1
MASTER THESIS
The Antecedents of Innovation, Slack Resources vs. TMT Diversity:
A Critical Appraisal and Contribution to Both the Behavioral Theory
and the Strategic Choice Theory of the Firm
Name: Fernant Bos
Student Number: 11219033
Supervisor: Dr. Marten Stienstra
Date of Submission: 31 August 2018
Version: Final version
Executive Programme in Management Studies – Strategy Track
Amsterdam Business School, University of Amsterdam
2
Statement of Originality
This document is written by Student Fernant Bos who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents
3
Table of Contents
1 Introduction ... 5
2 Literature review ... 10
2.1 Innovation ... 10
2.2 Behavioral theory and slack resources ... 13
2.3 Strategic choice theory and TMT diversity ... 20
2.4 Slack resources (behavioral theory), TMT diversity (strategic choice theory) and innovation ... 27 2.5 Conceptual model ... 30 3 Methodology ... 31 3.1 Type of research ... 31 3.2 Population vs. Sample ... 32 3.3 Measurement of variables ... 32 3.4 Data collection ... 37
3.5 Reliability and validity ... 38
3.6 Statistical analysis ... 39 4 Results ... 40 4.1 Univariate analyses ... 40 4.2 Bivariate analyses ... 42 4.3 Regression analyses ... 43 4.4 Hypothesis testing ... 49
5 Discussion and conclusion... 50
5.1 Discussion of main findings ... 50
5.2 Contributions ... 54
5.3 Limitations and future research ... 55
5.4 Conclusion ... 57
References ... 60
Appendix A: Calculation sample size ... 69
Appendix B: Codebook ... 70
Appendix C: Distribution and descriptive statistics variables ... 72
4
Abstract
This study aims to contribute to the debate how organizational innovation can be explained from multiple strategic renewal perspectives. Specifically, the focus is on how the adaptation perspectives behavioral theory and upper echelon theory (strategic choice) predict organization innovation. Both perspectives differ in their explanation how innovation is initiated in the firm. The behavioral theory of the firm argues that innovation is the responsibility of all the groups in the coalition and slack
resources are the enablers of these coalition to negotiate and innovate. Strategic choice theory argues that top management team (TMT) diversity is an important predictor for innovation. As strategic choice theory is partly founded on the behavioral theory of the firm, both perspectives may have complementary, but also conflicting effect on the degree of organizational innovation. This is reflected in the research question of how do the degree of slack resources and the degree of TMT diversity, both separately and jointly, influence the degree of innovation of the firm? Cross-sectional
data from the Osiris database is collected from a sample of 100 global fortune 500 firms in the year 2016. OLS regression results provide evidence that both theories explain a large part of variance in innovation. The constructs slack resources and TMT diversity independently show a positive significant effect on innovation. However, if the firm only depends on either slack resources or TMT diversity as antecedent of innovation, this mostly leads to suboptimal levels of innovation. A positive significant interaction term suggests that slack resources and TMT diversity reinforce each other in explaining organizational innovation. What leads to higher levels of innovation than both constructs separately and therefore better explains the innovation of the firm. Overall, the findings add to a deepened understanding of how various organizational attributes interact in explaining organizational innovation. Future research may adopt a longitudinal research design to increase causality. Another avenue for future research is to compare how selection perspectives in strategic renewal interact with either behavioral theory and/or strategic choice theory in explaining organizational innovation.
Keywords: behavioral theory, strategic choice theory, slack resources, TMT diversity,
5
1 Introduction
Imagine a world where all the products are the same, firms cannot make a performance
difference, everybody uses the same resources, all have access to the same technologies, and all firms earn zero economic profit. In the literature, this situation is the so-called neoclassical model of perfect competition. This theory was introduced in 1900 by Thorstein Veblen in the article 'Preconceptions of Economic Science’. In today’s world, such a situation is hard to imagine and occurs rarely. There are major differences between firms, and they can be distinguished from other firms in multiple ways. Moreover, any deviation from the assumptions of the model of perfect competition may lead to a possible source of competitive advantage that can lead to performance differentials. These differentials could also be created by continuously changing customer demands and disruptive technologies. For a firm to survive and prosper in such a competitive environment, it must respond to these changes and renew strategically to stay viable in the long run and sustain competitive advantage (Agarwal & Helfat, 2009). One way for a firm to distinguish itself from others is to innovate its products and services and to make sure it continuously adapts to new circumstances.
In the literature, innovation is often linked to exploration (Gupta et al. 2006; March 1991). March (1991, p. 72) describes exploration as including things captured by terms such as
“search, variation, risk-taking, experimentation, play, flexibility, discovery, innovation.” In building on March’s work, Jansen et al. (2006) use the term ‘exploratory innovation’ to described firms pursuing new knowledge and developing products and services for emerging customers and markets. So, for the long-term survival of the firm, the top management team should invest in exploration (innovation) (Levinthal & March 1993). However, results of exploration are often uncertain and do not show direct results (Burns & Stalker, 1961). Decisions about exploration can be considered as strategic choices of the firm (Child, 1972). This uncertainty makes it difficult for managers to pick and choose profitable and successful innovation projects, and
6 every innovation project needs to be considered in relation with a realistic estimation of its future benefits and transaction costs (Williamson, 1981).
To stay competitive and innovative In an environment in which firms need to adapt continuously to changing market conditions, and where trends and developments are diverse and uncertain, a firm needs multiple lenses and knowledge to detect emerging trends to be able to stay competitive and innovative (Menguc & Auh, 2005). In other words, to be able to renew itself a firm needs a sufficient degree of diversity. This is certainly true for a firm’s top
management team, which is responsible for its resource allocation and strategic direction (Hambrick & Mason, 1984). The top executives of the TMT have a direct influence on a firm’s direction, including its strategic and operational decisions. They must also attend to the firm's strategy on innovation and empower it (Castanias & Helfat, 1991; Hambrick & Mason, 1984). The upper echelon theory argues that a firm’s outcomes are partially predicted by the
managerial background characteristics of the top management team (TMT) (Hambrick & Mason, 1984). Diversity refers to the differences in the composition of a group of individuals, in this case the TMT. The idea is that in a complex and uncertain situation, diversity may lead to better problem solving (Menguc & Auh, 2005). According to Greve (2007), this problem solving is an essential element of a firm’s problematic search that lead to innovation. In other words, strategic decisions made by the TMT that do not "lend themselves to more calculable solutions"
(Hambrick & Mason, 1984, p. 195), may require a more diverse view of the TMT.
The strategic choice theory (Child, 1972) as described above is based on the behavioral theory of firms by Cyert and March (1964) and the agency theory (Lawrence & Lorsch, 1967). Therefore, the three theories show a considerable amount of overlap. The strategic choice theory and the behavioral theory differ in the way they perceive the function and role of the TMT in a firm. In contrast to the strategic choice theory, the behavioral theory considers a firm as a coalition of groups, thereby arguing that the TMT is not the only relevant group in it and that the
7 TMT is influenced by stockholders, customers, employees, suppliers, and so on. Therefore, according to behavioral theory, the TMT is not the atomistic decision maker in the firm. The various departments in the firm and other stakeholder influence decisions on the goals set by the TMT and their implementation. TMT decisions are made not only by financial considerations but also by general considerations towards the health of the organization, which are negotiated by the different stakeholders. Therefore, the organization needs excess resources to support this previously mentioned process of stakeholder negotiation. Also, according to Cyert and March (1964), the innovation is generated within the organization and between the coalition of groups. Therefore, innovation requires excess resources in the whole organization. According to Greve (2007), excess resources (slack) are an essential prerequisite for innovation in an
organization (slack search). Organizations without slack usually have stricter performance monitoring and fewer opportunities for experimenting (Lounamaa & March, 1987).
To improve a firm’s innovation strategy, there are two different approaches in the
literature as previously explained: firstly, from the behavioral theory perspective the relationship between slack resources and innovation, and secondly, from strategic choice theory the
relationship between TMT diversity and innovation. This debate is ongoing in the literature, and instead of regarding these theories as competing explanations for organizational innovation, this study investigates both the complementary and substitution effects between these theories. Moreover, TMT diversity may play an important role in deploying slack resources to innovation.
Both theories contribute to the empirical literature on a firm’s innovation strategy. Both theories show mixed findings and have different perspectives on what the best outcome is in relation to stimulating innovation in the organization. In the behavioral theory on firms, some authors argue that slack resources are positive (Bourgeois, 1981; Cyert & March,1963; Greve, 2007; Levinthal & March 1981; 1976), while others argue that they are negative (Jensen 1986, 1993;
8 Liebenstein 1969; Williamson, 1981) or posit an inverted U-shape with regard to the innovation performance of slack resources (Nohria & Gulati, 1997). With regard to the upper echelons (strategic choice) theory, some authors argue that TMT diversity has an positive effect on innovation (Bantel & Jackson, 1989; Talke, 2011). Still others argue that there may be a limit to TMT diversity, and that too much diversity has a negative effect on innovation (Hambrick et al., 1996; Hambrick & Mason, 1984; Østergaard et al., 2011)
Therefore, it is interesting to see how both theories interact: if they strengthen each other or cancel each other out. Furthermore, this study expects that a high degree of slack resources and a high degree of TMT diversity may lead to an increase in innovation. This suggests that in combination, the strategic choice and behavioral theories both partly explain innovation in a firm. Moreover, other interesting findings could be that a sufficient amount of slack resources is enough for a firm to innovate, and it can compensate for a low degree of diversity in the TMT (homogeneous TMT and is related and in favor of the behavioral theory regarding firms. Conversely, it also could be argued that a high degree of TMT diversity could compensate the lack of resources available regarding the innovation of the firm, this is in relation and in favor of the strategic choice theory of the firm
This study tries to contribute to the limited number of empirical studies available and to
contribute to the inconclusive findings from both behavioral theory of the firm (slack resources) and strategic choice theory (TMT diversity) on innovation. Furthermore, this study will contribute to the literature on TMT diversity by using it to moderate the relationship between slack
9 - How do the degree of slack resources and the degree of TMT diversity, both separately
and jointly, influence the degree of innovation of the firm?
To answer this question, this research is built up as follows. Chapter two discusses the
theoretical framework, and the relevant aspects associated; the hypotheses is also formulated and the conceptual model presented. The third chapter explains the methodology, the fourth chapter presents the results, and the last chapter contains the discussion and conclusion.
10
2 Literature review
This chapter will provide an overview of the literature and pinpoint relevant aspects. In particular, the relation between the behavioral theory (slack resources) and its effect on
innovation will be discussed, as well as the strategic choice theory (TMT diversity) and its effect on innovation. Furthermore, the interaction effect of the strategic choice and behavioral theories on a firm's innovation will be examined. This chapter is finalized with the conceptual model of this study, including the defined relations and hypotheses used to answer the research question.
2.1 Innovation
Innovation is a subject studied extensively in management sciences. One of the more popular theories in the field is Schumpeter’s theory of creative destruction (1942). According to him, the economy develops in cycles: new technologies and new firms use new technologies to destroy the existing economic structure by offering a product in an innovative way or by offering an entirely new one. In time, innovative firms themselves become part of the established order, to be outcompeted in turn by new innovators in a never-ending process of rising and falling. However, Schumpeter' theory focuses mainly on technical innovation as the only real source of economic growth. Successful innovation provides temporary market power, which affects the profits and market shares of firms based on previous techniques. After Schumpeter (1942) came many small adjustments to the way the literature defined innovation.
For example, Utterback and Abernathy (1975) use the terms product innovation and process innovation. Process innovation is the introduction of new elements to improve
productivity. Product innovation is the introduction of a new product or service to meet the demand of the external user or market. Dewar and Dutton (1986) use two different types of innovations within the framework of technology: radical and incremental innovations. Radical
11 innovations are fundamental changes that ensure revolutionary changes in technology.
Incremental innovations involve minor improvements or adjustments to the existing technology. The big difference is the degree of new knowledge that is embedded in the innovation. Another distinction is made by Damanpour (1991), who speaks of two types of innovations: technical and administrative. Technological innovations involve a new process or product, while
administrative ones introduce new procedures, policies, or organizational forms. He also argues that all types of innovation include a different decision-making process. March (1991)
distinguishes between exploration and exploitation. Exploration is often associated with the use of unfamiliar technologies and the creation of products with unknown demand or a search for new knowledge. These activities are not feasible in the short term, nor do they show direct results. Results are uncertain and may have distant benefits. Exploitation is associated with the renewal of existing products, technologies, and knowledge. The results are more convincing, and the direct benefits are clearer. However, both are necessary for the performance and survival of the firm, and the literature has produced many studies investigating the balance between exploration and exploitation. As argued by Levinthal and March (1993, p.105), "An organization that engages exclusively in exploration will ordinarily suffer from the fact that it never gains the returns of its knowledge. An organization that engages exclusively in exploitation will ordinarily suffer from obsolescence. The basic problem confronting an
organization is to engage in sufficient exploitation to ensure its current viability and, at the same time, to devote enough energy to exploration to ensure its future viability. Survival requires a balance, and the precise mix of exploitation and exploration that is optimal is hard to specify." (Levinthal & March,1993, p.105). This analysis suggests that according to Levinthal and March (1993) an organization needs excess (slack) resources to enable it to invest both in exploration and exploitation.
12 Thus, innovation is a complex balance between a firm’s short- and long-term innovation activities. In the arguments of Utterback and Abernathy (1975) and Levinthal and March (1993), the adoption of an innovative idea is implicitly included in the definitions and arguments.
Although as argued by Crossan and Apaydin (2010), the innovation literature tends to focus on the creativity phase, thereby ignoring the application or adoption element of innovation.
Damanpour and Evan (1984) had already pointed out that the adoption of a new idea an integral part is of the innovation process and its outcome. They argue that "a new idea was considered as an innovation when it actually began to be used" (Damanpour & Evan, 1984, p.393), and "the management's decision to adopt a new idea does not necessarily guarantee its implementation. Other requirements, such as the acceptance of the new idea by organizational members or necessary modifications in the procedures, must be satisfied. Only after organizational members had begun to use a new device or system, or a new policy or program, could we assume that the adoption had taken place" (Damanpour & Evan, 1984, p.393). Implicitly they imply that innovate ideas can occur in every part of the organization, but only after the
organization actually uses it can it be considered an innovation.
In contrast, the strategic choice (Child, 1972) and related upper echelon theories (Hambrick & Mason, 1984) argue that innovation mostly depends on the personal
characteristics and choices of the upper echelon / top management team. According to these theories, innovation leadership (Crossland & Hambrick, 2007) is a strategic choice. They argue that leaders have vital and multiple roles in promoting, supporting, and implementing innovation by giving guidance, but also by creating efficient and effective interactions between the different groups (departments) of the firm (Mumford & Licuanan, 2004; West et al., 2003). In their role of controlling the division and allocation of resources, they can also direct the emphasis and speed of innovation within the organization. According to Sternberg et al. (2003), the motivation of top executives to assume these roles depends on their perception of the environmental threats and
13 opportunities. Therefore, a diverse top management team enables the firm to perceive different opportunities and threats in the environment (different cognitions) in a timely manner and to act on them. These leadership theories thus assume that control of the organization is focused on the top management and that the organization can be controlled from the top.
Crossan and Apaydin (2010) argued that due to the fragmented definitions of innovation, a new more inclusive one was necessary. Crossan and Apaydin (2010) defined it as "production or adoption, assimilation, and exploitation of a value-added novelty in economic and social spheres; renewal and enlargement of products, services, and markets; development of new methods of production; and establishment of new management systems" (Crossan & Apaydin, 2010, p.1155). In line with those authors, this study looks at the organizational determinants of innovation because, according to them, these determinants are highly fragmented, and the various theories lack interconnectedness. This study focuses on the competing or
complementary aspects of the behavioral and strategic choice theories (upper echelon theory) of a firm. The motivation to do so is their competing views on the origin of innovation.
2.2 Behavioral theory and slack resources
Using the behavioral theory, this section focuses on firms’ "slack resources" in terms of their effect on innovation. A 1938 article by Bernard was one of the first to use the concept of organizational slack (he used the term “inducement-contribution ratio”). This ratio shows how much a firm is willing to pay for an employee than is strictly necessary. The idea of this ratio was to attract highly qualified employees to sustain growth and increase performance. The term ‘slack’ was used for the first time in 1958 in an article by March and Simon, who defined it as ‘‘the difference between total resources and total necessary payments”.
A definition of slack commonly used in the literature was introduced by Bourgeois (1981, p. 30): “Organizational slack is that cushion of actual or potential resources which allows an
14 organization to adapt successfully to internal pressures for adjustment or to external pressures
for change in policy, as well as to initiate changes in strategy concerning the external
environment.”
In the same article, Bourgeois (1981) argued that an organization has three reasons to reserve slack resources that can give the firm strategic advantages, divided into one internal and two external reasons. Internally, a firm maintains slack to be able to meet a possible increase in the level of activity and thus to be able to meet a possible increase in demand. The other two reasons are mainly related to external elements: (1) slack resources enable the firm to adapt to a changing environment with no major consequences for it or its stakeholders and (2) they offer opportunities to experiment with new product introductions, innovations, or other strategies.
In the literature, not all types of slack are presented as directly feasible or easy to recognize. Therefore, it uses various definitions to describe the types of slack. Moreover, different terms are often used to make the same distinctions between slack resources. The following terms are used in the literature; low discretion vs. high discretion slack (Sharfman et al., 1988), absorbed vs. unabsorbed slack (Singh, 1986), and the distinction between available, recoverable and potential slack (Bourgeois, 1981). Recoverable slack (also referred to as absorbed or
low-discretion slack) denotes resources already committed to the firm (e.g., excess overhead costs).
To be used again when efficiencies are increased, these slack resources offer managers the least room to relocate and are the least efficient in their usage possibilities. Examples include a surplus of capacity and specialized employees. Available slack (also referred to as unabsorbed or high direction slack) represents resources available within the firm but not yet reserved for particular allocations or expenses. This type offers managers many more possibilities for use and are therefore more efficient. The most obvious example of such a source is a surplus of
15 liquid assets, which can lead to an increase in dividend payments to owners or resources
available to fund innovation. Lastly, potential slack (also part of unabsorbed or high direction) indicates resources a firm could generate in the future (e.g., from the environment). Examples include taking on additional debt or implementing a share issue. These surplus resources can be used in almost every situation. Having such resources offers managers more security and safety when making decisions.
So, within this general description of slack, it can be argued that slack resources or excess slack are necessary for innovation and that having slack enables the firm to invest in
experiments and to generate new ideas, which may lead to important innovations and product introductions. Moreover, this excess slack is needed so that the firm does not use operational resources. Next, and in line with the general description of slack resources previously explained, follows the behavioral theory of a firm and its effect on innovation.
The effect of slack resources on innovation
Cyert and March (1964) adopt the term slack in discussing the behavioral theory of firms. They see firms as consisting of coalitions of various groups of stakeholders. To maintain such a coalition, an organization must succeed in compensating its stakeholders adequately. If the resources are available to meet the set requirements and are distributed so that they are ultimately met, then the chance of the coalition succeeding is high. This means that the
resources are distributed throughout the firm as a result of the negotiation process between the different stakeholders. Because the firm can often experience an imbalance between
requirements and received payments, this imbalance can be considered as a form of “organizational slack" (Cyert & March 1964).
16 Moreover, Cyert and March (1964) describe the influence of a decision on a surplus of resources (slack) when the environment changes. If the organization does better than expected, it can successfully use the additional slack available. Part of it slack will then be paid back as extra compensation to the members of the coalition. In a less favorable environment, slack enables the company to continue pursuing its set of organizational objectives. In such circumstances, negotiations arise between the various coalition members because fewer resources are available to compensate each member for what they require. Slack resources have a stabilizing role for an organization and for the aspirations of its coalition members. This happens in two ways. Firstly, the firm will absorb the extra slack resources available if it performs well in carrying out its various activities. In this way, coalition members will not soon increase their requirements. Secondly, slack resources ensure that in bad times the firm can still meet its requirements, so the aspirations of its members do not have to be revised downwards.
Furthermore, according to Cyert and March (1964), to support this process of negotiation and decision-making by the different stakeholders of the coalition, the firm needs excess
resources. Moreover, they argue that the innovation is developed within the organization and through negotiation by its coalition of stakeholders. So, the goals set by the TMT are influenced by the other stakeholders, and the TMT is not an atomistic decision maker, as suggested by the strategic choice and upper-echelon theories. Therefore, the innovation strategy is part of the entire organization, and the innovation can occur anywhere in the firm. As previously explained, they see excess resources as a necessary element providing the opportunity to innovate.
Slack resources and innovation are often linked together in the literature. Moreover, slack resources are often linked to financial slack or financial parameters. (Herold, et al 2006; Lewis & Lewis, 2013; Nohria & Gulati, 1996, 1997). Lewis and Lewis (2013, p.8) describe financial slack as; “cash possessed by the firm that is not committed to a specific foreseeable expense.” So,
17 this description is in line with previously explained ones: unabsorbed slack, available slack, and high discretion slack. The reason why financial slack is strongly related to innovation is that the degree to which firms can achieve high performance is strongly linked to both the successful management of financial resources (Chakravarthy & Balaji, 2011) and the successful
management of innovation (O’Brien, 2003). Therefore, this study will use financial slack as a variable for slack resources.
In the literature, the relationship between organizational slack and innovation shows different findings, and it remains an unanswered empirical question whether slack resources are positive or negative for innovation. Proponents of slack resources argue slack facilitates
innovation because it allows a firm to experiment with new strategies and innovation (Greve, 2007; Cyert & March 1963: Levinthal & March 1981; 1976). Opponents argue that slack inhibits innovation because it makes the firm less efficient, encourages staff to relax, promotes
undisciplined investment in R&D activities, and allows personal-interest innovation projects (Jensen, 1986; Leibenstein, 1969; Williamson, 1963). Moreover, it also could be argued that there may be an optimum level of slack resources, and that too much may cause the use of available resources in poor innovation projects. Too much will ultimately lead to less successful innovation and a waste of resources (Nohria & Gulati, 1997), whereas with too little no
resources would be available for experiments that could lead to important innovation (Nohria & Gulati, 1997). Alternative perspectives on slack resources and innovation will be discussed and substantiated hereafter.
Are slack resources positive or negative for innovation?
The starting point of a positive relationship between slack resources and innovation starts with answering the question of why slack exists. Cyrt and March (1963) provided an answer to this question in the behavioral theory of the firm. They argued that excess resources are needed to
18 negotiate in decision making between groups of stakeholders within the organization, to resolve conflicts within the coalition to prevent them from falling apart. Building on these insights, scholars have argued that organizational negotiation between stakeholders in decision making is an important stimulation for innovation for two reasons: (1) slack allows for the loosening of controls and (2) it represents resources that can be approved because of uncertainty in innovation projects. Slack resources allow individuals and departments to experiment with projects that might not succeed but may lead to important innovations and changes in the future. Therefore, slack resources promote a culture of experimentation in the organization (Bourgeois, 1981). Slack allows firms to experiment more safely with new strategies, such as entering new markets and introducing new products (Hambrick & Snow, 1977; Moses, 1992). Slack resources also promote innovation by allowing a “slack search,” or the pursuit of projects that do not appear to be justified regarding internal market controls, but which have great potential for scientists or other firm champions (Levinthal & March, 1981). Moreover, Greve (2007) argues that excess resources (slack) are an essential prerequisite for innovation in an organization (slack search). Although such projects often fail because of the uncertainty, they can create major positive results for a firm. An example of a successful innovation accidentally discovered because of the result of a slack search is the invention of Post-it notes at 3M (Mokyr, 1992).
In contrast, within the strategic choice and upper-echelon theories, slack resources are regarded as enablers for ambidexterity. In other words, if slack resources are present, a firm could pursue both an explorative and exploitative strategy (Alexiev et al., 2010; Lubatkin et al., 2006). This perspective assumes that if a firm chooses to focus on an exploration or exploitation strategy only, the strategic choice theory does not see slack resources as a prerequisite for innovation. This would suggest that the innovation strategy of the firm within the strategic-choice theory is built more on TMT characteristics than on slack resources.
19 The opposite of positive is negative, and some scholars and organizational economists argue a negative relation between slack resources and innovation. Leibenstein (1969) and Williamson (1963) have adopted a more adverse view of slack resources. They see them as being synonymous with waste and lack of ability more as a reflection of managerial self-interest (agency theory) than as a buffer necessary for organizational adaptation (Cyert & March,1963). Scientists that adhere to the behavioral theory of firms characterize them as coalitions of competing interests of different stakeholders that need to negotiate. However, they see these competing interests as being in favor of the principal–agent relationship theory (Jensen & Meckling, 1976), where agents (stakeholders) think only of their own interests rather than acting in the interests of the firm (Antle & Fellingham, 1990; Jensen & Meckling, 1976). Organizational economists do not see slack resources as the solution for solving these principal–agent
conflicts. They argue that resolving them correctly would be to structure incentives so that they are in line with the interests of principals and agents. They see slack as avoidable costs that should be excluded. Moreover, within the transaction cost theory, the use of slack resources in uncertain experimentation leads to higher transaction costs because not all of the invested resources will end to actual innovation, which leads to inefficient economic benefits (Williamson, 1981).
Jensen (1986, 1993) argued that firms that have a lot of slack resources often invest in dubious projects, such as unrelated acquisitions and research and development projects of interest to the manager. The availability of additional slack resources will boost the resource and development expenditure, but lacking control, these innovations will not add value to the firm. Based on the agency theory, investments with a negative net cash value will be implemented in this case because benefit the managers. Opponents of slack resources thus see a surplus of slack as a sign of inefficiency that decreases the general value of the firm. Leibenstein (1969)
20 even invented the term X-inefficiency to emphasize the discrepancy in slack resources between the actual output and maximum output for a given set of input.
Behavioral theory proponents argue that slack resources are essential for innovation and that they are needed for the creation and the execution of ideas. However, within the strategic choice and the upper-echelon theories, slack resources are not necessarily needed for
innovation. Moreover, this is because these theories assume that innovation comes mainly from TMT characteristics, and that they determine which innovation strategy will be followed, which is therefore independent of slack resources.
In sum, while both use arguments for whether slack resources are positive or negative for innovation. Proponents do recognize that slack resources come with costs, but the benefits outweigh them. Moreover, the behavioral theory sees slack resources as positive for innovation and slack resources as essential for the innovation strategy of the firm. The strategic choice theory sees slack resources, when available, as an opportunity for a firm to be ambidextrous, so that it could invest in both explorative and exploitative strategies. Thus, within the strategic-choice theory, the absence of slack will not influence innovation negatively. Furthermore, this study uses financial slack (unabsorbed slack) as an indicator of slack resources in relation to innovation. This leads to the following hypothesis:
Hypothesis 1: The degree of financial slack is positively related to the degree of innovation
2.3 Strategic choice theory and TMT diversity
This section provides a clear explanation of “top management diversity." Hereafter, the effect of TMT diversity in relation to innovation according to the strategic choice theory will be discussed. As in the previously cited behavioral theory by Cyert and March (1976), the innovation strategy occurs throughout an organization and involves all of its stakeholder groups, not just the TMT.
21 Moreover, within the behavioral theory there is nothing against diversity. In addition, they see the different groups of stakeholders as diverse, although it is not explicitly discussed that this diversity is needed for innovation. Child (1973) argues that within the strategic-choice theory — which is related to the upper-echelon theory (Hambrick & Mason, 1984) — the innovation comes from the TMT, which sees itself as the dominant coalition. Child sees a lack of attention to this dominant group in behavioral theory on firms. Moreover, Simon et al. (2011) argue that this is because in the end the TMT is responsible for the allocation of resources, and managers have an important role in effectively bundling leverage and structuring firm resources.
Diversity in teams has recently been a much-discussed topic in the literature (Carpenter et al. 2004; Certo et al. 2006; Milliken & Martins, 1996; Nielsen, 2010). According to Certo et al. (2006), the influence of diversity on team performance and the quality of team decisions is great. The diversity of the top management team has also had a large influence on team processes (Knight et al., 1999). Within the strategic choice theory, Child (1973) argues the importance of the role of leaders in influencing the firm by making choices through a dynamic political process. An exact definition of TMT according to Finkelstein and Hambrick (1996) is that it is the set of top executives who have a direct impact on the formulation and execution of a firm’s strategy. Diversity refers to the total amount of heterogeneity distributed among the aspects of individual members of the organization (Simons, et al. 1999). In other words, diversity means the differences in the composition of a group of individuals (Kearney et al. 2009). The idea is that diversity has a positive influence on the result when complex, non-standard issues have to be solved or complex decisions have to be made (Hambrick & Mason, 1984). Bringing together a variety of skills, knowledge, experience, and perspective results in better results when dealing with complex issues (Talke,et al. 2011). Also, diversity leads to a greater variety of outcomes and to better problem solving (Menguc & Auh, 2005). Also,
22 according to Greve (2007), this problem solving is an essential element of problematic search and innovation in a firm. He argues that these "problematic search[es]" trigger managers to invest in R&D expenditure (innovation) that may solve this "problematic search" and lead to an increase in innovation and performance. Moreover, it could be argued that more diversity in the TMT leads to more investment in slack resources. As mentioned by Alexiev et al., (2010) and Lubatkin et al., (2006), the availability of slack resources allows for investment in both
exploration and exploitation innovation. It could be argued that more diversity leads to higher reserves of slack resources, which are needed because of the multiple lenses present in the TMT to adapt continuously to changing trends and market conditions (Menguc & Auh, 2005). Harrison and Klein (2007) defined diversity in three ways: separation, variety, and disparity. Separation indicates the differences in positions of unit members. Variety is differences in categories, mainly focused on information, experience, or knowledge (demographic
characteristics). Disparity means differences in resources like pay or status between group members. In this study, the main focus will be on the category ‘variety,’ which aligns with the upper echelons theory explained below.
The effect of TMT diversity on innovation
An important theory in line with those previously explained that is often applied is that of the upper echelon, which was introduced in the first article by Hambrick in 1984 and further discussed in Cannella et al., 2008; Hambrick, 2007; and Nielsen, 2010. Built using the same arguments as the strategic choice theory from Child (1963), this theory was summarized in a paper by Hambrick and Mason (1984, page 1) as follows: "the organization as a reflection of its top managers." The central idea behind the upper echelon theory is that managers act
23 top managers, much of their personality, experiences, and values can be found (Hambrick & Mason, 1984).
The basic ideas of this theory are built on the principles of the coalition (Cyert & March 1963). In researching the realization of organizational objectives, Cyert and March concluded that these objectives were largely determined by the members of coalitions, each pursuing and negotiating their own interests. However, the difference is that within the upper echelon theory the focus is on the top management team as the main representative of the coalition in decision making.
As argued by Hambrick and Mason (1984), the upper echelon theory is seen as an information processing theory that explains how executives make decisions, depending on the different situations in which they find themselves. In other words, the strategic choices made by top management take place under bounded rationality. Moreover, the decisions of top
executives are bounded by the limitations of human nature and are therefore not always the most rational decision. Usually the major research on the upper echelons defines diversity as variety and sees team diversity as different demographic characteristics. Moreover,
demographic diversity includes characteristics that are often clearly observable, for example sex, race, nationality, educational level, and age (Harrison & Klein, 2007).
While there is a diverse body of literature about TMT diversity and performance outcome, the relation between TMT diversity and innovation has received less attention (Nielsen, 2010). This is quite surprising, considering the influence of top managers on the direction of a firm, the configuration of its project portfolio (Sirmon et al., 2011), and the allocation of resources
between innovation projects (Talke, 2011). Furthermore Talke (2011) argues that diversity leads to an increase in strategic intent towards addressing emerging customer needs and that it is more open to invention and new technologies.
24 TMT diversity is often conceived as the diversity of the educational, functional, industrial, or organizational background of the team members (Bantel & Jackson, 1989; Hambrick & Mason, 1984; Talke et al., 2011). Moreover, Bantel and Jackson (1989) and Talke (2011) found a positive relation between TMT diversity and innovation for the variables functional and
educational. In addition, Talke (2011) also found a positive relation for industrial and organizational background.
Less is known about the specific influence of gender diversity on innovation outcomes of the firm. Gender diversity could be interesting, because according to Miller and Del Carmen Triana (2009), diversity of gender leads to broader ideas and perspectives, which leads to combining knowledge within the team, which enables new innovative ideas. Therefore, it would be interesting to see if gender diversity contributes to a positive effect on innovation. A counter-argument to these authors could also be made, that too much gender diversity would lead to a variety of ideas and perspectives that makes it difficult to follow the firm’s strategy course.
However, existing studies show positive effects of gender diversity on innovation and other variables of a firm. For example, a study by Carter et al. (2003) showed a positive result with a firm’s performance outcome and the presence of women in a decision-making position. Moreover, results of research by Smith et al. (2006) show that the existence of women in TMT has a positive effect on firm performance outcome.
The general assumption of why diversity of the TMT leads to more innovation is because of bounded rationality: managers simplify decision situations that fit their view of the world based on their beliefs, knowledge, previous experiences, assumptions, and values (Finkelstein & Hambrick, 1990). Moreover, several reasons could be argued for why diversity should lead to a strategic intent that is more focused on addressing new customer needs and is more open to new technologies and innovation.
25 Firstly, diversity in the TMT brings together a wide range of knowledge, technical and management skills, capacities, and cross-border competencies (Simons et al., 1999). Therefore, it leads to information diversity, combining different and fresh visions that contribute to a more innovative attitude (Milliken & Martins, 1996). In this spirit, Bunderson and Sutcliffe (2002) believe that functional diversity is positively associated with information sharing when team members have broad experience in some functional areas (Dahlin et al., 2005). Believe that team diversity is linked to more information use. It could be argued that increased levels of information may reduce group thinking in heterogeneous groups compared to more
homogeneous teams (Hambrick & Mason, 1984, Janis, 1982). Educational diversity leads to diversity in knowledge, and access to a variety of knowledge is needed to create and combine new knowledge. Moreover, cognitive diversity also increases the chance that new information will be related to existing knowledge (Cohen & Levinthal, 2000), which strengthens the
assimilation force and thereby helps to identify probability. TMT diversity is therefore directly related to the knowledge creation capacity and innovation of a firm (Smith et al., 1994).
Secondly, TMTs with a higher level of cognitive diversity appeared to have more disagreement and conflicts when confronted with non-routine and complex problem solving (Bantel & Jackson, 1989). Therefore, diversity increases the motivation to debate and challenge the status quo (Michel & Hambrick, 1992). This factor should lead to more variation in decision-making alternatives, because task conflicts stimulate both diverse thinking processes and creative thinking. Moreover, confronted with arguments from team members with different backgrounds, top managers are forced to reconsider their point of view, to discuss alternatives critically, and to consider factors that have not been considered before. Therefore, a higher degree of TMT diversity must promote the identification of more new market and technology opportunities (Talke, 2011).
26 Thirdly, a higher degree of TMT diversity should lead to a broader influence of activities outside the firm, which may lead to a higher degree of innovation inside it. Moreover,
Geletkanycz and Hambrick (1997) found that extra-industrial tiers increase deviations from industry standards, whereas intra-industrial tiers of top managers tend to increase strategic conformity. TMT diversity in an industrial background will lead to an increase in extra-industrial links, which will make it easier to assume that the strategic orientation of the firm will deviate from current industry standards and focus more on new technologies and future customer needs.
In sum, most research on the upper echelon theory does not distinguish much between different characteristics; it is defined as a general construct, and diversity is hypothesized to have
uniform effects (Finkelstein & Hambrick 1996; Hambrick & Mason 1984). Moreover, as
previously explained, some empirical evidence suggests that TMT diversity is positively linked with higher levels of firm innovation (Bantel & Jackson, 1989; Talke et al., 2011). The
expectation is that all of the different diversity variables functional, educational, industrial, organizational background and gender may have a positive effect on firm innovation. Moreover, the strategic-choice theory sees the diversity of the top management team as positive in relation to innovation. Consequently, this leads to one global hypothesis regarding a positive linear relationship effect of TMT diversity on innovation. However, it is important to consider that the different variables also could have an effect on innovation separately, and that it is not strictly necessary (if data are not available) to measure all of the different variables.
27
2.4 Slack resources (behavioral theory), TMT diversity (strategic choice
theory) and innovation
In the previous paragraphs the two different theories are both hypothesized and analyzed separately. While both the behavioral theory (slack resources) and the strategic choice theory (TMT diversity) contribute to the empirical literature on a firm’s innovation strategy, few studies have attempted to integrate these two theoretical perspectives and investigate if these theories complement or substitute for each other. Moreover, in this paragraph the theories with respect to innovation are compared with each other. Referring to previous text, the two theories are important for organizations; however, they do not agree on the role of innovation or address where innovation in a firm is initiated. First, the option will be discussed of both theories being complementary; secondly, if the positive relationship between slack resources and innovation is moderated by a degree of TMT diversity.
As previously explained, both the behavioral (Cyrt & March,1963) and the strategic choice theory (Child,1972) contribute to the innovation strategy of a firm. The behavioral theory argues that innovation occurs through the whole organization from equal stakeholders (groups) and depends on the availability of slack resources. The behavioral theory does not recognize that the TMT is different from the other groups of the coalition and that it has special power. All the other groups of the coalition have the same power. So, these theory leaves no room for the hierarchical power of the TMT. The strategic choice theory argues that the top management team (TMT) is responsible for innovation and that diversity within this TMT lead to an increase in innovation. Whereas the other groups of the coalition follow the orders of the TMT, it is the only relevant group of the coalition with special power and is the source of the firm’s performance and innovation. Since both theories explain innovation in a firm, a combination of the two might do so better than both theories separately. Moreover, they may complement each other
28 a situation where the strategic-choice theory corrects the behavioral theory on the point that the TMT is recognized as a special coalition with special powers. The behavioral theory corrects the strategic-choice theory on the point that the TMT is not the only or most important coalition of a firm. Other coalitions within a firm also contribute to its innovation performance. Together, this forms a hybrid theory that recognizes the special powers and influence of some coalitions within a firm, but basically assumes that innovation arises from negotiations between the different coalitions (stakeholders). This suggests that financial slack (behavioral theory) and TMT diversity (strategic-choice theory) in firms are complementary and that combining both theories lead to a better explanation of the innovation in them. This leads to the following hypothesis;
Hypothesis 3: The degree of financial slack and the degree of TMT diversity are
complementary in explaining the degree of innovation.
TMT diversity moderates the relation between slack resources and innovation
A moderating variable changes the effect that the independent variable has on the dependent variable depending on the value of the moderator. The moderator thus changes the effect of the cause-and-effect relationship between the independent and dependent variables. This is called an interaction effect (Baron & Kenny, 1986). In this case it means that the degree of TMT diversity positively influences the relationship between the degree of slack resources and innovation, because in this study TMT diversity and slack resources are positively related to a firm’s innovation. In this study, the researcher argues that in line with the strategic choice theory (Child,1972), TMT diversity is an extension of the behavioral theory of the firm (Cyert &
March,1963). Therefore, this study perceives the slack resources within the organization as the primary source of innovation, and the TMT diversity as a moderating effect on the organizational
29 slack. In other words, a diverse TMT cannot innovate if there are no slack resources within the organization.
A high degree of TMT diversity should enable a higher degree of slack resources to be deployed for innovation within the firm. Moreover, the different types of TMT members lead to a more diverse viewpoint, because of their multiple perspectives. These multiple perspectives lead to more ideas and a better understanding of the market and customers. In terms of innovation, more diversity could lead to more strategic intent towards addressing emerging customer needs, and the firm being more open to invention, new technologies, and innovation (Talke et al., 2011). A low degree of TMT diversity (homogeneous TMT) should lead to fewer slack resources being deployed for innovation. Moreover, a homogeneous TMT means fewer different types of TMT members; therefore, a diverse view is not available. Homogeneous TMTs come up with the same ideas and pay attention to the same problems. What lead to a situation where less slack resources are deployed to innovation. Within a homogeneous TMT, the strategic intent towards innovation is less present (Talke et al., 2011). This aspect leads to the following hypothesis:
Hypothesis 4: The positive relationship between the degree of financial slack and the degree of
30
2.5 Conceptual model
Based on the expected relationships, the conceptual model of this thesis is presented in Figure 1.
Figure 1. Conceptual model
TMT Diversity
(Functional, Educational, Industrial, Organizational, Gender)Innovation
Slack
Resources
H1 H3,H4 H231
3 Methodology
In this chapter, the research methods are discussed. Paragraph 3.1 describes the design of the research. In paragraph 3.2, the population and sample are described. Paragraph 3.3 discusses the operationalization of the variables and the correlations between them. The data collection is discussed in section 3.4. Finally, in paragraph 3.5, the reliability and validity of the research are discussed.
3.1 Type of research
The research is quantitative and is suitable for explaining a particular phenomenon and for testing hypotheses. The goal of this project is to investigate the two different theories separately and jointly; the behavioral theory (slack resources) and the strategic choice theory (TMT
diversity) of a firm in relation to innovation. This research makes use of deductive reasoning. This choice indicates a top-down logic, based on the scientific principles, from theory to data, that are necessary to explain causal relations between variables and to test the existing scientific theory. The research philosophy used is positivistic. This philosophy is often used for research that consists of formulating hypotheses based on existing theories and testing these hypotheses on the basis of objectively observable or measurable data. This study makes use of secondary data only, and the overall study is archival. The reason for choosing secondary data (desk research) is because of the time constraint of five months. These data are directly
available and therefore do not depend on qualitative research. This study tries to contribute to the limited number of empirical studies available and to contribute to the inconclusive findings from both behavioral theory on firms (slack resources) and strategic choice theory (TMT
diversity) on innovation. Furthermore, this study will contribute to the literature on TMT diversity by using it as a moderator between slack resources and innovation.
32
3.2 Population vs. Sample
The population of this study is the Fortune Global 500; this list is a ranking of 500 companies worldwide, organized from large to small on the basis of their annual revenue. The list is drawn up yearly and published by Fortune magazine. This list can be seen as constituting the most powerful, well-known, and established firms worldwide. Moreover, the Fortune 500 is often used with data analyses (Rumelt, 2006; Thamhain, 2003). The sample of the population collected are 100 random firms of the Fortune Global 500. For the purpose of generalization, the minimum required sample size according to the Cochran sample size formula is 218, based on a 95% confidence interval and an error margin of 5% (Barlett et. al., 2001; Cochran, 1977). (See Appendix A for the calculation.) Therefore, it could be argued that because this study collects the data of 100 firms only the results are less reliable than if the minimum required Cochran sample size of 218 was met (Cochran, 1977). The firms were collected with Excel through simple random sampling. With this method, every individual firm of the Fortune Global 500 has an independent and equal chance of being selected. In addition, if some data for a firm are not available, a new firm is picked with Excel random sampling. Finally, for this research the data of year 2016 was gathered, as it was is the most recent year for which all of the data were
available.
3.3 Measurement of variables
The variables included in this study to test the hypotheses are operationalized as follows.
Dependent variable of innovation
Innovation is linked to the innovation strategy of the firm. Crossan and Apaydin (2010) define Innovation as: "production or adoption, assimilation, and exploitation of a value-added novelty in economic and social spheres; renewal and enlargement of products, services, and markets; development of new methods of production; and establishment of new management systems" (Crossan & Apaydin, 2010, p.1155). Innovation is measured by the research and development
33 (R&D) expenditures (Chiu & Liaw, 2009; Hufbauer, 1970; Mansfield, 1981). The sample of this study consists of firms of different sizes. Therefore, the R&D expenditures are measured by R&D intensity. Moreover, this study corrects the R&D expenditures by the sales revenues of the firm. This study multiplies the R&D to sales ratio by 100 to generate an R&D intensity
percentage. According to Adams et al. (2006), in their review of innovation measurement, this is the most accepted and frequently used measurement for innovation. The formula of the R&D intensity is calculated as follows:
Innovation = R&D spending / Sales revenue * 100
Independent variable of financial slack
The literature offers some usable measures and definitions to describe different types of slack. Moreover, this research uses financial slack as an indicator for slack resources. Financial slack is often linked to innovation in previous studies (Bourgeois, 1981; Lewis & Lewis, 2013). This study uses the available slack as a measurement for organizational slack because it represents resources that are available within the firm but are not yet reserved for particular allocations or specific expenses. Following (Geiger & Makri, 2006), this study operationalizes the available slack measurement via the current ratio, which can be calculated by dividing the total current assets by the current liabilities.
Financial slack = Current assets / Current liabilities
Empirical research has shown that slack recourses require a certain period of time to be used productively (Daniel et al., 2004). For this reason, it is expected that their current level is less relevant to influence innovation. That is why the analysis uses slack resources with a delay of a
34 year. Specifically, this means that for the forecast of innovation in 2016, the available slack data was collected of the 2015.
Independent variable of TMT diversity
An exact definition of TMT according to Finkelstein & Hambrick (1990) is that it denotes the set of top executives who have a direct impact on the formulation and execution of a firm’s strategy. Diversity refers to the total amount of heterogeneity distributed among the aspects of individual members of the organization (Simons et al., 1999). In other words, diversity means the
differences in the composition of a group of individuals (Kearney et al., 2009). According to Carpenter et al., 2004, the top management team are the top executives and the senior
managers of the firm who are directly responsible for its strategy, and their definition will also be used in this research. Following Hambrick and Mason (1984) and Hambrick (2007), this study measures TMT diversity as a formative construct. Hambrick and Mason (1984) defined managerial characteristics, and in extension the TMT diversity, as consisting of functional, educational, industrial, organizational, and gender variables. Moreover, each variable consists of different categories, which are also used in Bantel and Jackson, 1989 and Talke et al., 2011 (see Appendix B codebook for the different categories per group).
To calculate heterogeneity across the TMT on the five background measures previously explained, Blau’s index is used (Hambrick, 2007). Blau’s index is a measurement frequently employed for demographic heterogeneity in the literature (Bantel & Jackson, 1989; Smith et al., 2006). Blau’s index is measured as follows:
B =1 – Σ(pk^2)
Where B is the Blau index, and p is the percentage of TMT members in each category of the existing k-values. The higher the value of B, the greater the degree of diversity in the TMT. A low degree indicates a low degree of diversity. The minimum is always equal to zero. However,
35 the maximum score is not 100, and differs depending of the number of categories per variable. The maximum score can be calculated with the following formula: (K – 1) / K. Within this formula K refers to the number of different categories of the variable (Biemann & Kearney, 2010). This indicates that with an increase of categories, the maximum Blau’s index is higher. The
theoretical maximum and the number of categories per variable are presented in Table 1.
Variable Categories(K) Theoretical
Maximum
Index of Qualitative Variation (multiple factor)
Functional 8 0,875 1,142
Educational 4 0,750 1,111
Industrial 10 0,900 1,111
Organizations 10 0,900 1,333
Gender 2 0,500 2
Table 1. Theoretical maximum, categories per variable and multiple factor
As can been seen in Table 1, the theoretical maximum of the variables varies. For example, the theoretical maximum for functional is 0.750, while the theoretical maximum for gender is 0.500. Because of these differences in maximum, it is not possible to compare the variables. To compare them, they need to be corrected within this difference of number of groups. This correction can be done with the Index of Qualitative Variation introduced by Agresti and Agresti in 1978. The formula is Blau index per variable multiplied by K /(k-1). See Table 1 for the qualitative variation multiple factor per variable. Now the variable is standardized from zero to unity and no longer depends on the number of categories per variable, and now the different variables are comparable with each other. The last step is to add the five variables, and the average of the total sum is the TMT diversity number, which will be used as TMT diversity measure.
For TMT diversity, the data are collected for the year 2015, because the results of the
36 were responsible at the beginning of the year, and thus there is a delay of one year between the composition of the TMT and the innovation results.
Control variables
This study uses several control variables (firm age, firm size, team size, and geographic region) that are commonly used with innovation (Bantel & Jackson 1989; Jansen et al., 2006;
Srivastava & Lee, 2005).
Firm size
Larger firms could have access to more resources; however, due to their size, they may lack the flexibility for innovation. Firm size is included with the natural logarithm of the number of full-time employees within the firm (Bantel & Jackson 1989).
Firm age
Firm age may affect the firm's ability to pursue exploratory behavior, because firms may stick to firm strategy routines (Bantel & Jackson 1989). Firm age is included with the number of years from founding.
Team size
Team size may be positively correlated with diversity of the TMT. This is because in teams, an increase of one additional person substantially increases the maximum amount of diversity that is possible within the group. Therefore, the number of executives in the top management is measured (Bantel & Jackson 1989; Siegel & Hambrick, 2005).
37 Geographic region
Three dummy control variable regions are added to control for differences in regions, where Europe is the baseline variable and North-America and Asia are the control variables. Geographic region or location is a standard control variable (Bantel & Jackson 1989).
3.4 Data collection
This study made use of available public sources of data such as financial databases and organizational databases made available via the library of the University of Amsterdam. The study has a quantitative cross-sectional research design in which the researcher was mainly interested in the differences between firms and not over time. The data were sourced from listed and public firms. These firms are obligated to submit their financial and non-financial data regularly to the stock exchange, where is checked and approved by accountants. Due to this obligation and the fact that fraud in the annual reports can be prosecuted by the responsible stock-exchange and government institutions, the data can be considered highly reliable and trustworthy. The data were collected for 2016, which was chosen because it was the most recent year to assume that all the firms made their data available to the public.
In this study the researcher uses the data provided by the Osiris database from Bureau van Dijk for all of his variables. This database is relatively new one, which began data collection at the end of the 1990s, has a global coverage, and consists in 2018 of 80,000 companies. The Orisis database is already used in articles published in highly rated journals such as Journal of
Corporate Finance, Journal of World Business, Journal of Product Innovation Management,
Research Policy, and Global Strategy Journal. Therefore, the Osiris database is considered by
this study as a reliable and validated source of company information.
Furthermore, if the researcher encountered missing values, the researcher
38 website or on the website of the firm. Other sources used if data were missing are LinkedIn, Reuters, and Bloomberg.
3.5 Reliability and validity
Validity and reliability are determined by the testing methods and demonstrate the quality of the research data and the conclusions from the research. To minimize measurement errors, the properties that provide confidence that the correct type of measurements have been done need to be established. Therefore, insights into the meaning of validity and reliability are required.
The publicly available data have been audited by an accountant, and an accountant must comply with strict regulations. Therefore, these data are assumed to be more reliable and credible than data from other available sources that are nonpublic. If the data are not correct the firm loses credibility, which will have a deep impact on the firm and consequences for it. As previously described, fraud in data provided to the stock-exchange commission is seen as a criminally prosecutable offence. Therefore, these data, which are verified by an independent accountant, can be considered highly reliable.
To increase the internal validity of this research, several control variables have been added. This was done to make sure that the influence of any other variables next to the explanatory variables are ruled out. Leaving out the control variables means that the results of the research are less accurate (Bantel & Jackson, 1989).
Construct validity is about whether the variables measure what you want to measure. This study used measurements that were previously used in published scientific studies, herewith reinforcing the homogeneity and convergence of the measurements with previous studies (Heale & Twycross, 2015). The convergence is further investigated by comparing the results of the data analysis with the results of comparable studies. Furthermore, the criterion
39 validity is verified with a correlation analysis, which confirms if constructs have a convergent or divergent validity (Heale & Twycross, 2015).
External validity denotes the extent to which this research can be generalized (e.g., if the sample represents the entire population). Unfortunately, the data in this research were collected from 100 firms in total, while the minimum is 218 at a 5% margin error according to the Cochran sample size formula (Cochran, 1977). Therefore, it could be argued that this discrepancy affects the external validity negatively (Barlett et al. 2001).
3.6 Statistical analysis
For the data analysis, this study used the statistical program IBM SPSS Statistics Version 23. The SPSS 23 program is a widely used statistical application, mainly in social science studies. It is a flexible and reliable application because it was developed with social science datasets in mind and is more than capable of handling missing values and unbalanced datasets.
To reduce the effects of multicollinearity, this study has mean centered the interaction term. With mean centering, the average of a specific variable is subtracted from the individual scores. Hereafter, the scores of slack resources and TMT diversity are multiplied.
This study uses hierarchical simple and multiple regression analysis. According to Aiken (Aiken et al., 1991), this is suitable for measuring the relative contribution and the interaction of the variables in the total model. This study selected an ordinary least squares (OLS) regression analysis because it was mainly interested in the differences between companies and not in the change within a company. This study assumes that the amount of research that a company conducts, its organizational slack, and the composition of its board do not change considerably from year to year; therefore, a growth or change study would be inappropriate in this case. Furthermore, an OLS regression gives the best linear unbiased estimators in relation to a maximum likelihood (panel) regression, because a requirement for a maximum likelihood regression is that the distribution of the error terms must be known.
