MSC BA STRATEGIC INNOVATION MANAGEMENT THESIS
THE ROLE OF TOP MANAGEMENT TEAM HETEROGENEITY IN BUSINESS-LEVEL STRATEGIC CHANGE AND M&A PERFORMANCE
by
Bas van der Veen S2023636 Illegaliteitslaan 126
9727EE Groningen
E: basveen@gmail.com T: 06-41036644
June 2016
First supervisor: Prof. Dr. J. Surroca Second supervisor: Dr. P.M.M. de Faria
Word count: 16.224
2
ABSTRACT – Mergers and acquisitions (M&As) are popular in practice; yet, studies addressing the impact of an M&A have been quite equivocal. This study sheds new light on the literature on M&As and links business-level strategic change and top management team (TMT) heterogeneity to M&A performance. Conducting multiple analyses on a sample of 100 acquiring firms with a total of more than 700 top managers, it was found that business-level strategic change and TMT heterogeneity are related to M&A performance. In particular, TMT heterogeneity seems to have an indirect effect on M&A performance through strategic change.
Furthermore, some support was found for the belief that TMT heterogeneity might have a different effect when it comes to implementing strategic changes. These findings contribute to the current M&A literature and could assist companies with the decision-making process regarding the composition of the TMT.
Keywords: Mergers and acquisitions, top management team heterogeneity, business-level strategic change.
INTRODUCTION
In recent years, mergers and acquisitions (M&As) have been a popular topic in both practice and research. For several decades, M&As have been the preeminent corporate growth strategy leading to extraordinary levels of investments (Barkema & Schijven, 2008; Shimizu, Hitt, Vaidyanath, & Pisano, 2004). Just over 40.400 M&A deals were announced in 2014, equivalent to one contract every 13 minutes. This resulted in a total value exceeding 3.5 trillion US dollars, which is a 47% increase compared to 2013 levels and has been the strongest annual result since the financial crisis (Thomson Reuters, 2015).
Despite the prevalence of M&As in the corporate world, findings in the literature on the outcomes of M&As are quite equivocal. At the one hand, Siegel and Simons (2010) report a boost in productivity and efficiency after an acquisition, which results in greater M&A performance. Also, different studies claim that companies can significantly gain from an M&A (Houston, James, &
Ryngaert, 2001; Leeth & Borg, 2000). On the other hand, alternative studies found that acquisitions often eroded firm value and resulted in company losses (Selden & Colvin, 2013; Seth, Song, & Pettit, 2002). In between these two extremes are several studies arguing that an M&A is beneficial for shareholders of the target company but the investors of acquiring firms generally benefit less, or do not benefit at all, from making acquisitions (Cartwright & Schoenberg, 2006; Haleblian et al., 2009).
In explaining this diversity in findings, studies have identified several conditions that drive M&A performance, including related acquisitions (Capon, Farley, & Hoenig, 1990; Homburg & Bucerius, 2006; Palich, Cardinal, & Miller, 2000), method of payment (Hayward & Hambrick, 1997; Walker, 2000), conglomerate firms (Agrawal, Jaffe, & Mandelker, 1992; Anand & Singh, 1997), and prior acquisition experience (Haleblian & Finkelstein, 1998; Hitt, Harrison, Ireland, & Best, 1998; Zollo &
Singh, 2004). However, King, Dalton, Daily, & Covin (2004) found in their meta-analysis that none of
3
these variables are significant in explaining post-acquisition performance. Thus, even though a considerable amount of researchers focused on this topic, the outcomes of those studies still do not explain why some firms have better post-acquisition results than others (Gomes, Angwin, Weber, &
Yedidia Tarba, 2013). This clearly suggests that more research is needed to broaden our understanding of the reasons for success and failure of acquisitions.
To shed new light on the conflicting results, this study will investigate the role of business- level strategic changes on the outcomes of M&As. In particular, this study will introduce business- level strategy in the study of corporate-level strategy performance. Various studies have found that business-level strategic changes affect firm performance (e.g., Hambrick & Schecter, 1983; Haveman, 1992; Jauch, Osborn, & Glueck, 1980; Zajac & Kraatz, 1993) and it might have implications for M&A outcomes as well. Business-level strategy focuses on the question of how to compete in this business, while corporate-level strategy focuses on the question of what businesses the organization should be in (Beard & Dess, 1981; Hambrick, 1980). According to Schwartz and Davis (1981), it is of importance that corporate-level decisions fit with the business-level strategy. In other words, business-level strategies should be aligned with corporate-level strategies, including acquiring other companies.
Take, for instance, the congruence model for organizations developed by Nadler and Tushman (1980) on the importance of fit between organizational components. Merging with another firm or acquiring a firm could change organizational components and thus upset the internal fit. In an M&A, for instance, the goals of the target firm might not meet the demands of the acquiring firm’s formal organizational arrangements. This would hamper organizational effectiveness and efficiency leading to outcomes that are not optimal (Nadler & Tushman, 1980). Adapting your business-level strategy to make sure the components are internally aligned could thus enhance organizational performance. In this study, it is argued that the amount of post-acquisition strategic change on business level is a contingency factor of the outcomes of corporate-level strategic change. That is, business-level strategic change affects M&A performance.
In connecting business-level strategic change to M&A performance, the top management team (TMT) plays a central role. Ultimately, the TMT is the echelon that is responsible for the development, deployment, and execution of the strategy of the company (Carpenter, Geletkanycz, &
Sanders, 2004; Olson, Parayitam, & Twigg, 2006). After it is decided upon by the TMT to acquire another company, the corporate-level strategy, they are also responsible for making business-level strategic changes, such as the allocation of resources after an acquisition or how to compete in a specific industry after an acquisition. According to upper echelon theory, the intensity of these strategic choices is based on executive cognitions, values, and perceptions (Hambrick & Mason, 1984). Hence, the strategic choices made in an organization are reflections of the managers’
cognitions, which influence organizational outcomes, including M&A performance. Prior research has
shown that TMTs have a pivotal role when it comes to innovation, including organizational
reorientation, new product launches, and changes in R&D strategies (Alexiev, Jansen, Van den Bosch,
4
& Volberda, 2010). In particular, TMT heterogeneity has been found to be a significant predictor of the innovativeness and effectiveness of strategic choices (Hambrick, 2007). First, heterogeneous TMTs are less prone to groupthink than homogeneous teams due to differences in knowledge, perspectives, and information (Hambrick & Mason, 1984; Lant, Milliken, & Batra, 1992). In the same line of reasoning, Hoffman and Maier (1961) argue that a diversity of backgrounds and experiences allows the TMT to draw on different sources of knowledge, leading to possible superior alternatives which could not have been identified in a less diverse TMT. Second, since heterogeneity in TMTs is a proxy for cognitive heterogeneity, it may cause group members to disagree on task issues leading to greater variance in decision-making alternatives (Bantel & Jackson, 1989). Furthermore, task conflicts could also “stimulate creative thinking and divergent thought processes”, resulting in decisions that outrange the decisions that could have been made by a single individual (Talke, Salomo, & Rost, 2010, p. 910). In sum, the composition of the TMT seems to play a crucial role in strategic decision making and should be taken into account when studying business-level strategic change.
Overall, there is a lot of research on M&As, strategic change and TMTs but there are no studies connecting these constructs. In this paper, the effects of strategic change are examined in the context of M&As. Using a sample of 100 M&As and more than 700 top managers, this study will examine the role of heterogeneous TMTs in the context of business-level strategic change and M&As.
Furthermore, a comprehensive data collection method on the managers of the TMT allows the heterogeneity measure to consist out of eight different variables. The findings of this paper have several practical implications as well. In particular, this study could help managers understand the outcomes of their strategic decisions after an acquisition. Second, it presents the shareholders with some advice concerning the composition of the TMT. Particularly if it would benefit the organization, in terms of business-level strategic change and M&A performance, when a heterogeneous TMT would be in control of the organization. Finally, this study stresses the importance of internal fit between the business- and corporate-level strategies. That is, the success of corporate-level strategy is contingent upon the changes in strategy at the business-level.
This paper is structured in several sections. First, this paper will theoretically discuss the relationship between M&A performance, strategic change, and TMT heterogeneity. Following the theoretical discussion, several hypotheses will be developed accordingly. Subsequently, the methodology section will be discussed and the results of the analysis will be presented. In the final section, the theoretical and practical implications of the findings will be discussed, together with the main limitations and directions for future research.
THEORY AND HYPOTHESES DEVELOPMENT
This section will describe the current literature on business-level strategic change, TMT heterogeneity,
and M&A performance. Based on the topical literature, several hypotheses will be formed that will be
tested in this study.
5 Business-level strategic change
For successful outcomes of the M&A it is proposed in this paper that the company’s strategy needs to adapt accordingly, specifically changes in the business-level strategy after the acquisition. To focus on the process post-acquisition, and not the pre-M&A process, is also relevant since it is found to be
“more crucial and needs more attention” (Quah & Young, 2005). Strategic change, in general, is the means whereby an organization maintains alignment with changing circumstances (Kraatz & Zajac, 2001). Theoretical literature has distinguished two types of strategic change: (1) corporate-level strategic change, which is concerned with questions about what industry it should compete in and (2) business-level strategic change, which is concerned with questions about how to compete in a specified business (Beard & Dess, 1981; Hambrick, 1980). Unlike most research on M&As, which has been focused on corporate-level strategic change, this paper will also focus on business-level strategic change. In this study, earlier definitions of business-level strategic change (Carpenter, 2000; Zhang, 2006) are adopted resulting in the following conceptualization: business-level strategic change is “the extent to which a firm’s pattern of resource allocation in key strategic dimensions change over time”
(Zhang & Rajagopalan, 2010, p. 335). The M&A itself, in this case, would be part of the company’s corporate-level strategy.
An M&A represents a large change in resources, structures, and culture. Maintaining the old ways of operating, and not successfully integrating the two firms involved in the M&A, would not be appropriate (Zollo & Singh, 2004). For an M&A to produce synergies it is not enough to simply acquire other organizations’ resources, rather “the resources must be effectively integrated and managed to realize the synergy” (Harrison et al., 2001, p. 679). This can be done through changes in the business-level strategy. That is, the effectiveness of corporate-level decisions is contingent upon business-level strategic changes. Earlier studies also argue that changes in business-level strategy reflect the risk-taking aspects of a firm’s strategic choices (Carpenter, 2000; Finkelstein & Hambrick, 1990). Furthermore, strategic changes on the business level are found to promote bold thinking and novel strategic alternatives, leading to increased firm performance (Hambrick & Schecter, 1983).
It is argued that the combination of acquiring resources (corporate-level decisions) and reconfiguring resources (business-level decisions) is the source of value and innovation for a firm (Karim & Mitchel, 2004). Particularly, an organization should actively reconfigure its acquired units to create new value, rather than simply remain the old way of competing (Karim & Mitchell, 2004).
This implies a complementary relationship between corporate-level strategic change and business- level strategic change. That is, corporate-level strategic changes (e.g., M&As) will be more effective when it is followed up by a change in business-level strategy. In sum, an M&A implies a large change in an organization and adapting the way of operating to the new situation would be beneficial to the organization. Thus, the following is hypothesized:
Hypothesis 1: Business-level strategic change positively influences M&A performance.
6 Upper echelon theory
Over the last decades there has been a growing interest in the role of executive decision-makers. Since the study on upper echelon theory by Hambrick and Mason (1984), many studies have investigated features of the TMT, including antecedents, elements, and consequences of TMT composition (Carpenter, Geletkanycz, & Sanders, 2004). Building on the idea developed by Cyert and March (1963) of a dominant coalition, the upper echelon theory postulates that psychological factors, such as values and cognitions, can be measured through observable proxies of the TMT. The model of upper echelon has a dual role since it offers both a theoretical framework and a research methodology. The original upper echelon model by Hambrick and Mason (1984) explains how strategic choices made in firms are reflections of the executives’ cognitions and values, how this can be operationalized through observations, and that these observations result in organizational outcomes. In the literature on upper echelon theory, there are multiple studies reporting a significant relation between TMT characteristics and firm performance (Carpenter, 2002; Peng & Luo, 2000; Simons, Pelled, & Smith, 1999; Tushman
& Rosenkopf, 1996). According to the upper echelon perspective, the effectiveness of organizational responses (e.g., strategic changes) varies with how the TMT interprets strategic issues and triggers the response (Hambrick & Mason, 1984). During the strategic decision-making process, each manager has its own cognitive base that makes up their perceptions and interpretations, which make up the type and variety of the whole team’s cognitive base. Ultimately, it is the TMT’s cognition of the current situation that determines the strategic choices made by the company (Wiersema & Bantel, 1992).
Throughout this study, the corporation’s executives are referred to as the upper echelon or TMT, since they are responsible for the governance of the firm. The TMT has considerable power in the organization as it is allowed to evaluate managerial performance, it can allocate rewards and penalties to management, and it has the formal authority in the decision-making process (Fama &
Jensen, 1983). Decisions are likely to increase in complexity and difficulty when they have to be made at a higher level in the organization (Van Knippenberg et al., 2011). Therefore, it is particularly important for an organization to establish a TMT that is capable to deal with complex organizational issues.
Top management team heterogeneity
This section will first address the current literature on TMT heterogeneity before it will be linked to
business-level strategic change and M&A performance. Heterogeneity is defined as the quality or state
of being heterogeneous and not comparable. The more people differ from each other, the greater the
amount of team heterogeneity. A considerable amount of research has focused on the influence of
TMT heterogeneity on the performance of organizations and the composition of the TMT is likely to
influence the strategy of the firm (e.g., Barkema & Shvyrkov, 2007; Dwyer, Richard, & Chadwick,
2003). In studying strategic decision making processes, TMT heterogeneity plays an important role
(Papadakis & Barwise, 2002). For instance, heterogeneous TMTs are associated with high levels of
7
creativity and innovation (Bantel & Jackson 1989; Richard & Shelor, 2002) and heterogeneous TMTs are more likely to consider multiple sources and perspectives of information (Dutton & Duncan, 1987). Furthermore, TMT heterogeneity is found to be a proxy for cognitive variety within the TMT (Milliken & Martins, 1996). Consistent with prior literature (e.g., Bantel & Jackson 1989; Eisenhardt
& Schoonhoven, 1990; Michel & Hambrick, 1992), TMT heterogeneity will be measured through different kinds of attributes of the team’s demographic attributes, including age, gender, and nationality. Furthermore, some other aspects related to the executive will be taken into account, including tenure, education, functional background, and international experience.
Age can be viewed “as both a proxy for the extent of experience and a signal of a person’s propensity for risk taking” (Herrmann & Datta, 2005, p. 72). There are significant differences in behavior of managers from different age groups. For instance, Pegels and Yang (2000) found that older managers tend to be more risk averse and they are less capable of successfully handling new and creative ideas (Guthrie & Olian, 1991). Younger managers, on the other hand, were found to be more innovative and risk-taking (Grimm & Smith, 1991). Diversity in age within the TMT might cause disagreement and conflict (Auden, Shackman, & Onken, 2006). Furthermore, Ireland and colleagues have found that people of different age groups are more likely to have differing values and beliefs because their previous experiences in life are different (Ireland et al., 1987).
Diversity in gender, or sex, has been a growing topic of interest in the last decades but in most countries the women that reach top positions is still very low (Smith, Smith, & Verner, 2006). The numbers have been shifting though, and more and more women can be found in the board of directors.
Researchers have found that a greater diversity in gender in the TMT can significantly improve organizational performance (Krishnan, Park, & Kilbourne, 2006). In an extensive 19-year study on 215 Fortune 500 firms a strong correlation was found between profitability and women in the board of directors (Adler, 2001). Other authors also argue that firms can enjoy several competitive advantages when women would be part of the TMT, including reputation and greater organizational learning (Burke, 2003; Krishnan & Park, 2005).
Diversity in nationalities can be defined as “the distribution and number of team members’
national backgrounds” (Dahlin, Weingart, & Hinds, 2005, p. 1108). Nowadays, because of globalization, it is of greater importance to take diversity in nationality into account. Especially, since the people from the TMT consist of multiple nationalities and backgrounds (Harris & Ghauri, 2000).
Also, when the TMT has more different nationalities it also better matches the demographic characteristics of more nationalities, which could help the organization achieve a competitive edge in the market (Cox & Blake, 1991). From a sensemaking perspective, diversity in national origin is required to make sense of contemporary complex environments (Weick, 1993).
Heterogeneity in organizational tenure deals with “the differences among team members in
terms of the amount of time spent with the organization” (Chi, Huang, & Lin, 2009, p. 700). A top
team’s tenure in the organization determines whether the organization will likely stick to the status
8
quo and what its attitude will be towards risk (Finkelstein & Hambrick, 1990). Horwitz and Horwitz (2007) found that TMT tenure diversity is positively related to innovation. Likewise, Simons, Pelled, and Smith (1999) found that diversity in company tenure was positively influencing the quality of discussions and decision-making. Lastly, Zenger and Lawrence (1989, p. 372) found that tenure diversity provides good training for new employees since “longer-tenured employees have strong incentives for getting new employees up-to-speed”. That is, managers are less likely to develop new knowledge when the other managers in the TMT have the same tenure.
Diversity in education is the difference among team members in sets of task-relevant skills, knowledge, and abilities represented by their educational background and level (Dahlin, Weingart, &
Hinds, 2005). Educational heterogeneity can be separated into two categories: educational background and educational level. As the educational background of the TMT gets more diverse, it allows the decision-makers to draw from a broader knowledge base. This could lead to more dispersed views on the current ways of doing business and ultimately to a change in doing business (Ostergaard, Timmermans, & Kristinsson, 2011). Executives from TMTs also differ in their educational levels.
Where some have a bachelor’s degree, others have a master’s degree or even promoted in their area.
One might expect that the highest level of education for each manager would lead to the optimal result. Auh and Menguc (2005), however, argue that a TMT composed of members all possessing, for instance, a doctorate degree is not necessarily beneficial for the firm. This could limit the creativity and innovative solutions due to the uniformity in structure and mental maps of the members of the TMT. Overall, educational heterogeneity has also been found to positively affect firm performance (Smith et al., 1994).
Diversity in functional background holds similar arguments as diversity in educational backgrounds as to why it could be advantageous for the company. Namely, as with educational background heterogeneity, it allows the TMT to tap into diverse knowledge bases. Bantel and Jackson (1989) also found that TMT functional heterogeneity leads to a greater level of administrative innovation. Functional background has been found to positively influence both the firms’
diversification strategies (Smith & White, 1987) and the firms’ competitive strategies (Chaganti &
Sambharya, 1987).
Lastly, differences in experience abroad of the managers are expected to positively influence the organization. Sambharay (1996) found that heterogeneity of managers with foreign business experience was related to diversification strategies of the organization. Moreover, heterogeneity in international experience could give the TMT a competitive advantage in identifying new market opportunities (Hitt, Hoskisson, & Ireland, 1994).
Previous research on the influence of TMT heterogeneity has not been equivocal and as Mello
and Ruckes (2006) argue, to address the composition of teams in organizations one needs to look at
both the effect of heterogeneity on the decision itself as the effect of heterogeneity on the
implementation of the decision. Defining strategic change and implementing strategic change are also
9
two distinct aspects. As defined earlier, business-level strategic change is about changing the allocation of resources over strategic dimensions over time. This definition can be broken down into two parts: the formation of a strategic change and the enactment or implementation of such a change in strategy. Therefore, this study will differentiate between the effect of TMTs on (1) the likelihood and amount of business-level strategic change and (2) the implementation of business-level strategic change.
Level of business-level strategic change
In the upper echelon theory, Hambrick and Mason (1984) state that characteristics of key executives are related to strategic decisions. The TMT can influence the likelihood and amount of business-level strategic change, depending on its heterogeneity. Strategic change means, as defined earlier, that the firm modifies its pattern of resource allocation and that the current status quo is not maintained.
According to Finkelstein and Hambrick (1990), demographic heterogeneity of the TMT is related to an increase in the number of strategic alternatives considered. Similarly, Dutton and Duncan (1987) argue that differences in cognitive structures will increase the diversity of information, interpretation, and solutions generated by the TMT. In turn, this will increase the likelihood of finding a strategic alternative that appears to be superior to the status quo. If the TMT would be more homogeneous it could lead to greater social cohesion, decreasing the likelihood that the status quo will be questioned by members of the TMT (Michel & Hambrick, 1992). Heterogeneous TMTs are more likely to break with past patterns and practices and are more open to reconfigure the business strategy. Besides the expectation that business-level strategic change is more likely for firms with more diverse TMTs, it is also expected that heterogeneous TMTs increase the level of strategic change. Heterogeneity within a TMT leads to diverse opinions and is associated with greater changes to the business strategy (Boeker, 1997). It has been found that several forms of diversity in the TMT stimulate action and innovativeness more quickly and with a greater magnitude, including heterogeneity in age, tenure, gender, function, and nationality (Carpenter, Geletkanycz, & Sanders, 2004). More homogeneous teams could also decide to adjust strategies but will have lower levels of creativity and are more likely to rely on information and perspectives that are closely related to the current strategy (Wiersema &
Bantel, 1992). Thus, more diversity within a TMT should lead to more diverse opinions and greater changes in strategy:
Hypothesis 2: TMT heterogeneity increases the level of business-level strategic change.
Considerable prior research has investigated the direct effects of TMT heterogeneity on strategic
choice or performance (e.g., Simons, Pelled, & Smith, 1999; Tushman & Rosenkopf, 1996). However,
putting together Hypothesis 1 and Hypothesis 2, it is expected that strategic change is mediating the
relationship between TMT heterogeneity and M&A performance. This follows the line of reasoning of
10
Olson, Parayitam, and Twigg (2006), who argue that certain TMT characteristics only indirectly influence firm performance, through strategic choice. They found that functionally diverse teams will provide positive results to the firm because they “will engage in strategic inputs such as innovation”
(Olson, Parayitam, & Twigg, 2006, p. 116). This means that TMT heterogeneity positively influences the change in business strategy which, in turn, would have a positive influence on performance of the M&A. Therefore, the following hypothesis is tested:
Hypothesis 3: Business-level strategic change mediates the relationship between TMT
heterogeneity and M&A performance.
Implementation of business-level strategic change
Heterogeneity and diversity may be needed for a successful formation of a strategy, but “homogeneity and efficiency may be the critical factor in successfully executing what was formed” (Menguc & Auh, 2005, p. 5). As proposed in this section, TMT heterogeneity could well be a double-edged sword and negatively influence the performance of the firm as well. More heterogeneous TMTs are found to have more issues coping with stressful problem situations, have lower understanding amongst each other, reduced communication, and increased internal conflict (Ancona & Caldwell, 1992; Mello & Ruckes, 2006; O’Reilly, Caldwell, & Barnett, 1989). Implementation of strategic changes can be described as the critical link between strategy formulation and the achievement of superior organizational performance (Noble & Mokwa, 1999). This definition underscores the importance of not only forming strategic change but also implementing it and linking it to performance. For successful implementation it is claimed that less diversity is needed within the TMT and that higher levels of heterogeneity actually hamper effective implementation (Menguc & Auh, 2005). It could well be the case that homogenous teams might be more effective in implementing such strategic changes. O’Reilly, Caldwell, and Barnett (1989), for instance, found that homogeneity of tenure in the team was positively related to the group’s social integration, rather than heterogeneity. Moreover, Ancona and Caldwell (1992) argue that homogeneous teams can overcome problems associated with heterogeneous teams, such as poor communication and excessive conflict of interest. Besides, the highest level of consensus can be found in TMTs that are less heterogeneous (Priem, 1990). For the aforementioned arguments the second hypothesis goes as follows:
Hypothesis 4: TMT heterogeneity negatively moderates the relationship between business-
level strategic change and M&A performance, so that for increased levels of TMT heterogeneity the positive effect of business-level strategic change will be weakened.
From the hypotheses presented above, a conceptual model was constructed (see Figure 1). This model
shows that it is expected that TMT heterogeneity affects M&A performance in two ways: (1)
11
indirectly through a higher level of strategic change, and (2) by moderating the relationship between strategic change and M&A performance.
Figure 1 Conceptual model.
METHODOLOGY Research sample
For this study, data will be used that is of secondary nature; the Orbis database by Bureau van Dijk,
Reuters, Bloomberg, and company websites. The companies selected are companies that have engaged
in an M&A in 2013 as acquiring company. From those companies, specific data was required to
measure the dimensions of strategic change. Filtering the companies without the required data left a
population of 1,319 firms. The data on managers from the TMT were collected through company
websites, the Reuters website, and the Bloomberg research database. Measuring TMT heterogeneity
through secondary data on directors and managers has been done before (e.g., Boeker, 1997; Gordon
et al., 2000; Hambrick, Cho, & Chen, 1996; Tushman & Rosenkopf, 1996) and is applied in this study
as well. CEOs and senior executives were selected for inclusion in the TMT and the reason for this is
twofold. First, decisions about the business strategy are not made by the supervisory board or the
board of directors, rather by the executive team. Second, these managers belong to the apex of the
organization and are thus most knowledgeable and influential on the context of strategic changes
(Menguc & Auh, 2005). Therefore, only the members of the executive team (or senior management)
were taken into account. A minimum of five executives was a requirement since a lower amount of
executives could give biased outcomes when it comes to diversity. As many firms in the sample had
small executive teams, or reported information on just a few executives, a great deal of the population
was filtered out. Finally, 127 firms had at least five executives and information available on the
strategic dimensions needed for measuring business-level strategic change. When manually collecting
data on the 127 firms, another eighteen firms were deleted since they either did not exist anymore or
no information on the executives was available through the sources mentioned. This left a final sample
of 109 firms and 778 top managers.
12 Variables and measurement
M&A performance (independent variable) – In the literature there have been many different ways of measuring M&A performance. The most common method uses financial measures to assess performance but it has received criticism because it fails to consider “differences in systematic risk, temporary disequilibrium effects, tax laws, and accounting conventions regarding R&D and advertising” (Benston, 1985; Wernerfelt & Montgomery, 1988). According to Wernerfelt and Montgomery (1988), these issues can be avoided through the use of Tobin’s q, which measures the capital market value of the firm divided by the replacement value of its assets. Data required to measure Tobin’s q is collected from Thomson Reuters Datastream. As an increase of performance due to changes in strategy is expected to have a delay, the Tobin’s q at the end of 2015 is used for analysis, which is one year after the point business-level strategic change was measured. Data for the Tobin’s q was collected through Datastream and calculated using the following formula:
𝑚𝑎𝑟𝑘𝑒𝑡 𝑐𝑎𝑝𝑖𝑡𝑎𝑙𝑖𝑧𝑎𝑡𝑖𝑜𝑛 + 𝑚𝑎𝑟𝑘𝑒𝑡 𝑣𝑎𝑙𝑢𝑒 𝑡𝑜𝑡𝑎𝑙 𝑙𝑖𝑎𝑏𝑖𝑙𝑖𝑡𝑖𝑒𝑠 + 𝑡𝑜𝑡𝑎𝑙 𝑠𝑡𝑜𝑐𝑘
Due to lack of data on market capitalization on nine firms, they were left out of the sample. These restrictions decreased the sample to 100 firms, which would be the final sample size of this study.
Strategic change – This study will focus on the acquiring firm since post-acquisition decisions are
“housed primarily within the acquirer’s corporate development department” (Zollo & Singh, 2004, p.
1233). The degree of business-level strategic change by each company will be measured after the acquisition was completed. Finkelstein and Hambrick (1990) have operationalized a company’s business strategy, which was later utilized by other authors (e.g., Carpenter, 2000; Zhang, 2006; Zhang
& Rajagopalan, 2010). To operationalize strategic change, a composite measure of six strategy dimensions is used. The six key strategic dimensions are: (1) research and development intensity (R&D/sales); (2) plant and machinery newness (net P&M/gross P&M); (3) cost efficiency (cost of goods sold/sales); (4) inventory levels (net stated inventory/sales); (5) overhead costs (other operating expenses/sales); and (6) financial leverage (liabilities/shareholder funds). Finkelstein and Hambrick (1990, p. 491) explain that the choice for these dimensions is fourfold: “(a) they are potentially controllable by top managers; (b) they may have an important effect on firm performance; (c) they are complementary, each focusing on an important but specific aspect of the firm's strategic profile; and (d) they are amenable to data collection and have relatively reliable comparability across firms within an industry.” Furthermore, all dimensions have been used in prior studies and have been found accepted measures of strategic dimensions (Finkelstein & Hambrick, 1990; Schendel & Patton, 1978).
Data on these dimensions were drawn from Orbis.
13
The approach by Zhang and Rajagopalan (2010) was adapted to obtain a composite measure of strategic change from the six aforementioned dimensions. First, the differences in the ratios between the current and prior year were calculated. For instance, Δfirm cost efficiency = (firm cost efficiency
t– firm cost efficiency
t-1). Next, the measure was adjusted for the industry effect. This was done through subtracting the industry median changes in these ratios. In the above example it would mean the following: industry-adjusted Δcost efficiency = (firm cost efficiency
t– firm cost efficiency
-1) – (industry median cost efficiency
t– industry median cost efficiency
t-1). Then, the absolute values of the industry-adjusted changes in the sample were calculated and standardized within the sample.
Finally, the average of the six standardized values made up the composite measure of strategic change.
Top management team heterogeneity – The focus of this study on the role of the management team instead of a single individual (i.e., the CEO) is consistent with strategic management literature since it is unlikely that managerial decisions are made exclusively by one person (e.g., Amason, 1996;
Hambrick, 2007). TMT heterogeneity was measured by eight variables: (1) diversity in age, (2) in gender, (3) in tenure, (4) in educational levels, (5) in educational backgrounds, (6) in functional backgrounds, (7) in international experience, and (8) in nationalities. Some data is categorical (e.g., education and international experience) while other data is continuous (e.g., age and tenure). In Appendix A, the definition and codes for each of the eight diversity variables can be found. For educational background it was sometimes found that the executive had degrees in different areas. In those cases, the educational background was reported in which the highest degree was obtained (Wiersema & Bantel, 1992).
Measuring averages would be contradictory to measuring diversity and thus an index of diversity is applied, which measures variation in categorical data through a mathematical analysis. For the measurement of diversity, both the standard deviation and the Herfindahl index method have been adopted. Measuring diversity through a mathematical analysis such as the Herfindahl index, is “the most widely accepted measure for providing information on a distribution” (Chemers & Murphy, 1995, p. 196; Gibbs & Martin, 1967). Moreover, this kind of analysis was adopted in other demographic research (Blau, 2000; Hambrick, Cho, & Chen, 1996; Richard et al., 2004). For variables consisting of interval-level data, such as age and tenure, the preferable method to capture diversity is the coefficient of variation, defined as the standard deviation divided by the mean. (Allison, 1978;
Harrison & Klein, 2007). The Herfindahl index for diversity is measured through the following equation:
𝐻 = 1 − ∑ 𝑝
𝑖2𝑁 𝑖=1
14
Where H is the heterogeneity measure, N is the number of categories, and p
iis the percentage of TMT members in each of the categories. The value of H can range from 0 to 1, with higher values indicating that the TMT is more heterogeneous.
1Finally, the diversity measures of each company were adjusted to the industry. For instance, gender diversity in company A was calculated as follows: gender diversity company A – industry average gender diversity. Then, the absolute values of the industry- adjusted changes in the sample were calculated and standardized within the sample. This was done for all eight diversity categories. In the end, the average of the eight standardized values made up the composite measure of TMT heterogeneity.
Control variables
TMT size – The control variable TMT size will be included in the model. The size of the TMT could influence both the decision making process regarding strategic change and the implementation of the strategic change (Amason & Sapienza, 1997; Pegels, Song, & Yang, 2000). Furthermore, previous studies found a positive relationship between TMT size and a firm’s financial performance (Certo, Lester, Dalton, & Dalton, 2006; Simons, Pelled, & Smith, 1999). The size of the TMT is an additive measure. Calculating the size was a matter of identifying and counting the members on a team (Hoffman, Lheureux, & Lamont, 1997).
TMT tenure – TMT tenure has been identified in the literature as a variable that can influence strategic decision processes and outcomes (e.g., Hambrick & Mason, 1984; Wiersema & Bantel, 1992). For instance, short team tenure was found to be related to receptivity to change and willingness to take risk (Wiersema & Bantel, 1992). The average number of years TMT members had belonged to the TMT was used for TMT tenure (Simons, Pelled, & Smith, 1999).
Firm size – Following Puranam and Srikanth (2006), this study also controlled for firm size. Larger firms typically have access to more resources, which could positively influence performance (Li, 2013). According to the authors, the number of employees is an accurate proxy for firm size and thus firm size is included as a control variable. Firm size is measured as the natural algorithm of the number of employees. Information on the number of employees was obtained from Orbis. This information was checked and complemented through company websites.
1 To illustrate the Herfindahl index, imagine that in a category four different colors are represented: yellow (25%), orange (0%), blue (50%), and red (25%). The percentages are first squared before the sum is computed:
0.252 + 02 + 0.52 + 0.252 = 0.375. The total diversity of this category is then 1 – 0.375 = 0.625. The maximum amount of diversity in a category with four options is ¾ = 0.75. To scale the diversity in colors in percentages the diversity needs to be divided by the maximum amount of diversity: 0.625/0.75 = 83.33%. The diversity score for colors would have a score of 83.33% on the diversity index.
15
Firm age – Older firms, like larger firms, typically have access to a greater pool of resources.
Furthermore, it could provide the firm with certain historical experiences that can be adopted in handling the M&A (Li, 2013). On the other hand, Loderer and Waelchli (2009) found that older firms slowly lose their ability to compete, resulting in a decline in Tobin’s Q ratios. From Orbis, information was gathered on the date of corporation, which was then deducted from 2016. The natural logarithm of the number of years in corporation represented the variable firm age.
Prior firm performance – Previous research has found that prior firm performance can be a significant influencer of both strategic change and firm performance (Zhang & Rajagopalan, 2010). Therefore, it is included as a control variable in this study. Prior firm performance was operationalized as the Tobin’s Q in the year before the M&A, 2012, to ensure the M&A was not of influence (Zhang &
Rajagopalan, 2010). The data for prior firm performance were obtained from Thomson Reuters Datastream.
Industry – Lastly, in order to control for industry specificities regarding M&A performance, a control variable was added following Waddock and Graves (1994, 1997). They found clear differences in performance and strategies among different industries. Controlling for industry takes these differences into account. The industry was determined by the 4-digit US SIC as can be seen in Table 1 (Waddock
& Graves, 1997). In the model, the industries are represented through industry binary variables.
Table 1
Specification and frequency of industries.
Industry SIC N
1 Mining, construction 100-1999 13
2 Food, textiles, apparel 2000-2390 5
3 Forest products, paper, publishing 2391-2780 5
4 Chemicals, pharmaceuticals 2781-2890 7
5 Refining, rubber, plastic 2891-3199 4
6 Containers, steel, heavy mfg. 3200-3569 14
7 Computers, autos, aerospace 3570-3990 14
8 Transportation 3991-4731 5
9 Telephone, utilities 4732-4991 15
10 Wholesale, retail 4992-5990 5
11 Bank, financial services 5991-6799 4
12 Hotel, entertainment, hospital management 6800-8744 9
Analysis
In this study, several analyses were performed in order to test the different hypotheses. To clearly
structure the analyses the test were performed in a certain sequence, following the hypothesis. To test
16
the first hypothesis a multiple linear regression was performed, since it is expected that strategic change influences M&A performance. However, since control variables were included the appropriate method of analysis was through a multiple linear regression using the following formula:
Y = β
0+ β
1X
1+ β
2C
2+ β
3C
3+ β
4C
4+ β
5C
5+ β
6C
6+ … + β
kC
kWhere Y is the outcome (M&A performance), β
0is the regression intercept or regression constant, β
1is the regression coefficient for predictor variable X
1(strategic change), and C
2until C
kare the control variables in this study (TMT size, firm size, firm age, and industry). Lastly, β
2until β
kare the regression coefficients for the control variables.
Mediation analysis
The second and third hypothesis, which state that TMT heterogeneity positively influences the amount of strategic change (H2) and that TMT heterogeneity positively influences M&A performance through strategic change (H3). In other words, Hypothesis 3 expects strategic change to be a mediating variable. These hypotheses will be tested through a mediation analysis. Since a mediator analysis consist of multiple effects, both the constituent components of the indirect effect must be estimated, meaning the effect of TMT heterogeneity on strategic change as well as the effect of strategic change on M&A performance. Since there are two consequent variables (i.e., strategic change and M&A performance), two equations represent this analysis:
M = β
1+ aX
1Y = β
2+ c’X
1+ bM
Where β
1and β
2are the regression intercepts, a, b, and c’ are the regression coefficients given to the
predictors in this model. Strategic change is represented by M, TMT heterogeneity is represented by X,
and Y stands for M&A performance. In testing the mediating effect of strategic change, the direct
effect of TMT heterogeneity on M&A performance has to be taken into account as well. This is to
determine whether there are indirect effects, full mediation effects or partial mediation effects
(Mathieu & Taylor, 2006). The indirect effect in this case is the effect of TMT heterogeneity on M&A
performance through strategic change. The full mediation effect hypothesizes that strategic change is
fully accountable for the significance of TMT heterogeneity on M&A performance. That is, once
strategic change is found to significantly influence M&A performance, the direct effect of TMT
heterogeneity on M&A performance is no longer significant. Lastly, partial mediation occurs when
strategic change is partly accountable for the effect of TMT heterogeneity on M&A performance. So
both a significant relationship is found between TMT heterogeneity and strategic change and between
TMT heterogeneity and M&A performance (Mathieu & Taylor, 2006). For a visualization of the full
17
mediation effect, partial mediation effect, and indirect effect, see Appendix B. The mediation analyses in this study will differ slightly from the most popular method by Baron and Kenny (1986). The mediation analyses as described by Baron and Kenny (1986) suffers from low statistical power (MacKinnon et al., 2002) and in smaller samples their method produces biased results (Preacher &
Hayes, 2004). Instead, a formal significance test of the indirect effect gives more accurate results and overcome the “shortcomings inherent in the Baron and Kenny method” (Preacher & Hayes, 2004, p.
719). To estimate the indirect effect with a formal significance test as described by Preacher and Hayes (2004), the PROCESS macro from Hayes (2014) will be used. Besides the fact that it increases the accuracy of the analyses, it also allows the user to control for variables during the mediation analysis. Additionally, the bias-corrected bootstrap method was used to produce the confidence intervals. Bootstrapping can best be described according to the following passage from Hayes (2009, p. 412): “bootstrapping generates an empirical representation of the sampling distribution of the indirect effect by treating the obtained sample of size n as a representation of the population in miniature, one that is repeatedly resampled during analysis as a means of mimicking the original sampling process”. Once the resample is constructed, the coefficients are estimated for this resampled dataset. This is repeated for typically at least a total of 1000 times. After this process, there will be at least 1000 estimates of the indirect effect from which a confidence interval is estimated based on the sample of size n from the original population.
2According to MacKinnon, Lockwood, and Williams (2004), the bootstrap method produces the most accurate confidence intervals for mediation effects.
Bootstrapping is superior to, for instance, the distribution of the product method when testing hypotheses about indirect effects as “it makes fewer unrealistic assumptions” (Hayes, 2012, p. 6).
Furthermore, the bootstrapping method produces a test that is not based on large-sample theory, meaning it can be applied to small samples with greater confidence (Preacher & Hayes, 2004). Thus, for the indirect effects of the mediation analysis, significance was tested by investigating the bootstrap confidence intervals.
Moderation analysis
To test the moderating effect of heterogeneity (H4), a new model was constructed including TMT heterogeneity, a so called moderator analysis. When TMT heterogeneity can predict the effect of strategic change on M&A performance it may be called a moderator, or that TMT heterogeneity and strategic change interact in their influence on M&A performance (Hayes, 2014). The following equation can be used to test the moderating effect of TMT heterogeneity:
Y = β
0+ β
1X + β
2Z + β
3X *Z
2 Assume for the sake of illustration that 1,000 bootstrap samples have been requested of the original sample size. To derive the 95% confidence interval, the lower limit of the confidence interval is defined as the 25th score and the upper limit as the 976th score in the distribution.
18
In the equation, Y will be M&A performance, β
0is the constant, β
1is the regression coefficient for predictor X (strategic change), and β
2is the regression coefficient for the moderator (TMT heterogeneity). Finally, β
3is the regression coefficient for the interaction between X and Z, which is the interaction between strategic change and TMT heterogeneity. To prove the moderating effect of TMT heterogeneity, β
3has to be significantly different from 0. Otherwise, there is no interaction of a moderating effect. To mitigate the potential threat of multicollinearity, the variables in the interaction term were mean-centered (Kopalle & Lehmann, 2006). That is, for TMT heterogeneity and strategic change the mean was subtracted from each score prior to analysis.
RESULTS Descriptive statistics and correlations
Table 2 provides the descriptive statistics for the main variables used in this study. From the initial descriptive statistics, it appeared that some variables had an excessively high amount of skewness and kurtosis, pointing to non-normality. However, for analysis with continuous variables and continuous- like variables, such as categorical variables with many categories, Tabachnick and Fidell (2007) recommend normally distributed variables. Univariate skewness and kurtosis have a threshold of respectively 2.00 and 5.00 and therefore the variables firm performance, strategic change, TMT size and firm size had to be transformed into natural log (normal) variables. Variables are given normality through mathematical transformation and only those in the sample with valid natural logs can be maintained. Fortunately, all variables in the sample could be given normality and thus the sample size (N = 100) was preserved. When checking the variance inflation factors (VIFs), no multicollinearity was found between the variables of interest since it is well below the threshold of 10. In Table 2, the descriptive statistics for the industry control dummy variables are omitted. The division of companies in each industry can be found in the methodology section in Table 1.
Table 2 Descriptive statistics.
N = 100 Variable Mean SD VIF Min Max Skewness Kurtosis 1 M&A performance 0.10 0.60 3.00 -1.16 1.91 0.25 2.82 2 Strategic change -2.13 1.76 1.61 -7.01 1.66 -0.21 2.96 3 TMT heterogeneity 0.00 1.00 1.58 -2.58 1.76 0.14 2.76
4 TMT size 1.91 0.30 1.30 1.61 2.94 1.02 3.79
5 TMT tenure 1.45 0.67 1.31 0.00 3.14 -0.01 2.76 6 Firm size 8.79 1.93 1.31 2.71 12.77 -0.14 3.12
7 Firm age 3.60 0.94 1.31 1.10 6.21 0.32 2.52
8 Prior performance -0.42 0.79 1.67 -2.02 1.78 0.14 2.54
In Table 3 below, the Pearson correlation matrix can be found for the variables in this study.
Interestingly, M&A performance is significantly correlated (on a 1% significance level) to strategic
change, TMT heterogeneity, TMT size, TMT tenure, and prior performance. Strategic change is in its
19
Table 3 Correlation matrix.1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
1 M&A performance 1.00
2 Strategic change 0.38*** 1.00
3 TMT heterogeneity 0.36*** 0.39*** 1.00
4 TMT size 0.38*** 0.00 0.22** 1.00
5 TMT tenure 0.29*** 0.06 -0.07 0.06 1.00
6 Firm size 0.15 -0.13 0.01 0.24** -0.07 1.00
7 Firm age 0.13 0.06 0.07 0.06 0.12 0.22** 1.00
8 Prior performance 0.76*** 0.33*** 0.38*** 0.35*** 0.13 0.16 0.06 1.00
9 Industry 1 -0.22** -0.20** -0.02 -0.08 -0.05 -0.10 -0.14 -0.17* 1.00
10 Industry 2 0.03 0.12 0.07 0.03 -0.13 0.15 0.13 0.01 -0.09 1.00
11 Industry 3 0.06 0.05 0.01 0.04 0.17* 0.15 0.29*** 0.05 -0.09 -0.05 1.00
12 Industry 4 0.03 0.04 0.03 -0.04 0.09 0.11 0.09 0.04 -0.11 -0.06 -0.06 1.00
13 Industry 5 -0.03 0.00 -0.03 0.05 -0.15 0.10 0.16 0.02 -0.08 -0.05 -0.05 -0.06 1.00
14 Industry 6 -0.15 -0.05 -0.12 -0.02 0.03 -0.09 0.02 -0.17* -0.16 -0.09 -0.09 -0.11 -0.08 1.00
15 Industry 7 0.36*** 0.15 0.11 0.20** 0.13 -0.07 -0.13 0.31*** -0.16 -0.09 -0.09 -0.11 -0.08 -0.16 1.00
16 Industry 8 -0.07 -0.02 -0.05 0.03 -0.09 0.10 0.01 -0.16 -0.09 -0.05 -0.05 -0.06 -0.05 -0.09 -0.09 1.00
17 Industry 9 -0.06 -0.20* -0.03 -0.04 -0.20* 0.03 0.02 -0.07 -0.16 -0.10 -0.10 -0.12 -0.09 -0.17* -0.17* -0.10 1.00
18 Industry 10 -0.07 -0.02 0.02 -0.04 0.00 0.06 -0.03 0.03 -0.09 -0.05 -0.05 -0.06 -0.05 -0.09 -0.09 -0.05 -0.10 1.00
19 Industry 11 -0.10 0.16 0.00 -0.12 -0.08 -0.15 -0.05 -0.08 -0.08 -0.05 -0.05 -0.06 -0.04 -0.08 -0.08 -0.05 -0.09 -0.05 1.00
20 Industry 12 0.18* 0.13 0.03 -0.03 0.23** -0.13 -0.18* 0.17* -0.12 -0.07 -0.07 -0.09 -0.06 -0.13 -0.13 -0.07 -0.13 -0.07 -0.06 1.00
Note: Pearson’s correlation coefficients.
Standard errors in parentheses. Significance: ***p<0.01, **p<0.05, *p<0.1.
20
turn related to TMT heterogeneity and prior firm performance. Furthermore, TMT heterogeneity and firm size are both correlated to TMT size. This seems plausible since a greater board size allows more heterogeneity in the executive board. An explanation for the relationship between firm size and TMT size could be that larger firms employ more senior managers to lead the organization. The age of the firm and the firm size are also significantly correlated. Again reasonable since older firms have had more time to grow and thus are more likely to employ a greater amount of people. Lastly, M&A performance, strategic change, TMT size, TMT tenure, firm age, and prior performance are all correlated to some of the industry control variables.
Tests of hypotheses
In this section, the results of the analysis will be presented and the hypotheses will be tested. Table 4 presents the main findings of the first regression analyses. The baseline model (Model 1), which only includes control variables, shows that TMT tenure and prior performance have a positive influence on M&A performance on a 5% and 1% significance level. TMT size, firm size and firm age do not turn out to be significant. This is in line with previous findings, since conflicting results have been presented on the effect of size and age on performance (Li, 2013; Loderer and Waelchli, 2009).
The correlation matrix in Table 3 provides preliminary support for Hypothesis 1 since a positive relationship was found between strategic change and M&A performance (r = 0.38, p < 0.01).
To examine this relationship more thoroughly, strategic change was included in the second model as an independent variable (Hypothesis 1). Model 2 shows that strategic change positively influences M&A performance (β = 0.05, p < 0.05). Therefore, the first hypothesis can be accepted. Interesting to see is that with the inclusion of strategic change as the independent variable, the second model represents a significant improvement over the first model (ΔR
2= 0.02, p < 0.01). Besides strategic change, the control variables TMT size, TMT tenure, and prior performance significantly affect M&A performance.
Mediation analysis
Through a mediation analysis, the second and third hypotheses were tested. This is to estimate the
relationship between TMT heterogeneity and strategic change (Hypothesis 2) and to see whether TMT
heterogeneity indirectly influences M&A performance, through strategic change (Hypothesis 3). As
stated in the methodological section, the mediation analysis consists of two formulas. To test for
indirect effects, the effect of TMT heterogeneity on the mediator (strategic change) has to be assessed
first. This relationship is highly significant, as can be seen in Table 5. In the table, the baseline model
(Model 1) again includes just the control variables. Here, the control variables firm size and prior
performance are significantly related to strategic change. Firm size is negatively related to strategic
change, which could be explained by Chen and Hambrick (1995) who argue that changes are relatively
much greater and visible in smaller firms. Furthermore, as expected, prior firm performance is a
21
Table 4Results of regression analyses for M&A performance.
M&A performance
Independent variables Model 1 Model 2
Hypothesis-testing variable
Strategic change 0.0540** (0.0245)
Firm-level controls
TMT size 0.2052 (0.1432) 0.2339* (0.1406)
TMT tenure 0.1598** (0.0644) 0.1652** (0.0630)
Firm size 0.0104 (0.0223) 0.0198 (0.0222)
Firm age 0.0469 (0.0461) 0.0403 (0.0452)
Prior performance 0.4829*** (0.0577) 0.4419*** (0.0594) Industry-level controls
Industry 1 -0.0157 (0.2267) 0.0241 (0.2224)
Industry 2 0.1871 (0.2555) 0.1296 (0.2511)
Industry 3 0.0210 (0.2627) -0.0032 (0.2570)
Industry 4 0.0574 (0.2446) 0.0429 (0.2392)
Industry 6 0.0460 (0.2229) 0.0482 (0.2179)
Industry 7 0.2854 (0.2308) 0.2709 (0.2257)
Industry 8 0.2093 (0.2595) 0.1888 (0.2539)
Industry 9 0.1225 (0.2172) 0.1627 (0.2131)
Industry 10 -0.1055 (0.2613) -0.0956 (0.2555)
Industry 11 0.0562 (0.2771) -0.0102 (0.2726)
Industry 12 0.2059 (0.2500) 0.1863 (0.2446)
Constant -0.6806 (0.4208) -0.7044 (0.4115)
Model statistics
N 100 100
Goodness-of-fit of the model (F) 10.11*** 10.24***
R² 0.6609 0.6799
ΔR² 0.02
Standard errors in parentheses. Significance: ***p<0.01, **p<0.05, *p<0.1.
significant indicator for strategic change (Zhang & Rajagopalan, 2010). In Model 2 of Table 5, TMT heterogeneity is added and found to be significant and positively related to strategic change (β = 0.5430, p < 0.01). Including TMT heterogeneity also significantly increases the fit of the model (ΔR
2= 0.08, p < 0.01). Therefore, it can be concluded that heterogeneous TMTs increase the amount of strategic change and Hypothesis 2 can be accepted.
Now, for Hypothesis 3, the indirect effect of TMT heterogeneity on M&A performance
becomes important. The results of the mediation analysis can be found in Table 6. In Model 1 of Table
6, the control variables were first related to M&A performance. In Model 2, the total effect of TMT
heterogeneity on M&A performance is shown. This includes all control variables and TMT
heterogeneity as an independent variable. Interestingly, TMT heterogeneity was not found to be
significantly related to M&A performance. This is rather unexpected when looking at the correlation
matrix in Table 3 where TMT heterogeneity and M&A performance were significantly correlated (r =
22
Table 5Results of regression analyses for strategic change.
Strategic change
Independent variables Model 1 Model 2
Hypothesis-testing variable
TMT heterogeneity 0.5430*** (0.1779)
Firm-level controls
TMT size -0.5310 (0.6272) -0.7796 (0.6034)
TMT tenure -0.1008 (0.2818) 0.0155 (0.2713)
Firm size -0.1746* (0.0975) -0.1367 (0.0937)
Firm age 0.1217 (0.2018) 0.0670 (0.1932)
Prior performance 0.7600*** (0.2527) 0.5103** (0.2544) Industry-level controls
Industry 1 -0.7352 (0.9935) -0.9662 (0.9501)
Industry 2 1.0650 (1.1192) 0.7725 (1.0713)
Industry 3 0.4470 (1.1509) 0.2869 (1.0985)
Industry 4 0.2687 (1.0715) 0.0203 (1.0248)
Industry 6 -0.0412 (0.9764) -0.1233 (0.9313)
Industry 7 0.2676 (1.0111) 0.1177 (0.9652)
Industry 8 0.3801 (1.1371) 0.2130 (1.0854)
Industry 9 -0.7445 (0.9516) -0.8738 (0.9082)
Industry 10 -0.1842 (1.1448) -0.4136 (1.0940)
Industry 11 1.2280 (1.2142) 1.0042 (1.1599)
Industry 12 0.3645 (1.0955) 0.1942 (1.0459)
Constant 0.4424 (1.8438) 0.6738 (1.7595)
Model statistics
N 100 100
Goodness-of-fit of the model (F)
1.75* 2.36***
R² 0.2526 0.3289
ΔR² 0.0763
Standard errors in parentheses. Significance: ***p<0.01, **p<0.05, *p<0.1.
