Effects of market growth on the relationship between EO and performance

Effects of Market Growth on the Relationship between EO and Performance

Name: Reinald Mast R.A. Student number: 10317457

Date: 30, August 2014

Version: Final

Study: Executive Programme in Management Studies –Strategy Track) University: Amsterdam Business School, University of Amsterdam

Abstract

In recent decades, Entrepreneurial Orientation (EO) has been researched in many different set-tings and has become a popular topic for scholars of entrepreneurship. Many scholars state that the relationship between EO and performance is unmistakable and focus on mediating and mod-erating effects to clarify and deepen the EO-performance relationship. This study focuses on the influence of market growth or decline on the relationship between EO and performance. It specifically addresses the separate factors of EO: innovativeness, proactiveness, and risk-tak-ing, in growing and declining markets. This leads to a main research question: How does market growth influence the relationship between entrepreneurial orientation (risk taking, innovative-ness, and proactiveness) and performance? Through an online survey 606 respondents of busi-nesses in the Netherlands provided input for hypothesis testing. Regression analyses showed that the relationship between EO and performance is stronger in declining markets than in grow-ing markets. Furthermore, the results suggest that the different factors of EO must be examined separately in order for managers to work with these findings under different types of market conditions. Managers might wonder if they are able to alter their EO and what influences this adaptability—an unexplored area of research. Can companies really alter their EO, and does this adaptability influence performance?

Keywords: Entrepreneurial Orientation, EO, Innovativeness, Proactiveness, Risk-taking, Performance, Market growth, Processes of entrepreneurship

Table of contents

Table of contents ___________________________________________________________ 2 1. Introduction _____________________________________________________________ 4 2. Theoretical overview and hypothesis _________________________________________ 7 2.1 Entrepreneurship and entrepreneurial orientation _________________________________________ 7 2.2 Performance ______________________________________________________________________ 8 2.3 Entrepreneurial orientation and performance ____________________________________________ 9 2.4 Innovativeness, risk-taking, proactiveness and performance _________________________________ 9 2.5 Market growth ___________________________________________________________________ 11 2.6 Entrepreneurial orientation, market growth and performance _______________________________ 11 2.7 Risk taking, market growth and performance ___________________________________________ 13 2.8 Innovativeness, market growth and performance ________________________________________ 15 2.9 Proactiveness, market growth and performance _________________________________________ 17 3. Research method ________________________________________________________ 21 3.1 Procedure _______________________________________________________________________ 21 3.2 Measurements ___________________________________________________________________ 22 3.2 Sample _________________________________________________________________________ 23 3.4 Data preparation _________________________________________________________________ 26 3.5 Reliability analysis _______________________________________________________________ 26 4. Results ________________________________________________________________ 29 4.1 Correlation ______________________________________________________________________ 29 4.2 Regression ______________________________________________________________________ 31 4.3 Multiple regression _______________________________________________________________ 32 4.4 Moderation analysis _______________________________________________________________ 35 5. Conclusion and Discussion ________________________________________________ 38

5.1 Conclusions _____________________________________________________________________ 38 5.2 Limitations ______________________________________________________________________ 41 5.3 Further research __________________________________________________________________ 43 Appendix ________________________________________________________________ 45

Appendix 1: Measures and constructs ____________________________________________________ 45 References _________________________________________________________________________ 48

List of tables and figures

Percentage of Distribution of Occupations and Markets of the Sample

Alphas, KMO, Eigenvalues, Explained Variance, Items, Items Removed, and Loadings

Means, Standard Deviations, Correlations and Reliabil-ities

Regression results, R², β, level of Significance and F-val-ues

Multiple Regression Analysis

Conceptual model

Scatterplot of EO – Performance for Low, Moderate and High Market Growth

Multiple Regression with Moderation Effect of Market Growth (CBS/Respondent)

1.

Introduction

Why did Blockbuster lose the battle for home video to Netflix? The lack of entrepre-neurship on the part of Blockbuster is often put forward as one of the main reasons it went bankrupt in 2012 (Graser, 2013). With the proper entrepreneurial posture, would Blockbuster have survived and outperformed competitors such as Netflix? This study does not answer this particular question, but it does asses the problems created by a lack of entrepreneurial orienta-tion (EO) and how to cope with these issues under different market condiorienta-tions.

Head and Kirchoff (2009) claim that the U.S. economy owes most of its growth to en-trepreneurial ventures. Because of situations like the struggle between Netflix and Blockbuster and the urge to understand how entrepreneurship and performance relate, entrepreneurship has become a hot topic on corporate and academic agendas. Many scholars have studied a particular aspect of entrepreneurship; namely, entrepreneurial orientation (EO).

What exactly is EO? EO is a firm’s strategic posture regarding proactiveness, innova-tiveness and risk-taking. EO can be described as the way companies undertake new entry. The differences in processes, practices, and decision-making regarding new entry actually lead to differences in performance. “Firms with high levels of entrepreneurial orientation tend to

con-stantly scan and monitor their operating environment in order to find new opportunities and strengthen their competitive positions” (Covin & Slevin, 1991). Thus EO is about the processes

of entrepreneurship rather than the content (Peters & Waterman, 1982). A great deal of research in the field of EO and its effects on performance has been performed in recent decades (Covin & Slevin, 1991; Dimitratos, Lioukas, & Carter, 2004; Miller, 1983; Miller & Friesen, 1978; Rauch, Wiklund, Lumpkin, & Frese, 2009; Venkatraman, 1989; Wiklund & Dess, 2003; Wiklund & Shepherd, 2003; Zahra & Covin, 1995; Zahra & Garvis, 2000). These studies con-cluded that EO can positively influence performance.

The question then arises as to how a changing market environment influences this rela-tionship between EO and performance. An example of such a changing market is the financial market. The global financial crisis has had a great impact on how the financial sector operates. All of a sudden, faith was lost in the financial system and banks failed. This crisis forced banks to reconsider their current business model, and has had a very large impact on the strategic decision-making of these firms (Watts, 2009). However, there has been a lack of research into the differences between EO and its separate factors before the crisis and after, even though managers might wonder which separate factors of EO became more important after the crisis. Although EO has been studied before, there is lack of a theory that provides insight into how the separate factors relate to performance in growing or declining markets. The current literature has researched EO in relation to performance (Covin & Slevin, 1991; Dimitratos, Lioukas, & Carter, 2004; Miller, 1983; Miller & Friesen, 1978; Venkatraman, 1989; Wiklund & Dess, 2003; Wiklund & Shepherd, 2003; Zahra & Covin, 1995; Zahra & Garvis, 2000) and market growth in relation to performance (Slater & Narver, 1994; Greenley, 1995) as separate factors. Covin and Slevin as well as Wiklund and Shepherd investigated EO itself but did not make a distinction as to which factors are more important under certain market conditions. It is quite likely that different market conditions require focus on different factors of EO. These studies draw no conclusions as to how innovativeness, risk taking, and proactiveness differ in a declining market as compared to a growing market. Companies can focus on different factors of EO and still achieve the same performance outcome. This implies that is important to study the influence of the different factors of EO separately and understand how they relate to perfor-mance. This study provides insight into how innovativeness, risk-taking, and proactiveness re-late to performance under different market conditions, focusing specifically on market growth and decline. In other words, how are the key entrepreneurial processes of different firms influ-enced by the current state of the market with regard to its growth or decline, and how does this

relate to the performance of those firms? This leads to the main research question that is ad-dressed in this study: How does market growth influence the relationship between

2. Theoretical overview and hypothesis

This chapter will present an overview of the most relevant literature about EO, perfor-mance, market growth, and their interrelatedness. Initially, the main differences between entre-preneurship and EO will be discussed as well as the unit of analysis of this particular study. Secondly, the hypotheses proposed in earlier studies of EO will be further examined and will lead to the first set of four hypotheses (H1, H1a, H1b & H1c) of this study. All three factors, innovativeness, risk-taking, and proactiveness, will be briefly evaluated as to their relation to performance. Finally, this chapter will discuss the market growth construct and interrelate it with the elements of EO and performance. This generates eight more hypotheses (H2, H2-, H2a, H2a-, H2b, H2b- & H2c, & H2c-) that examine the moderating effect of market growth on the rela-tionship between EO and performance. All the hypotheses with a minus sign (-) at the end of its designation indicate a negative moderating effect, since literature provides argumentation for the existence of both positive and negative moderating effects.

2.1 Entrepreneurship and entrepreneurial orientation

EO is focused on how a firm is organized with regard to exploring and exploiting op-portunities (Wiklund & Shepherd, 2003). EO, as proposed by Lumpkin and Dess (1996) and Wiklund and Shepherd (2003), consists of three factors: innovativeness, risk taking, and proac-tiveness. There is a clear distinction between EO and entrepreneurship; “Entrepreneurship

ex-plains what new entry consists of, and EO describes how new entry is undertaken” (Lumpkin

& Dess, 1996). EO focuses more on the strategic posture of the firm: the processes of entrepre-neurship rather than the content. Lumpkin and Dess (1996) state that successful entrepreneur-ship can be influenced by the actions taken regarding new entry. Thus EO is not about new entry itself, but the processes leading up to new entry. In addition, EO is more about a firm’s overall culture than about cherry-picking which attributes of EO are the best fit for the current

situation (Lee & Peterson, 2001). Therefore, even though managers might have more insight into which of the separate factors are more important, changes cannot be made overnight.

The appropriate unit of analysis for studies concerning EO is often debated. Kilby (1971) stated that, since most revolutionary inventions come from individuals, the unit of anal-ysis in entrepreneurship studies should be the individual. Guth and Ginsberg (1990), on the other hand, focus on corporate entrepreneurship and argue that strategic growth and renewal is more difficult without the appropriate resources. This study will use the same level of analysis used by Lumpkin and Dess (1996); namely, the firm-level, because that is where entrepreneurial processes (EO) best take place.

2.2 Performance

Murphy (1996) points out that performance measurement in the entrepreneurial field varies widely. Studies use different dimensions of performance such as efficiency, growth, and profit. He examined 51 empirical entrepreneurship studies that used performance as a depend-ent variable for their study. If the performance of businesses is measured by a narrow set of questions focused on short-term profitability, this will fail to generate an adequate indicator of performance. For example, heavy investments in innovation and R&D might cause a drop in profitability in one year and an increase in profitability in another year. Thus the type of per-formance measurement can have a great influence on the results of a study. In this study, the measurement of performance will be based on seventeen different dimensions of performance, which will be explained later in the method section of this thesis. Since Murphy (1996) explic-itly stated that different entrepreneurial studies might require different forms of performance measurement, this study uses the same measurement used by Wiklund and Shepherd (2003), since they also researched EO and its relation to performance. Since these measures focus pri-marily on how the performance of a firm compares to its two biggest competitors, another set

of questions was added to the survey to measure performance relative to the firm itself (Steensma, Tihanyi, Lyles, & Dhanaraj, 2005).

2.3 Entrepreneurial orientation and performance

Scholars have studied the relationship between EO and performance over and over again, including moderators such as firm size and flexibility (Ha-Brookshire, 2009; Lumpkin & Dess, 1996; Zahra & Covin, 1995; Zahra & Garvis, 2000). Ha-Brookshire (2009) as well as Zahra and Covin (1995) and Zahra and Garvis (2000) found a strong positive relationship tween EO and performance. Lumpkin and Dess (1996) also acknowledge the relationship be-tween EO and performance and suggest further studies focusing on moderating effects. Ha-Brookshire (2009) found that firm size did not have a statistically significant effect on the rela-tionship between EO and performance. To clarify that the direct relarela-tionship between EO and performance as found in previous studies is also supported by the results of this study, the first hypothesis will simply focus on the relationship between EO and performance.The EO con-struct was also researched in the study performed by Wiklund and Shepherd (2003). Since Wiklund and Shepherd (2003) already proved that there is a relation between EO and perfor-mance, the results for this study should confirm this. Nevertheless, all factors, including the three factors of EO, will be separately tested for a direct effect in order to confirm that this sample generates the same results as did prior studies regarding EO.

Hypothesis 1: Entrepreneurial orientation has a positive effect on a firm’s performance.

2.4 Innovativeness, risk-taking, proactiveness and performance

The three separate factors of EO are defined as innovativeness, proactiveness and risk-taking. Innovativeness is explained by Lumpkin and Dess (1996) as a tendency to emphasize novelty and new ideas, as opposed to existing technologies and current practices. Kleinschmidt

and Cooper (1991) argue that the relationship between innovativeness and performance is un-mistakable, since successful innovations lead to greater performance. Proactiveness could be seen as “a firm’s tendency to influence the environment and even initiate change” (Sandberg, 2002); in other words, a firm’s level of control over the environment. Risk-taking can be defined as the “willingness to commit large amounts of resources to projects where the cost of failure

may be high” (Miller & Friesen, 1978). This study will research all elements of EO in relation

to performance so that the moderating effect of market growth can be measured and translated into more specific managerial implications.

The research of Lumpkin and Dess (1996) originally suggested five factors of EO, whereas Miller (1983) only focused on three. This research uses Miller’s (1983) original three factors because this gives managers enough factors to differentiate in times of market growth or decline, but it does not overload the EO construct with too many factors. Five EO factors could make it difficult for managers to alter their EO in case of market growth or decline.

Many studies concerning EO have shown that there is a positive relationship between EO and performance. Moreover, taken separately, innovativeness, risk-taking, and proactive-ness all have a positive relationship with performance. These previous studies supply sufficient evidence for a positive relationship between these separate factors of EO and performance to predict that the same relationship will be found in this study. Nevertheless, this study will test the separate factors of EO in relation to performance as well, to confirm that it will yield the same result. This leads to the following three hypotheses:

Hypothesis 1a: Risk-taking has a positive effect on a firm’s performance. Hypothesis 1b: Innovativeness has a positive effect on a firm’s performance. Hypothesis 1c: Proactiveness has a positive effect on a firm’s performance.

2.5 Market growth

Market growth, as explained by Narver and Slater (1994) in their research concerning market orientation, is the growth of sales in a market1_{. Kohli and Jaworski (1990) predicted that}

the business environment would have a great deal of influence on the relationship between market orientation and performance. This prediction is based on the premise that companies change their market orientation if their business environment changes, a premise that also ap-plies to the relationship between EO and performance. If the business environment changes, a company’s EO might also change. Slater and Narver found moderation of the market growth variable on the relationship between market orientation and performance. A previous study by Bass, Cattin, and Wittink (1978) also suggests that market growth is an important factor that influences performance.

This study uses the same construct of market growth as initially proposed in the research of Narver and Slater (1994). It will provide an indicator of the relative growth of the size of the market over the past three years. Since this is not a longitudinal study, this measure slightly obscures the time-effect between EO and performance, as an average of the past three years is measured.

2.6 Entrepreneurial orientation, market growth and performance

Lumpkin and Dess (1996) suggested that future research should focus on market char-acteristics as a moderating variable. Market growth, as discussed by Bass et al. (1978), is one of the most important market structure variables in relation to profitability. Their research sug-gests that market growth has a strong positive relation to profitability. Therefore, firms in grow-ing markets are more likely to be profitable. Kohli and Jaworski (1990), on the other hand, state

that in times of high market growth and prosperity, customers will be less critical of their ex-penditures so there is “easy money” to be made. In times of weak economic prosperity, con-sumers consider every dollar before spending it. Although market growth and economic pros-perity are not the same thing, it is likely the same reaction will occur in times of market growth as with economic prosperity, because market growth and economic prosperity are closely re-lated (Headd & Kirchhoff, 2009; Osterman, 1999). Given the assumption that high perfor-mance is more likely in growing markets, one could conclude that there is less need to act entrepreneurially in these types of markets. This conclusion is also supported by Davidsson (1991) who claims that there are three determinants of performance: (1) need, (2) ability and (3) opportunity. Need-based entrepreneurship usually occurs in firms that are active in declining markets. Companies in declining markets are highly motivated to act entrepreneurially because overall performance is lower (Bass et al., 1978). There is simply more need to improve the company’s current performance. This would influence the strength of the relationship between EO and performance in different types of markets, as proposed earlier. Therefore, the relation-ship between EO and performance might be stronger in a declining market, which suggests that the relationship between EO and performance is negatively moderated by market growth, lead-ing to the followlead-ing hypothesis:

Hypothesis 2-: Market growth has a negative moderating effect on the relationship be-tween entrepreneurial orientation and firm performance.

However, the causal direction between EO and performance must not be forgotten in this discussion. Does low financial performance cause firms to increase EO, or does the lack of EO lead to lower financial performance, and vice versa? For the most part, firms owe their financial performance to their EO (Covin & Slevin, 1990; Rauch et al. 2009). This is also

acknowledged by Darroch (2005) who claims that EO is more of an embodiment of entrepre-neurial processes in a firm, rather than characteristics that can be exchanged whenever desired. Hax and Majluf (1984) add that the effect as described by Kohli and Jaworski (1990) only occurs under extreme market conditions where market demand transcends supply. In addition, they agree that slow-growing or declining markets follow a predictable path regarding customer needs. This generally leads to a main focus by customers on price (Hax & Majluf, 1984). There-fore, the margins in a slow-growing or declining market are thinner, which automatically leads to lower financial performance. This would suggest that the relationship between EO and per-formance is stronger in times of market growth where margins are higher and the financial effect is larger (Slater & Beck, 1994). Davidsson (1991) agrees with this premise, and suggests that firms in a growing market can see more opportunities and are able to convert these oppor-tunities into performance. Logically, EO firms would be able to detect and utilize market op-portunities better given their proactive posture, no matter what market they are in. However, in growing markets the relationship between EO and performance should be more present than in a declining market, due to the difference in margins and the fact that there are more opportuni-ties to be detected (Davidsson, 1991). Therefore, a hypothesis contrasting with H2- is proposed that indicates that market growth positively moderates the relationship between EO and perfor-mance.

Hypothesis 2: Market growth has a positive moderating effect on the relationship be-tween entrepreneurial orientation and firm performance.

2.7 Risk taking, market growth and performance

How do the separate factors of EO relate to market growth or decline? Risk-taking and its relationship to performance is often debated. Brealy and Meyers (1981) state that these two

factors are positively correlated. The results of Bromiley (1991) suggest that low average mar-ket and industry performance indicate greater uncertainty about expected income. Could it be that risk and performance might have an interaction effect that suggests a vicious circle? Bro-miley states that “poor performance has a negative influence on risk, which has a negative

influence on future performance”, where risk indicates the willingness of a firm to take risks.

Vice versa, more risk leads to higher performance, which provides more financial space for risky activities. Thus firms in a growing market might take more risk, but this will lead to greater performance. This leads to H2a that posits that market growth has a positive moderating

effect on the relationship between risk-taking and performance.

Hypothesis 2a: Market growth has a positive moderating effect on the relation between risk-taking and firm performance.

Bromiley adds that the effect of this vicious circle of risk and performance is rather small and that there are other factors that might eclipse the effect of this feedback loop. But how do declining markets differ from growing markets in this respect? Thornhill and Amit (2003) show that market decline is positively correlated (β = .15) with the total number of bankruptcies. Assuming that most bankruptcies result from financial distress, this indicates that it is more difficult to perform in a declining market than in a growing market. Entrepreneurial firms that take risks in a declining market might have a greater chance of bankruptcy, but ex-pected performance is higher in comparison with the competition (Singh, 1986). This suggests that there might be a sampling bias in this study. Since there are fewer bankruptcies in growing markets, the influence of risk on performance is more present for the part of the sample that operates in a growing market. Both flawed and successful risky decisions are represented in measures of performance in growing markets, whereas in declining markets the flawed risky decisions are less present because firms are more likely to go bankrupt. Since this study focuses

on active firms only, bankrupt firms are not included in this research. Risk-taking in a growing market does not lead to as many bankruptcies as in a declining market, since overall perfor-mance is higher in a growing market. Firms in a growing market that took risks and yet failed to improve performance still comprise part of the sample used for this research, in contrast to firms in a declining market that went bankrupt because of a flawed risky decision, which are not included in the sample. Firms with flawed risky projects in declining markets might have a greater chance of bankruptcy. In growing markets flawed risky projects lead to lower perfor-mance but are not followed by bankruptcy since there is a larger financial cushion. In other words, the effect of EO on performance is larger in declining markets. Therefore, market growth should have a negative moderating effect on the relationship between risk-taking and perfor-mance. Since firms that operate in a declining market are more often confronted with financial distress (Thornhill & Amit, 2003), they are also more likely to take risky decisions. They pre-sumably have a more all-or-nothing kind of posture. Since only firms that succeed are included is this sample, market growth will have a negative moderating effect on the relationship be-tween risk-taking and performance. This leads to a hypothesis contrasting with hypothesis H2a concerning the moderating effect of market growth.

Hypothesis 2a-: Market growth has a negative moderating effect on the relationship be-tween risk taking and firm performance.

2.8 Innovativeness, market growth and performance

Leiponen (2000) discusses the link between innovativeness and performance. He sug-gests that there are different factors that influence the profitability of innovating companies than for non-innovating companies. In markets where innovativeness is expected and more common among competitors, the lack of innovativeness might have a greater impact on performance

(Audretsch, 1995). Several authors have discussed the influence of the external business envi-ronment on the relationship between innovativeness and performance and acknowledge the value of innovativeness (Zahra, 1996; Zahra and Bogner, 1999). Fast-growing markets dictate shorter product life cycles with a greater chance of obsolescence of current products or services (Jansen, Van Den Bosch, & Volberda, 2006). Logically, these two factors are interrelated; more innovations create shorter product life cycles which in turn ask for more innovations. Darroch (2005) adds that companies owe their financial performance to their innovative posture and that innovativeness is not a light switch that can be turned on or off; it is a part of the culture of a firm. One of the determinants of innovativeness is the amount of revenue that flows back into new innovations. This is acknowledged by Capon, Farley, Lehmann and Hulbert. (1992), who state that growing markets have more financial resources to invest in innovations and R&D: “Moore and Tushman (1980) suggest that major product innovations are likely to occur in the

introductory and growth stages of the product class life cycle, whereas production process in-novations are more important in maturity and decline. They argue that product innovation in the introductory stage is frequently radical and discontinuous but that later in the life cycle a dominant design emerges to become the basis for product standardization. Firms that compete in rapidly growing markets should thus be active product innovators” (Capon et al., 1992).

These findings indicates that in declining markets, firms should focus more on process innova-tions rather than disruptive product innovainnova-tions. As stated by Damanpour & Gopalakrishnan (2001), process innovations are less correlated with performance than product innovations. If growing markets focus on product innovations, which are more correlated with performance than process innovations, than market growth should have a positive moderating effect on the relation between innovativeness and firm performance.

Hypothesis 2b: Market growth has a positive moderating effect on the relationship be-tween innovativeness and firm performance.

Damanpour and Evan (1984) studied innovativeness and its relationship with perfor-mance. Results showed a positive relationship between these two variables. A similar effect was found in the research of Subramanian and Nilakata (1996). They state that the pressure to be innovative might be more present in a declining market since companies are worried about their survival and future existence. If companies do not intervene in their current situation, their future existence might be in jeopardy. Cho and Pucik (2005) mention this effect in their research and suggest further research on the influence of external factors on the relationship between innovativeness and performance. They report that certain market conditions might very well lead to a stronger effect of innovativeness on performance. For the same reason that there might be a negative moderating effect of market growth on the relationship of risk-taking and perfor-mance, this effect might also occur for innovativeness. Thus firms in a declining market that foresee financial problems in the future might act more innovatively than other firms, leading to better performance.

Hypothesis 2b-: Market growth has a negative moderating effect on the relationship between innovativeness and firm performance.

2.9 Proactiveness, market growth and performance

The term proactiveness as used in studies regarding EO can be described as the forward-looking perspective of a firm in anticipating or shaping the environment (Miller & Friesen, 1978). Lumpkin and Dess (1996) argue that competitive aggressiveness and proactiveness should not be mistaken for the same construct. The definition of proactiveness used for this study is that first suggested by Covin and Slevin (1989), because its relation to and importance for the EO construct is proven. Covin and Slevin (1989) argue that forceful proactiveness can be extremely dangerous in the presence of certain economic conditions or market pressures.

The logical assumption is that proactiveness in a growing market generates better performance because there are simply more opportunities to chase in a growing market. Also, proactive firms tend to benefit more from first-mover advantages than firms with a less proactive stance (Lieberman & Montgomery, 1988). This assumption is also supported by Wiklund and Dess (2003). Ward et al. (1994) claim that proactiveness is closely linked to performance because proactiveness helps firms to seek, see, and seize opportunities. Once more, this is reasoned from the angle of ability and opportunity. But how does market growth or decline influence this relationship? Growing markets provide more space to be proactive as a firm, while on the other hand, declining markets increase the urgency for firms to be proactive. “For entrepreneurial

opportunities to exist, people must not agree on the value of resources at a given point in time”

(Eckhardt & Shane, 2003). Therefore the question to be answered is, where does this disagree-ment occur more often, in growing or declining markets? Since margins are higher in growing markets (Hax & Majluf, 1984), the disagreement about the value of the resources, in combina-tion with the price of the service or products, is more likely to occur in growing markets than in declining markets. Therefore, opportunities generally present themselves more often in grow-ing markets, which is directly linked to the definition of proactiveness that firms can influence the environment and initiate change. This leads to the following hypothesis:

Hypothesis 2c: Market growth has a positive moderating effect on the relationship be-tween proactiveness and firm performance.

However, Davidsson (1996) addresses the differences between the need, ability, and opportunity for growth, and these differences also apply to proactiveness. Firms that operate in a declining market might have greater urgency to be proactive and alter their current perfor-mance. The disagreement about the value of resources might just as well occur in a declining market since the market itself is declining, indicating less need for the product. In order to

change the shrinking effect of the market, firms must be proactive and alter their business en-vironment so performance can be improved. Fast-growing markets also indicate greater uncer-tainty about future development, which makes it more difficult for firms to actively shape their business environment (Miller & Friesen, 1978). On the other hand, declining markets tend to have a clearer path for the development of the market, leading firms to be able to be more proactive. Since the relationship between innovativeness and performance is already proven, this would lead to the premise that firms in declining markets can be more proactive, which indicates greater performance.

Hypothesis 2c-: Market growth has a negative moderating effect on the relationship be-tween proactiveness and firm performance.

3. Research method

This chapter describes the methodology used to acquire the data to test the proposed hypotheses. In addition, the procedure, the measurement, and the sample are further explained. To test the proposed hypotheses, a deductive and explanatory study was conducted trying to clarify the relationship between existing variables. This chapter focuses on the details of the data and lays the groundwork for hypothesis testing.

3.1 Procedure

Data was collected through an online survey, and secondary data from a database from the Central Bureau of Statistics (CBS) was also used. Survey participants were invited by email to participate in this study. To obtain email addresses for prospective survey participants, a database was purchased in collaboration with another business student. The data from the online survey covered all of the variables as proposed in the conceptual model. The secondary data from the CBS was used as an objective measure of market growth next to the subjective measure of the survey.

This study used online questionnaires to gather data, utilizing Qualtrics software. A total of 26,886 potential respondents were invited to participate in this study. These respondents met the following criteria: 1) work in a firm active in the Dutch market, 2) speak Dutch and 3) work in a company that has at least 10 employees. Occupations in the research sample range from executive directors to product managers and marketers. Respondents were asked if they would like to receive the results of this research, and as a small gesture of gratitude a gift certificate was raffled off among participating respondents.

The questionnaire was pre-tested with a group of 15 acquaintances to test whether the survey was adequate to distribute to the total sample. Results showed that the questionnaire was

clear, the time estimate was accurate, and that the survey could be distributed to all potential respondents.

3.2 Measurements

The dependent variable in this study is performance. This research used the performance indicator used by Wiklund and Shepherd (2003) and also a performance measure from Steensma et al. (2005). The survey question regards the development of the firm over the past three years in comparison with their two biggest competitors on ten different sub-measures of performance. Responses about performance relative to a firm’s competitors could depend on the specifics of the competition in the particular market. Furthermore, respondents from the same firm might assign different competitors due to their understanding of how far competition goes. For instance, a newspaper might look at other newspapers as competitors or look at online-based newspapers as competitors, depending on the understanding of the respondent as to what competition is. Therefore, seven questions from Steensma et al. (2005) were incorpo-rated to provide another indication of performance that is more focused on the increase in per-formance in relation to the firm’s previous perper-formance. Both series of questions allow answers on a 5-point Likert scale and concern the years 2011, 2012 and 2013.

EO is the independent variable in this study. To establish data about the EO of a firm, this study incorporated questions as used in the research of Covin and Slevin (1996), in which they state that EO consist of three items as factors of EO: research innovativeness, risk-taking, and proactiveness. These measures are common in EO research, and are studied for their rela-tionship to performance. Covin and Slevin’s (1996) results on these separate factors lead to an indicator of EO (α = .87). Several other researchers have used the same measure (Lumpkin & Dess, 2001; Wiklund & Dess, 1996; Wiklund & Shepherd, 2003). All three separate factors of EO were measured through a series of three 7-point Likert scale questions per factor.

Market growth is the moderating variable in this study. Market growth is measured via three questions from Narver and Slater (1994) and one question from McDougall, Covin, Rob-inson and Herron (1994). The questions from Narver and Slater (1994) are open questions with a required entry of the annual growth percentage of their firm in the years 2011, 2012 and 2013. Answers for annual growth percentages were limited for response (-25% – +25%). The meas-urement from McDougall et al. (1994) is a 5-point Likert scale that measures the growth or decline of demand in a specific market. Respondents were asked in which market they are cur-rently working before the answers concerning market growth were posed. Questions from Nar-ver and Slater as well as McDougall were chosen because they refer to a three-year time period and are commonly used in other studies. Additionally, the objective data of the CBS was added to measure economic conditions in different markets. Since certain submarkets can have differ-ent market growth than others, questions are included about market growth in the questionnaire. To connect the questionnaires with the external database of the CBS, participants were asked in which industry they work. Industry distribution was based on the four-number SBI 2008 Index, which is equivalent to the NACE index of the European Union. This is the same distri-bution used in the secondary data provided by the CBS. Control variables were added in the form of number of years worked at the company and the total number of employees of the company.

3.2 Sample

The distribution list consisted of a total of 26,886 potential respondents. 5,745 respond-ents opened the email, of which 1,593 started the survey. After a week, a reminder was sent to the respondents who had not already filled out a survey. This generated 202 extra completed surveys in addition to the 404 already obtained from the first mailing, which resulted in a total of 606 questionnaires with all the obligatory questions answered. The total of 606 completed

surveys from the two email distributions resulted in a response rate of 2.25%. With a confi-dence level of 5%, this sample had a conficonfi-dence interval of 3.94%

The respondents in the sample were mostly male (82%) with only a small portion being female (18%). This distribution is similar to the distribution of the potential respondents (male = 93%, female = 7%). It is a commonly accepted fact that women tend to be more willingly to fill in a survey (Sax, Gilmartin, & Bryant, 2003). On average the respondents took 6 minutes, 8 seconds to complete the survey. Furthermore, respondents were asked about their current occupation and how long they had been currently active at their firm. The validity of the re-sponses increases when employees have worked longer for their firm. The total number of years working at their current company for respondents who completed the survey ranged from 1 to 46 years, with an average of 19.27 years (SD = 10.54).

Respondents from different industries participated in the survey. Most were active in manufacturing (32%), commodities trading (14%), construction (14%), services (9%) or transport (6%). The remaining respondents (26%) were active in other industries (Table 1). The occupations of respondents included executive directors (70%), managers (11%) and marketers (3%). Assuming that executive directors and managers have a good understanding of how the market and the firm have developed over the past three years, these numbers indicate sufficient reliability of the data. The sample that was used for this study contained employees of firms in the Netherlands with 10 or more employees. This ensured that the firms are actually large enough to have a view as to their strategic posture. Companies in the sample had an average of 332.86 employees (SD = 5481.65), with a minimum of ten employees.

Table 1

Percentage of Occupations and Markets of the Sample

Occupation Percent Market Percent

Executive director 69.90 Manufacturing 32.20

Manager 10.50 Commodities trading 13.94

Marketeer 3.00 Construction 13.37

Sales 2.60 Services 8.47

Administrator 2.40 Transport and storage 5.65

Controller 2.40 Specialized business services 5.46

Branch manager 1.90 Information and communication 4.71 Human resources 1.30 Agriculture. forestry and fishing 4.71

Personnel department 0.80 Financial services 2.26

Policy Advisor 0.60 Health and welfare 1.69

Product manager 0.60 Energy 1.32

Communications advisor 0.40 Catering 1.32

CTO 0.40 Education 1.13

Product Specialist 0.40 Rental and other business services 0.94 Assistent management 0.20 Water supply and waste 0.75

Boardmember 0.20 Mining and quarrying 0.56

Chief Technology Officer 0.20 Public administration and services 0.56 Executive assistant 0.20 Culture, sport and recreation 0.56

Finances 0.20 Rental and commercial property 0.19

Sales 0.20 Households 0.19

Research and 0.20 Total 100.00

Information technology 0.20 Quality coordinator 0.20 Management 0.20 Receptionist 0.20 Surveyor 0.20 Publisher 0.20 Chairman governing 0.20 Observer 0.20 Total 100.00

3.4 Data preparation

Since a many of the questions in the survey asked about the performance of the firm over the last three years, it is logical that the answers of respondents working fewer than three years for their current company would not present an accurate image of their firm’s perfor-mance. Therefore, the results of the 51 respondents that had worked less than three years for their current firm were removed. After these respondents were removed, a data set of 555 com-pleted surveys remained. Since only surveys are used that filled out all the obligatory questions, no incomplete surveys needed to be deleted from the dataset nor were data replacement tech-niques used.

The data from every question was transformed into a z-score and analyzed for outliers. A total of 24 surveys were removed because z-scores of questions were lower than -3.29 or higher than 3.29 (Field, 2009). Eleven of these outliers occurred on the sub-measure of the performance measure (Wiklund & Shepherd, 2003) regarding relative lower costs per unit. Nine outliers occurred on the sub-measure of the quality of the product or service provided in the performance measure (Steensma et al., 2005) in relation to the firm’s two biggest competitors. Four outliers occurred on the sub-measure of increase of productivity of employees in the rela-tive performance measure of Wiklund & Shepherd (2003). After deleting these surveys, the dataset contained data from 531 respondents. The annual percentages of market growth for the past three years from the CBS database were added as three new variables (2011, 2012 & 2013) for the market growth construct and linked with the industry of the respondent’s choice.

3.5 Reliability analysis

The actual data was analyzed using quantitative methods. This research employed ques-tions from other studies in order to ensure a high amount of validity and reliability. At first a principal component analysis (PCA) was performed with orthogonal rotation (varimax). The

Kaiser-Meyer-Olkin (KMO) measure verified the sampling adequacy for the analysis. All KMO values for individual items were > 0.5, which is acceptable according to Field (2009). Χ2

(136) = 4583.30, p <.001, indicated that correlations between items were sufficiently large for PCA. All five components had eigenvalues greater than Kaiser’s minimum of 1. In total, four components were retained. For the performance variable, one question about overhead costs was not included on the basis of the PCA results. The EO construct is the aggregation of inno-vativeness, risk-taking and proactiveness. There was no separate PCA performed to because it consists of three different measures.

A reliability analysis was performed to test whether the answers yielded consistent re-sults. Cronbach’s αs show that EO, risk-taking, innovativeness, and performance all have a high internal consistency. All αs are greater than .7, which is indicated by Kline (2013) as a generally accepted cut-off point for business studies. The Cronbach’s α of proactiveness is .641 if all three questions as measured are included. When the third question about proactiveness was deleted, a Cronbach’s α of .745 resulted. In general a Cronbach’s α of .700 is acceptable. Rules of thumb indicate a minimal delta of Cronbach’s α for deletion of .1. The reason that this question negatively influenced the internal reliability of the proactiveness variable, might be that it focused more on a firm’s competitive aggressiveness rather than on its proactiveness. The reliability analysis of the market growth construct resulted in a Cronbach’s α of .160, which

Table 2

Alphas, KMO, Eigenvalues , Explained Variance, Items, Items Removed, and Loadings

α KMO EV %Var Items Removed L Low L High

EO .822 .824 3.820 42.44% 8 1 .531 .745 Innovativeness .730 .615 1.968 65.60% 3 0 .666 .876 Proactiveness .745 .564 1.783 59.42% 2 1 .575 .872 Risk-taking .808 .709 2.170 72.34% 3 0 .835 .871 Market growth .160 .500 1.252 62.61% 2 0 .791 .791 Performance .872 .873 5.888 34.64% 16 1 .358 .807

is extremely low. Since the objective market growth data from the CBS database and the sub-jective market growth measure of the questionnaire are not aligned, this researches continues with both variables separately. This results in the variables MarketgrowthCBS and Market-growthRespondent. Further analysis will incorporate both types of variables. A possible reason that these two items do not yield the same results could be that respondents lack the willingness to fill in accurately in which market their firm is currently active. Since the data about the mar-ket is used to link the objective marmar-ket growth data from CBS to the individual surveys, this might influence the results of this measure. All items were analyzed for statistics regarding the mean and standard deviations. Performance reported the lowest mean (M = 3.42, SD = 0.48) and firm size reported the highest mean (M = 349.39, SD = 5654.16).

4. Results

This chapter discusses the statistical analysis of the data from the sample. First, a corre-lation analysis was done. Second, a linear regression analysis was performed on the independent and dependent variables to accept or reject hypotheses H1, H1a, H1b and H1c. Third, a multiple regression analysis was performed to provide insight into the market growth variable. Finally, a moderation analysis was executed to test hypotheses H2, H2-, H2a, H2a-, H2b, H2b- & H2c, & H2c-.

4.1 Correlation

For all the variables, a Pearson correlation coefficient was calculated. Table 3 shows that most of the variables are significantly correlated. Obviously, this analysis does not provide insight into the direction of causality; it merely states a relationship between variables. The relationship between the variables Yearsworking and Performance was unexpected. It is a neg-ative correlation, and therefore shows that the longer someone has been working at their current company, the lower the performance of their company, or vice versa. Whether this has to do with the fact that people that have worked at a company for a longer time tend to be more restrained when evaluating their company’s performance, or that people that have worked at a company for a shorter time are more positive about their company’s performance, is unknown. Such a conclusion cannot be drawn from this data.

Since the EO construct consist of the separate factors (Lumpkin & Dess, 1996; Wiklund & Shepherd, 2003), it is no surprise that innovativeness, proactiveness and risk-taking all are significantly correlated with EO. Also, innovativeness, proactiveness, and risk taking are sig-nificantly correlated with market growth.

EFFECTS OF MARKET GROWTH ON THE RELATIONSHIP BETWEEN EO AND PERFORMANCE

Table 3

Means, Standard Deviations, Correlations and Reliabilities

Variables Number of items M SD 1 2 3 4 5 6 7 8 9 1. EO 8 3.89 0.99 (.82) 2. Innovativeness 3 3.83 1.31 .84 ** (.73) 3. Proactiveness 2 4.36 1.29 .78 ** .56 ** (.75) 4. Risk-taking 3 3.63 1.16 .75 ** .37 ** .41 ** (.81) 5. Market growth CBS 1 0.09 1.48 .14 ** .17 ** .11 ** .05

-6. Market growth Respondent1 0.24 8.33 .20 ** .21 ** .15 ** .10 ** .25 **

-7. Years working 1 19.42 10.60 -.04 -.01 -.05 -.04 -.02 -.03

-8. Firm size 1 328.36 5481.72 -.02 -.02 .04 -.05 .04 .02 .03

-9. Performance 16 3.44 0.46 .50 ** .39 ** .48 ** .34 ** .16 ** .41 ** (.87) .03 (.87)

Note: N = 531. Reliabilities are reported along the diagonal

** Correlation is significant at the 0.01 level (1-tailed).

Table 4

Regression results, R², β, Level of Significance and F-values

Variables R² β Sig. F 1. EO .253 .503 .000 179.331 2. Innovativeness .154 .392 .000 96.153 3. Proactiveness .234 .484 .000 161.442 4. Risk-taking .119 .344 .000 71.219 5. Market growth CBS .026 .162 .000 14.308 6. Market growth Respondent.169 .411 .000 107.757

7. Yearsworking .029 -.170 .000 15.741

8. Firm size .001 .027 .535 0.385

4.2 Regression

The measurements were further analyzed to pretest the hypotheses as proposed in Chap-ter 2. All the constructs incorporated in the hypotheses have significant correlations with each other. Regression analysis indicates if one variable can be predicted by another. Logically, the relationship between the independent variable and the dependent variable is researched as pro-posed earlier in the hypotheses.

For parametric tests such as regression, the assumptions of normal distributed data have to be met. The central limit theorem (Field, 2009) states that with a large sample size (N > 30) normality can be assumed. Since the sample size is large enough (N = 531) for the variables to be distributed normally, no data transformations have been performed to correct for skewness or kurtosis.

First, a linear regression analysis was performed to determine if and how the inde-pendent variables predicted the deinde-pendent variable. Analysis showed that EO explains 25% of the variance of performance (β = 503, p < .001). A total of eight linear regression tests were performed. These regressions were performed on both measures of market growth and perfor-mance to analyze the direct effect of market growth on perforperfor-mance. Results indicate that market growth as measured in the database of CBS and market growth as measured in the sur-vey can both significantly predict performance. However, the measure of market growth in the CBS explains merely 3% of the variance in performance, while the measure provided by the survey responses explains 17%.

Results also show that innovativeness explains 15% of the variance in performance. Proactiveness explains 23% of the variance of performance and risk-taking explains 11%. This confirms that EO itself explains most of the variance in performance. The results for the

sub-were further analyzed in a multiple regression analysis to indicate whether H1, H1a, H1b, and H1c can be rejected or accepted.

4.3 Multiple regression

Since there are multiple factors that can predict performance, a multiple regression anal-ysis was conducted. This analanal-ysis provides insight into how the three components of EO relate to performance and which has the greatest effect in the model. The analysis also develops input for an equation that can predict performance more accurately than can single predictors. Results show that innovativeness, proactiveness, and risk-taking could explain 27% of the variance in performance (Table 5). Since tolerance was at least .640, the average VIF was close to 1, and the largest VIF did not exceed 10, there was no collinearity within the data (Bowerman & O’Connell, 1900; Myers, 1990; Menard, 1995). Results therefore indicated that H1 can be ac-cepted, meaning that EO has positive relationship with performance. Innovativeness also ex-hibited a significant positive relationship with performance (β = .148, p = .001), which means that H1a can be accepted. The same positive relationship with performance was found for pro-activeness (β = .339, p < .001) and risk-taking (β = .151, p < .001), therefore H1b and H1c can also be accepted. With all factors of EO showing a significant positive relationship with perfor-mance, this results of this analysis are similar to those of other studies presented in the past.

The analysis also makes clear that market growth as an independent variable is a good predictor of performance. When the variable of market growth of respondents was added to the multiple regression analysis, innovativeness, proactiveness, risk-taking, and market growth were able to explain 39% of the variance of performance. Again, tolerance was at least .640, the average VIF was close to 1, and the largest VIF does not exceed 10, so there was no collin-earity within the data. As suspected, market growth is an important factor in predicting perfor-mance in this particular model.

In order to be able to draw a conclusion about the population based on the regression analysis of the sample, assumptions for regression analysis must be met (Field, 2009). All pre-dictor variables and the outcome variable are quantitative measures and have variances greater than 0. Since none of the variables have a correlation coefficient of 1 or near to 1 there is no perfect multicollinearity.

EFFECTS OF MARKET GROWTH ON THE RELATIONSHIP BETWEEN EO AND PERFORMANCE

Table 5

Multiple regression

R2 β Sig

Firm size, Yearsworking* 0,030 -.171 .000

Firm size, Yearsworking, Innovativeness, proactiveness*, risk-taking 0,296 .330 .000

Firm size, Yearsworking, Innovativeness, proactiveness*, risk-taking, market growth CBS (moderation) 0.298 .329 .000

Firm size, Yearsworking, Innovativeness, proactiveness*, risk-taking, market growth Respondent (moderation) 0.304 .320 .000

* = β and Sig

4.4 Moderation analysis

To test if and how market growth has a moderating effect on the three factors of EO, moderating analyses were performed for all the three factors. In order to test the moderating influence of market growth, a SPSS Marco named process from Preacher and Hayes (2008) was used. Both the moderation effect of the variable market growth as measured by respondents and market growth as indicated by the data from the study were examined to test for moderation (Table 6).

Moderating analyses were executed to test whether hypotheses H2, H2-, H2a, H2a-, H2b, H2b-, H2c and H2c- can be rejected or accepted. The first regression analysis was performed to test whether market growth as measured by the CBS has a moderating effect on the relationship of EO and performance. Results indicate that there is a small negative moderating effect (β = .008, p = .84) but the result was not significant. The moderation analysis of market growth as measured in the survey also showed a negative moderating effect but was significant, in con-tradiction to the previous analysis (β = -.105, p < .003). This indicates that H2 can be rejected and that H2- can be accepted on the basis of the market growth measure from the survey. Market growth has a negative moderating effect on the relationship between EO and performance.

A second linear regression was performed on the moderating effect of market growth on innovativeness and performance. Analysis of the market growth variable from the CBS shows a negative moderation effect but the effect is far from significant (β = .008, p = .84). Results show that there is a negative moderating effect of market growth as measured in the survey on the relationship between innovativeness and performance (β = -.090, p = .016) and the effect is significant. This indicates that H2a can be rejected but H2a- can be accepted. In other words, there is reason to accept the hypothesis that market growth has a moderating effect on innovativeness on the basis of the respondent data about market growth.

The same linear regression analysis was also conducted to determine whether market growth has a moderating effect on proactiveness and performance. Results indicate that there is a positive moderating effect of market growth as measured by the CBS on the relationship with performance (β = .007, p = .849), but the effect is not significant. The analysis of the market growth measure of the survey indicates a significant negative moderating effect of mar-ket growth on the relationship between proactiveness and performance (β = -.100, p = .005). Thus H2b can be rejected and H2b- can be accepted on the basis of the market growth measure of the survey.

The regression analysis was also performed for Risk taking. Results show that market growth has a positive moderating effect on the relationship between risk-taking and perfor-mance (β = .028, p = .487) for the CBS market growth measures but the result is not significant. The results of the moderation analysis of the survey respondents’ measure of market growth show that there is negative moderating effect on the relationship between risk-taking and per-formance (β = -.072, p = .060), but it is not significant. Therefore H2b and H2b- can be rejected.

The data was divided into three groups to analyze the strength of the relationship be-tween EO and performance in different market conditions. The surveys were divided into groups on the basis of market performance as measured in the survey, because it generated the

Table 6

Multiple Regression With Moderation Effect of Market Growth (CBS/Respondent)

Variables R² β Sig. R² β Sig.

1. EO, Market growth .366 -.105 .003 .261 .008 .840

2. Innovativeness, Market growth .276 -.090 .016 .163 .008 .840 3. Proactiveness, Market growth .361 -.100 .005 .246 .007 .849 4. Risk-taking, Market growth .267 -.072 .060 .141 .028 .487

* = β and Sig

Market growth CBS* Market growth respondent*

most significant results as a moderator. This generated groups with low market growth, mod-erate market growth, and high market growth. The scatterplot (Figure 1) clarifies the relation-ship between EO and performance for the three different groups. The steepness of the line in-dicates that this relationship is stronger within low growth markets as compared with moder-ate and high growth markets.

Figure 2

Scatterplot of EO – Performance for Low, Moderate and High Market Growth

5. Conclusion and Discussion

This chapter will discuss the most evident findings from this study and present the im-plications of the results for practice and literature. The chapter will begin with answering the research question on the basis of the accepted and rejected hypotheses, and connect these con-clusions with literature. It will also discuss the limitations of this study. Finally, suggestions for further research will be proposed.

5.1 Conclusions

This study investigated the influence of market growth and decline on the construct of EO, its separate factors, and their relationship with firm performance. All the hypotheses that tested the direct relationship between EO, its separate factors, and performance were accepted. The quantitative analysis of the relationship without moderation showed that proactiveness has the greatest influence in predicting performance in the model of the separate factors of EO. Innovativeness had the second-greatest level of influence, and risk taking, the third. These re-sults are consistent with existing conclusions about EO and performance: namely, EO is an important construct with regard to performance. However, the results of this study also provide valuable insight into how the different factors of EO influence performance. The main research question focused on the moderating effect of market growth in the relationship between EO and performance. The analysis showed that market growth had a negative moderating effect on the relationship between EO, innovativeness, proactiveness and performance. Therefore, the rela-tionship between EO and performance appears to be stronger in declining markets than in grow-ing markets.

Almost all of the hypotheses suggesting a negative moderating effect of market growth on performance were accepted, with the exception of the hypothesis regarding risk-taking. This despite the fact that the hypotheses suggesting a positive moderating effect of market growth

were quite logical or could at least be predicted according to results published in the current literature. The acceptance of H2- indicated that market growth does have a negative moderating

effect on the relationship between EO and performance. Since the linear regression analysis showed that EO can predict performance, this outcome acknowledges that EO is an important factor for companies to consider, regardless of whether the firm operates in a growing or a declining market environment. Other regression analyses do however show that EO as a con-struct is negatively moderated in its relationship with performance, but not all separate factors of EO show the same moderation effect. These findings suggest a negative moderating effect of market growth on innovativeness and proactiveness and their relation to performance, but not for risk-taking. So, where one might expect that firms have to take risks in order to outper-form competitors in a declining market, this study did not find a quantitative basis for this con-clusion. This demonstrates that it is important to evaluate the factors of EO separately, depend-ing of the type of market in which a firm is active.

From the study results, it can also be concluded that market growth is a significant pre-dictor of firm performance. Logically, one can expect that firms who operate in a growing mar-ket perform well almost automatically, which has been proven before by Bass et al. (1978). Therefore the results of the analysis of market growth and its direct relation to performance in this study suggest that certain market conditions might call for a different composition of EO factors. These findings do not only directly add to the field of research on EO, but also on entrepreneurship itself. Since entrepreneurship focuses on determining in which markets to be active or which to enter, and EO focuses on the process of entrepreneurship rather than the content, these findings could provide a small bridge between the processes and content of en-trepreneurship. The findings and conclusions could help firms to choose which markets to enter on the basis of their EO.

Analysis of the relationship between innovativeness and performance also confirmed the prior conclusions of researchers who studied EO and its separate factors (Lumpkin & Dess, 1996). Innovativeness is an important factor in predicting performance. The moderation analy-sis results of this study were significant enough to indicate a moderating effect of market growth on this relationship. In other words, innovativeness is more positively related to performance in declining markets than in growing markets.

The hypotheses regarding risk-taking proposed that market growth has either a positive or a negative moderating effect on the relationship between risk-taking and performance. Re-sults show that this is not the case. There is a moderating effect, but it is not significant. Thus risk-taking is related to performance but market growth or decline does not affect this relation-ship. Since all the past literature has incorporated risk-taking in their EO construct (Zahra & Covin, 1995, Wiklund & Dess, 2003; Covin & Slevin, 1991; Miller, 1983; Miller & Friesen, 1978; Venkatraman, 1989; Wiklund & Shepherd, 2003) and based their conclusions on the EO construct as a whole, this study would suggest the factors should be analyzed separately. Wiklund and Shepherd (2003) emphasized the practical and academic importance of EO but did not conclude how the separate factors should be valued. For managers that want their firm to act more entrepreneurially in order to increase performance this study supports the previous findings about EO and suggests that they should focus on the separate aspects of EO. The results of the moderation analysis emphasized that is important to keep the single factors in mind since not all factors of EO were negatively moderated by market growth. As mentioned before, EO and its separate factors are not light switches that can be turned on and off quickly and easily. Managers who want to alter their strategic posture regarding EO can use these findings to create a game plan for altering their firm’s EO.

The analysis of proactiveness also generated interesting results. Proactiveness is nega-tively moderated by market growth as related to performance. Thus, proactiveness has a larger impact on a firm’s performance in declining markets than in growing markets. This is the op-posite of the result of the study by Lumpkin and Dess (2001), who found that that

“Proactive-ness, a response to opportunities, is an appropriate mode for firms in dynamic environments or in growth stage industries where conditions are rapidly changing and opportunities for ad-vancement are numerous”. What could be a plausible explanation for this result? The pressure

to survive might be greater in declining markets because bankruptcy might be just around the corner. Without a proactive stand, companies could lose it all. Also, since declining markets are often maturing markets, consolidation is a regular phenomenon (Mendelson, 1987). In other words, firms in these markets seek to take over other firms for economies of scale to increase performance by lowering costs. It’s eat or be eaten. For managers that operate in a declining market, a focus on proactiveness and innovativeness might just do the trick. However, the mod-eration results also indicated that firms who operate in a growing market have lower perfor-mance effects from proactiveness than firms in declining markets. Proactiveness in a growing market still has a positive effect on performance, but not as large an influence as in a declining market.

5.2 Limitations

The findings of this study are subject to a number of limitations. First, since the data that is analyzed came from self-reports, a possible self-enhancement bias might influence the outcome of this study. Since the EO construct is built on data from respondents, this effect is common when studying EO. To counter this bias, researchers could consider finding objective measures of innovativeness, proactiveness, risk-taking and market growth.