Firm performance in a dynamic, complex and hostile task environment: the potential mediating role of firm innovation

Firm performance in a dynamic, complex and hostile task environment: the

potential mediating role of firm innovation

Robyn Genee

18-02-2015

Master Thesis

MSc. Small Business and Entrepreneurship

Faculty of Economics and Business

University of Groningen

Kraneweg 45A

9718 JG Groningen

06 11 08 58 48

r.k.genee@student.rug.nl

Studentnumber: 1923811

Supervisor: A.J. Rauch

Co-assessor: O. Belousova

- 2 -

Abstract

This thesis aims to answer the following research question: To what extent does the task environment affect firm innovation and what is the effect on firm performance? In the literature, there is a need to find variables which influence the relationship between the task environment and the performance of firms. A mediating relationship is proposed, in which firm innovation can potentially serve as a mediator. The environment has been identified into three main dimensions: dynamism, complexity and hostility. A sample of 109 respondents working in SMEs in Germany and the Netherlands has been used to test the different relationships. Results indicated that firm innovation does not perform as a mediator in the environment-firm performance relationship. Only environmental hostility has a negative influence on firm performance, but has no relationship with firm innovation. Dynamic and complex environments turn out to be fruitful environments for firms to innovate, however the link to firm performance is missing. The support for an firm innovation to outperform other firms did not convince.

Keywords

SMEs, innovation, task environment, dynamism, complexity, hostility, firm performance, mediation, regression, quantitative analysis

Research theme

- 3 -

Introduction

In the early literature of management, the structure-conduct-performance paradigm has been of importance (Bain, 1956). This paradigm is built on a S-C-P framework presented by Learned, Christensen, Andrews and Guth (LCAG), according to Porter (1981). The LCAG framework (Andrews, 1971) has become the foundation of business policy, whereas it describes that “a successful firm has to match its internal competences and values to its external environment” (Porter, 1981, p.610). The traditional Bain/Mason Industrial Organization paradigm (Bain, 1968) has similar arguments, whereas “the essence of this paradigm is that a firm's performance in the marketplace depends critically on the characteristics of the industry environment in which it competes” (Porter, 1981, p.610). The industry structure should determine how firms behave. The joint behavior (or conduct) of these firms together represents the embodied performance in their marketplace (Bain, 1968). The conduct can be seen as the economic dimensions of a firm's strategy, because conduct was the firm's choice of key decision variables as price, advertising, capacity and quality (Porter, 1981, p.611). Porter (1981) tried to solve the limitations he perceived with respect to the Bain/Mason paradigm. One of the most important elements of industry structure, barriers to entry, has been missed in this paradigm (Porter, 1981). Barriers to entry are important, because without these barriers, profits are eliminated whenever firms entry, which moves the industry to a long run equilibrium (McWilliams and Smart, 1993). In that case, above-normal cannot exist and therefore it is hard for firms to extract profits and thus to perform well (Scherer, 1980).

The aforementioned researchers describe the same S-C-P framework with similar arguments. The essence of the framework is that the economic performance of an industry is a function of the industry's structure and the economic conduct (McWilliams and Smart, 1993, p.64). In other words, "the S-C-P-paradigm implies that the structural characteristics of an industry, particularly the level of concentration of firms and the height of entry barriers, have a significant influence on the ability of firms within an industry to price above the competitive price" (McWilliams and Smart, 1993, p.65). As a consequence, it can be expected that these structural characteristics give direction to the potential performance of firms (McWilliams and Smart, 1993).

- 4 -

the academic perspective of organizational theory argues similarly. Lumpkin and Dess (1996) have provided a contingency framework, in which it is argued that a fit between the environment and a firm’s strategy leads to optimal performance. Depending on the environments, firms should adapt the right structure (Bourgeois, 1985). When a firm’s conduct matches with the environment properly, it can lead to firm performance. The environment to be a strategically critical determinant of firm’s performance and success (Child, 1972: Paswan et al., 1998; Thompson, 1967). The choice of strategy should be dependent on the level of these environmental elements. Therefore, the a firm’s environment can potentially serve as a moderator, strengthening the relationship between a firm’s conduct and its performance.

However, a main criticism regarding the S-C-P framework is that it is a static analysis (Porter, 1981; Grant, 1991; McWilliams and Smart, 1993). Optimal conditions could not always sustain, because environments are characterized by some degree of change (McWilliams and Smart, 1993). For example, entry barriers or concentration in the environment rise and fall (Porter, 1981). In such environments, firms need to choose a proper strategy, otherwise profits will evade. Schumpeter (1942) explained this already. Through innovation, firms gain a temporary quasi-monopoly position that enables them to extract rents. Through either competitors which copied the innovation, or competitors which brought up new innovations which makes the original innovation obsolete, this quasi-monopoly position could be vanished. More recently, researchers argue similarly. For example, in a dynamic environment, firms may need to be able to respond rapidly to unforeseen change in order to survive (Covin and Slevin, 1989). “Firms that explore and exploit opportunities in changing environments can outperform their rivals” (Rosenbuch et al., 2013, p.637). Freel (2005, p.51) argues similarly, “a less predictable (dynamic) environment generates more innovation through a greater scope for opportunity seeking and adaptive behavior”. For these reasons, innovation is in this context the most important strategy for firms. In certain environments, innovation can be a motivating force that drives the environment to firm performance. Therefore, a different mechanism than moderation is proposed within this thesis, which is a mediating mechanism.

- 5 -

(Dill, 1958; Shortell, 1977). This includes the interaction of firms with customers, competitors, suppliers and other stakeholders (Castrogiovanni, 1991). The task environment affects decisions, actions and performance of organizations (Rosenbusch et al., 2013). While a firm’s behavior, with a focus on firm innovation, is examined in this thesis, the task environment is relevant. Therefore, the construct task environment will be used. In this thesis, key dimensions of a firm's task environment are examined: dynamism, complexity and hostility (Dess and Beard, 1984; Rosenbusch et al., 2013).

According to the arguments above, there are reasons to assume that firm innovation could be an important mediator in the relationship between the task environment and firm performance. “In business, innovation is something that is new or significantly improved, done by an enterprise to create added value either directly for the enterprise or indirectly for the customers” (Business Council of Australia, 1993, p.3). Empirical tests have generally supported the expectation that the more innovative typologies of Miles and Snow (1978) outperform the less innovative typologies. (e.g. Conant et al., 1990; Dyer and Song, 1997). However, Desarbo et al. (2005, p.51) indicate that more research is still warranted, particularly with regard to environmental factors which might affect the performance achieved by different strategic types. Freel (2005) recommends further research on this subject by obtaining richer data collection.

With this thesis, the researcher tries to make a contribution to the literature in three ways. Firstly, the S-C-P framework, as described, focusses on industry structure (f.e. Porter, 1981). This thesis will replace that focus to the task environment.

Furthermore, this thesis proposes alternate relationships with respect to the environment-strategy-performance relationship. In this thesis, the proposition of different researchers that a fit between the task environment and a firm’s strategy leads to optimal performance (f.e. Bourgeois, 1995; Lumpkin and Dess, 1996) will not be followed. Instead of that, this thesis proposes that the relationship between the task environment and performance is mediated by firm innovation. Innovation could be an important mediating mechanism in the environment-performance relationship which is missing yet in the literature. This thesis tries to fill in this gap.

above-- 6 above--

average performance levels” (Rosenbusch et al., 2013, p.634). However, the description of the construct EO is rather vague and not well defined in this article. Rosenbusch et al. (2013) suggest for further research, as they state that additional mediators to the relationship environment – firm performance could be found by scholars and that a more detailed picture of the relationships among a firm’s environment, strategy making and firm performance is needed. Rather than adopting an EO, this thesis proposes that firm innovation is a mediating mechanism in the environment – performance relationship.

As a final note, this thesis could be of managerial interest, because the outcomes of this study could inform managers whether firm innovation is a wise strategy to follow, while operating in a given environment. This thesis could also give answers to the question whether this affects the firm performance. The general importance of firm innovation within small- and medium-sized firms will be examined, which can contribute to managerial decisions about choosing the right type of strategy.

Research question

The proposed relationship between the environment, innovation as a strategy and firm performance, which is described in the introduction, needs to be further investigated. This research aims to examine the importance of innovativeness for small and medium-sized firms, comparing this to its task environment.

The research question of this master thesis is formulated as followed:

- 7 -

Literature Review

First, the different concepts which are important within this thesis, have to be clarified. After that, different relationships between the constructs are provided from which the conceptual model has been built.

Defining constructs

The task environment

In the literature, the characteristics of a firm’s environment have been identified by researchers in different ways. Firstly, some scholars elaborate about ‘environmental uncertainty’, in which they describe an uncertain environment for instance as complex or dynamic (f.e. Duncan, 1972; Anderson and Tushman, 2001; Tushman and Nadler, 1978; Freel, 2005). Furthermore, some scholars recognize a firm somewhere in between extremes, such as the ‘simple-complex’ dimension and the ‘static-dynamic’ dimension (Duncan, 1972). Another example of this has been presented by Ansoff (1987), who argued that the task environment is somewhere between two extremes. The one extreme is an environment which is placid, where nothing changes and there is absolute predictability. At the other extreme, “the environment is an environment with major technological change and social political upheavals” (Uzkurt et al., 2012, p.8).

- 8 -

and Reimann, 1973). For these reasons, task environment is used in this thesis to characterize a firm’s environment.

In order to embody the task environment, this thesis follows the most important dimensions, identified by Dess and Beard (1984). Different dimensions of the task environment are used, because different dimensions can generate more perceptions about the proposed mediating influence of innovation (Rosenbusch et al., 2013). However, to remain a good overview, this thesis examines only the most important dimensions of Dess and Beard (1984). According to Dess and Beard (1984), all the relevant environmental factors such as customers, suppliers or competitors are widely documented. However, based on an extensive review by Aldrich (1979), three dimensions with similar values can be applied to define the task environment. These dimensions are dynamism, complexity and munificence (Dess and Beard, 1984).

Dynamism is defined as “change that is hard to predict and that heightens uncertainty for key organizational members” (Dess and Beard, 1984, p.56). The rate and amount of change in the task environment is important regarding dynamism (Miller and Friesen, 1983). Change of products, marketing practices, demands and tastes of customers and modes of production/services are important examples which can predict change (Miller and Friesen, 1982/1983). Regarding the second dimension, Dess and Beard (1984) cite Child’s (1972, p.3) definition of environmental complexity, as “the heterogeneity and range of an organization’s activities”. Complexity implies the numerosity of inputs and outputs (Dess and Beard, 1984). The greater the complexity in a firm’s task environment, the more information information-processing requirements are needed (Dess and Beard, 1984; Damanpour, 1996). Munificence, the last dimension identified by Dess and Beard (1984), describes the favorability of the firm’s task environment in terms of the existence of opportunities and the availability of resources (Dess and Beard, 1984; Rosenbusch et al., 2013, p.635). According to Castrogiovanni (1991, p.542), “munificence is about the scarcity or abundance of critical resources needed by (one or more) firms operating within an environment”.

- 9 -

(Zahra and Garvis, 2000, p.475) and legal, political and economic constraints (Miller, 1987). Environmental hostility contains threats with respect to the survival and growth of a firm, rather than the external changes to which a firm must align. (Özsomer et al., 1997).

Recalling, munificence represented favorability of the firm’s task environment in terms of the existence of opportunities and the availability of resources (Dess and Beard, 1984; Rosenbusch et al., 2013). Hostility is the scarcity of external resources and opportunities in a specific environment (Dess and Beard, 1984; Zahra, 1993). Therefore hostility can be seen as an opposite term in line with the definitions given above (Lumpkin and Dess, 2001). Following Lumpkin and Dess (2001), who perceive hostility as an task environmental construct consistent with earlier research, hostility is further used in the remainder of this thesis to describe the influence of external forces of the task environment on the firm innovation and its performance.

Firm innovation

Every firm must adopt a strategy, in order to run a business properly and to overcome problems. A firm could for instance choose to react to the market or reduce costs to drive away competitors (Miles et al., 1978). However, in the introduction, the importance of innovation within a firm was indicated. Of the different strategies from which a firm can choose, innovation is claimed to be the most important (f.e. McEvily et al., 2004; Hitt et al., 1996; Roberts, 1999; Zahra and Covin, 1994). This is because firm innovation leads a firm to a sustainable competitive advantage (Zahra and Covin, 1994). For example, Roberts (1999) finds support for a relationship between high firm innovation and sustained superior profitability. Those firms which response and react to changes in the environment on time, turn out to be the winning firms (Teece et al., 1997). This is often guarded by accumulating technological assets and an aggressive intellectual property stance. Kim and Mauborgne (1999, p.43) argue that “to achieve sustainable profitable growth, companies must break out of the competitive and imitative trap. Rather than striving to match or outperform the competition, companies must cultivate value innovation”.

- 10 -

differentiate themselves successfully in their marketplace” (Baregheh et al., 2009, p.1334). This definition is similar to the definition of Garcia and Calatone (2002: p112), whereas they define firm innovation as “an iterative process initiated by the perception of a new market and/or new service opportunity for a technology-based invention which leads to development, production, and marketing tasks striving for the commercial success of the invention”.

Firm innovation can take various forms, which are indicated by different scholars. For instance, firm innovation can lead to new organizational forms, new products, new services, new materials or new management practices (Ettlie and Reza, 1992; Baregheh et al., 2009, p.1323; O’Regan and Ghobadian, 2005). Firm innovation can be either radical or incremental (Garcia and Calantone, 2002; Kleinschmidt and Cooper, 1991).

To recall the definition of Baregheh et al. (2009; p.1334): “Firm innovation is the multi-stage process whereby organizations transform ideas into new/improved products, services or processes, in order to advance, compete and differentiate themselves successfully in their marketplace”. Two parts in this definition emerge. The first part is the multi-stage process whereby firms transform ideas into new/improved products, services or processes. This includes stages from the start of an idea until the implementation of the product, service or process (Baregheh et al., 2009). This part of the definition will be examined in this thesis.

- 11 - Firm performance

The last construct of the research model is firm performance. The measurement of firm performance is of importance in many areas of management research, because it investigates the ultimate aim of each company (Wall et al., 2004). Van der Stede et al. (2006) made a distinction between objective and subjective performance measures.

“Where accurate objective measures of performance are available, their use is strongly supported and encouraged” (Van der Stede et al., 2006; p. 270). “Objective measures increase the credibility of research results among practioners” (Strecker, 2007, p. 209). Also, objective measures of performance are less sensitive for potential respondent biases, such as perception bias (Muckler, 1992). Therefore, objective measures are considered to be an important dimension of firm performance.

However, traditional financial (objective) measures, such as return of investment, return on sales, price variances, sales per employee, productivity and profit per unit of production are claimed not to determine properly a firm's performance, and therefore give not enough direction for strategic advice. (f.e. Ghalayini and Noble, 1996; Kaplan and Norton, 1996). This is for instance because these traditional financial measures are inflexible and expensive. Besides, these measures are often not incorporated with a firm’s strategy and are not always relevant for practical purposes, such as product quality or customer satisfaction (Ghalayini and Noble, 1996).

Besides these weaknesses of objective financial measures, there might be reasons also to determine firm performance in a subjective way (Wall et al., 2004; van der Stede et al., 2006). For instance, such data can be collected through questionnaires or interviews, which is a practical advantage. Also, with some firms (mostly small firms), there are no appropriate financial records (Wall et al., 2004). Furthermore, managers may reject to provide hard facts about sales, investments or profit, due to sensitivity reasons (Dess and Robinson, 1984). Another reason could be that there is a greater risk of error when measuring small firms only with respect to objective measurements (Dess and Robinson, 1984). Because of the strengths of subjective measures and the weaknesses of objective measures, subjective measures are claimed to be important as well in this thesis.

- 12 -

1993; Dawes, 1999). There is consensus among researchers that sales growth is the best growth measure, because these numbers are both easy to obtain and as well short-term as long-term goals are reflected (Hoy et al., 1992). Furthermore, sales increase at first when a firm performs well, for instance when their services or products are of a high quality level (Wiklund, 1999). For these reasons, regarding objective firm performance, the focus within this thesis is on sales growth.

Regarding subjective firm performance, researchers describe this as the performance of a firm relative to its competitors, often described using a scale with anchors such as “very good” or “very poor” (Birley and Westhead, 1990; Dawes, 1999; van der Stede et al., 2006; Voss and Voss, 2006). “Subjective measures of performance captures managers’ perceptions of how well they perform comparing to competitors” (Voss and Voss, 2006, p73). Overall firm performance/success, total sales growth and return on total assets compared to competitors, are proposed as a subjective measure of firm performance (Dess and Robinson, 1984). Wiklund and Shepherd (2003) indicate ten important dimensions of performance: sales growth, revenue growth, growth in number of employees, net profit margin, product/service innovation, process innovation, adoption of new technology, product/service quality and – variety and customer satisfaction. Following these researchers, the elements of these subjective performance measures are examined within this thesis.

- 13 -

Hypotheses

The task environment and firm performance

To recall, firm innovation has been proposed in the introduction as a mediator in the task environment – firm performance relationship. Therefore, this relationship will be further elaborated in this section. The general idea is that a firm performs at its maximum in an environment which is described as ‘the surprising environment’, in other words, an environment which has some degree of change (Ansoff, 1987).

Some arguments for this proposition are provided in the literature. A first argument is that in a complex and dynamic task environment, there is a higher need for information, a more extensive network and a strong capability of information processing to improve firm performance (He et al., 2009). The argument is that firms need better managers in this situation, which will lead to higher firm performance (Waldman et al., 2001). Furthermore, a dynamic and complex task environment enables creativity (Stemberg, 2005) and amplifies the effects of dynamic capabilities (Wilden, 2010).

Bradley et al. (2011) also bring forward an opposite argument. Less dynamic and complex task environments are usually mature task environments. “In such environments, firms are better allowed to establish patterns in production processes, product designs and modes of services from their competitors” (Bradley et al., 2011, p.1078). Therefore, less profits could be extracted.

“In dynamic task environments, where technology, demand and competitor behavior change quickly, existing opportunities and resources can quickly become redundant” (Rosenbusch et al., 2013, p.637). Because of this, managers become more focused and active and therefore firms that explore and exploit opportunities in such task environments can outperform their rivals (Sanyal and Sett, 2011).

- 14 -

In general, firms in a more dynamic task environment have greater access to capital (Wiklund and Shepherd, 2005), because the higher potential opportunities attracts more financial capital. “Access to financial capital provides the resource slack necessary to encourage experimentation within the firm, allowing it to pursue new opportunities” (Wiklund and Shepherd, 2005, p.73). Frank et al. (2010) confirm this argument empirically, whereas they found that a firm’s performance was higher in dynamic task environments due to the higher access of capital. Needed know-how can also be more easily acquired due to greater access of capital (Frank et al., 2010).

Zahra (1993) empirically tested firm performance in different task environments. In a dynamic task environment, Zahra (1993) found that firms were more consistently associated with superior financial performance. Other scholars also found dynamism to be significantly related to operating performance (Keats and Hitt, 1988; Miles et al., 2000).

A complex task environment can also influence a firm´s performance. Osborn (1976) provides the main argument that a complex task environment could lead to higher performance. “As complexity increases, firms should become more flexible, adaptive and organic and thus more effective” (Osborn, 1976, p.179). This is because technologies develop quickly and are frequently on an unpredictable track (Branzei and Thornhill, 2006). Firms in complex task environments are in a higher need to proactively seek for new combinations or resources that can be applied to transform opportunities into above-average performance levels (Rosenbusch et al., 2013, p.638). However, Bourgeois (1980) argues that complexity remains a relatively constant factor in task environments, therefore the impact on firm performance will be relatively lower than for instance dynamism.

- 15 -

Based on the arguments presented, the first three hypotheses can be presented:

H1: Task environmental dynamism has a positive relationship with the performance of this firm.

H2: Task environmental complexity has a positive relationship with the performance of this firm.

H3: Task environmental hostility has a negative relationship with the performance of this firm.

The task environment and firm innovation

The literature suggests that firms are more likely to pursue more proactive and more aggressive strategies as complexity and dynamism increases (Freel, 2005; Ozsömer et al., 1997). This is because firms in such environments are more likely to adapt to the task environment and to search for opportunities (Freel, 2005). Firms must be able to accommodate to environmental change (Miller and Friesen, 1983) and therefore must respond to the task environment with a proper strategy. In a highly changing task environment, firms often choose to implement firm innovation. The main argument is that “environmental change or uncertainty stimulates changes in the strategy and/or structure of an organization, which in turn lead to the implementation of innovations” (Damanpour and Evan, 1984, p.407). Similarly, “the greater variability in the external task environment is likely to compel incumbent firms to safeguard their market position through the introduction of new products or new processes” (Freel, 2005, p.52). For these reasons, firm innovation might be a proper strategic choice for firms facing a changing and varying task environment. In these type of task environments, firm innovation can be a tool to outperform competitors (Miles et al., 1978).

- 16 -

either been argued or empirically been found by different scholars (f.e. Miller and Friesen, 1982; Koberg et al., 2003; Brown and Eisenhardt, 1997).

Regarding the task environmental dimension complexity, firm innovation is very important to respond to this complexity (Christensen, 1997). A firm in a complex task environment should, besides being aware of a proper quality of their product/service, need to go beyond this and search for new opportunities for value creation to exploit (Dervitsiotis, 2012). “In complex task environments, companies generally have to develop their knowledge of the task environment internally because rapid technological advancement makes it very difficult to wait to adopt what others firms launch in the market” (Pérez-Luno and Cambra, 2013, p.183). Therefore, the need for firm innovation is high. Furthermore, “a firm in a complex task environment is more likely to learn from their experience with competitors and customers, i.e. they will borrow ideas from one market and apply them in another” (Miller and Friesen, 1982, p.3). The relationship between task environmental complexity and firm innovation has also been found by different scholars (f.e. Damanpour, 1996; Pérez-Luno and Cambra, 2013).

- 17 -

Based on the arguments above, new hypotheses can be formulated:

H4: Task environmental dynamism has a positive relationship with firm innovation.

H5: Task environmental complexity has a positive relationship with firm innovation.

H6: Task environmental hostility has a negative relationship with firm innovation.

Innovational strategy and firm performance

In the innovation literature it becomes clear that firm innovation plays a key role on firm performance and firm’s competitiveness. Through innovation, firms gain a temporary quasi-monopoly position that enables them to extract rents (Schumpeter, 1942). Furthermore, innovation yields creativity at firms, which in turn leads to a high level of performance (Bratnicka and Bratnicki, 2013). By having a continuous stream of successful innovations, superior performance can be achieved, which can positively contribute to firm value of firm performance (Sharma and Lacey, 2004). Scholars link a higher level of innovation to a higher level of performance (f.e. Damanpour and Evan, 1984; Damanpour et al., 1989; Han et al., 1998) Damanpour and Evan (1984, p.407) conclude in their article that “a balanced implementation of administrative and technical innovations would help to maintain the equilibrium between social and technical systems, which in turns will lead to high performance”. It has been argued that firm innovation stimulates new products, services or processes, which contributes to firm performance (Dibrell et al., 2011). This is because firm innovation leads to higher output, higher sales and therefore higher (economical) performance (Klomp and van Leeuwen, 2010)

- 18 -

Based on the literature and arguments above, the last hypothesis can be presented:

H7: Firm innovation has a positive relationship with the performance of this firm.

Firm innovation as a mediator

The relationships between the different constructs are clear. As stated in the introduction, firm innovation plays a crucial role in the task environment – firm performance relationship. A dynamic task environment contains a high degree of change, especially with respect to products, marketing practices, demands and tastes of customers or modes of production/services (Miller and Friesen, 1982/1983). In these type of environments, profitable conditions could not be sustained (McWilliams and Smart, 1993). Without firm innovation, firms could not perform optimal in such task environment, because a temporary quasi-monopoly position could not be created (Schumpeter, 1942). In that case, firms could not stay at a sufficient competitive level (O’Regan and Ghobadian, 2005). Profits could therefore not be extracted by firms and the relationship between a dynamic task environment and firm performance evades. However, firm innovation causes these firms to exploit opportunities, leading to firm performance. (Covin and Slevin, 1989; Rosenbusch et al., 2013). Therefore, firm innovation should be the driving force regarding the relationship between a dynamic task environment and firm performance and firm innovation is expected to act as a mediator.

The same reasoning can be applied to the relationship between a complex task environment and firm performance. As stated before, a complex task environment implies the heterogeneity and range of organization’s activities (Child, 1972). The more activities of firms, the less secure a task environment becomes to an individual firm. Quasi-monopoly positions to extract profits are more difficult to sustain, without firm innovation. Therefore, also in here, innovation is expected to act as a mediator in the relationship between a complex task environment and firm performance.

- 19 -

profits. Therefore, it is expected that (the lack of) firm innovation will mediate the negative relationship between environmental hostility and firm performance.

H8: Firm innovation will mediate the positive relationship between environmental dynamism and firm performance

H9: Firm innovation will mediate the positive relationship between environmental complexity and firm performance

H10: Firm innovation will mediate the negative relationship between environmental hostility and firm performance

Model building

Now that the hypotheses are formulated, the following modes can be presented.

- 20 -

Methodology

Research method

This research is a quantitative research in which theory is tested through statistical analyses. In order to perform the analyses, the statistical program SPSS was used. Data was collected due to field research, partly by this researcher and partly by other students. This dataset has been performed as the basis for statistical analyses. Regressions and a mediating analysis were used as statistical techniques. This will be discussed later in this method section.

Research design

Data collection

To collect enough data for a statistical analysis, information was gathered from 109 respondents who are (partly) in charge of a small- and medium sized firm (SME) in Germany and the Netherlands. The information was collected in two ways: through either a questionnaire and an interview. The questionnaire was used to collect data about the firm’s task environment, the strategic part of firm innovation and subjective measures of firm performance. Questionnaires have practical advantages; for example questionnaires are not time-consuming, less costly and results of the questionnaires can be easily quantified by the researcher (Ackroyd and Hughes, 1992). Since this thesis is a quantitative research, questionnaires are used in this thesis.

- 21 -

Entrepreneurs within the sample should meet the following criteria. The entrepreneurs had to be in charge of the firm, either individually or together with other individuals, because in this case useful information can be derived from the interview. Following the definition of Van Praag and Versloot (2007), these firms should be market entrants. Therefore, the firm age should have be at a minimum of 1 year and a maximum of 8 years. Furthermore, these firms should have a minimum of 1 part-time or permanent employee, because this excludes freelancers or independent contractors, who could not be seen as an SME.

As part of a greater research, the data was partly collected by the researcher, but the greatest part of the data was collected by other German students under the supervision of mr. Rauch. To be precise, 10 respondents were interviewed by this researcher and the information about the other 99 respondents was gathered by German students. Of the 10 respondents who were interviewed by this researcher, six of them were approached on the basis of personal contacts and network of the researcher. These six persons all accepted the request. Four other respondents were found in the database of the Rijksuniversiteit Groningen, ‘Orbis’, after firms were selected for the necessary criteria. The database ‘Orbis’ was used because it contains a extensive dataset of Dutch (small) firms which is based on information of the Chamber of Commerce. Due to practical (travelling) reasons, only entrepreneurs in the North of the Netherlands were approached (Friesland, Groningen and Drenthe). 60 of these entrepreneurs were approached by e-mail, which can be found in Appendix 1. One of them did not accept the request, while four of them accepted the request (response rate of 8.3%).

To recall, the database contains information about 109 respondents. 99 of them came from Germany. The other ten respondents were from the Netherlands. A short, anonymous overview of these respondents (interviewed by this researcher) can be found in Appendix 2. All these 10 respondents were owners of a firm in the Northern region of the Netherlands (1x Groningen, 2x Drenthe, 7x Friesland). Almost all respondents run a firm in a different industry.

- 22 -

between 2012 and 2014. Despite that, these respondents were included in the dataset. The mean of these founding years is 2007. On average, these respondents went to school for 12 years and received a degree at the (Dutch) education level MBO/HBO.

Measurement

The constructs and their dimensions were measured using different items in the dataset. An overview of these constructs and its dimensions is presented in table 1. Below table 1, the overview is explained and a description of the different variables is given. The complete description of the different variables from the interview and questionnaire can be found in Appendix 3.

Table 1: measurement of constructs

Construct Dimensions Measurement Scale

Task environment Dynamism Mean (SPS1 to SPS5) 7 point scale Complexity Mean (CPL1 to CPL2) 7 point scale

Hostility Mean (EH1 to EH3) 7 point scale

Firm Innovation Degree of innovation Mean (Innovation [Product] and Innovation [Processes])

5 point scale

Strategic Posture Mean (SP1 to SP9) 7 point scale

Firm performance Objective Relative Growth Sales Not applicable Subjective 1 Mean (SU1 to SU10) 5 point scale

Subjective 2 SW1 7 point scale

The task environment

In the literature review it became clear that the task environment is divided in three elements: dynamism, complexity and hostility. Information about these dimensions were captured in the questionnaire. As stated before, the information which is needed to measure the task environment was captured in the questionnaire.

- 23 -

and are based on the dynamism scale from the article of Miller and Friesen (1982). The items form a pooled variable dynamism. The information is for instance based on the perceived rate of change of marketing practices, customer demand and tastes, modes of production/service, as well as the rate of obsolesce in the market and the rate of prediction of competitors’ actions. The items were reliable on a 0.786 Cronbach’s Alfa coefficient.

Environmental complexity. Similar to environmental dynamism, this variable was measured through the questionnaire. The items (CPL 1 and 2) can also be found in Appendix 3. The items are based on the research by Covin et al. (1990). Two items together form the pooled variable complexity, in which the respondents were asked whether the task environment is technological sophisticated and complex, as well as which grade of R&D is typical for their task environment. The items were reliable on a 0.801 Cronbach’s Alfa coefficient.

Environmental hostility. The last dimension of the task environment has also been measured by the questionnaire. Questions were asked to respondents to which degree they perceived their task environment as safe, rich in opportunities and degree of control. This is based on Khandwalla (1976/1977). These three items (EH1 to EH3) together were pooled as the variable hostility and can be found in Appendix 3. The items were reliable on a 0.717 Cronbach’s Alfa coefficient.

Firm Innovation

The proposed mediating variable in this thesis is firm innovation. Information about an innovative strategy was collected as well in the questionnaire as in the interview. As mentioned before, the interview was used to get an in-depth insight of the level of firm innovation of the respondents.

In the literature review, firm innovation was defined based on the definition of Baregheh et al. (2009). Since the definition contained two separate parts, two different variables of firm innovation will be used within this thesis.

- 24 -

firm innovation (both product- and process innovation). The mean of these two values has been used to measure the degree of firm innovation. A coding scheme was used by the researcher in order to value the degree of firm innovation (both product- and process innovation). This coding scheme is based on dimensions of firm innovation by Calantone et al. (2002). Based on a fit between information from the interview similar to the quotes from the coding scheme, a degree of firm innovation has been determined by the researcher. With a base score of 1 (“no innovation at all”), after ticking the boxes within the coding scheme, the highest possible degree of innovation was 5. The coding scheme has also been added in Appendix 3. The Cronbach’s Alfa coefficient of these two items together is 0.507.

In the dataset, the degree of firm innovation was not always filled in regarding the German respondents. However, the dichotomous question whether that particular respondent complied to firm innovation, was available. Therefore, this was adapted in the dataset. When a respondent did not comply to firm innovation, a ‘1’ was filled in at the column of firm innovation (“no innovation at all”). Therefore, the data was slightly adapted with respect to the dataset that had been acquired from the German students.

Recalling the literature review, strategic posture (Covin and Slevin, 1989) was used to capture the second part of the definition of Baregheh et al. (2009; p.1334): “in order to advance, compete and differentiate themselves successfully in their marketplace”. Therefore, 9 items based on the dimensions innovation, proactiveness and risk-taking (Covin and Slevin, 1989) were used to measure firm innovation. The items of this second variable can also be found in Appendix 3. The Cronbach’s Alfa coefficient of these items is 0.851.

Firm performance

The dependent variable in this thesis is firm performance. As stated in the literature review, dimensions of firm performance are divided in two streams: ‘objective’ and ‘subjective’ measurements (Dess and Robinson, 1984). Reasons why both objective and subjective measurements were used, were already presented in the literature review section.

- 25 -

or digital overviews. A problem was that not all entrepreneurs were either able or wanted to provide ‘hard facts’ about the performance of their firm. Information about sales were given by more or less 50% of the respondents.

The dataset contained ‘objective’ information about employment, investments, profits and sales. This thesis has used sales growth, because sales growth is a popular tool to measure firm performance objectively (Dess and Robinson, 1984). There is a consensus between researchers that sales growth is the best growth measure, because these numbers are both easy to obtain and as well short-term as long-term goals are reflected (Hoy et al., 1992). Also, another reason why sales growth is a popular measure is because sales increase at first when a firm performs well, for instance when their services or products are at a high quality level (Wiklund, 1999).

In total, information in the dataset was available about the sales of the firms during the years 2011 to 2014. Relative growth was calculated with the formula (Relative sales growth = (sales 2013 – sales 2012) / sales 2012). Growth was measured between the years 2012 and 2013, because in this case there were the most valid options (62 valid, 47 missing). The sales growth 2011-2012 had only 50 valid entries (59 missing), the sales growth 2013-2014 had 51 valid entries (58 missing) and the total sales growth during the period 2011-2013-2014 had only 35 valid entries (74 missing). This is because the data was mostly gathered in 2013 by German students and therefore there was yet no information available about 2014. Therefore, the relative growth formula was used between the sales of 2012 and 2013. More information about this measurement of objective firm performance can be found in Appendix 3.

- 26 -

basis for the second subjective measurement for firm performance, which can also be found in Appendix 3.

Control variables

In order to conduct different regression analyses, the relationships have to be controlled for certain control variables. The first control variable is country-specific. The literature identified country-specific cultural values as individualism/collectivism, power distance or uncertainty avoidance/uncertainty acceptance as values which influence firm innovation (Hofstede, 1980; Shane et al., 1995). Also, countries may differ in economic or regulatory environments (Zhu et al., 2006). Based on this, firm innovation could be different between countries and therefore the dataset was controlled for country differences.

The type of industry could also be a variable which influences firm innovation. Different types of industry could for example differ in the degree of technology which definitely influences innovation (Buganza et al., 2011). Of course, in line with the literature section, industries also differ in terms as dynamism or complexity. Therefore, there is a need to control for the type of industry. For example, firm innovations in service industries are totally different from the more traditional industries (Gadrey et al., 1995). The dataset tried to capture the industry type in four dimensions: trade, craft, service and manufacturing. These type of industries were based on the article of Rauch et al. (2005). These industry types has been recoded into dummy variables. This can be found in table 2.

An overview of the control variables can be found in table 2.

Table 2: Control Variables

Control variables Measurement

1: Country The Netherlands = 1, Germany = 0

2: Service industry Service = 1, Other industries (Trade, Craft, Manufacturing) = 0 3: Trade industry Trade = 1, Other industries (Service, Craft, Manufacturing) = 0 4: Craft industry Craft = 1, Other industries (Trade, Service, Manufacturing) = 0

- 27 -

In the analysis, the industry dummy variables are controlled for the ‘dummy variable trap’ (Bech and Gyrd-Hansen, 2005). “Dummy coding involves that an attribute with L

quantitative levels is transformed into L-1 dummy variables in which each dummy is set equal to 1 when the qualitative level is present and set equal to 0 if it is not” (Bech and

Gyrd-Hansen, 20005, p.1079). The Lth level is excluded in order to avoid perfect collinearity, in other words the dummy variable trap. “This means that the level of the Lth level can be perfectly predicted from the L-1 dummy variables” (Bech and Gyrd-Hansen, 2005, p.1079). To avoid this, one of these dummy variables have to be excluded. Therefore, the dummy variable manufacturing industry has been excluded a priori from the regressions, because this variable has the smallest N within the dataset (N=8). In that way, the most data would still be used. This has also been done by Lazaridis and Tryfonidis (2006).

Reliability analysis

Cronbach’s Alfa coefficients already have been presented. The Cronbach’s Alpha is the most widely used measure of scale reliability (Peterson, 1994). In his meta-analysis, Peterson (1994) concludes that 75% of the investigated Cronbach’s Alpha coefficients were higher than 0.7, and therefore 0.7 is considered to be an appropriate minimum level of the Cronbach’s Alpha coefficient, in order to include a variable in the analysis. The Cronbach’s Alpha coefficients of scales which were used in this thesis are presented in table 3.

Table 3: Reliability analysis

Construct Dimensions Measurement Cronbach’s Alpha Task environment Dynamism Mean (SPS1 to SPS5) .786

Complexity Mean (CPL1 to CPL2) .801

Hostility Mean (EH1 to EH3) .717

Firm Innovation Degree of innovation Mean (Innovation

[Product] and Innovation [Processes])

.507

Strategic Posture Mean (SP1 to SP9) .851

Firm performance Objective Relative Growth Sales Not applicable Subjective 1 Mean (SU1 to SU10) .851

- 28 -

The internal consistency between most of the items within the variables were at an

appropriate level (>0.7). Only the variable degree of firm innovation attracted attention in a negative way, whereas the coefficient of this variable was 0.507. However, when the

correlation between two items is relatively high and significant, the variable can be taken into account (Shevlin et al., 2000). The correlation between innovation (product) and innovation (processes) was significant (β = .47, p<0.01), and therefore this variable will be examined within this thesis.

Analysis

In total, ten hypotheses are formulated in the literature section. These hypotheses has been tested using multivariate statistical techniques in SPSS. Hypotheses 1 to 7 has been tested using stepwise linear regression, thereby including the control variables. Eventually, the model searched for a mediating relationship during the last three hypotheses (H8 to H10). This required a statistical technique which is called mediator analysis. Baron and Kenny (1986) proposed a mediating technique which has already been cited over 10.000 times since 1986 (Zhao et al., 2010).

Figure 1: Mediating relationship (Baron and Kenny, 1986)

- 29 -

- 30 -

Results

In this section, the results of the statistical analyses are presented. First, descriptive statistics are provided to give a clear overview about the different variables.

Descriptive statistics

Table 4 : Descriptive statistics per variable

Variables N Missing Mean SD Minimum Maximum

- 31 -

The descriptive statistics are presented in table 4. The means of the environmental variables imply that the entrepreneurs valued their environment on a 7 point scale as an average task environment concerning dynamism, complexity and hostility. The two indicators of firm innovation differ from each other, whereas on the one side the degree of firm innovation is fairly low (degree: 2.043 on a 5 point scale), while on the other side the strategic dimension of firm innovation is relatively higher (strategic posture: 4.204 on a 7 point scale). Furthermore, the variable objective firm performance attracts attention, since sales grew on average by 0.643 in the years 2012-2013. This means that on average firm’s sales in the dataset grew by more than 1.5 times with respect to the previous year. Referring to the fact that starting firms were investigated, this number meets the expectation. Only 8 of the 62 firms, from which objective performance was measured, showed actually a negative sales growth. The rest of the firms performed in 2013 similarly or better than the previous year (2012). Even without two great outliers (sales growth of 13.4 and 14.0), the average sales growth was positive (0.20). These outliers were although included in the analysis. Regarding the variable subjective firm performance (1 and 2), respondents rated their firm on average as a well performing firm (3.620/5 point scale and 5.481/7 point scale). Regarding the control variable industry type, by far the most entrepreneurs worked for a firm in the industry of services (67 of 106).

Intercorrelation analysis

In order to examine the mutual influence between the different variables used in the analysis, an intercorrelation table (table 5) is presented, which contains all the Pearson’s correlation coefficients between these variables.

- 32 -

Table 5: Intercorrelation table

- 33 -

Multivariate analysis

As mentioned in the methodology section, four steps are needed in order to conduct a mediating analysis (Kenny, 2014). These steps are presented in different tables, based on statistical output of SPSS. Step 1 is presented in table 6, control variables are included. Since firm performance is measured in three different ways, results of three models are presented.

Step 1

Table 6 presents the regression results in order to test the first three hypotheses. The first hypothesis expected a positive relationship between environmental dynamism and firm performance. Controlled for the control variables, just the result on subjective firm performance (2) was significant, but negative (β = -2,24 p<0.05). The relationships with objective firm performance and subjective firm performance (1) were not significant. Therefore, hypothesis 1 can be rejected.

The second hypothesis expected environmental complexity to have a positive relationship with firm performance. The results show that the relationship with subjective firm performance (1 and 2) were not significant (β = -.015 and β = -.054). The relationship of environmental complexity with objective firm performance was significant (β = .303, p<0.05) However, the model was not significant (∆ R2 = .152, F = 1.333) Therefore, the results do not support hypothesis 2 and can therefore be rejected.

- 34 -

Table 6: Step 1 multivariate analysis

** Significant at the p < 0.01 level (2-tailed). * Significant at the p < 0.05 level (2-tailed).

Variables Firm performance Objective Firm performance Subjective 1 Firm performance Subjective 2 Model β β β 1 (Constant) .380 3.592** 5.628** Country -.057 .038 -.007 Service industry .053 .018 .026 Trade industry -.035 .033 -.257 Craft industry .194 -.009 -.157

Manufacturing industry excluded excluded excluded

R2 .042 .003 .091* F 0.607 .055 2.506* 2 (Constant) -1.256 4.421** 7.127** Country -.053 -.017 -.013 Service industry .208 .039 .018 Trade industry .070 .061 -.323* Craft industry .284 .000 -.161

Manufacturing industry excluded excluded excluded

- 35 -

Step 2

The next step of the multivariate analysis was to test the influence of the environmental dimensions on the degree of innovation within a firm. Table 7 presents the results of the regressions which test hypotheses 4 to 6. Since firm innovation was measured in two different ways, results of two different models are presented. Also here, regressions were controlled for the control variables country and different types of industry.

The fourth hypothesis proposed a positive relationship between environmental dynamism and firm innovation. Results show that the relationship between dynamism and firm innovation (degree) was positive (β = .213) but not significant. Furthermore, the relationship between dynamism and firm innovation (strategic posture) was either positive as well as significant (β = .248, p < 0.05) The model was significant (∆ R2

= .191, p < 0.01 and F = 3.247, p < 0.01). Control variable country had also a positive effect on the dependent variable firm innovation (degree) (β = .480, p < 0.01). Regarding the results, hypothesis 4 is supported.

The fifth hypothesis expected a positive relationship between environmental complexity and firm innovation. Referring to the results, this relationship has been found in this analysis only in relationship with firm innovation (degree). Results indicated a positive relationship (β = .265), which was significant on p < 0.05. Also in here, the model was significant ((∆ R2

= .400, p < 0.01) Hypothesis 5 is therefore supported. A positive effect was also found (β = .188) with respect to innovation (strategic posture), however not significant.

- 36 -

** Significant at the p < 0.01 level (2-tailed). * Significant at the p < 0.05 level (2-tailed).

Table 7: Step 2 multivariate analysis

Variables Firm Innovation (degree) Firm Innovation (strategic posture) Model β β 1 (Constant) 2.106 4.248** Country .498** .115 Service industry -.077 -.124 Trade industry -.178 -.296 Craft industry -,091 -.161

Manufacturing industry excluded excluded

R2 .274** .065 F 8.286** 1.742 2 (Constant) .781 2.860** Country .480** .088 Service industry .003 -.051 Trade industry -.004 -.150 Craft industry -.035 -.116

Manufacturing industry excluded excluded

- 37 -

Step 3

The last step before examining a mediating relationship, is to test hypothesis 7. This hypothesis proposed that a the relationship between firm innovation and firm performance was expected to be positive. Different regressions were run, because firm performance was measured in three different ways. Results are presented in table 8.

With respect to firm innovation (degree) and its relationship with firm performance, one result was and two results were not significant. The only significant result was the relationship between firm innovation (degree) and objective firm performance (1) (β = -.348, p < 0.05), however with an insignificant model (∆ R2 = .171 and F = 1.549). Results about the influence of firm innovation (degree) and subjective firm performance (1 and 2) were not significant.

- 38 -

Table 8: Step 3 multivariate analysis

** Significant at the p < 0.01 level (2-tailed). * Significant at the p < 0.05 level (2-tailed).

Variables Firm performance Objective Firm performance Subjective 1 Firm performance Subjective 2 Model β β β 1 (Constant) .380 3.649** 5.858** Country -.164 .028 -.001 Service industry -.250 -.069 -.097 Trade industry -.273 -.024 -.342 Craft industry -.392 -.193 -.210

Manufacturing industry excluded excluded excluded

R2 .094 .030 .086 F 1.213 .550 2.020 2 (Constant) .471 3.013** 5.943** Country -.013 .113 .060 Service industry -.128 -.020 -.100 Trade industry -.218 .090 -.348 Craft industry -.296 -.110 -.209

Manufacturing industry excluded excluded excluded

- 39 -

Step 4

The last step in the analysis is to examine the potential mediating relationships. To recall, three necessary conditions were important before conducting a mediating analysis (Baron and Kenny, 1986). There should be a significant relationship between the independent variable (task environment) and the dependent variable (firm performance). Furthermore, both the dependent and independent variables should have a significant relationship with the proposed mediator. Only in that case, the relationship between the task environment and firm performance can be mediated by a variable.

Hence, recalling table 6, a mediating relationship can only be found at the relationship between environmental dynamism and subjective firm performance (2) and the relationship between environmental hostility and subjective firm performance (1 and 2). However, the relationship between environmental dynamism and subjective firm performance (2) has a different (negative) sign than expected and therefore it is useless to examine a mediating relationship.

- 40 -

Discussion and implications

Summary of results

First, a brief summary of the results is given. Pointing out the first three hypotheses, it can be concluded that only environmental hostility has a significant negative relationship with firm performance. Due to for instance a lack of opportunities, increased competition and low prices in this type of task environment (Miller and Friesen, 1983), results show that firms perform worse than firms in a less hostile environment. The influence of dynamism and complexity on firm performance is in general lower and not significant. The only significant result was that firms in a more dynamic task environment had a negative relationship with subjective firm performance. This finding is against expectations (f.e Frank et al., 2010).

The relationship between a dynamic and complex task environment and firm innovation within these firms was expected to be positive (f.e. Freel, 2005). Results support this expectation. A hostile environment, however, does not have any significant influence on firm innovation, despite the expectations. An interesting result is the difference between Germany and the Netherlands. The Netherlands turned out to have significantly higher firm innovation within firms than in Germany, following the results. However, since only 10 Dutch firms were included in the dataset, the researcher should be careful in interpreting this result.

Although overwhelming evidence by other researchers (Rubera and Kirca, 2012; Keskin, 2006; Bowen et al., 2010; Rheea et al., 2010) was presented in the literature review, stating that firm innovation has on average a positive relationship with firm performance, this result has only partially been found within this thesis. Partially, because results indicate that firm innovation has a positive significant relationship with subjective firm performance, however not with objective performance. Surprisingly, there is a negative relationship with another measure of subjective firm performance. Due to the less convincing results, firms do not perform better with high levels of firm innovation.

- 41 -

only a hostile task environment influences a firm’s performance, but has less influence firm innovation. Differently, both a dynamic or a complex task environment have a positive relationship with firm innovation, but has little effect on its performance. The overall conclusion regarding the analyses is that a firm's task environment definitely determines the firm innovation, however this does not mean that firm innovation will lead to higher firm performance.

Discussion

Despite the expectations in the literature review, a mediating relationship has not been found. All three hypotheses which expected innovation to mediate in the task environment - firm performance relationship could not even be conducted with statistical analyses, due to the fact that criteria for mediation were not met. This can possibly stimulate a discussion whether examining innovation as a mediator of the relationship between the task environment and firm performance was appropriate.

- 42 -

It could also be discussable that firm innovation does not derive from operating in a certain task environment, yet firms can choose to operate in a certain environment depending on their level of firm innovation. Following the results, managers should avoid hostile environments because of its negative relationship on firm performance. The discussion about which construct influences another construct, can be further investigated and argued by other researchers.

In the literature, the S-C-P framework was described. Industrial economists claim that industry structure influences conduct or strategy, which could lead to firm performance (Miller, 1988). This thesis has replaced industrial structure with the task environment. Results indicate that the task environment can also be claimed to have a relationship with firm innovation. The role of the task environment can be a point of discussion, because the world is changing, hence the task environment can have changed over time (Castrogiovanni, 2002). Therefore, the influence of the task environment can have changed over time. Further research can obtain more data over time to examine the role of task environment more appropriate.

The significant finding that a hostile environment has a negative relationship with firm performance, can provide food for discussion. It can be argued that highly regulated task environments (high hostility) limit the importance of firm behavior (Tan and Litschert, 1994). These hostile task environments can delegate a passive role to managers, hereby determining the firm performance (Tan and Litschert, 1994). It is furthermore concluded that, when facing a hostile task environment, managers often are afraid of firm innovation, due to high risks with respect to their position (Tan and Litschert, 1994). However, on the other side, it could be argued that in these task environments, a manager can play a critical role (Hrebiniak and Joyce, 1985). Following this discussion, the role of firm innovation is therefore not very clear. It might be the case that firm innovation of firms facing hostile task environments are of less value.

- 43 -

task environment - firm performance relationship, with respect to the EO construct of Rosenbusch et al. (2013).

However, following the results, both firm innovation (degree) and firm innovation (strategic posture) had a significant relationship with subjective firm performance (1). However, the former had a negative sign, whereas the latter had a positive sign. These contradictory results are notable, particularly because firm innovation (degree) and firm innovation (strategic posture) correlate significantly and positively with each other. It raises the question whether firm innovation has been measured in an appropriate way. The same applies to measurements of firm performance. At least two elements of firm performance are needed to measure subjective firm performance properly (Dess and Robinson, 1984). Therefore, the second measurement of subjective firm performance in this thesis could not be an appropriate measure, because this variable is measured on the basis of a single question (one element).

Theoretical and managerial implications

It is important to examine what the outcomes of this thesis add to the literature. To start, this thesis can temper the seemingly important role of firm innovation (Child, 1972). As expected by researchers, results show that firm innovation can be important in certain task environments (f.e. Damanpour and Evan, 1984; Özsomer et al., 1997; Freel, 2005). Obviously, firm innovation plays still an important role, especially at younger firms (f.e. Audretsch, 1995). However, the effect on firm performance has not been found. Firm innovation is surely not the main reason for firms to perform better than their competitors in certain environments, despite claims by researchers (f.e. Conant et al., 1990; Dyer and Song, 1997).