The Impact of Environmental Munificence on the Performance of Multinational Enterprises

(1)

Master’s thesis

The Impact of Environmental Munificence on

the Performance of Multinational Enterprises

Supervisor: Dr Gjalt de Jong

Phan Thai Binh (1544985)

MSc International Economics and Business

Faculty of Economics

University of Groningen

(2)

Acknowledgements

My thesis would not have been completed without the support of several people. First of all, I would like to express my special thanks to Dr Gjalt de Jong, my supervisor. His invaluable advice on my thesis and many other issues concerning study and work styles has been of great importance to the thesis, as well as to my knowledge generally and future job prospects. Many thanks to Professor H.W.A. Dietzenbacher who has given me much useful advice related to the research method section of my thesis.

I am overwhelmingly grateful to Dr G.J. Lanjouw, Ms G.A.M. Lakerveld and Ms A. Veltmaat who helped me on numerous occasions during my studies at Groningen. I would like to thank the NPT programme, University of Groningen, and the University of Can Tho for the financial support for my study.

I am forever indebted to my parents and my whole family for their love and for always being with me. Finally, I would like to express my deep appreciation for the great support and encouragement provided by my friends in Groningen and at home.

(3)

Abstract

This study investigates the effects of environmental munificence, which is perceived as consisting of production factors and institutions, including formal and informal institutions, and the enforcement system, on multinational enterprise performance. The panel data sample and two-step procedure model are employed to estimate these effects. The study finds that for production factors, while country-level technology capability and infrastructure are found to have positive effects on multinational enterprise performance, no support can be found for the impact of the efficiency of financial markets on multinational enterprise performance. For formal institutions, in contrast to trade promotion policy which is found not to be related to multinational enterprise performance, investment promotion policy and the flexibility of labour regulations are suggested as the vital determinants of the location choices and performance of multinational enterprises. For informal institutions, while individualism (IDV) is found to have a negative effect on multinational enterprise performance, the impact of the two remaining informal institutions, particularly masculinity (MAS) and uncertainty avoidance (UAI), on multinational enterprise performance is found to be unclear. The other institutional element, the efficiency of

the law enforcement system, is suggested to have a positive effect on multinational enterprise

performance.

(4)

List of Tables and Figures

Pages

Table 1: Descriptive statistics ... 30

Table 2: Variables, measures and data sources... 33

Table 3: Correlation matrix of the variables ... 34

Table 4: The results of the fixed and random effects models ... 35

Table 5: The results of the second step of the two-step procedure ... 39

Table 6: The results of the two-step procedure regression model ... 40

Figure 1: Histogram and normality test of the residuals − ROA (fixed effect model) ... 59

Figure 2: Histogram and normality test of the residuals − EBITOA (fixed effect model) ... 59

Figure 3: Histogram and normality test of the residuals − ROA (the original OLS model)... 60

Figure 4: Histogram and normality test of the residuals − EBITOA (the original OLS model).... 60

(6)

1. Introduction

Multinational enterprises (MNEs), companies with substantial international business operations and investments (Goehle, 1980), have become popular and key players in the global market. The important role of MNEs is illustrated by the increasing share of foreign subsidiaries in total manufacturing employment and sales for G5 countries (Navaretti and Venables, 2004)1_{. For} example, the sales of French and British firms abroad have accounted for approximately 30 percent of total sales for France and the United Kingdom respectively. Similarly, over 20 percent of total American sales come from its foreign subsidiaries (Navaretti and Venables, 2004). The recent impressive development of MNEs has been significantly assisted by the supportive measures in environmental munificence which countries have attempted to introduce in order to attract the investment of MNEs.

Environmental munificence is the scarcity or abundance of critical resources needed by one or more firm operating within an environment (Dess and Beard, 1984; Castrogiovanni, 1991). The resources available within an environment influence the survival, growth and performance of firms sharing that environment. They also affect the ability of new firms to enter that environment (Randolph and Dess, 1984). Environmental munificence consists of two factors: production resources and institutions (Wan and Hoskisson, 2003). Production resources are defined as comprising the production factors used to produce goods and services (Wan and Hoskisson, 2003). Porter (1990) distinguishes two groups of production resources: basic and advanced resources. Basic resources include natural resources, climate, location and unskilled labour. They are widely available and readily accessible to all multinational firms through foreign activities or sourcing on international markets. Thus, the competitive advantages of MNEs are only marginally based on these resources. Drawing on this argument, the basic factors should not be taken into account in studies on multinationals. Advanced resources refer to such elements as physical infrastructure, financial resources and technology capability. The advanced factors have been perceived as the most necessary for achieving high-order competitive advantages which determine the success of multinationals in the global market. The second element of environmental munificence is institutions, which are defined as ‘the rules of the game in a society or, more formally, are the humanly devised constraints that shape human interaction’ (North, 1990).

(7)

Although some studies on the effects of environmental munificence on multinational firm performance have been undertaken, they have examined the effects of the host and home countries’

environments on multinational firm performance separately. In other words, they do not take into account the joint effect of the home and host country environments on multinational firm performance. For example, in his study on the competitive advantages of multinationals, Porter (1990) emphasizes the role of home country resources in generating and sustaining the competitive advantages of multinational firms, but ignores the role of host country resources. He argues that the competitive advantages of multinational firms ultimately result from an effective combination of home country circumstances and company strategies, whereby the home country conditions may create an environment in which MNEs can attain international competitive advantages (Porter, 1990). By contrast, other economists, such as Chang (1995) and Branstetter (2000), emphasize the effect of the host country environment on MNEs. They suggest that firms considering investment abroad need to assess their capacity to overcome the liabilities of foreignness and to use local factors of production in a cost-efficient manner. Thus, their long-term success depends on the stability of the business environment, as well as the existing technology and capabilities of host countries (Branstetter, 2000). Generally, existing studies on the effect of the host country environment on MNEs mainly use subsidiary-level data to examine how an individual subsidiary of an MNE is separately affected by the environment of the host country in which it is located. The effect of the environment of the other host countries of other subsidiaries within the same MNE on this individual subsidiary is often ignored.

(8)

munificence of all countries where an MNE operates should be taken into account in studies of multinational enterprise performance.

This study aims to examine the effects of environmental munificence on multinational enterprise performance. This study is able to make some significant contributions to the current literature.

First, in contrast to the existing studies of multinational firm performance, this paper examines the joint effects of the environmental munificence of all the countries (consisting of home and all host countries) in which a multinational operates. Second, informal institutions, such as cultural factors, which are almost forgotten by the existing studies, are also appropriately taken into account. Third, by employing panel data, this study is able to control for the omitted individual and period-specific effects, thus obtaining more reliable results than do previous studies of multinational firm performance which employ either cross-sectional or time-series data. To achieve the research objective, this study attempts to investigate the two following main research questions:

- How do production factors affect multinational enterprise performance?

- How do institutions affect multinational enterprise performance?

(9)

2. Theoretical Background and Hypotheses

2.1. Theoretical background

2.1.1. Production resources

The first element of environmental munificence is production resources. There are numerous studies assessing the undoubted impact of national production resources on firm performance. Studies taking a resource-based view emphasize the importance of firm capabilities in creating and sustaining competitive advantages (Barney, 1991). As such, a firm has a competitive advantage if it is in the process of implementing a value-creating strategy that is not simultaneously being implemented by any current or potential competitor. A firm’s competitive advantage is sustainable over a long period of time as it is not easily imitated by other firms. Porter (1990) argues that the capabilities for creating competitive advantages in firms greatly depend on home country resources. Innovation is one of the best ways for firms to seek competitive advantages, and the home country resources are integral to the process of perceiving the opportunity for innovation and its successful implementation. Innovation takes place due to stimulation by home country resources. Innovation succeeds because it is supported by the home country resources. Indeed, a country with high technological capabilities, advanced infrastructure and plentiful financial resources would provide firms with great opportunities to pursue innovation strategies. Porter (1990) also stresses the crucial role of home country resources in sustaining the competitive advantages of firms. An advantage once gained is only sustained by a continuous search for different and better ways of operating and through ongoing modifications in firm behaviour. Abundant advanced resources are necessary for this ‘upgrading’ process. In addition to Porter, Barney (1988) also suggests that the firm’s capacity becomes stronger when its internal resources are integrated with external resources. Barney (1991) introduces the concept of a strategic factor market which is defined as a market where the resources necessary to implement a strategy are acquired. He asserts that this market seems to be imperfectly competitive. In such a factor market, firms are able to generate above normal economic performance by acquiring strategic resources such as highly skilled personnel and technology, which may only exist in countries with munificent resources. Moreover, Grant (1991) argues that national resources have a powerful influence on the competitive advantages of companies and industries. Similarly, Castrogiovani (1991) also suggests that abundant resources encourage the entry of firms, while resource scarcity prevents firms from entering the market.

(10)

advantages. For instance, firms have increasingly shaped factor creation in educational institutions by sponsoring students or sending their personnel to study at such institutions (Porter, 1990). Firms also provide financial support to research institutions. Nestle, for example, has funded and supported the Geneva-based IMI, which has become a leading European business school. Of course, Nestle has benefited from a steady flow of talented management students and a vehicle for ongoing management training (Porter, 1990). The increasing financial support offered by firms to research institutions (Blumenthal et al., 1986) is due to the fact that the research performed at these institutions is an important resource in firm innovation (Nelson, 1986).

(11)

country-level resources would offer firms opportunities to access superior inputs at low costs, and thus improve their performance.

In summary, the current literature has shown that production resources have a significant role in determining economic organization performance. However, surprisingly, they do not pay adequate attention to how these resources impact the performance of multinational firms. In particular, their objects of study are mostly national firms, not multinational firms. Although there has been some research into the effects of country-level resources on multinational firm performance, these studies only examined the effects of host or home country resources but not both. For instance, Porter (1990) studied the competitive advantages of multinationals, emphasizing the role of home country resources but ignoring the role of host country resources whereas other economists, such as Chang (1995) and Branstetter (2000), emphasize the role of host country resources but ignore the role of home country resources. They argue that MNEs favour investment abroad because it allows them to exploit local production factors which are rare or expensive in their home countries. Thus, their long-term success depends greatly on the stability of the business environment, as well as on the existing technology and capabilities of host countries. Moreover, these studies mainly employ subsidiary-level data to examine only how an individual subsidiary is impacted by the resources of the country in which it operates. The effects on the individual subsidiary of the resources of other countries in which other subsidiaries of the same MNE operate are virtually ignored.

However, as mentioned, the exchanges of resources between units in home and host countries and between units in different host countries, plays a vital role in multinational activities. By exchanging resources between an organization’s units, MNEs are able to overcome insufficiencies in the resources of individual countries. For instance, the theory of the multinational suggests that multinational firms can engage in arbitrage across factor markets and exert leverage using their market power to reduce input costs (Kogut, 1985). They have opportunities to achieve greater returns on resources and to seek less expensive inputs and less price-sensitive markets (Kim et al., 1993). Moreover, Wan and Hoskisson (2003) suggest that when home country-level resource abundance is low, firms are able to maintain a significant competitive advantage by investing in host countries that supply the factors lacking in the home country. Hence, the inter-supports2 between units within a multinational firm are crucial and relevant, with the performance of a unit in a home country being significantly affected by the environmental munificence of not only the home country but also the host countries where other units operate, and vice versa. Therefore, it is not sufficient for us to

2_{Inter-supports refer to the circumstances that allow subsidiaries within an MNE to support each other.}

(12)

separate the effects of the resources of home and host countries when studying the effects of environmental munificence on multinational firm performance. Instead, the joint effect of the environmental munificence of all countries, including the home country as well as all the host countries in which a multinational operates, should be taken into account in studies on multinational firm performance. This thesis makes such a contribution to the current literature.

2.1.2. Institutions

The second aspect of environmental munificence is institutions. North (1990) defines institutions as ‘the rules of the game in a society or, more formally, [institutions] are the humanly devised constraints that shape human interaction’. More specifically, North (1990) and Yeager (1999) propose three components of institutions: formal rules, informal rules and law enforcement mechanisms. First, formal rules are the written rules of society, such as laws, regulations, government policies and bank regulations. This thesis considers three formal institutions: labour regulations (flexibility level of labour regulation), investment promotion policies and international trade promotion policies, all of which may affect the economic performance of multinational firms. Second, informal institutions are the unwritten rules of society which include culture, norms of behaviour and codes of conduct. This thesis examines the effect of social cultures on multinational firm performance. Hofstede (1980) conceptualizes social culture as existing across four dimensions: individualism (the preference for behaviour that promotes one’s self-interest), power distance (the tolerance of social inequality), masculinity (the distribution of roles between the genders) and uncertainty avoidance (intolerance of ambiguity and uncertainty). In this thesis, three of the four dimensions most likely to affect firm performance are taken into account: individualism (IDV), masculinity (MAS) and uncertainty

avoidance (UAI). The remaining dimension, power distance, is not considered in this study as it

(13)

system, which refers to government mechanisms used to ensure that the formal rules are enforced (Yeager, 1999).

Organizations, including MNEs, are constrained and supported by institutional forces (Scott, 1995). There are numerous studies suggesting that the institutional environment significantly affects the performance of economic organizations. Again, studies using a resource-based view also emphasize the effect of the institutional framework in generating and sustaining firms’ competitive advantages. First, formal institutions, such as labour regulations, investment and international trade promotion policies can bring considerable advantages or disadvantages to firms. For example, strict labour regulations have a sizable and significantly negative effect on firm employment decisions (Mondino and Montoya, 2000). Trade protection policies such as high import tariffs and quotas are strong motives for MNE investment, since by setting up local production, multinational firms are able to eliminate the trade costs associated with high trade-barrier policies (Navaretti and Venables, 2004: 136). Second, in addition to formal institutions, informal institutions are also crucial determinants of a firm’s competitive advantage (or disadvantage). For instance, while highly individualist societies maintain three particular beliefs − in freedom, outward orientation and the importance of

contact with senior managers (Hofstede, 1980) − which have been found to encourage innovation

(Shane, 1993), high uncertainty-avoidance societies restrict firms’ capabilities in innovation and product development (Nakata and Sivakumar, 1996). Third, the law enforcement system influences a firm’s ability to sustain competitive advantages. In particular, multinationals, as well as national firms that engage in the development of new technology and innovation in order to obtain competitive advantages, face difficulties in sustaining these advantages as competitors always attempt to imitate successful innovations and adapt them to their own use (Helpman, 1993). This problem has been exacerbated in developing countries where intellectual property rights and enforced domestic law and regulations are lax (Benko, 1987). The crucial role of the law enforcement system in sustaining a firm’s advantages is pointed out by Lee and Mansfield (1996), who find that a relatively weak intellectual property protection and law enforcement system in a developing country may lower the probability that multinational firms will invest there, and even if they do so, they may only be willing (because of weak protection) to invest in wholly-owned subsidiaries rather than entering into joint ventures with local partners, or only to transfer older technology, due to the risk of losing competitive advantages. More recently, Besanko et al. (2004) have also appreciated the crucial role of intellectual rights and the law enforcement system as impediments to imitation.

(14)

Similar to the resource-based view, economists with a transaction-cost view also stress the role of effective institutions in reducing transaction costs and thus increasing economic organization performance. Institutional economists, most notably Williamson (1975) and North (1990), have contended that without the institutions, or ‘the rules of the game’, that prescribe a country’s incentive structure and economic specialization, complex interfirm transactions would become too costly to complete (Greif, 1993). Without sufficient institutions in place, firms have to engage in tedious transactional processes such as contract negotiation and dispute resolution (Wan and Hoskisson, 2003). Yeager (1999) argues that a nation’s institutional framework, including formal rules, informal rules and the law enforcement system, determines the transaction costs. Formal rules determine the level of transaction costs through their impact on measurement and enforcement costs.3_{Informal rules are much more important than formal rules in determining transaction costs} because they influence the nature of formal rules and the way formal rules are enforced (Yeager, 1999). For instance, high uncertainty avoidance lowers transaction costs (Brouthers and Brouthers, 2000). Meyer (2001) found that institutions reduce transaction costs by reducing uncertainty and establishing a stable structure to facilitate interactions. To support this argument, Meyer suggests that inefficient institutions in transition economies are the main cause of high transaction costs incurred by firms in these countries. In particular, in transition economies with inefficient institutions, the new economic agents lack knowledge, tacit or otherwise, both in how to use the market mechanism and of potential partners and competitors. They need to identify potential types of business and the preferences of potential business partners, and they need to learn to assess the composition of demand and supply to set the right prices. This would increase the search, negotiation and contracting costs of new businesses. The lack of knowledge of markets and the functioning of a market economy magnifies transaction costs in the transition context (Meyer, 2001).

(15)

MNEs prefer their local subcontractors to invest in relation-specific assets since these investments are necessary for the efficiency of particular transactions (Williamson, 1985).4 Writing a complete contract for a relation-specific investment is impossible (Navaretti and Venables, 2004: 102), thus the so-called ‘hold-up’ problem is likely to occur if firms operate in countries with low trust and weak legal and enforcement systems.5 The ‘hold-up’ problem raises transaction costs in four ways: more difficult contract negotiation and more frequent renegotiations, investment to improve ex post bargaining positions, distrust and reduced investment in future relationship-specific investments (Williamson, 1985). Especially for MNEs, the transfer of tacit knowledge, technology and modern managerial skills between head office and subsidiaries and amongst subsidiaries is one of the core activities in sustaining competitive advantage (Ghoshal et al., 1994). In markets with an inefficient institutional environment, multinational firms favour internalizing this transfer process because the inefficient institutional framework does not provide efficient protection of intellectual property rights (Meyer, 2001). This internalized transfer process requires multinationals to become involved in management training and establish internal knowledge transfer systems (Meyer, 2001), which increases the transaction costs incurred by MNEs.

To sum up, the current literature has shown significant effects produced by institutional environments on national as well as multinational firms’ strategies and their performance. However, although several previous studies on the effect of the institutional environment on firm performance take into account the formal institutions and law enforcement systems, the informal institutions, such as cultures and norms, seem to have been forgotten. The vital role of informal institutions in determining the behaviour and performance of an economic organization has only been identified by a few economists. For example, Adler (1986) shows that national culture has a great impact on employees. Puffer (1993) and House et al. (1999) emphasize the influence of national culture on managerial decision-making, leadership style and human resource management practices and firm performance. Societal culture can also be seen as a part of a firm’s resources, leading to a competitive advantage (Porter, 1991; Dunning and Bansal, 1997). Therefore, it is necessary to take into account informal institutions together with formal institutions and law enforcement systems in studying multinational firm performance. This thesis also makes such a contribution to the current literature.

4_{A specific-asset relationship cannot be redeployed to another transaction without some sacrifice of the productivity of} the asset or some cost in adapting the asset to the new transaction (Besanko et al., 2003: 129)

(16)

2.2. Hypotheses

2.2.1. Production factors

The increasing number of MNEs has created a competitive environment for firms, especially for MNEs who mainly operate in the global market. To be successful in such a market, the resource-based view argues that firms have to generate and sustain their own competitive advantages (Besanko et al., 2004). Technology and innovation are the most important sources for creating and sustaining competitive advantages. First, the role of innovation in creating competitive advantages was discovered very early in a study by Schumpeter (1950), who points out that a firm can only renew the value of its asset endowments through innovation. Scholars have emphasized innovation as a crucial mechanism through which firms generate competitive advantages in order to secure a position in the marketplace (Ven and Andrew, 1986). In addition, according to McGrath et al. (1996), innovation is also a mechanism through which a firm gains superior insight into, and access to, firm-specific resources with positive future value. Second, in addition to generating competitive advantages, innovation and new technologies are also able to act as crucial mechanisms allowing firms to sustain their competitive advantages. For instance, through implementing innovations, firms can establish ‘flows’ of resources, which can lead to ‘stocks’ of specific assets that others will find impossible to replicate quickly (Dierickx and Cool, 1989). Recently, D’Aveni (1994) stated that competitive advantages are eventually eroded and economic profitability declines, and that firms can only sustain positive economic profits by continually developing new sources of advantages (innovation). In conclusion, innovation and technology development would be important determinants of multinational firm performance.

(17)

that investing in the US, a country with high technology capabilities, does increase Japanese firms’ propensity to patent. Consequently, it is argued that the technology capabilities of home and host countries would positively affect multinational firms’ innovation capabilities, and that these enable them to create and sustain competitive advantages and thus improve their performance. Therefore, we propose the following hypothesis:

Hypothesis 1a: Country-level technology capabilities are positively associated with multinational

firm performance.

The quality of transport, telecommunications and information infrastructure, as well as energy availability in countries where an MNE operates could be a vital factor determining an MNE’s productivity (Mody and Srinivasan, 1998). In fact, high-quality infrastructure creates opportunities for interaction among firms and customers. It enables firms to reduce the cost of obtaining inputs from suppliers and shipping finished goods to customers. In addition, multinational firms that are located in highly accessible areas are exposed to severe competition and thus are forced to improve productivity (Deichmann et al., 2002). Moreover, an outstanding infrastructure would encourage inter-support between units within a multinational firm in terms of such elements as technology, management skills and inputs. The role of infrastructure in determining multinational firm performance has been found in numerous previous studies. For instance, Wheeler and Mody (1992) emphasize country-level infrastructure as a crucial determinant of the location choice of multinationals. More recently, Kumar (2001) suggested a significant sensitivity of MNEs to infrastructure availability for locating their investments designed to feed the global, regional or home country markets. Therefore, we propose the following hypothesis:

Hypothesis 1b: The quality of infrastructure is positively associated with multinational firm

performance.

The financial resource abundance (financial market efficiency) of countries where a multinational firm operates could be beneficial to the multinational. In fact, due to inflation, exchange rate variability and high capital transfer costs resulting from transferring capital between units in different countries, MNEs tend to employ local financial resources in order to reduce capital costs. Moreover, foreign affiliates extensively access financial markets in their host countries for the purpose of utilizing the tax deductibility of interest expenses and as a natural hedge against fluctuations in future local currency earnings (Feldstein, 1994). In addition, using local financial resources by selling shares to local employees at priority prices may effectively motivate them as they have the incentive of improved firm performance. Hence, the dependence of MNEs on local financial resources is relevant, and as a result

(18)

their performance is highly dependent on the efficiency of local financial markets. The reason for this is that an efficient financial market would reduce capital costs resulting from such elements as uncertainties and transaction costs. Therefore, we propose the following hypothesis:

Hypothesis 1c: The efficiency of a financial market is positively associated with multinational firm

performance.

2.2.2. Institutions

Formal institutions

The investment promotion policies of countries where a multinational operates may have a great impact on its location choice and performance. In fact, the benefits that multinational firms bring to home and host countries are enormous. For home countries, they improve their own and national technology capabilities as well as the skill of the national labour force by transferring new technologies from host to home countries (Navaretti and Venables, 2004). For host countries, multinational firms generate employment, increase demand for domestic intermediates, transfer technology to local firms and improve the skill of the labour force, all of which could work as catalysts for the development of domestic industries (Navaretti and Venables, 2004). Therefore, most countries have attempted to design investment promotion policies to attract the investment of multinationals. UNCTAD (1996) classifies the incentive policies into three categories: fiscal incentives, financial incentives and other incentives. Fiscal incentives include various types of direct and indirect subsidies or tax relief. The most common financial incentives are government subsidies aimed at lowering the capital, production, or credit costs of MNEs. Other incentives comprise a wide range of measures that could give advantages to multinational firms. These incentives are undoubtedly beneficial to MNEs and therefore, it is argued that investment promotion policies would give multinationals competitive advantages and positively affect their profits.

Hypothesis 2a: Investment promotion policies are positively associated with multinational firm

performance.

In addition to investment promotion policies, international trade promotion policies are considered an effective mechanism for countries to attract the investment of multinational firms. Firms may choose to supply a market through exports (national firms) or by setting up local production (multinational firms). If a country imposes high international trade barriers resulting in high trade costs, firms favour the severing of contact with this country’s market, through local production (becoming multinationals) rather than exporting to this country (remaining national firms). In

,

(19)

contrast, in a country which heavily promotes international trade and imposes low international trade barriers, the imported goods from foreign countries, which do not have to incur high trade costs, are able to compete with the goods produced locally by domestic firms or MNEs. Thus, the market share of domestic firms or MNEs in such a market would be reduced by the imported goods. Hence, it is argued that high trade promotion (or low trade barriers) brings many disadvantages to MNEs. The role of high local trade barriers in generating competitive advantages for multinationals over national firms is clearly proven by the model of Markusen and Venables (1998) which deals with the choice between supplying the particular local market through local production (multinational firms) and exporting to the local market (national firms). The model shows that multinationals are preferred over national firms in general as transport and tariff costs are high. Therefore, we propose the following hypothesis:

Hypothesis 2b: International trade promotion policies are negatively associated with multinational

firm performance.

(20)

face difficulties in firing inefficient employees or incur high firing costs, thus negatively affecting their performance (Addison et al., 2000). Consequently, it is argued that multinational firms operating in countries with a rigid labour market would have certain disadvantages compared to those operating in countries with a flexible labour market. Therefore, we propose the following hypothesis:

Hypothesis 2c: The flexibility level of labour regulations is positively associated with multinational

firm performance.

Law enforcement system

According to the resource-based view, to be successful firms not only have to create their own competitive advantages but also sustain them. When operating in several countries, multinationals face a high risk of losing competitive advantages resulting from imitation by local as well as international rivals. This problem is exacerbated, particularly in developing countries, where intellectual property rights and domestic law enforcement mechanisms are lax (Benko, 1987). Thus, it could be argued that weak intellectual property rights and law enforcement mechanisms have negative effects on a multinational firm’s performance for the following reasons: first, the profit of MNEs is negatively affected by losing competitive advantages to rivals. Second, MNEs incur high transaction costs, which may result from the time and expense of negotiating, writing and enforcing detailed contracts for transactions with local agencies (Besanko et al., 2003). Such contracts are required to prevent the loss of competitive advantages through imitation by these agencies and due to the lack of appropriate means to retaliate in countries with weak intellectual property rights and law enforcement mechanisms. Third, in such countries, multinationals are likely to transfer only old technologies (Lee and Mansfield, 1996), thus their performance would not be high. Fourth, in markets with an inefficient intellectual property rights and enforcement system, MNEs choose to internalize most of their production activities, such as input and distribution, to prevent the leakage of private information, even if outsourcing to local agents would be more effective (Besanko et al., 2003). Therefore, we propose the following hypothesis:

Hypothesis 3: The efficiency level of the law enforcement system is positively associated with

multinational firm performance.

Informal institutions

Individualism refers to the ‘relationship between the individual and the collectivity which prevails in a

(21)

among members, that is, everyone looks after his or her own interests and those of the immediate family. In contrast, societies with low levels of individualism share group values and beliefs and seek collective interests (Hofstede 1980: 222). Multinationals tend to be in favour of operating in individualist rather than collectivist societies for the following reasons: first, in collectivist societies the relationship between employee and employer is seen in moral terms, thus the poor performance of an employee gives no reason for dismissal (Hofstede, 2001:237). Conversely, in individualist societies the relationship between employee and employer is conceived of as a business transaction, a calculative relationship between actors on a ‘labour market’, and thus poor performance by the employee or a better pay offer from another employer are legitimate and socially acceptable reasons for terminating a work relationship (Hofstede, 2001: 237). Second, individualism presents three beliefs − in freedom, outward

orientation and the importance of contact with senior managers (Hofstede, 1980) − that have been found

to encourage innovation (Shane, 1993). The role of freedom granted to managers in taking the actions they see as most worthwhile has been found to be important for successful innovation in organizations in the United States (Kanter, 1982) and Japan (Jolly and Kayama, 1990). Outward orientation is important to innovation since contact with outsiders stimulates creativity (Feldman, 1988; Nonaka, 1990). Innovation requires the support and interest of senior managers (Fast and Pratt 1981; Maidique and Hayes, 1984). Therefore, we propose the following hypothesis:

Hypothesis 4a: Individualism is positively associated with multinational firm performance.

Masculinity refers to the distribution of the roles between genders. In the masculine countries, women

are somewhat assertive and competitive, but not as much as men, so that these countries show a gap between men’s values and women’s values. In contrast, women in feminine countries have the same modest, caring values as men (Hofstede, 2001: 297). Low-level masculine societies emphasize people, the quality of life, helping others, preserving the environment and not drawing attention to oneself (Hofstede 1980: 261−277).6_{Due to these characteristics, high-level feminine (or low-level} masculine) societies would grant firms more competitive advantages, compared to high-level masculine societies. First, low-level masculinity, as expressed by managerial attention paid to people and their interrelations (Nakata and Sivakumar, 1996), would be positively related to firm performance. Thwaites (1992), for example, finds that people-related aspects such as trust, good communication, team spirit and low conflict are associated with superior performance. Similarly,

6_{Individualism versus Masculinity: Individualism/collectivism is about ‘I’ versus ‘We’, independence from versus} dependence on in-groups. Relationships in collectivist cultures are basically predetermined by group ties. By contrast, masculinity/femininity is about ego enhancement versus relationship enhancement, regardless of group ties (Hofstede,

(22)

Barczak and Wilemon (1992) have found that, among leaders of a new product team, a combination of human relations and technical skills affects outcomes more positively than do technical skills alone. Second, another dimension of low-level masculinity that possibly fosters firm activities is the creation of a supportive climate (Nakata and Sivakumar, 1996). For example, in a study of innovation, Bums and Stalker (1961) point out that supportive internal environments, characterized by high degrees of inter-support among their members, foster firms’ innovation performance. Similarly, a study by Chakrabarti and Rubenstein (1976) examining NASA’s (National Aeronautics and Space Administration) technology adoptions and utilizations identifies organizational climate as being positively and significantly correlated with successful projects. In particular, companies characterized by ‘warm’ climates (that is, personal and supportive) experience greater success. Consequently, it is expected that multinational firms favour operating in low-level masculine societies.

Hypothesis 4b: Masculinity is negatively associated with multinational firm performance.

Uncertainty avoidance refers to how societies deal with unknown aspects of the future. On the

one hand, in low uncertainty avoidance societies such as the US, the UK and Canada, people work to meet basic needs, are tolerant of different behaviour and feel relatively secure (Hofstede 1980: 165). On the other hand, in high uncertainty avoidance societies, people are anxious about the future, actively avoid risk and are continually devising means, such as technologies, laws and social plans, to create a sense of control. Examples of countries with high uncertainty avoidance are France, Belgium, Greece and Portugal (Hofstede 1980: 165). Low uncertainty avoidance societies would give multinational firms more competitive advantages compared to high uncertainty avoidance countries. First, uncertainty acceptance is associated with firms’ capabilities in innovation and product development (Nakata and Sivakumar, 1996). Managers and employees in low uncertainty avoidance countries are less likely to resist changes and more in favour of new ideas (Philips and Wright, 1977). The managers tend to persuade other members of the organization to provide support for innovation. They also allow the innovators autonomy from the rules, procedures and systems of organization, so that they can establish creative solutions to existing problems. Second, consumers in low uncertainty avoidance countries are likely to be more open-minded when searching for information and in accessing innovation (Hofstede, 2001: 170). Since MNEs’ competitive advantages are very much based on innovation, the open-mindedness of consumers fosters the development of multinationals. Hence, it is expected that multinationals operating in low uncertainty avoidance societies have better innovation capabilities

(23)

and thus achieve better performance than those operating in high uncertainty avoidance societies. Therefore, we propose the following hypothesis:

Hypothesis 4c: Uncertainty avoidance is negatively associated with multinational firm performance.

3. Research method

3.1. Panel data models

This study employs the so-called panel data method which refers to the pooling of observations on a cross section of firms over several time periods. Panel data is able to overcome the shortcomings of cross-sectional data often used in multinational firm performance studies. Previous studies on firm performance, especially multinational firm performance, have employed different approaches and found numerous determinants of firm performance. Economists taking an industry structure view (associated with Porter, 1980) argue that firm performance is primarily a function of a firm’s membership of an industry with favourable structural characteristics (for example, relative bargaining power, barriers to entry, etc.), thus they focus on the industry as the relevant unit of analysis. By contrast, the resource-based view suggests that differential firm performance is caused by firm heterogeneity rather than industry structure (Barney, 1991). The relational view argues that firm performance is determined by inter-organizational competitive advantages resulting from dyad/network routines or relationships with other firms. In addition, firm performance is also affected by management capability, ownership structure (Davies, et al., 2005) and institutional factors such as labour regulations (Addison and Belfeild, 2001), law and law enforcement institutions (Chhibber and Majumdar, 1999). Hence, it is argued that multinational firm performance is affected by several factors, which can be classified into three groups: time-invariant, individual-invariant and individual time-varying factors. Time-invariant factors (individual heterogeneity) are those which remain the same over time but vary across firms, such as management capability, location, culture, education of employees, gender and socioeconomic background. The individual-invariant factors (time-specific heterogeneity) are factors which vary over time but are the same across firms, such as prices, interest rates and wide spread optimism or pessimism. The individual time-varying factors are those that vary across individuals and time, such as firm profits, sales, capital stock, internationalization level and production resources.

An individual study of multinational firm performance is not able to take into account all of these factors. Moreover, several of them are not easily observed or measured, such as management capability,

(24)

socioeconomic background and the strategic position of a firm in the market. As a result, using either cross-sectional or time-series data in studies on multinational firm performance would lead to the so-called omitted-variable problem, which results in biased least square estimators, as the omitted variables are correlated with the included explanatory variables (Hill et al., 2001: 185). The use of panel data is able to provide a means of resolving or reducing the magnitude of this problem. By using information on both the intertemporal dynamics and the individuality of the entities under investigation, that is, multinationals, this study is able to control, more naturally, for the effects of omitted or unobserved time-invariant and individual-invariant variables (Hsiao, 1986: 3).

The two models widely used in panel data analysis are fixed effects and random effects models. Each of them has different strengths and shortcomings. The choice of the appropriate model is primarily based on the characteristics of the data. The following section introduces the two models and explains the choice of one or the other models.

3.1.1. Fixed-Effects Model (FE): Least Squares Dummy Variable Approach

The most restrictive model used in panel data analysis is a pooled OLS model with N*T observations.7_{Like the OLS model used in cross-sectional analysis, this model specifies constant} coefficients (intercept and slopes). However, if individual and period-specific heterogeneities are present, OLS estimates would be biased if the omitted variables are correlated with the included explanatory variables (Cameron and Trivedi, 2005: 699). In panel data analysis, this problem can be solved by introducing dummy variables for every multinational and time period to account for the effects of those omitted variables representing individual and periodic heterogeneities (Meurs, 1991). This approach is called the ‘fixed effects model’.

it t i it it x y =α+β +δ +γ +ε i = 1, ….., N t = 1,…..,T (1) E( it) = 0 Var( t) = 2

E(Xit it) = 0 E( it i,t-s) = 0 (s 0) where,

yit = the performance measures of multinational firm i at time t;

xit = a K x 1 vector of individual time-varying dependent variables: national technology capability, infrastructure quality, financial market efficiency, investment promotion policy, international trade promotion policy, labour regulation, enforcement system efficiency, revenue and sales growth. i indexes the individual (multinational firm) in a cross section and t indexes time.

(25)

The parameters i and t represent cross-sectional individual or period-specific effects which capture individual and time-specific heterogeneities. The econometrics packages specify a constant, , and [(N-1) + (T-1)] dummy variables to capture the individual and period-specific heterogeneities ( i and t). The idiosyncratic residual, it, represents the effects of all remaining omitted variables that vary across individuals and time periods.

Like the OLS model, the FE model is able to give unbiased results only if the assumptions of no autocorrelation, no heteroscedasticity, no correlation between the included explanatory variables, xit, and the error term, it, and the normality of the residuals all hold. First, the autocorrelation problem, which arises if the residuals are correlated with one another, can be detected by using the Durbin-Watson test or Lagrange Muliplier (LM) test. These procedures test the null hypothesis of no autocorrelation against the alternative hypothesis of autocorrelation. If there is autocorrelation, the problem can be overcome by introducing a first-order autoregressive errors term – AR(1) – into the model (Reiman and Hill, 2001: 127−128). Second, the heteroscedasticity problem, which arises

(26)

assumption can be relaxed in the FE models, as all of the other assumptions hold and the sample size, with 1,030 observations, is sufficiently large (Hill et al., 2001: 80).

To sum up, the main advantage of the FE model is that it allows for correlations between the omitted individual and period-specific variables with the included explanatory variables. However, the fixed-effect model still has a number of shortcomings. First, the estimation of coefficients of time-invariant regressors, particularly IDV, MAS and UAI in this thesis, is impossible because they are absorbed into the individual-specific effect, i (Cameron and Trivedi, 2005: 715). In other words, the time-invariant variables and the individual dummy variables are perfectly collinear. Second, by introducing dummy variables for every individual and period, the degree of freedom decreases by the number of dummy variables, and thus the power of the statistical tests would decrease. These shortcomings may disappear in the random-effects model.

3.1.2. Random- Effects Model (RE): an Error Component Model

In the FE model, the effects of omitted individual and period-specific variables ( i and t) are treated as fixed constants over time and individual, respectively. In the RE model, these variables are treated, like uit, as random variables (Hsiao, 1986: 33).

it it it x u y =α+β + i = 1, ….., N t = 1,…..,T (2) it t i it u =δ +γ +ε E( i) = 0 Var( i) = 2 E( t) = 0 Var( t) = 2 E( it) = 0 Var( t) = 2

E(xit it) = 0 E( it i,t-s) = 0 (s 0) and,

y2 = u2 = 2 + 2 + 2 E( i t) = E( i it) = E( t it) = 0 E(xit i) = E(xit t) = 0

where, xit includes not only individual time-varying variables but also time-invariant variables: IDV, MAS and UAI.

To obtain better estimates the generalized least squares estimator (GLS), which applies to a transformed model with appropriately transformed error term, is employed. GLS yields standard errors that are appropriate for interval estimation and hypothesis testing (Hill, 2001: 360). The GLS model can be obtained as the OLS of ∗

(27)

.. 3 . 2 . 1x x x x xit∗ = it −θ i −θ t+θ Where, 2 1 σε λ = , 2 2 2 σδ σε λ = T + , 2 2 3 σγ σε λ = N + , 2 2 2 4 σδ σγ σε λ =T +N + ) / ( 1 1/2 2 1 σ λ θ = − ε , θ2 =1−(σε /λ13/2), θ3 =θ1+θ2+(σv/λ14/2)−1 = − = T t it i T y y 1 1 . , = − = N n it t N y y 1 1 . , = = − = N n T t it y NT y 1 1 1 .. ( ) = − = T t it i T x x 1 1 . , = − = N n it t N x x 1 1 . , = = − = N n T t it x NT x 1 1 1 .. ( )

Like the OLS model and the FE model, the RE model is able to give the unbiased results only if the assumptions of no autocorrelation, no heteroscedasticity, no correlation between the included explanatory variables, xit, and the error term, it, and the normality of the residuals all hold. The statistical techniques and procedures employed to test the assumptions of no autocorrelation, no heteroscedasticity, normality of residuals, and to correct the problems in the RE model are the same as those in the FE model. However, the error term, uit, contains the effect of omitted individual and period-specific variables ( i and t). Thus, the RE model is not able to control for the omitted and unobserved individual and period-specific effects. As a result, there is more likely to be a correlation between the error term and the included explanatory variables in the RE model than the FE model. Therefore, the assumption of no correlation between the error term and the included explanatory variables, which is ignored in the FE model, has to be tested in the RE model. The ‘Hausman Test via Artificial Regression’ is employed to test this assumption.

To sum up, the main disadvantage of the RE model relative to the FE model is that it does not allow for the correlations between the omitted individual and period-specific variables ( i and t) and the included explanatory variables, xit. In other words, the RE model is biased if the correlations exist. However, compared to the FE model, the RE model has several advantages. First, without introducing dummy variables for every individual and time period, the RE model is able to save the degrees of freedom, hence maintain the power of statistical tests. Second, without introducing the dummy variables for every individual (MNE), the RE model does not suffer the multicollinearity between these dummy variables and the time-invariant variables, particularly IDV, MAS and UAI. Thus, the effects of these time-invariant variables on the dependent variables can be estimated in the RE model.

The choice between the FE and RE models is based on the characteristics of the data. If the omitted individual and period-specific variables ( i and t) are not correlated with xit, the estimates of the

(28)

-RE model would not be statistically different from those of the FE model. In this case, the -RE model would be preferred since it is able to save the degrees of freedom and estimate the effects of time-invariant variables. By contrast, if the RE model is biased due to the correlations of i and/or t with xit, the FE should be chosen. The F-test will be employed to test for the choice between the FE and RE models. This procedure tests the null hypothesis that the RE model would be consistent and efficient against the alternative hypothesis that the RE model would be inconsistent and the FE model thus preferred.

3.1.3. Two-step procedure

As mentioned above, the time-invariant variables, IDV, MAS and UAI, cannot be included in the FE model. Hence, if the FE model is preferred over the RE model, a well-known model proposed by Hsiao (1986: 50) will be employed involving a two-step procedure. The two-step procedure combines the FE model (first step) and the OLS model (second step).

it i it it x z u y =α+β +η + i = 1, ….., N t = 1,…..,T (3) it t i it u =δ +γ +ε where,

xit = a K x 1 vector of individual time-varying explanatory variables;

zi = a P x 1 vector of time-invariant explanatory variables: IDV, MAS and UAI

First step: is obtained by applying the original FE model (1)

Second step: , can be estimated by applying the following OLS model:

i i i i x z u y −β =α+η + i = 1, ….., N (4) E(u ) = 0 i var(u ) = i 2 cov (u , ) = 0 i s i us cov(z, u ) = 0 where, T y y T t i i = =1 , _T x x T t i i = =1

The second step employs the OLS model to estimate the effect of the time-invariant variables, particularly IDV, MAS and UAI, on the dependent variable, multinational firm performance. This model regresses the residuals of the original FE model (first step), containing the time-invariant (or individual) specific effects, on the time-invariant explanatory variables. The OLS estimator is unbiased only if the assumptions of no autocorrelation, no heteroscedasticity, no correlation between included explanatory variables, zi, and the residuals,u , and the normality of residuals all

(29)

hold. The statistical techniques and procedures which are employed to test the assumptions and correct the problems in the RE model are also applied in this model.

It is noted that the two-step procedure is consistent only if the sample size tends to infinity (Hsiao, 1986: 51). Since the number of observations in this study is relative large, applying the two-step procedure in this thesis would be appropriate and provide unbiased outcomes.

The two-step procedure not only has the advantage of the original FE model in allowing for the correlations between the explanatory variables and the omitted individual and time-specific variables, but is also able to measure the effect of the time-invariant variables (Plumper and Troeger, 2005). However, the two-step procedure still suffers from a relevant disadvantage found in the original FE model, the loss of degrees of freedom. Therefore, the two-step procedure is only appropriate if the RE model is biased. In other words, if the F-test, used to test for the choice between the FE and the RE models, suggests that the RE model, which is able to save degrees of freedom and to estimate time-invariant variables, is unbiased, it would be preferred over the FE model. In this case, we would not need to use the two-step procedure as the RE model would be preferred. By contrast, if the F-test suggests that the RE model is biased, the two-step procedure model will be chosen.

3.2. Data collection, measures and description

The data used in this study is derived from the Amadeus Database, the Global Competitiveness Reports, the World Bank and Hofstede’s database on cultural dimensions. The firm-level data, collected from the Amadeus Database, consists of aspects of the annual financial reports, such as returns on assets (ROA), earnings before interest and taxes divided by assets (EBITOA), revenue and sales growth, the number and location of the subsidiaries of 206 large European MNEs in high knowledge-intensive industries, such as petroleum refining, metal, electrical equipment, industrial machinery, transportation, post and telecommunications, during the 2000−2004 period. Country-level

data are derived from the other sources. Country-level technology capability, infrastructure quality, financial market efficiency and government policies seem to change over time, so the indices calculated by several different organizations to measure these variables also vary across time. Therefore, it is judicious to assume these variables to be individual time-varying, and as a result the panel data of 150 countries over five years (2000−2004) was collected for these variables. By

(30)

IDV, MAS and UAI, to be time-invariant. Thus, cross-sectional data from 150 countries was collected for these variables.

The dependent variable is multinational firm performance, which is measured in two ways: (i) return on assets (ROA), and (ii) earnings before interest and taxes divided by assets (EBITOA). ROA is widely used to measure the performance of firms. However, since taxation rules as well as capital structure seem to vary across countries, we also take into account EBITOA, which is a useful measure of the performance of multinational firms coming from and operating in different countries. This study constructs two sets of equations: an equation which measures multinational firm performance in terms of ROA and an equation which measures multinational firm performance in terms of EBITOA. Table 1 shows that the value of EBITOA is on average 6.59 percent, while that of ROA is on average only 3.79 percent.

Independent variables: as mentioned, one of the most interesting contributions of this study is that it examines how the environmental munificence of all countries (home and host countries) in which an MNE operates jointly impact on the MNEs’ performance. Therefore, the joint environmental munificence indices have to be constructed to measure the joint environmental munificence variables. In their study of the effect of institutional quality on FDI, Daude and Stein (2004) construct two institutional indices, government efficiency and political stability, each of which absorbs the joint effects of several institutional dimensions. The government efficiency index, which absorbs the joint effect of rule of law, control of corruption, government effectiveness and regulatory quality dimensions, is measured by the average of the indices of all of these institutional dimensions.8_{Similarly, Tridico (2006) constructs six governance indices, each of which comprises} several dimensions. For instance, the extra-cost corruption index, which represents the joint effect of the cost of the creation of new market institutions, laws, governmental agencies, competition authorities and offices for consumer protection, is measured by the average of the indices of all of these dimensions. In this study, the joint environmental munificence indices are constructed based on this method.

Technology Capability refers to those factors which facilitate and enable the technological capacity

of a country. This includes the general availability of technologies and the penetration rate of information and communication technologies, as these tools are seen as critical indicators of the overall technological readiness of a country (Lopez-Claros, 2005). The joint technology capability index (TCit) at time t, which represents the joint country-level technology capability of all countries

(31)

j (home and all host countries) in which MNEi operates at time t, is calculated as the weighted average of the country-level technology capability indices (

j i ijt

ij_TC

n n

) of all of the countries j in which MNEi operates at time t. The weight is the ratio of the number of subsidiaries of MNEi in country j (nij) to its total number of subsidiaries (ni) – (see the formula in Table 2). The country-level technology capability indices (TCijt) are measured by the national technology readiness indices of the Global Competitiveness Reports, which range from 1 to 7. The higher the national technology readiness index of a country, the better its technological capability. As seen from Table 1, the values of the joint technology capability index (TCit) range from 3.25 to 6.25, with an average value of 4.87.

It is noted that the investment level of an MNE varies across economies and is based on their size. In particular, an MNE may invest more in large economies than in small. Hence, the technology capability indices of large economies have to be adequately weighted with respect to small economies in the weighted average of the country-level technology capability indices. Drawing from this argument, the share of the investment of MNEi in country j in relation to its total investment should be taken into account as the weight of the technology capability index of country j in the weighted average of the country-level technology capability indices of all countries j in which MNEi operates. However, this method cannot be applied in this thesis since the information on the country (or subsidiary) segment investments of MNEs is not fully available. There are two alternative methods. The first is that the share of the sales of MNEi in country j in relation to its total sales is used as the weighted measure. The drawback of this method is that information on subsidiary-level sales is not fully available. Furthermore, the share of the sales of MNEi in country j, which receives a small investment from MNEi, may be large relative to the share of the investment of MNEi in country j in terms of its total investment if the subsidiaries of MNEi within country j import a large number of resources and finished goods from other subsidiaries within MNEi operating in other countries. In this case, the share of the sales does not fully reflect the share of the investment.

(32)

country group using the number of subsidiaries in the advanced-country group and the developing-country group, respectively. With this measure, the author also ignores the differences in size between subsidiaries, treating all subsidiaries equally. In addition, since the information on the country (or subsidiary) segment investments of MNEs is not fully available, many researchers also use the total number of foreign subsidiaries of an MNE as a proxy for the level of its multinationality (Barkema and Vermeulen, 1998; Wan and Hoskisson, 2003; Berry, 2006). This measure also ignores the difference in size between subsidiaries. Following these studies, this thesis takes into account the ratio of the number of subsidiaries of MNEi in country j (nij) to its total number of subsidiaries (ni) as the weight of the technology capability index of country j in calculating the joint technology capability index (TCit). This weight is also applied in calculating the other joint environmental munificence indices described below (see the formulas in Table 2).

Infrastructure refers to the quality of the country-level overall infrastructure. It includes several aspects

of infrastructure, for instance, transportation facilities such as road network, ports and airports, communication infrastructure covering the telecommunication network, information infrastructure and energy availability, (Kumar, 2001). The joint infrastructure index (IFit) at time t, which represents the joint country-level infrastructure quality of all countries j in which MNEi operates at time t, is calculated as the weighted average of the country-level infrastructure indices (

j i ijt

ij _IF

n n

) of all of the countries j in which MNEi operates at time t (see the formula in Table 2). The country-level infrastructure indices (IFijt) are measured by the national infrastructure indices of the Global Competitiveness Reports, which range from 1 to 7. The higher the national infrastructure index of a country, the better its overall infrastructure. The maximum, minimum and average values of the joint infrastructure index (IFit), which are presented in Table 1, are 1.20, 6.80, and 3.69, respectively.

Financial Market Efficiency refers to the efficiency of the country-level financial market. It is defined

as the availability of capital for business investment from resources such as credit from a sound banking sector, well-functioning equity markets, or venture capital (Lopez-Claros, 2005). The joint financial market efficiency index (FMit) at time t, which represents the joint financial market efficiency level of all countries j in which MNEi operates at time t, is calculated as the weighted average of the country-level financial market efficiency indices (

j ijt i ij FM n n

The Impact of Environmental Munificence on the Performance of Multinational Enterprises

Master’s thesis