlations between corruption and the control variables, we can conclude that they tend to be weak and statistically insignificant. Only low corruption levels and firm size had a relatively low, but significantly, positive correlation.
28 TABLE 2. DESCRIPTIVE STATISTICS N Mean S.D. 1 2 3 4 5 6 1. SHARE OF EQUITY 1338 75.67 35.96 _ 2. RETURN ON EQUITY 1245 35.82 89.04 0.08** _ 3. FIRM SIZE 1325 4.9 0.92 -0.12** 0.05* _ 4. FIRM EXPERIENCE 1338 73.01 53.72 -0.10** 0.00 0.19** _ 5. LOW LEVEL OF CORRUPTION 1338 0.61 0.49 -0.19** -0.03 0.09** -0.01 _ 6. MODERATE LEVEL OF CORRUPTION 1338 0.14 0.35 0.06* 0.05 -0.02 0.03 -0.52** _ 7. HIGH LEVEL OF CORRUPTION 1338 0.43 0.43 0.17** -0.01 -0.09 -0.02 -0.71** -0.23** **. CORRELATION IS SIGNIFICANT AT THE 0.01 LEVEL (2-TAILED) *. CORRELATION IS SIGNIFICANT AT THE 0.05 LEVEL (2-TAILED)
29
4.2. Regression analysis
To examine the four hypothesises proposed in the theory development section of this study, five different models have been tested with five regression analysis’s. The results of these can be found in Table 3. Model 1 shows the effects of the control variables on the dependent variable, excluding all explanatory variables. Model 2 shows the results when all the explanatory variables are included. Models 3 to 5 illustrate the effect of each individual explanatory variable on the dependent variable, including the control variables.
When comparing Model 1 and Model 2, the results indicate that the incremental R2 between the two models was 0.041, suggesting that corruption levels explain a
significant percentage of the overall variance. Also, when regarding models 3 to 5 which test the proposed hypothesises, each will explicitly explain a significantly higher percentage of the total variance than Model 1 does. This becomes clear when comparing the value of R2 from Model 1 (R2 = 0.027) with those of Model 3 (R2 =
0.065), Model 4 (R2 = 0.030) and model 5 (R2 = 0.058).
The overall results of this study were expected to show that there would be a strong positive effect of the explanatory variables in Model 3 and 5 and a negative effect in Model 4. This is not the entirely the case. Hypothesis 1 suggests a positive association between low corruption levels and MNE’s host country capital commitment decision. This proposition was tested in Model 3, which however established that this is not the case. Model 3 does not show a positive effect on capital commitment when corruption levels are low. Rather, it establishes with statistical relevance this effect to be negative (t = -6.981, p < 0.001). Hypothesis 2 suggests that
30 moderate corruption levels will be negatively associated with a MNE’s host country capital commitment decision. This proposition was subjected to a test in Model 4 which resulted in a rejection of the hypothesis. The results show that moderate corruption levels will have a neutral- to very small positive effect on the capital commitment decision when corruption levels are moderate (t = 1.804, p < 0.071). Hypothesis 3 proposed that a positive association between high corruption levels and a MNE’s host country capital commitment decision exists. After subjecting this proposition to a test in Model 5, we can indeed establish with statistical relevance that this effect will be positive when corruption levels are high (t = 6.298, p < 0.001).
To test hypothesis 4, the sample included a number of preselected companies, representing multiple industries and having foreign subsidiaries in a set of preselected countries. The idea for this was that when establishing the proposed corruption-capital commitment relationship using cross-industry data, we can reject the idea that this relationship, as proposed by Petrou and Thanos (2014) is industry specific. To accept this hypothesis, it was necessary for hypothesizes 1 to 3 to be accepted. Hypothesis 1 and 2 however, could not be accepted. These findings cause us to not be able to accept hypothesis 4.
To conclude, this study proposed the idea that there will be a U-shaped relationship between host country corruption and a firms’ market commitment decision. A relationship which would explain that firms will commit less capital when host country corruption levels are moderate and will commit more capital when host country corruption levels are low or high. However, the condition for this relationship to be substantiated in this study was an acceptance of all 4 proposed hypothesises, which did not happen in this study.
31
TABLE 3. REGRESSION ANALYSIS
VARIABLES Model I (Control Variables)
Model II (Full Model) Model III (Low Corruption) Model IV (Moderate Corruption) Model V (High Corruption) RETURN ON EQUITY 0.093** 0.088** 0.086** 0.090** 0.093 FIRM SIZE -0.106** -0.083** -0.086** -0.105** -0.088 FIRM EXPERIENCE -0.076** -0.082** -0.083** -0.077 -0.79 LOW LEVEL OF CORRUPTION -0.392 -0.194** MODERATE LEVEL OF CORRUPTION -0.183 0.051° HIGH LEVEL OF CORRUPTION -0.148 0.175** F 11.56** 14.82** 21.191** 9.502** 18.859** R2 0.027 0.068 0.065 0.030 0.058 N 1231 1231 1231 1231 1231 **. CORRELATION IS SIGNIFICANT AT THE 0.01 LEVEL (2-TAILED) *. CORRELATION IS SIGNIFICANT AT THE 0.05 LEVEL (2-TAILED) °. CORRELATION IS SIGNIFICANT AT THE 0.10 LEVEL (2-TAILED)
RESULTS ARE REPORTED USING STANDARDISED COEFFICIENTS
32
5. Discussion
The goal of this research is to investigate the relationship between host country corruption and a firm’s capital commitment decision. This section will start with discussing these major findings of this study, followed by a discussion of the contributions this study provides for the literature on the topic. Finally, the main limitations will be discussed and recommendations for future research will be made. Overall, the results of the regression analysis’s have rejected 3 out of 4 proposed hypothesises that taken together, suggested a U-shaped relationship between the dependent and independent variables. This leads us to conclude that this proposed relationship cannot be substantiated with the results of this study.
5.1. Major findings
In recent literature, there has been a considerable debate about the nature of host country corruption and how it affects the market entry strategy of MNEs. One argument used is that corruption leads to undesirable and costly outcomes because it negatively affects investment and growth (Brunetti et al., 1998; Lee & Oh, 2007; Rodriguez et al., 2006). Proponents of this view argue that corruption causes costs, which negatively impact a firm’s operations in a certain market, and as such influences its decision of how to enter this market (Agarwal & Ramaswami, 1992; Klapper et al., 2006; Pan & Tse, 2000; Zhou & Guillén, 2015). This argument is contrasted by research that showed that corruption can be considered as the “grease” that makes commerce possible and profitable (Dreher & Gassebner, 2013; Egger &
33 Winner, 2005; Lee & Oh, 2007; Rodriguez et al., 2005). Especially in regions with a lower quality of public goods and governance, this effect was measured (Egger & Winner, 2005; Lee & Oh, 2007). The results of this study point in this direction as well, suggesting with statistical relevance that corruption levels do have a positive effect on a firm’s capital commitment decision.
The results demonstrated an expected positive association between high corruption levels and capital commitment. Whilst at the same time demonstrating an unexpected positive effect when corruption levels are moderate. The proposition in this study was that when corruption was moderate, market commitment would be negatively affected. However, the findings resulted in a rejection of this hypothesis. An explanation for these findings could be that in highly corrupt countries, it’s more obvious and straightforward for a firm to calculate and minimize its entry risks, making it more feasible for them to commit more resources (Dreher & Gassebner, 2013; Lee & Oh, 2007; Rodriguez et al., 2005). This explanation illustrates that it’s more important for host countries having moderate corruption levels, like Spain and Poland, to address the corruption issue than it is for countries like Argentina and Russia, which are classified in the category of highly corrupt countries, if they want to minimise potential loss of FDI. The result here are in line with the research of Klapper et al. (2006), which suggest that regulatory barriers do tend to adversely affect firms when entering in less corrupt countries, but they do not adversely affect firms when entering in highly corrupt countries. This suggests that corruption is a facilitator of commerce. Nevertheless, this proposition has to be nuanced a little. The results here indeed show that high corruption levels will have a positive effect on capital commitment, but the suggested negative effect when corruption is moderate has not been found. More specifically, this study found a neutral to positive effect on
34 the capital commitment decision of firms when faced with moderate corruption, but this effect is very small and less statistically relevant when compared with the results at high corruption levels.
Furthermore, contrary to what hypothesis 1 suggests, the results show a clear and statistically relevant negative effect of low corruption levels on capital commitment. A possible explanation of this result may be linked with the concepts of risk, cost and trust. The research of Blomqvist, Hurmelinna-Laukkanen, Nummela and Saarenketo (2008) for example suggests that in a globally competitive industry, business tends to be very dynamic and fast passed. This can often render traditional ways of operating too time-consuming to manage all the risks involved. They suggest that in this case, firms will put a stronger emphasis on the role of trust, contracts and partnerships when expanding abroad, reducing the need for capital commitment (Blomqvist et al., 2008). Applying these finding on the results of this study, their research focusses on firms operating in technology intensive industries, industries which are typically located in highly developed countries. The latter are usually ranked in the low corruption levels category (Transparency International, 2016). This may provide a possible explanation of the negative effect on capital commitment measured in this study.
As such, the results of this study failed to substantiate the work of Petrou and Thanos (2014). They suggested a negative effect of corruption on market commitment “up until a threshold point” where it will become a facilitator of business. The result indeed point towards this facilitator effect on corruption. But contrary to their U-shaped view, the results of this study indicate a linear view. For their idea to be substantiated, all 4 proposed hypothesises had to be statistically accepted, something which has not happened. Mainly because first of all, no positive effects have been
35 measured between low host country corruption and capital commitment, the effect was demonstrated to be negative. And second of all, the hypothesised negative effect of moderate corruption levels on capital commitment were demonstrated to be weakly positive.
5.2. Contributions of the study
This study makes a number of contributions. Firstly, it contributes to the debate on how host country corruption influences FDI. The goal was to substantiate a unifying theory which suggested that corruption and market commitment would be negatively associated up until a threshold point where it will become a facilitator of business. However, the findings here mainly support a “grease the wheels” perspective on corruption. This suggests that when corruption levels are low, capital commitment is discouraged and suggests an opposing result when corruption levels are high. This inclines us to conclude that, depending on the corruption levels in the host country, MNEs have to make different choices when managing corruption in a multitude of countries. As such, it opens up new ways for research on MNEs to investigate how new host countries are selected when firms decide to expand abroad.
Secondly, this study contributes to the internationalisation literature by including the institutional theory and the resource dependency perspectives to explain how a firm’s market commitment decision is influenced by host country corruption. This study draws on both the institutional theory and the resource based theory by including the institutional setting in which firms operate and their firm specific capacities to deal with these settings. As such, a firm is considered a non-independent
36 entity operating within its business environment, influenced by normative pressures that arise from both external and internal sources. In this study, these normative pressures take the form of host country corruption, affecting both an MNE’s operations by influencing its market entry decision, and its subsidiary’s performance by facing it directly with corruption costs. As such, the findings of this study contribute to current studies which employ the institutional theory to evaluate entry mode choice and acknowledges that host country corruption can significantly influence foreign entry decisions.
Finally, this study contributes to the cross-industry analysis literature of the impact that corruption has on firms’ operations abroad. We suggested that corruption has the ability to create both potentially high costs when foreign operations are impacted by host country corruption, but also has the ability to be beneficial for a MNE. This study provides a view on how firms manage different levels of corruption across industries by managing their capital commitment decision. Suggesting that it is more straightforward for a firm to calculate and minimise its entry risks when corruption levels are high, making it more feasible for them to commit more resources.
5.3. Limitations and future research
This study has provided some important insights on the market entry literature using an aggregate determinant to measure corruption and using it to determine its effects on market entry. The way in which the CPI is used is by creating an aggregate of the corruption variable in order to make it useful for research on the topic, but this
37 has also its drawbacks. In general, the use of perception indices, like the CPI measure, can raise concerns about perception biases when creating these indices. The study Wei (2000) already mentioned this issue. They explained that three types of corruption measures are available, but that all of these are subjective, since corruption has an elusive nature which renders quantification difficult. Another limitation is that, even though this study takes into account a positive relationship between industry experiences and the number and variety of opportunities identified by a MNE, it does not extend this view towards their subsidiary themselves. Research has suggested that firms not only transfer knowledge from their country of origin to the host country, but that a firm’s subsidiaries combine this knowledge with knowledge gained through experiences in the host country itself, in order to enhance their performance (Barkema & Drogendijk, 2007; Zhou & Guillén, 2015). This ability would influence a MNE and its subsidiaries’ operations and explain why some are better prepared when faced with corruption than others who don’t have any experience regarding this matter (Gruber et al., 2012). This could influence both subsidiary- and MNE performance and as such, influence the capital commitment decision.
The recommendations for future research this study wants to put forward are twofold. It has first of all already been mentioned here that due to the aggregate nature of the measurement of corruption, this study may not have been able to tells us as much about the relationship between corruption and individual agents as planned. One of the reasons for not finding support for the idea of a non-linear relationship between host country corruption and a MNE’s capital commitment decision may be because not only corruption and industry factors determine this relationship. For a firm to determine how much capital to invest in a foreign venture, many variables are at play. Example are market regulations, industry competition levels, level of host
38 country technological progress, labour costs, host country climate, firm specific assets etc. When all of these variables are filtered out, it may be easier to establish a clear relationship between the dependent and independent variable chosen in this study. A second reason why there is no clear support here for the U-shaped relationship proposed in this study is that this relationship may be industry specific. Indeed, it is a possibility that the relationship between the dependent and independent variable depends on industry specific factors. By taking cross industry data to investigate this effect, this study had the ambitious goal to establish this relationship across industries once and for all. It may also have been its main limitation however. It is a real possibility that the effects between the variables in one industry may have been blurred by an opposing effect in another industry. Future research should therefore find a way to filter out this blurring of one industry specific effect by another.
6. Conclusion
Corruption is a concern that has been shared by a substantial number of international investors. For MNEs its vital to know how to react when faced with corruption in the host environment. To do this correctly is an ability that only increases in importance when these firms raise the significance of their commitment levels during international expansions. We have started out by summarising the recent literature on the topic and described two distinct views of the relationship between market commitment and host country corruption. Next, this study built further on the research of Petrou and Thanos (2014) in order to further substantiate their unifying theory. In order for this theory to be substantiated here, all 4 proposed hypothesises had to be accepted. However, only one of these could be accepted resulting in a
39 failure of this study to substantiate this idea. Instead, it provided new evidence that corruption does not necessarily has to be a cost inducing factor for MNEs. The results here lend support to the idea that in some cases corruption can be considered as the “grease” that makes commerce possible and profitable. In so doing, this study makes a number of contributions to the debate on how host country corruption influences FDI. It suggests that MNEs have to make different and specific choices when managing corruption in a multitude of countries. It further contributes to the internationalisation literature by including the institutional theory and the resource dependency perspectives to explain how a firm’s market commitment decision is influenced by host country corruption. It finally contributes to the cross-industry analysis literature of the impact that corruption has on firms’ operations abroad. In all, this study provides a glance on how firms manage different levels of corruption across industries by managing their capital commitment decision.
40
7. References
Agarwal, S., & Ramaswami, S. N. (1992). Choice of Foreign Market Entry Mode: Impact of Ownership, Location and Internalization Factors. Journal of
International Business Studies, 23(1), 1–27.
https://doi.org/10.1057/palgrave.jibs.8490257
Audia, P. G., & Greve, H. R. (2006). Less Likely to Fail: Low Performance, Firm Size, and Factory Expansion in the Shipbuilding Industry. Management
Science, 52(1), 83–94. https://doi.org/10.1287/mnsc.1050.0446
Auerbach, A. J., & Gorodnichenko, Y. (2012). Fiscal Multipliers in Recession and Expansion. NBER, 63–98.
Barkema, H. G., & Drogendijk, R. (2007). Internationalising in small, incremental or larger steps? Journal of International Business Studies, 38(7), 1132–1148. https://doi.org/10.1057/palgrave.jibs.8400315
Bausch, A., & Krist, M. (2007). The effect of context-related moderators on the internationalization-performance relationship: Evidence from meta-analysis.
Management International Review, 47(3), 319–347.
https://doi.org/10.1007/s11575-007-0019-z
Blomqvist, K., Hurmelinna-Laukkanen, P., Nummela, N., & Saarenketo, S. (2008). The role of trust and contracts in the internationalization of
technology-intensive Born Globals. Journal of Engineering and Technology Management,
25(1), 123–135. https://doi.org/10.1016/j.jengtecman.2008.01.006
Boddewyn, J. J., & Brewer, T. L. (1994). International-Business Political Behavior: New Theoretical Directions. Academy of Management Review, 19(1), 119– 143. https://doi.org/10.5465/AMR.1994.9410122010
Brady, D., & Lee, H. Y. (2014). The rise and fall of government spending in affluent democracies, 1971–2008. Journal of European Social Policy, 24(1), 56–79. https://doi.org/10.1177/0958928713511281
Brouthers, K. D., & Hennart, J.-F. (2007). Boundaries of the Firm: Insights From International Entry Mode Research. Journal of Management, 33(3), 395–425. https://doi.org/10.1177/0149206307300817
Brunetti, A., Kisunko, G., & Weder, B. (1998). Credibility of Rules and Economic Growth: Evidence from a Worldwide Survey of the Private Sector. The World
41
Bank Economic Review, 12(3), 353–384.
https://doi.org/10.1093/wber/12.3.353
Chan, C. M., Isobe, T., & Makino, S. (2008). Which country matters? Institutional development and foreign affiliate performance.
https://doi.org/10.1002/smj.705
Chari, M. D., & Chang, K. (2009). Determinants of the share of equity sought in cross-border acquisitions. Journal of International Business Studies, 40(8), 1277–1297. https://doi.org/10.1057/jibs.2008.103
Cole, M. A. (2007). Corruption, income and the environment: An empirical analysis.
Ecological Economics, 62(3–4), 637–647.
https://doi.org/10.1016/j.ecolecon.2006.08.003
Contractor, F. J., & Lorange, P. (2002). The growth of alliances in the knowledge-based economy. International Business Review, 11(4), 485–502.
https://doi.org/10.1016/S0969-5931(02)00021-5
Cuervo-Cazurra, A. (2006). Who cares about corruption? Journal of International
Business Studies, 37(6), 807–822.
https://doi.org/10.1057/palgrave.jibs.8400223
Damania, R., Fredriksson, P. G., & List, J. A. (2003). Trade liberalization, corruption, and environmental policy formation: theory and evidence. Journal of
Environmental Economics and Management, 46(3), 490–512.
https://doi.org/10.1016/S0095-0696(03)00025-1
Dreher, A., & Gassebner, M. (2013). Greasing the wheels? The impact of regulations and corruption on firm entry. Public Choice, 155(3–4), 413–432.
https://doi.org/10.1007/s11127-011-9871-2
Duanmu, J.-L. (2011). The effect of corruption distance and market orientation on the ownership choice of MNEs: Evidence from China. Journal of International
Management, 17(2), 162–174. https://doi.org/10.1016/j.intman.2011.01.003
Egger, P., & Winner, H. (2005). Evidence on corruption as an incentive for foreign direct investment. European Journal of Political Economy, 21(4), 932–952. https://doi.org/10.1016/j.ejpoleco.2005.01.002
Fredriksson, P. G., Vollebergh, H. R. J., & Dijkgraaf, E. (2004). Corruption and energy efficiency in OECD countries: theory and evidence. Journal of
