UNIVERSITY OF GRONINGEN
FACULTY OF ECONOMICS & BUSINESS
MASTER’S THESIS IE&B
Topic: The impacts of national cultural fit and structural compatibility on the performance of the combined firm: evidences from USA organizations.
Thesis supervisor : Prof. Gjalt de Jong
Student's name : Lien Mai Pham (s2076365)
Email : l.pham@student.rug.nl
Acknowledgements
I would like to express my sincere thanks to the people who have supported me enormously in completing this thesis. First and foremost, I would like to thank Prof. Gjalt de Jong for his useful instructions and orientation which help to keep me on the right track regarding the scope of this study; and for his faith in the feasibility of this thesis, which inspired me to undertake the study.
Secondly, I wish to thank Prof. dr. Erik Dietzenbacher for his useful recommendation on the research methodology of the thesis. Also, many thanks to Mr. Dut Vo and Ms. Trang Doan for their guidance and helpful comments.
Moreover, I would like to thank the University of Groningen and the Huygens Scholarship Program for giving me the chance and financial support to study in The Netherlands.
Additionally, I would like to thank my dear friend and sister, Loan, for spending her time to help me proofread for my language limitation.
Abstract
This study examines the impact of national cultural fit and structural compatibility on the performance of the combined firm, which is merger and acquisition (M&A) outcomes. We also explore the interrelationship between national cultural fit and the compatibility of firm’s age or size concerning the two combining firms. Using OLS multiple regressions with interaction variables, we analyse the sample which consists of 88 cross-border M&A deals of 23 countries in which USA is the default country. Our results show that there is a significant relationship between firm’s performance and national cultural fit or structural compatibility but no association between the two factors has been reported.
Table of content
1. Introduction ... 1
2. Theoretical background and hypotheses ... 4
2.1. Effectiveness of cross-border mergers and acquisitions ... 4
2.1.1. Drivers for cross-border M&As ... 5
2.1.2. Post-merger integration: reason for success and failure. ... 8
2.2. Cultural fit in mergers and acquisitions ... 11
2.2.1. National culture ... 11
2.2.2. Hofstede’s national cultural dimensions ... 13
2.2.3. Other frameworks ... 15
2.2.4 National cultural fit ... 17
2.3. Structural compatibility. ... 18
2.4. National culture fit and structural compatibility ... 20
3. The models... 22
4. Methodology ... 23
4.1. Data and sample ... 23
4.2. The variables ... 25
5. Analysis and results ... 29
5.1. Description ... 29
5.2. Diagnostic check ... 30
5.3. Regression results and discussions ... 35
6. Conclusion ... 42
References ... 44
Appendix A ... 51
Appendix B ... 52
1 1. Introduction
Over the last decades, mergers and acquisitions (hence forth: M&As) have no longer been considered as a brand new method business collaboration introduced to businessmen and scholars in the era of globalization. The rationale of M&As is to accomplish synergies between integrating business units in a combination corporation that will gain competitive advantages, and thus higher profitability (Porter, 1985). Today, to account for the dominant share of Foreign Direct Investment (FDI) flows in the modern world (from 66.3% to 76.2% over the 1998-2001 period), M&As have contributed to financing new investment projects in developing countries, as well as the scarcity of take-over targets in these countries (Navaretti & Venables, 2004). Regrettably, regarding financial performance and social issues, many M&As “have not lived up to expectations” (Weber et al., 1996, p.2) with statistically estimated rate of failure at 30% (Bruner, 2002 & 2005) or 50-75% (Jansen, 2002). Concerning financial performance, about 40% of M&As deals are considered successful against original targets, set beforehand, within three-year corporation, regardless of financial or accountancy measures used (Cartwright & Cooper, 1993). Apart from finance, it is observed that traumatic problems between employees and managers occur frequently in combined firms, especially in human and cultural issues, often leading to unproductiveness and attitudinal conflicts (Schwieger & DeNisi, 1991; Hambrick & Cannella, 1993). Such social difficulties might add further costs to the integration process and weaken the capacity of firms to gain synergies, thereby offset expectations of higher benefit in a merger.
2 potential cause for M&A collapse because it creates problematic human issues between two hierarchies originated from different cultures with dissimilarities in languages, legal systems, regulatory obstacles and national cultural barriers (Weber & Schweiger, 1992; Hambrick & Cannella, 1993; Olie, 1994; Vaara, 2003; Aguilera & Dencker, 2004; Shimizu et al., 2004; Bundy & Hukins, 2009). Consequently, these differences cause clashes of organizational values, conflicts between managerial practices, contradictory ways of thinking among employees of the involving firms, as well as dearth of collaboration, missing opportunities and ineffectiveness of the combined incorporation.
However, empirical findings on the impact of cultural fit on the performance of the combined firm are mixed and inconsistent. In most recent studies, the concept of cultural fit has been implicitly defined via “cultural difference” or “cultural distance” because these terms reflect two levels of interaction between two specific cultures: supportive and destructive to the effectiveness of the combined firm. Some scholars report a positive effect of cultural distances on the outcomes of cross-border M&As (Weber et al., 1996; Larsson & Risberg, 1998; Morosini et al., 1998), while other conclude a negative one (Datta & Puia, 1995). Despite the contradictory findings, in general, most current research focuses only upon the impact of cultural differences on the survival of M&A deals, ignoring other factors which could play a significant role to both cultural issues and the effectiveness of performance during post-merger integration such as human resources, employee satisfaction, institutional structure and strategies, etc. (Czarniawska, 1999; Grant et al., 1998). The mixed and inconsistent findings of this complex phenomenon suggest further theoretical and empirical studies investigating unidentified variables rather than cultural fit concerning the performance of the combined firm.
3 Thus, the main research question that will be answered in the thesis is: What is the impact of national cultural fit and structural compatibility on the performance of the combined firms after cross-border M&A completed deals?
The study has derived the following sub- questions:
- What is national cultural fit and how can it be measured? What is the impact of national cultural fit on the combined firm’s performance?
- What is relative size and how can it be measured? What is the impact of relative size on the combined firm’s performance?
- What is age similarity and how can it be measured? What is the impact of age similarity on the combined firm’s performance?
- Does national cultural fit affect relative size between the partners in the performance of the combined firm?
- Does national cultural fit affect age similarity between the partners in the performance of the combined firm?
We use a cross-sectional study with a number of 88 completed M&A deals from 22 countries over the globe in the period 1998-2009 to examine the effect of national cultural fit, age similarity and relative size on the performance of the combined firm. In our dataset, the United States of America (USA) is chosen as a default country, which means that American is the nationality of the bidder or the target. The 88 deals refer to the deals that American companies engaged with 88 foreign firms all over the world. We also obtain our present data with common method of performance measurement in recent studies (see Zollo & Meier, 2008) to build up econometric model as well as attain econometric evidences on the impacts of the three aspects mentioned above on the degree of effectiveness.
4 2. Theoretical background and hypotheses
2.1. Effectiveness of cross-border mergers and acquisitions
As indicated in the Introduction, a M&A is a crucial channel supporting firms to gain higher profitability and expand their power in new markets with a wide range of customers. In a specific incorporation, the terms “merger” and “acquisition”, which are parts of M&A phrase, are used interchangeably but contain a few differences in meaning. Merger and acquisition, which currently regard as a part of “business combination” and “takeover” respectively involve a switch from one group of shareholders to another. A merger, which is synonymous with “consolidation”1, occurs when two firms agree to combine into a new single legal entity, in which the shareholders of the two firms are the joint owners. Apart from a merger, an acquisition proclaims in an arm’s length contract, in which one firm purchases either assets or shares of another, then ceasing the ownership of the target’s shareholders. As a result, the target normally becomes a subsidiary of the bidder. It is thus normally known that a takeover is similar to an acquisition, but Sudarsanam (1995, p.1) states that “the acquirer is much larger than the acquired”.
Introduced at the end of the nineteenth century in the USA, M&As have facilitated rapidly increasing levels of corporate business activities throughout the 1940s, 1950s, 1980s and up to the most recent two decades, in which the number of values and deals are vastly exceeding all of the preceding ones (Andrade et al.,2001). In the 1940s and 1950s, the USA was seen as a country, in which most of M&A deals initiated as at that time, the US stock market underwent a drastic development with the appearance of conglomerate2 (Owen, 2006). The number of conglomerate corporate fluctuated from “8.3% of Fortune 500 firms in 1959 to 18.7% in 1969” (Rumelt, 1974, p.7). The 1980s witnessed the turn of conglomerate in the previous period into horizontal takeovers in which firms chose to concentrate its activities on the most profitable and cost-effective field of business. During this decade, more than 50% of all major US companies were “the recipients of an unsolicited takeover bid” (Owen, 2006, p.8). Most of M&A activities focus on sectors of internet, telecommunications, healthcare, banking and semiconductors (Kleiner & Nguyen, 2003). Moreover, the most recent two decades has noticed the drastic
1
“Consolidation is known as the fusion of two existing companies into a new company in which both the existing companies extinguish. Thus, consolidation is mixing up of the two companies to make them into a new one in which both the existing companies lose their identity and cease to exist. In other words, all the assets, liabilities and stocks of the consolidating companies stand transferred to new company in consideration of payment in terms of equity shares or bonds or cash or combination of the two or all modes of payments in proper mix.” (Shrivastava et. al, 2007, p.4)
2
5 popularity and involvement of international elements in a new type cross-border M&A. In 1999, it was recorded that cross-border M&A deals totaled at approximately USD1.4 trillion, about 40% of the overall M&A value for that year and doubling the value of the previous year (Hitt et al., 2001a, b). According to the 2004 world’s investment report, published by the United Nations Commission on Trade and Development (UNCTAD), the annual growth of cross-border M&As in the globe reached 22.5% over the last decades. Also in that year, international M&As contributed larger values to firms and the world’s economy in which totaled up to approximately 30000 M&A deals and USD 1900 billion in value, equivalent to one completed deal every 18 minutes and exceeding the GDP of several developed countries (Cartwright & Schoenberg, 2006). With the more incredible expansion in number of deals and values, M&As in general and international M&As in particular are likely to prove their importance in helping companies to achieve strategies and financial objectives by combining cultures, corporate characteristics and value systems (Sudarsanam, 1995).
2.1.1. Drivers for cross-border M&As
6 Obviously, each manager of a specific firm has different reasons to engage in international combination, depending on their individual strategies and financial strength. Current and relevant literature offers comprehensive insights into diversified rationales behind cross-border M&As, but most of them are categorized into three main drivers: access to new markets, access to new strategic assets and finally, access to learning and learning ability.
Access to new markets
To search for other markets beyond domestic market is considered one of the most important strategies in any corporation. The view that international M&As are a new potential mode of entry into foreign markets has attracted a vast number of studies (Kogut & Singh, 1988; Andersen, 1997; Harzing, 2002). Despite a variety in a number of other methods to explore new markets such as Greenfield investment, setting up joint ventures or strategic alliances with local companies in the host country, M&As have been considered as a quicker vehicle to access a new market entrance (Wang& Boateng, 2007) because of the advantages of existing partners in market shares, customers and marketing researches in the local market.
Access to strategic assets
Due to differences in geography and country-level characteristics, some resources in production and services cannot be easily transferable across countries, contributing to major incentives of companies to seek out M&As in order to get control of these assets in faster and effective ways (Amit & Schoemaker, 1993). This argument has been supported that the bidder cannot only obtain existing organizational intangible assets (i.e. knowledge, reputation, technologies) of the target but also generate valuable resources during integration (Harrison et al., 2001; Meyer et al., 2009).
Access to learning capabilities
7 gain benefits from being attached in the routines of well-organized institution characterized by market competition and customer concentration (Gubbi et al., 2009).
Theoretically, most scholars focus on the perspectives of internalization theory and theory of industrial organization to explain the motives behind cross-border M&As. Another theory taken into account is the theory of revealed comparative advantage based on the classic view of comparative advantage long time ago. These theories have not indicated explicitly the motivations behind current waves of cross-border M&As but have been considered an essential base to understand the appearance of these international deals because cross-border M&As are a part of international investment and corporation.
Theory of internalization
Rooted by Coase (1937) and advanced by Dunning (1973), Hymer (1976), etc., the theory of internalization suggests that the rationale behind overseas investment of firms is the extensive exploitation of tangible as well as intangible firm-specific assets, and market opportunities, defined by high cost of transaction (Markides & Oyon, 2003). The firm-specific assets, such as technological and research capacities, human and managerial sources, brand names, etc., is based on proprietary information, and thus cannot be transferred easily across firms or nations. One way to overcome this difficulty is by internalizing the markets for such assets, thereby rising their values by abroad expansion. According to the theory, cross-border M&As are one of various vehicles to help acquiring firms successfully create more values if such companies tap into expertise and know-how in globalized markets.
Theory of industrial organization
8 Theory of revealed comparative advantage
Comparative advantage of a company in the production is a lower opportunity cost, at which the company can produce goods or provide services, than other companies in the same country. With respect to international corporation, companies from different countries in the world could gain benefit if they exchange their comparative advantages of their own specialization of production within alliances and corporation (Madura, 2006). The theory of comparative advantages has been essentially a foundation for trade researches to explain the motivation for any business activities in the long history of academic research on foreign trade (Neary, 2003). To further the basic theory of comparative advantages, scholars develop the concept of revealed comparative advantages to explain the motivation of foreign M&As (Brakman et al., 2010; Neary, 2003, 2007). Generally, revealed comparative advantage theory suggests that firms, which are from the industry that has a comparative advantage, can gain cost-benefit with lower marginal cost than ones from other industries, consequently producing and exporting at higher level than other firms. These firms are so-called strong companies in the economy. Particularly, for cross-border M&As, the appearance of revealed comparative advantages encourage strong firms in the world’s economy to attempt at other targets in order to make use of a wide range of resources all over the world. Strong firms (namely the bidder) could offer the targets more benefits in human resources or technologies than other weak firms in the economy, thus level up the increase of attractiveness of the offers as well as lessen the worries of the targets about the future of corporation. The target, hence, with the advantages they can attain from the bidder, can be motivated to engage in the M&A. In research so far, the theory has been widely adopted in studies regarding trade (Halpern, 1995) or competitiveness of particular industries and countries (Cooper, 2007; Havrila & Gunawardana, 2003).
2.1.2. Post-merger integration: reason for success and failure.
9 cultures, beliefs, values and assumptions within cross-border deals. That is a reason why the target and the bidder have to be mutually depended on each other if they seek for long-term success corporation in the future. This interdependence is determined by fitted targets, in which each side needs each other to fulfill their individual goals and by the outcomes of each individual party (Ritov, 1996).
In recent years, due to economic uncertainty background in the world’s economy, success in integration process is hardly achieved. The rate of failure in cross-border M&A deals is reported at high level, up to 70% in most of relevant academic studies (Boateng & Lodorfos, 2006). Many M&As “have not lived up to expectations” (Weber et al., 1996, p.2) with statistically estimated rate of failure at 30% (Bruner, 2002 & 2005) or 50-75% (Jansen, 2002). Moreover, about 40% of M&As deals are considered successful against original targets set for them within three-year corporation regardless of financial or accountancy measures used (Cartwright & Cooper, 1993).
To attempt at explaining the massive failure, the number of academic research has been switching interests into post-merger organizational integration process, which has been considered as crucial factors to understand organizational outcomes of M&As (Haspeslagh & Jemison, 1991; Larsson & Finkelstein, 1999). Literature about post-merger integration is predominant in managerial perspectives, however little attention has been paid to the involvement of social elements constructed in M&A integration (Czarniawska, 1999; Grant et al., 1998).
10 prioritize to accomplish cost efficiencies. Thus, the need to attain synergies within organizations and in a newly combined cooperation should be realized as quickly as possible after a merger or an acquisition.
To seek for answers regarding these questions from managerial perspectives, in terms of international M&As, scholars have listed various elements leading to the success or failure of cross-border corporation, but most of them focus on the relationship between potential opportunities resulting from the combination and certain reactions within the organizations. For example, “liability of foreignness”, which is considered a unique risk within organizations in terms of uncertainty and information asymmetry in foreign markets, hinders both the target and the bidders from organizational adjustment and learning capabilities (Kogut &Singh, 1988; Zaheer, 1995). Moreover, in a survey for CEO of cross-border M&A, 50% interviewees said that the success of international M&As, which is endangered by conflicts in management styles, can be improved by clear mutual communication in early stages of corporation (Merrill, 2009). Following this survey, other elements have contributed to challenges to cross-border mergers including conflicting cultural values (25%), communication difficulties between geographically separated companies (21%) and language barriers (4%).
The pursuit of successful cross-border M&As is also viewed from other perspectives with the involvement of social construction such as human issues, employee problems, social added values, etc. Regarding this perspective, scholars investigate human consequences of M&As and focus on organizational resistance as an essentially social but problematic force for integration and cooperation. However, a number of empirical studies in this field have received little attention. Some examine employees’ reaction after the merger (Napier, 1989; Cartwright & Cooper, 1990; Lohrum, 1996; Schweiger & DeNisi, 1991). Some concentrate on managers’ attitude (Hambrick & Canella, 1993; Greenwood et al., 1994). Thus, there is still a lack of knowledge concerning socially constructed perspectives, hindering the process of organizational integration (Greenwood et al., 1994).
11 organizational management studies (Datta, 1991; Chattetjee et al., 1992; Larsson, 1993; Cartwright & Cooper, 1993, Webber et al., 1996). Besides organizational culture, the national cultural distance plays an important role during post-acquisition integration (Very et al., 1997; Webber et al., 1996; Gertsen et al., 1998; Morosini et al., 1998), especially the process of acculturation (Nahavandi & Malekzadeh, 1988; Gertsen et al., 1998; Vaara, 1999) due to the fact that during the first two years after the merger was announced, cultural clashes approximately led to the loss of seven billion US dollars per year (Friedman, 1997). To explain the outcomes of international M&As, Barkema et al. (1996, p.4) proposes the problems of “double – layered acculturation”, defined as differences in national cultures, business practices, customer preferences and institutional barriers, such as government legal systems hampering companies to fully realize their strategic targets. To sum up, with reasons from either organizational or national culture, from the cultural perspectives, integration process is “cumbersome and long-lasting” (Vaara, 2002, p.6) and consequently affected by cultural incompatibility and conflicts.
From the theoretical perspectives above, regardless of motives or reasons for cross-border M&As, international integration is recognized in creating added values for corporations or, in other words, realizing synergism, in order to maintain the longevity of international combinations. With the involvement of such complex bundles of resources in financial strength, labour force, knowledge, cultures, languages, etc., the integration process, in which the role of organizational management and social influences could joint, needs to be paid more attention in further studies. The perspective of culture is currently dominant in explaining the integration process and reasons for the success or failure of cross-border M&As via the impact of cultural incompatibility. However, there are few studies investigating how distances in culture can affect both organizational management and social interactions in the combined firms, leading to the effectiveness or ineffectiveness of the performance of cross-border M&As. Thus, our research tries to fill up this gap, and focuses on the importance of cultural compatibility accompanied with organizational elements within companies, with respect to both management and social perspectives.
2.2. Cultural fit in mergers and acquisitions 2.2.1. National culture
12 classified into two categories: at national level and at organizational level. In the early phase of popularity in the 1980s, the term “organizational culture” was defined as the assumptions and values, which were commonly shared among members within an organization (Schein, 1985) or “commitment, satisfaction, productivity and longevity” of employees and employers within the firm (O’Reilly et al., 1991, p.8). Besides, national culture includes the sets of commitments, obligations and suggestions for every citizen within a particular country or, in other words, national culture is a “collective programming of the mind” (Hofstede, 1980).
The national culture with its long and stable history often generates the organizational culture. In terms of international collaboration, national culture makes the country unique with national characteristics, which are more apparently recognized by foreign investors than local people. These characteristics are also driving forces attracting international investment flows to local companies in particular and to the host country in general. Differences in the national culture is defined as the level to which the cultural norms behave in a different way to those in other countries (Kogut & Singh, 1988). In other words, it can be referred to routines and repertoires for organizational structure, new product development and other aspects of management that are found in the bidder's and the target's mother countries (Kogut & Singh, 1988). An organization, which is a part of a community, can be considered as a small society with separate and specific social structure containing sets of values, assumptions, norms and beliefs. In a broader context of internationalization, an organization represents the unique values of its own business and its national characteristics of its home country as well. Thus, it is where the role of national culture can be elaborated. Moreover, different organizational procedures, rules or behaviours resulted from different national cultures cannot easily be transferred within organizations, especially in the process of integration when both the targets and the bidders have to work together and become interdependent. As a result, understanding influences of national cultural differences on effectiveness of cross-border M&As has been considered one of the most strategic priorities for managers of acquiring firms in order to shape mutual corporate culture.
13 assumed, thus greater attention should be paid towards cultural distances in the context of international M&As, but first, basic national cultural differentiation needs to be clarified.
2.2.2. Hofstede’s national cultural dimensions
In this study, we are determined to follow the approach of Geert Hofstede, which broadens the terms of national culture. The model in a survey in 66 countries, 60000 responses to 116000 questionnaires, suggested by Hofstede (1980), includes four primary dimensions of culture contributing significantly to the understanding of basic cultural differences in various studies over the last twenty years. In this part of the thesis, the overview of Hostede’s work on the four dimensions of national culture, which are power distance, uncertainty avoidance, masculinity / femininity and individualism / collectivism, will be presented. Accordingly, I will explain the reason why I choose this approach as well as some limitations of Hofstede’s framework compared to other cultural measurements proposed later on.
To propose original four dimensions of culture, Hofstede (1980) worked on the survey collecting data obtained between 1967 and 1973 from about 117.000 IBM employees working in 66 different countries in the globe. His analyses were based on the responses of these employees, explaining inter-country variations related to demography, geography, economics and politics of societies (Soares et al., 2007). Hofstede named these four dimensions as power distance, uncertainty avoidance, individualism and masculinity which were ranged from 0 to 100. Another dimension added to the model is known as “Short-term versus long-term orientation” or “confuscian dynamism” in correspondence with Asian culture (Hofstede & Bond, 1988). The fifth dimension was created from the survey in over 23 countries around the world in which answers to a questionnaire designed by Chinese researchers and related to the teachings of Confucius3 had been analyzed. Long-term orientation is defined as values relative to perseverance and thrift, while short-term orientation deals with implementing social obligations, respecting and protecting traditional values of firms. Due to the relation with Confucius, which is a much lesser extent to Western cultures, I will not take this fifth dimension into account but the four original ones.
Power Distance Index (PDI) is defined as the degree of people’s beliefs and acceptance about inequity of distribution in power and positions or titles of employees and employers within the organization. In countries with higher power distance culture, individuals with high titles stand to benefit from significant power whilst employees who depend on these more powerful individuals
3
14 tend to accept such centralization of power because of their little participation in decision making process. In contrast, in countries with lower power distance culture, employees are more active in decision making process, meaning that they do not agree easily with the centralization power.
Uncertainty Avoidance Index (UAI) is considered as “extent to which the members of a culture feel threatened by uncertain or unknown situations” (Hofstede, 1991, p.113), or in other words, a degree to which people in a nation are likely to favour structured over unstructured circumstances. In countries with high level of uncertainty avoidance in terms of culture, individuals within organizations tend to do their tasks in a tradition way of thinking, refusing new technologies, not expectedly expressing their own emotions, thus resulting in suffering strict laws and security measures. On the contrary, in culture with low uncertainty avoidance, people are more adaptive and welcomed to the involvement of new technologies.
Individualism versus Collectivism (IDV) refers to the degree to which people within a country act as individuals rather than as a member of cohesive communities. Culture of individualism indicates a society in which individuals are likely to look after themselves and their family. Opposite to individualism, in collectivist culture, people are connected by incorporation from birth onwards into strong cohesive in-groups and loyalty protection. To sum up, people place the thought of “us” importantly in collectivism, whereas in individualism, it is the thought of “me”.
Masculinity versus Femininity (MAS) elaborates the degree to which values of masculinity such as assertiveness, effectiveness and success triumph over values of femininity such as standard of life, maintenance of good interpersonal relationships and gesture of solidarity. The studies about IBM revealed that there was a difference in values between women and men in the work place, showing the gap among gender within a country. Masculine countries contain assertiveness, competiveness and value the importance of material success. Otherwise, feminine countries value tenderness, modesty and quality of life.
15 computer manufacturing, with the data collected about 40 years ago. These data has been criticized as outdated and incompatible with the modern world’s economy today. For example, in Uncertainty Avoidance index, Singapore has the lowest score at 8. However, in the 1970s, the Singapore government carried out reforms in order to enhance economic growth and prevent potential risks. As a result, 30 years later, the country is known as one of the best countries in the world having the strictest institutional system, which means that unnecessary risks are likely to be removed. Mead (1998) argued that the case of IBM not only limited the study to one group of specific people with similarities in business and culture backgrounds but also did not reflect other social communities in the nation. Moreover, four or five dimensions of culture proposed by Hofstede is argued as being insufficient to explain the whole basis of culture, and just only classifying the most visible components of national culture (Lees, 2003).
Despite these criticisms, Hofstede’s model is still in significance. Culture results from a long history of each specific nation, thus the change in culture is generally supposed to be slow, thus the longevity and validity of Hofstede’s dimensions are still in use despite of nearly forty-year distance (Nakata & Sivakumar, 2001). People living in different countries with different norms will have dissimilar management style or, in other words, the country, in which we live will shape the dimension of managerial, organizational and strategic perspectives and actions within the companies. These four or five Hofstede’s dimensions of culture depend on management approaches related to national boundaries, therefore, they would make management board of the company seek for the root of these differences, thus enhance the effectiveness of international incorporation. To answer the debate on dimensions of culture, Hofstede (2002) indicates that nations are considered the best measurement for cultural distance and comparison. Moreover, data from IBM just works with cultural differences among countries, including appropriate samples of a huge sample size. Following these reasons, we will apply Hofstede’s model in the thesis with the four original cultural dimensions.
2.2.3. Other frameworks
After the model of Hofstede was introduced, many researchers have tried to obtain more appropriate approaches as well as definitions of cultural components. Bjerke (1999) indicates some possible dimensions allowing potential classification of a national culture as follows:
- Class structure: Vague versus distinct. - Power distance: Short versus long.
16 - Decision-making style: Democratic versus Autocratic.
- Concepts of achievement: Unbalanced versus Balanced.
- Interpretation of communication: Low context versus High context. - Personal orientation: Individualistic versus Collectivistic. - Social status of managers: Rank-dependent versus Class-dependent.
- Kinship: Weak versus strong.
Trompenaars (1994) suggests seven cultural dimensions in terms of cross-cultural communication based on an argument that each culture is differentiated by the way of solving problems. These seven dimensions, accordingly, are categorized by specific solutions in cultural clashes in each country, which are defined as follows:
- Universalism versus Particularism: Higher universalism means that formal regulations are more vital to the society, while higher particularism is defined as more important when relationships are more worthy than regulations.
- Individualism versus Communitarianism: Do people in the society react as an individual or an element of a group?
- Specific versus Diffuse: Specific cultures separate work from privacy, while in diffuse cultures, people tend to share their privacy when they come to work together.
- Neutral versus Affective: Neutral cultures prevent employees from showing their emotions, while in affective cultures, people tends to show their emotional expressions easily.
- Achievement versus Ascription: In achievement cultures, people are positioned by their performance and records, whilst by birth, family origins, gender or age in ascription cultures.
- Sequential versus Synchronic (time): In sequential cultures, being on time is placed as a priority to succeed, while in synchronic cultures, it is likely to be of less importance, meaning that people may possibly change or cancel their scheduled meetings at the last minutes.
- Internal versus External control (environment): It is assumed that people tend to either control whatever happens in their environment (internal control) or let everything happen in its own ways (external control).
17 2.2.4 National cultural fit
As indicated in previous parts, the important role of national culture in the performance of the combined firms has been emphasized in the literature on cross-border M&As. However, when investigating the concept of cultural fit in M&As, most scholars conclude that “culture fit and culture compatibility are well-used but ill-defined expressions” (Cartwright and Cooper, 1993, p.60). They often conceptualize cultural fit as cultural distance, pertaining to cross-national culture, to investigate supportive or destructive levels of cultures on the effectiveness of international incorporations. Generally, fit or congruence in culture is not only similarities but also complementarities between firms involved in a long-term contract or, in other words, diversified or specific products, market access and particular knowledge can fit and stimulate one another (Larsson &Finkelstein, 1999; Haspeslagh & Jemison, 1991). Webber et al. (1996) indicates that the terms of cultural distance implicitly assume the incongruence or lack of fit in cultures.
Distance in national culture can be understood as extra -costs, financial or organizational risks, conflicts in strategy, human or management, etc. associated with cross-cultural connection between or within groups and organizations (Hofstede, 1980). Cultural difference in cross-border acquisitions can be a source of target firm management resistance at the time of acquisition, thereby increasing the costs of the acquisition (Cartwright & Cooper, 1992). Shane (1992) posits cultural difference in trust levels resulted in higher transaction costs and perceptions that in turn moderates business strategy across borders. In addition, cultural difference in cross-border M&As brings about difficulties in post-acquisition assimilation and in the transfer of distinctive competencies between the acquiring and the acquired entities (Geringer et al., 1989). The distinctive competencies, particularly those in functional areas such as marketing and labor relations, are often specific cultural elements. Yet, in case of cultural differences, these distinctive competencies cannot be easily transferred. In addition, the greater the cultural distance, the less the likelihood that knowledge about local markets is readily available to the acquiring firm management. This increases coordination costs associated with monitoring foreign operations, adversely affecting a firm's market value at the time of acquisition (Doukas & Travlos, 1988; Hisey & Caves, 1985). From the perspectives above, national cultural fit and distance can be obtained at any level of integration with a crucial role in the effectiveness of the combined firm.
18 the performance of M&A (Datta & Puia, 1995) while another trend in current research protects the idea of positive relationship concerning the performance of a newly combined firm (Weber et al., 1996; Larsson & Risberg, 1998; Morosini et al., 1998).
In spite of these different effects, national culture often enhances the outcome of M&As through learning and specialization. First, some firms in some certain nations find it hard to develop advantageous routines, thus by interacting with another firm of a different national culture, they can learn from each other at any level of operation (Haspeslagh & Jemison, 1991). Second, through specialization, combination with a firm from different national culture allows access to country-specific routines for foreign firms with the local firm’s context (Lincoln et al., 1981). Normally, the combined firm locates in the country of the bidder or the target, and maintains the parallel operations in the two countries. Thus, to make use of existing employee resources at local location is a very important task to the survival of the combined firm. For instance, within the combined firm, with of the freedom of organizational operation varying across different culture and nations, it is less costly to implement a joint task in consistent ways with employees’ culture than to force them to commence incompatible methods with their cultural awareness. Thus, national cultural distance leads to the adoption of advanced country-specific routines in operations and management, thus resulting in task accomplishment (Fiol, 1991), positively stimulating the performance of the combined firm. In other words, national cultural fit level dissipates effectiveness of the firm. Hence, we conclude with the following hypothesis:
Hypothesis 1: When the cultural fit between partners is higher, the performance of the combined firm will be lower.
2.3. Structural compatibility
19 With respect to organization, organizational structure defines the level of coordination between members and split their responsibilities. Thus, it is considered as a vital channel to expose the effect of knowledge distribution among its members (Hayek, 1945). Hayek (1945) argues that business activities often include the integration of different and contradictory knowledge of each individual in the corporate. That is why knowledge transfer within the organization plays a critical role in the development of the firm. Over the last two decades, scholars have further developed this perspective to explain extensive topics concerning international collaborations, such as M&As (Bresman et al., 1999), alliances (Simonin, 1999) and joint ventures (Kogut, 1988). Regarding cross-border M&As, Bresman et al. (1999) examines 42 international M&As undertaken by Swedish firms and suggests that due to different phases of M&As, there would be differences in the type, the direction and the quality of knowledge transfer. In early stage of integration process, knowledge transfer is mostly from the bidder to the target, while the later stage witnesses a high degree of mutual knowledge transfer within both firms. Thus, in order to assess the effectiveness of knowledge transfer, it is necessary to examine the relative organization structure within the two firms. This relationship can be evaluated by comparing their firm size and organizational age, which elaborate the absorptive capabilities, the richness of experiences and the length of business history of each company.
Relative size
20 Hypothesis 2: When the relative size between partners is higher, the performance of the combined firm will be better.
Age similarity
Like size, organizational age is academically indicated as important characteristics related to structure of the firm. Age of an organization usually calculates its longevity or survival. It also indicates the length of company’s history and their wealth of business experiences. Thus, the similarity between two firms can also be defined as the convergence on their longevity and survival in the economies. The effect of organizational age between companies on the performance of alliances, clusters or incorporations has been found in many studies over the decades yet with inconsistencies. In the field of M&A, scholars, when finding motivations for M&A, mainly focus on age of corporations from perspectives of the bidder who can acquire or buy out the target (Fowler & Schmidt, 1989). The bidder, who is considered as bigger and more mature firm in the economy, usually acquires the smaller and younger target whose internal rates of return are lower, the bidder believes that they can gain greatest profit via huge cash supply for outside investment, enhancing the performance of M&A transaction (Mueller, 1969). However, a positive relationship between the performance of the collaboration and the similarity between their ages has been sometimes reported. Fowler & Schmidt (1989) state that when the involving companies tend to be founded at the same period of time, their founders seemingly adopt the convergence on institutionalized practices at the time of foundation, thus undertake the same organizational practices.
Based on the above arguments, we predict:
Hypothesis 3: When the age similarity between partners is higher, the performance of the combined firm will be better.
2.4. National culture fit and structural compatibility
21 well as tap into cultural differences implicitly. Thompson (1966) argues that structure is the number of methods, by which the firm establishes limitations and boundaries for effectiveness of performance, divides separate responsibilities within organizational individuals, and controls corporate resources, especially human capital. As indicated in previous parts, culture in organizations is defined by ways of thinking, routines, beliefs of members and interactions between them. Thus, in the phase of integration, in which two separate companies combine, structure related to culture might delineate the extra-differences beyond the basically existing cultural distances in thoughts, working or intercommunication among organizational members. For instance, when managers of corporation could not afford appropriate changes in structure on time as expected, some human conflicts and managerial misguides can lead to more severe distances between involving firms. Thus, in organizational context, especially in the context of international involvement, in which differences in culture matter, the compatibility in firms’ structure needs to be clarified along with the existing influences of culture on the performance of the newly-combined firm.
The impact of structural compatibility on the degree in which cultural fit affects the effectiveness of the newly combined firm is paid little attention by researchers and scholars, yet the findings are still in contrary and at a micro level. On the one hand, the effect of differences in organizational age on patterns of culture differences could not be confirmed by scholars, especially in terms of transitional nation context, in which differentiations between the old and new organizations is still blurred because of rapid changes in the society. On the other hand, it is recorded that differences in organizational size might reveal less the relationships with cultural distances within organizations. Organizations with different sizes are likely to obtain differences in cultures, behaviours, people, procedures and contacts (Chan & Chan, 2005; Ghobadian & Gallear, 1997). One of the most severe problems concerning cultural difference within organizations is capabilities of transferring organizational values during integration process. In cross-border deals, large foreign partners behave systematically and posses a long-term approach, while their smaller local ones are often without established operational systems and policies (Lane & Beamish, 1990). Thus, in case the local target could not adapt to the big structure of the foreign bidder in a meantime, the two organizations might suffer from a longer distance in cultural styles as well as losses. Based on the above arguments, we propose:
22 Hypothesis 4b: The effect of the age similarity between partners on the performance of the combined firm will be stronger at higher degree of national cultural fit and weaker at lower degree of national cultural fit.
3. The models
As indicated throughout this thesis, the hypotheses will be tested by examining two levels of impact: one is the separate impact of national cultural fit, relative size and age similarity on the performance of the combined firm, and the other one is the impact of interaction between either relative size or age similarity and cultural fit on the outcomes of cross-border M&As. For the following reasons, we use three regressions to analyze the relationships indicated above, two of which are multiple linear regressions. The third one is a multiple regression with interaction variables. First, our dependent variable, the performance of the combined firm, is constructed as a numeric variable (the value of ROA). Second, most previous research uses OLS estimation while the dependent variable of the present study is a numeric variable. Third, three separate models are taken into account to observe changes of independent and exogenous variables over the models in order to attain better general results and conclusions. The models examine the OLS estimate with the following specifications:
Model 1: Performancei = β0 + β1 X1i control+ β2 X2i control + β3 X3i control + εi
Model 2: Performancei = β0 + β1 (cultural fiti) + β2 (relative sizei) + β3 (age similarityi) + Xcontrols + εi
Model 3: Performancei = β0 + β1 (cultural fiti) + β2 (relative sizei) + β3 (age similarityi) + β4 (cultural fiti x relative sizei) + β5 (cultural fiti x age similarityi) + Xcontrols + εi
In the three models, performance stands as the dependent variable, β (from β0 to β5) represents estimated regression coefficients. Other descriptions related to these variables can be found in Table 1 in Appendix B. Our dependent variable also depends on the error term (εi), thus in order to ensure that the least square estimations are the best Linear Unbiased Estimators (BLUE), the six following assumptions need to be guaranteed:
Assumption 1: Yi = β0 +β1 xi1 + β2 xi2 +…+ βk xik + εi i = 1,…, N
23 Assumption 3: Var(Yi) = var (εi) = (No heteroscedasticity) The variance measures the uncertainty of the model, assumed to be similarly for each observations and the errors in which homoskedasticity exits.
Assumption 4: Cov(Yi , Yj) = cov (εi,εj) (No autocorrelation). There is no correlation among the size of error terms of separate observations. This means that for the dependent variable, any two observations are uncorrelated as well.
Assumption 5: The values of xik are not random and are not exact linear functions of the other explanatory variables. It is assumed that the explanatory variables are valued prior to observing the values of the dependent variable. If this assumption is violated, collinearity exists and the least squares regression would fail.
Assumption 6: Yi ~ N [β0 +β1 xi1 + β2 xi2 +…+ βk xik, ] ↔ εi ~N(0, ). It is assumed that the random errors are normally distributed. For the dependent variable, its values have a normal distribution about its mean.
If the first five assumptions are held, according to the Gauss-Markov Theorem, the least squares estimators are BLUE. If one of these assumptions is violated yet the least squared estimator is still used to estimate the regression coefficients, the least squares estimator is still a linear regression and unbiased estimator, but:
- The least squared estimations are no longer BLUE. - The standard errors are incorrect.
- Confidence intervals and hypothesis test using these standard errors might be misleading.
4. Methodology 4.1. Data and sample
24 of over 200000 private and public companies all over the world. Moreover, necessary data relating to the year of foundation and the number of employees of involving firms can be found on official websites of these organizations based on information from both Zephyr and Orbis. Our data for measuring dependent, independent and control variables is extracted from collective income statements, balance sheets and other supplementary sources.
Given the lack of availability of disclosed financial information of the combined firm, the bidder or the target, Zephyr and Orbis yielded 88 completed cross-border M&A deals between U.S firms and foreign firms from 22 countries during the period 1998 - 2009. In the sample, the United Stated of America (the USA) is taken as the default country for either an acquirer or target. Thus, the 88 deals refer to those that U.S companies engaged with 88 foreign firms across 22 countries in the world. The list of firms’ nationality is inferred from Hofstede’s indices (66 countries including USA which can be seen in the Appendix A). We will examine the performance of the combined firm within the first three following years after the merger or the acquisition, thus our single unit of observations is restricted with criterion that the combined firm was born between 1998 and 2006. Thus, the data makes up a cross-sectional data4. Other collection for specific variables is elaborated in Table 1 in Appendix B. As can be seen from Table 2, in the sample, 66 percent of the total cross-border M&A deals are combinations between the US firms and European ones while 15 percent coming from Canadian firms representing North American geography.
Table 2: Frequency of continents which originate foreign partners with US firms during the period 1998-2009
Continents
Number of the
combined firms Percent
Europe 66 75 North America 15 17.05 South America 1 1.14 South Africa 1 1.14 Asia 5 5.68 Total 88 100 4
25 4.2. The variables
A summary of all variables in the models can be seen at Table 1 in Appendix B. In this part, we will elaborate how these variables are measured in academic research, and specifically in the thesis.
Performance of combined firm
Performance of the combined firm is built as a dependent variable in the models. The concept of firm performance is much likely to be broad and complicated in the wide range of literature, spreading from accounting performance to customer or employee retention, and when it comes particularly to M&As, it is also obtained at three main levels of integration: task level, transaction level and firm level (Zollo & Meier, 2008). According to Zollo & Meier (2008), the performance of the combined organization is defined as “the variation in firm performance that occurred during the period of relevance for the execution of the business plan connected to the acquisition” (p.58).
Due to theoretical complexity, there is a variety in measuring and analyzing performance of the combined firm after the announcement of completed M&A deal. These measurements are categorized into two levels: short-term and long-term. The short-term measurement focuses on the bidder’s stock market returns (Carow et al., 2004; Datta & Puia, 1995) in the most recent three to six months after the announcement of completed deals. Meanwhile, long-term measurement relies on accounting indicators related to assets, equity, capital and profit of the combined firm. Three of the most popular accounting indicators which are used academically (Weber, 1996; Palich et al., 2000; Singh & Montgomery, 1987; Zollo & Singh, 2004) are return on equity (ROE), return on asset (ROA) and return on capital (ROC). ROE measures how much profit a company makes with the money invested by shareholders, while ROA tells how much revenue is achieved based on assets the company holds, or how efficient the management uses the assets. Then, ROC measures the profit made with money invested by shareholders and the money borrowed.
26 within the first three following years after the mergers will be calculated to quantify the dependent variable.
National cultural fit
As mentioned before, national cultural fit is defined through national cultural distance and considered as the first independent variable in the models. Thus, this independent variable is measured by the level of national cultural distance which is calculated by Kogut and Singh’s (henceforth: K&S) formulation (1998). K&S formulation quantifies national cultural distance between USA firms and a firm from other countries. This formulation describes a multidimensional measure based on Hofstede’s (1980) aggregate scores which has frequently been used in foreign entry studies (Agarwal & Ramaswami, 1992; Barkema et al., 1997) with four facets: Power Distance Index (PDI), Individualism (IDV), Masculinity (MAS) and Uncertainty Avoidance Index (UAI). K&S (1988) formulate:
= ∑ { −
/Vi }/4
where CDj is cultural difference of the jth country from the USA.
Iij is the index for the ith cultural dimension and jth country, derived from Hofstede’s indices.
Vi is the variance of the index of the ith dimension.
In the thesis, the high numeric value of CDj indicates the low level of national cultural fit between the USA firm and the firm from another country or, in other words, weak degree of fit in national culture in the combined firm. The values of Iij and IiUSA are collected from Hofstede’s publication "Culture's consequences: International differences in work-related values" (1980) which studied cultural dimensions in 66 different countries from Asia, Europe and America. The scores of national cultural distance proposed by Hofstede (1980) have been adopted in academic research (Shane, 1992; Morosini & Singh, 1994; Kogut & Singh, 1988), and evidently proved their reliability and validity over the decades.
27 from the idea that the bidder has higher position than the target in the transaction. As the matter of fact, Hofstede measured the scores with external source, which was not related to the sample itself and was independent from interviewed managers’ memory, thus the scores is taken into account to measure the independent variable of national cultural fit in this study.
Relative size
Relative size is considered as the second independent variable in the models. There is a number of ways to measure it academically. Some scholars prefer the ratio related to market value of both firms (Sirower, 1997), others prefer the ratio of the target’s total assets to the bidder’s in the year preceding the announcement of completed deals (Fowler & Schmidt, 1989). As indicated above about the definition of relative size, which is the similarity on the managerial capacity of the bidder and the target, we measure this independent variable as the ratio of the bigger’s annual sales to the annual sales of the smaller a year prior to the announcement date of legal combination. If sales are not available, we can use total assets or total number of employees. The reason why I chose this measurement is that sales, assets or number of employees in accountancy illustrate the most obvious indicator for managerial capacity and abilities of firms in the long term. The way to collect data for the ratio is described in Table 1 in Appendix B. The value of relative size ranges from 0 to 1. The more it closes to 1, the more similar between sizes of the two companies. Table 3 shows the frequency of three measures above, which can be applied in the sample.
Table 3: Frequency of three measures for the independent variable of relative size
Measure
Frequency of the
combined firms Percent
Annual sales 65 74
Total assets 14 16
Number of employees 9 10
Total 88 100
Age similarity
28 foundation year from the immediately preceding year before the legal combination. Thus, age similarity is the ratio of the age of the younger firm to the older firm’s age. The ratio ranges from 0 (meaning always bigger than 0) to 1 (the maximum level reached when the bidder and the target were founded in the same year). The way to obtain data for this independent variable is illustrated in Table 1 in Appendix B.
Control variables
To control potential endogeneity, the study use some control variables pertaining to characteristics of nations and firms.
Geographical distance: This variable is measured as geographical distance between the USA and other countries with its unit in kilometer. This control variable reflects the geographic differences between the home and the host country by creating difficulties in the delay timing transfer. We can obtain this variable by calculating the distance between two capitals of the two countries.
Economic distance: This variable determines the differences in the degree of economic development between the USA and other countries, measuring in US dollars. It is measured by the difference in GDP per capita between the USA’s and other countries’ in the year when the announcement is made. Data of this variable is obtained from The World Bank’s official website.
Industry: We control the type of industry sector in which the combined firm operates. Because of the difference in industry effect, each combination gains various propensities for on-going survival. Due to the fact that manufacturing and service companies account for 49% and 51% respectively in the sample, we code a dummy variable for industry to put all firms into manufacturing industry (1) and non-manufacturing industry (0). Regarding this thesis, we use The North American Industry Classification System (NAICS) initiated by Federal statistical agencies for classifying industry sectors in the U.S business economy. Details of NAICS code for manufacturing sectors can be found in Appendix C.
29 Cash payment: We control the payment method of the deal via cash payment which is equal to 1 if the bidder paid by cash and 0 otherwise. It is viewed that the method by cash payment made the deal less complicated than one financed by other ways because it lessens the effect of uncertainty (Weston & Jawien, 1999).
5. Analysis and results 5.1. Description
Table 4 describes the descriptive statistics of all variables presented in our regression models with the means, minimum and maximum values, standard deviations, skewness and kurtosis. The number of observations for the dependent variable, the performance of the combined firm, is 88. The dependent variable is obtained with the mean of 0.061, ranging from -0.407 to 0.68, the skewness of 1.016 and kurtosis of 4.884. Thus, the variable is highly right skewed.
Cultural differences show an average value of 0.643 with a minimum of 0.019 (with Australia) and maximum of 3.293 (with Colombia). This means that in our sample, there is a sufficient difference between USA and other countries on how culture and business activities interact.
30 Table 4: Descriptive statistics
1 2 3 4 5 6 7 8 9 Mean 0.061 0.643 0.334 0.477 7470.398 8804.057 0.489 0.841 0.409 Maximum 0.680 3.293 0.950 1 16486 43806 1 1 1 Minimum -0.407 0.019 0.002 0.005 3335 -12969 0 0 0 Std. Dev. 0.192 0.842 0.305 0.289 2349 9666.653 0.503 0.368 0.494 Skewness 1.061 1.487 0.428 0.428 0.243 2.005 0.045 -1.864 0.370 Kurtosis 4.884 4.151 1.754 1.754 5.285 8.538 1.002 4.475 1.137 Observations 88 88 88 88 88 88 88 88 88
(1) Performance of combined firm (2) Cultural differences (3) Relative size (4) Age similarity (5) Geographical distance (6) Economic distance (7) Industry
(8) Previous M&A experiences (9) Cash payment
5.2.Diagnostic check
As indicated above, in order to ensure the validity of the multiple regression models, the first five assumptions for BLUE models need to be checked and ensured. In this part, we will discuss how well our data fit the OLS assumptions. The assumptions we will check as follows are whether (1) the regression function is linear, (2) error terms have a constant variance or homoskedasticity exists, (3) error terms are independent or autocorrelation exists, (4) there are outliers in the data, (5) error terms are normally distributed and (6) the model is properly specified.
(1) Assumption 1 & 2 - Linearity
31 related to firm performance is negatively expected for national cultural fit and positively for relative size and age similarity. One common way to detect linearity or nonlinearity is evidently in a plot of residuals versus predicted values, which are outputs of STATA. It is likely to be linear if the points are symmetrically distributed around a horizontal line in the plot. In the thesis, we examined three scatter plots between the standardized residuals against each independent variable in the regression models. Figure 1 in the Appendix B illustrates these scatter plots with the predicted shaped line. Thus, we conclude that it is likely to a linear relationship between the dependent variable and the independent variables except cultural differences. It means that nonlinearity will exist in our case or, in other words, it is predicted that Model 3 will be more appropriate than the other two models.
(2) Assumption 3 - Homoskedasticity
Homoskedasticity occurs when the error terms have a constant variance, otherwise it is heteroskedasticity. Heteroskedasticity, if often found in a panel data, includes the issues of time variance. The consequence of heteroskedasticity is that one of the least squares assumptions is violated (var (ei) = ), leading to the results that the least squares estimator is still a linear and unbiased estimator but no longer BLUE. To check for heteroskedasticity, we performed the White’s test in which the null hypothesis is the existence of homoskedasticity. If heteroskedasticity happens, to report robust standard errors in multiple regressions is taken into account in order to prevent biased estimation. The result of White test shows that p-value is 0.554. Thus, we fail to reject the null hypothesis of no heteroskedasticiy in the model at the significant level of 10%. It can thus be concluded that there is no heteroskesdasticity in our models.
(3) Assumption 4 – Autocorrelation
32 sample is independent from each other. We can conclude that autocorrelation is not a problem in our case.
(5) Assumption 5 – Collinearity
Collinearity happens when there is a linear relationship between two explanatory variables in the simple regression model. In the multiple regression models, which are the case in the study, collinearity can be referred to multicollinearity, in which two or more explanatory variables can be highly related. In case of multicollinearity, each explanatory variable’s impact on the dependent variable is not valid and precise because there are two traumatic problems involved. First, p-value of each variable can be misleading, which means a higher p-value, although the variable is really important. Second, when it comes to coefficients, the confidence intervals on the regression can be very wide, reducing the reliability and accuracy of the test. Hence, excluding or adding a new subject into the model can change the coefficients dramatically and in some cases, even alter their signs. According to Dikova et al. (2010), in the test of multicollinearity, if a correlation coefficient is larger than 0.7 or smaller than -0.7, multicollinearity exists among independent variables.
