Environmental Conditions and the Debate
on Low Performance
influencing Risk Behavior
- Master Thesis -
Author: Amanda S. Bernoster / Student No. 11395869 Msc. Business Administration – Strategy Track
University of Amsterdam, Faculty of Economics and Business
Supervisor: MSc. B (Bernardo) Silveira Barbosa Correia Lima University of Amsterdam, Amsterdam Business School
Statement of originality
This document is written by Amanda Sascha Bernoster, who declares to take full responsibility for the contents of this document. I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
INDEX
ABSTRACT 4
INTRODUCTION 5
THEORY AND HYPOTHESES 9
THE EFFECT OF PERFORMANCE BELOW ASPIRATION LEVEL ON RISK-BEHAVIOR AND THE DEBATE 9
ENVIRONMENTAL CONDITIONS 18
THE MODERATING EFFECT OF ENVIRONMENTAL DYNAMISM 19
THE MODERATING EFFECT OF MUNIFICENCE 22
ASPIRATION LEVEL PREFERENCES UNDER ENVIRONMENTAL CONDITIONS 24
METHOD 27 SAMPLING STRATEGY 27 DEPENDENT VARIABLE 28 INDEPENDENT VARIABLE 28 MODERATING VARIABLES 29 CONTROL VARIABLES 30 DATA ANALYSIS 31 RESULTS 32
DESCRIPTIVE STATISTICS AND CORRELATIONS 32
REGRESSION MODELS 34
T-TEST AND INTERACTION FIGURES 37
DISCUSSION 42
THEORY REVIEW,MAJOR FINDINGS AND CONTRIBUTIONS 42
LIMITATIONS AND FUTURE RESEARCH 48
CONCLUSION 51
REFERENCES 52
ABSTRACT
There is a relationship between performance feedback and the risk behavior of firms. Previous research illustrates an active debate about this relationship between performance feedback and risk behavior when the performance is below a firms’ aspiration level. The behavioral theory of the firm and prospect theory predict that performance below an aspiration level increases risk taking behavior, although research on organizational decline proposes that performance below an aspiration level decreases risk taking behavior. Researchers expect that this contradiction can be explained by different boundary conditions of the firm, which impact this relationship. This research investigates the moderating effect of environmental conditions on this relationship. I posit that firms who operate under different environmental conditions interpret performance feedback in another way and thus influence risk behavior. Additionally, I posit that decision makers who operate under different environmental conditions have a preference for an aspiration level, because one aspiration can reflect goals and opportunities for a firm in this environment better then the other aspiration. Using data on research and development spending of U.S. Electronical Hardware Industry and the environmental conditions dynamism and munificence, I find that firm risk behavior varies significantly depending on whether a firm experiences environmental conditions. However, I cannot imply risk seeking or risk adverse behavior because this varies between different aspiration levels. I also find that the differences within the environmental conditions impact preference for an aspiration level.
INTRODUCTION
Over the last 50 years, performance feedback theory is one of the popular research programs in strategic change in organizations. The central argument in this theory is that decision makers use an aspiration level to evaluate firm performance (Cyert & March, 1963; Kahneman & Tversky, 1979; March & Shapira, 1987, 1992; Shapira, 1986; Audia & Greve, 2006). There are two performance aspiration levels; historical aspiration and social aspiration. Historical aspiration is based on past performance of the firm. Social aspiration is based on performance of similar firms (Cyert & March, 1963). It is found that performance aspirations influence the risk behavior of firms. Most studies in organizational learning adopting this theory suggest that when performance is above the aspiration level of a firm, increases in performance decrease risk-taking actions, so firms show risk averse behavior (Bromiley, Miller & Rau, 2001; Nickel & Rodriguez, 2002; Audia & Greve, 2006). Decision makers of high performing firms are expected to take status quo preserving measures and to avoid actions that might negatively influence performance (Cyert & March, 1963; Audia, Locke & Smith, 2000).
Although, when performance falls below firms’ aspiration the effect on risk behavior remains subject to active debate (Lopes 1987; March & Shapira 1987; Ocasio 1995; Sitkin & Pablo 1992; Audia & Greve, 2006; Shinkle, 2012; Schimmer & Brauer, 2012). This debate shows two opposing arguments. On one hand, research in traditional behavioral decision theory and prospect theory argues that performance below aspiration level of a firm increases risk taking and changes, which implies risk seeking behavior (Cyert & March, 1963; Kahneman & Tversky, 1979; Bromiley, 1991; Greve, 1998; Audia & Greve, 2006). When performance is below aspirations, managers may engage in problemistic search in order to improve the firms’ performance (Cyert & March, 1963). On the other hand, research on organizational decline suggests that firm performance below the firm's aspiration level leads to less risk taking and rigidity of firms, which implies risk averse
behavior (Staw, Sandelands & Dutton, 1981; Lopes, 1987; Sitkin & Pablo, 1992; Wiseman & Bromiley,1996; Audia & Greve, 2006). This contradictory theoretical perspective and conflicting empirical evidence makes this a fertile area for research and immediately brings numerous questions. Researchers have suggested that the opposing arguments regarding risk behavior may be due to unobserved heterogeneity and have proposed many contingencies and moderators, or also called boundary conditions. Firms under different boundary conditions possibly will interpret performance feedback in another way, and this can lead to differences in risk taking behavior (Greve, 2003; Audia & Greve, 2006; Shimizu, 2007; Short & Palmer, 2003; Vissa et al., 2010, Shinkle, 2012). Prior literature studying boundary effects provides insight under which conditions risk aversion or risk seeking prevails. Slack search (Greve, 2003), firm size (Audia & Greve, 2006), operating experience, poor legitimacy, and structural inertia (Desai, 2008), and corporate structure (Gaba & Joseph, 2013) affect the relationship between performance feedback and risk behavior. Studying boundary conditions of firms and its affect on the relationship between performance feedback and risk behavior extends our knowledge about this topic. Moreover it may contribute in solving the debate about differences in risk behavior when performance is below aspirations. Some studies describe that environmental conditions may be responsible for the disunity in the risk behavior studies (Mone, McKinley & Barker, 1998; Shinkle, 2012; Schimmer & Brauer, 2012). Whereas many studies argue that organizations have to change and, are motivated to change because of the changing environment (Sadler & Barry, 1970; Morgan, 2006; Kacperczyk, Beckman & Moliterno, 2015), and to remain viable as a firm (Duncan, 1972), researches have overlooked the moderating impact of environment conditions on the relationship between performance below aspiration level and risk behavior. While decision makers in search for better practices and strategies, scan the business environment in close proximity to firms’ current positioning (Cyert & March, 1963; Kacperczyk et al., 2015).
This study attempts to clarify the effect of performance on risk behavior by identifying environmental boundary conditions under which performance below the aspiration level leads to risk seeking or risk averse behavior. It is found that the environmental condition dynamism increases information uncertainty (Dess & Beard, 1984), this makes it much more difficult and stressful for managers to interpret firm performance (Schimmer & Brauer, 2012). Additionally, avoiding failure is more crucial in dynamic environments and therefore will worsen managers’ ability to look for opportunities to increase performance. (Brown & Eisenhardt, 1997; Eisenhardt & Bourgeois, 1987, 1988a; Audia & Greve, 2006). Munificence, which is another environmental condition, provides slack resources that support sustained growth and possibilities for diversification (Starbuck, 1973; Dess & Beard, 1984). Also, in munificent environments there is stability, which makes it easier for firms to predict and respond to changes in the environment. These characteristics of munificence may instigate that decision makers look more at opportunities when they are underperforming. Based on the above arguments, I suppose that environmental conditions, dynamism and munificence, impact the decision makers’ interpretation of aspiration levels. As a consequence, firms will act differently under dissimilar environmental conditions and perform different risk behavior. So, studying the moderating effect of environmental conditions may yield new insights in the debate about the relationship between performance feedback below aspirations and risk behavior.
Furthermore, previous research on performance aspirations assumed that historical and social aspiration work in parallel and influences strategic behavior in a similar manner. While the two aspiration levels emerge from different sources of performance feedback. So it may be logical that decision makers interpret the two aspiration levels in another way, which leads to differences in risk behavior. Previous research shows that firms’ risk behavior vary significantly depending on the aspiration used (Kim, Finkelstein & Haleblian, 2015). However this research does not study under what conditions managers are more likely to attend to one or the other aspiration. In response to this
gap, researching the preference for an aspiration under environmental conditions may clarify the effect of performance below aspiration on risk behavior even more.
This study makes important contributions to existing literature. First, by considering the moderators of the aspiration-strategic behavior relationship (Shinkle, 2012). This extends the behavioral theory of firms since researching the moderating effect may generate new insights for the debate in the literature about the relationship between performance below aspiration levels and risk behavior. Also, it will extend the research of Schimmer and Brauer (2012) who focused on the moderating influence of industry dynamism and industry munificence on the relationship between performance and strategic divergence-convergence, where this study will focus on the risk behavior by R&D investments of the firm. At last, this study also enhances the research on differential effects of historical and social aspirations by Kim et al. (2015), by examining under what environmental conditions decision makers are more likely to attend to one or the other aspiration level.
THEORY AND HYPOTHESES
This chapter discusses the main insights of existing literature on instances of firm performance below aspiration level, risk behavior and environmental conditions; dynamism and munificence. First, I explain the behavioral perspective of performance to the aspiration level and how this influences risk-behavior and more specifically R&D investments. Then, I address the literature gap in this debate and present the research question of this study. Third, I describe environmental conditions and present hypotheses for dynamism and munificence. Finally, I argue on which aspiration level the low performing firms under different environmental conditions rely.
The effect of Performance below Aspiration Level on Risk-behavior and the Debate
The central argument of the behavioral perspective on organizational learning is that decision makers use an aspiration level to evaluate performance, and that aspirations level influence the decision maker to take risk and make changes (Cyert & March, 1963; Kahneman & Tversky, 1979; March & Shapira, 1987; Shapira, 1986; Audia & Greve, 2006; Shinkle, 2012; Schimmer & Brauer, 2012). So, performance relative to aspirations is widely acknowledged to influence the risk behavior as well as the strategies undertaken. Research concludes that decision makers learn from performance feedback, which is based on aspiration levels (Levinthal & March, 1981; Mezias & Glynn, 1993; Shinkle, 2012). This research is based on the behavioral theory of the firm, which argues that decision makers are bounded rational (Cyert & March, 1963; Shinkle, 2012). Rationality means that people base their decision on all information available and judge all possible courses of action. However, this is considered impossible and therefore it is argued that humans are bounded rational. This means that an individual’s rationality is limited by manageability of the decision, the cognitive limitations of the mind, and the time available to make the decision (Simon, 1972; Eisenhardt, 1989). The consequence is that individuals look for the course of action that is
satisfactory and leave out the aspects that are not relevant. Given that humans are bounded rational, they will concentrate on relevant information in order to make a satisfactory decision. Based on this information, decision makers of organization, which are bounded rationally, focus on the boundary of the firm between success and failure through the use of aspiration levels to reflect organizational goals and serve as a benchmark to make their decisions.
Aspiration levels are desired performance levels in specific organizational outcomes and are also called reference points to set firm goals (Cyert & March, 1963; Greve, 1998; Shinkle, 2012). Organizational aspirations are central to strategic decision making, by setting goals and objectives for the firm. This is based on psychological processes of risk perception and preference (Kahneman & Tversky, 1979; Audia & Greve, 2006) and organizational processes of search (Cyert & March, 1963). When managers perceive performance below their aspirations, they engage in problemistic search to increase their performance. Problemistic search is search stimulated by a problem and is directed toward finding a solution to that problem (Greve, 2003). These solutions could be found in further developing their products and processes. To pursue these innovations firms can invest in research and development (R&D). R&D Investments are expenditures of the firm based on development of new product lines or increased support for existing projects. The outcome of R&D expenditures is uncertain, and thus a risky investment for firms (Greve, 2003; Vissa et al., 2010; Lim & McCann, 2014). Therefore, risk behavior of firms can be expressed by the investments in research and development (R&D), or R&D intensity (Hitt, Hoskisson, Ireland & Harrison, 1991; Hitt, Hoskisson, Johnson & Moesel, 1996; Lim & McCann, 2014). Following this prior research, I use R&D intensity as the proxy for firms’ risk behavior. I suppose that the higher firms investments in R&D the more risk seeking this firm behaves.
The central argument of behavioral theory is that firms establish an aspiration level based on two antecedents: social comparison and historical (self) comparison (Cyert & March, 1963). Previous research on performance aspirations assumes that historical and social aspirations work in
parallel and influence risk behavior in a similar manner. However, recent research shows that firms’ risk behavior varies significantly depending on the aspiration levels used (Kim et al., 2015). This may assume that firms prefer one aspiration to another, or that one aspiration has a stronger relationship with risk behavior. Kim et al. (2015) shows that firms’ acquisition behavior varies significantly depending on whether historical or social aspirations are used.
Concluding, the behavioral theory of the firm suggests that decision makers use aspirations to evaluate performance, and their interpretation of performance triggers search and risk behavior. Aspirations are desired performance levels based on two antecedents, and consequently decision makers evaluate their performance and take actions based on how they have performed relative to their aspirations. Differences in risk behavior can vary between the reliance on the social aspirations or the historical aspirations. When performance is low, managers engage in problematistic search in order to find solutions for their problems and so to increase their firm performance. In order to find these solutions, firms invest in R&D. These investments are considered uncertain since the return on investments is unclear and unpredictable. Due to this uncertainty, a R&D investment is considered, as a risky decision. As a result, risk behavior is expressed in expenditures in R&D.
The behavioral theory of the firm suggests that when performance is above the aspiration level of the firm, risk-taking actions, such as R&D expenditures, decrease (Miller & Bromiley, 1990; Wiseman & Bromiley, 1996). When performance exceeds aspirations, managers are expected to maintain the status quo (Bromiley, et al., 2001), avoid actions that might negatively influence performance (March & Shapira, 1987), and strive for even higher performance (Cyert & March, 1963; Audia et al., 2000; Bromiley et al., 2001; Audia & Greve, 2006; Schimmer & Brauer, 2012; Shinkle, 2012). In short, when performance exceeds aspiration level, researchers predict that the decision maker is risk averse and decreases its R&D expenditures.
Although, risk aversion is widely accepted when performance exceeds the aspiration level, research is in active debate about the relationship between performance feedback and risk behavior
when performance falls below the aspiration level (Lopes 1987; March & Shapira 1987; Ocasio 1995; Sitkin & Pablo 1992; Audia & Greve, 2006; Shinkle, 2012; Schimmer & Brauer, 2012). Previous research illustrates two opposing arguments. On one hand, research in traditional behavioral decision theory and prospect theory argue that organizations with performance levels below aspiration level increase risk taking and change (Cyert & March, 1963; Kahneman & Tversky, 1979; Bromiley, 1991; Gooding et al., 1996; Wiseman & Bromiley, 1996; Greve, 1998; Palmer & Wiseman, 1999; Miller & Chen, 2004; Chen & Miller, 2007). When performance problems increase, managers are found to engage in more extensive and more complex search processes, to find (strategic) solutions, such new innovations, to increase performance (Cyert & March, 1963; Levinthal & March, 1981). Searching for new innovations initiate R&D expenditures, which is an action that is argued as risk seeking behavior. Concluding, when performance is below aspiration level organizations show risk seeking behavior.
On the other hand, research on organizational decline suggests that performance below aspirations leads to risk averse behavior. Firms who decrease or eliminate their risk taking become risk averse and rigid (Staw et al., 1981; Lopes, 1987; Miller & Bromiley, 1990; Sitkin & Pablo, 1992; Wiseman & Bromiley, 1996). Decision makers interpret performance not as a repairable gap, as the prospect and the behavioral theorists assume, but rather as a threat to their vital interest (Milliken & Lant, 1991; Sitkin & Pablo, 1992, Ocassio, 1995, Mone et al, 1998; Audia & Greve, 2006). Experiencing threat leads to psychological stress and anxiety. This restricts information processing and reduces behavior flexibility (Audia & Greve, 2006). Scholars studying organizational decline, argue that firms who experience declining performance reduce or eliminate their riskier activities such as R&D investments (Wiseman & Bromiley, 1996). Probably, because underperforming firms are anxious to take risk, which can lead them closer to bankruptcy. Decision makers have a strong need for security and are motivated to avoid bad outcomes and additional losses (Audia & Greve, 2006).
To summarize, it is generally accepted that performance aspiration levels influence risk behavior. Additionally, it is widely accepted that performance above aspiration levels leads to risk averse firm behavior. However, when performance is below aspiration level, previous research illustrates an active debate. Researchers suggest that the opposing arguments regarding risk behavior in low performing firms may be explained by unobserved heterogeneity or organizational boundary conditions. Some argue that the moderators of this relationship are understudied (Greve, 2003; Shimizu, 2007; Short & Palmer, 2003; Vissa et al., 2010, Shinkle, 2012). Firms under different boundary conditions possibly will interpret performance feedback in another way, and this can lead to differences in risk taking behavior (Greve, 2003; Audia & Greve, 2006; Shimizu, 2007; Short & Palmer, 2003; Vissa et al., 2010, Shinkle, 2012). Furthermore, experiencing low performance can threaten normal functioning of a firm and even its survival (Audia & Greve, 2016). So, boundary conditions may explain when risk aversion or risk seeking prevails (Audia & Greve, 2006; Shinkle, 2012). It is also given that risk behavior is influenced by numerous internal and external determinants of the firm, such as danger, slack, the neighbourhood of an aspiration level, assimilation of resources and self-confidence (Sitkin & Pablo, 1992; March & Shapira, 1992; Shinkle, 2012).
Previous studies in the relationship between performance feedback and risk behavior, which took boundary conditions in consideration, found some interesting results regarding the debate. These findings extend the knowledge about the relationship between performance feedback. Especially it extends the knowledge about the relationship between performance below aspirations and risk behavior. Furthermore, these findings extended the knowledge of the boundary effects of the relationship and gained arguments for the debate in the literature. For example, Greve (2003) found that when a firm has high slack resources, R&D intensity increased, thus implies risk-seeking behavior. Audia and Greve (2006) argue that firm resources affect decision makers’ risk tolerance; so, they study the moderating effect of firm size. They find that performance below the aspiration
level reduces risk taking in small firms, but either does not affect risk taking, or increase risk taking in large firms. Desai (2008) also elaborates on the debate of the effect of performance below aspiration level on risk behavior and its boundary conditions. He argues that the relationship between performance feedback and risk behavior is dependent on its organizational context. The study finds that organizations with limited operating experience, poor legitimacy, and structural inertia are less able to support risk taking behavior when experiencing shortfalls. However, these studies focus on conditions of the firm itself and ignored or overlooked the environmental conditions of the firms, while environmental conditions may be responsible for the disunity in the risk behavior studies (Mone et al., 1998; Shinkle, 2012; Schimmer & Brauer, 2012). Firms may interpret performance differently when experiencing distinct environmental conditions, and this leads to differences in risk behavior.
Environmental conditions are factors outside the boundaries of the firm that affect the firm and/or the firm has to deal with. Environmental conditions are of great influence on organizational behavior and matter to the relationship, as it is found that in periods of economic growth, stability, and decline, firms change within their strategic groups (Mascarenhas, 1989). This extends the arguments that organizations have to change with the environment in order to survive (Sadler & Barry, 1970; Morgan, 2006) and remain viable (Duncan, 1972). Moreover, decision makers in search for better practices and strategies scan the business environment in close proximity, to find the firms current position (Cyert & March, 1963; Kacperczyk et al., 2015). Therefore, this study attempts to clarify the effect of performance below aspiration level on risk behavior by identifying environmental conditions. Environmental conditions are characteristics of the environment, such as dynamism and munificence. A dynamic environment is an environment where there is rapid and discontinuous change in demand, competitors, technology, or regulation, so that information is often inaccurate, unavailable or obsolete (Eisenhardt & Bourgeois, 1998a;b). Furthermore, dynamism makes changes among environmental elements unpredictable (Child, 1972; Starbuck, 1973;
Randolph & Dess, 1984; Castrogiovanni, 2002). This makes it much more difficult and stressful for managers to interpret firm performance and thereby makes it harder to find opportunities to increase performance. Additionally, avoiding failure is more crucial in dynamic environments and therefore will worsen managers’ ability to look for opportunities to increase performance (Brown & Eisenhardt, 1997; Eisenhardt & Bourgeois, 1987, 1988a; Audia & Greve, 2006). A traditional way to avoid failure and cope with uncertainty is to simply wait to see how the events unfold or to imitate others; this results in homogenous firms (March & Simon, 1958; Duncan, 1972; Dess & Beard, 1984; Eisenhardt & Bourgeois, 1987). Environmental dynamism destroys existing competencies and capabilities within industries and organizations (Tushman & Anderson, 1986; Leonard-Barton, 1992). Threatened by the chance of additional losses that can endanger the survival of the firm, thus bring the firm closer to bankruptcy, decision makers become risk averse. Based on this information, I expect that environmental dynamism influences managers’ perception on low performance as a step closer to firm failure. Consequently, low performing firms in a dynamic environment will show dissimilar organizational risk behavior than low performing firms in less dynamic environments.
A munificent environment is more stable, which provides slack resources that support sustained growth and possibilities for diversification (Starbuck, 1973; Dess & Beard, 1984). Firms operating in these environments may be less concerned about the risk of incurring additional losses, because of the opportunities given by the environment to grow and diversify. Munificence has been regarded as a source of stability, and a lack of munificence as a source of uncertainty and unpredictability (Dess & Beard, 1984). So, managers in more munificent environments can predict the changes in the environment better, which makes it easier for firms to predict and respond to the changes in the environment. Also, firms that operate in munificent environments are buffered from external threats (Finkelstein, Hambrick & Cannella, 2009), and thus have more space available to respond to low performing aspirations than firms that operate in less munificent environments. This
may instigate that decision makers in munificent environments look more at opportunities when they are underperforming, than decision makers in less munificent environments. So, the moderating effect of munificence likely results in distinctive observations of risk behavior based on low performance.
Additionally, research on performance aspirations assumes that historical and social aspirations work in parallel and influence the strategic behavior in a similar manner. While the two aspirations emerge from different sources of performance feedback. The historical aspiration is based on the firms’ own performance. The social aspiration is based on the performance of a reference group of firms (e.g., Cyert & March, 1963; Baum, Rowley, Shipilov & Chuang, 2005; Harris & Bromiley, 2007). Recent research finds that firms’ risk behavior varies significantly depending on the aspiration level used (Kim et al. 2015). It is argued that social aspirations motivate risk and change within the firm (Schimmer & Brauer, 2012). However, prior studies who tried to reconcile the debate in the literature study the boundary conditions of the firm (e.g. Audia & Greve, 2006; Desai, 2008), but overlook the possibility that firms’ risk behavior may vary depending on whether historical or social aspirations used. Considering that historical and social aspirations emerge from different sources, they may give rise to divergent risk behavior (Kim et al., 2015) and thus leads to differences in R&D investments. Due to bounded rationality managers may prefer one aspiration level to another in certain environmental conditions. Researching both historical and social aspiration individually under different environmental conditions may clarify the effect of performance below aspiration on risk behavior even more.
To conclude, studying the influence of environmental conditions will extend our knowledge about the behavior theory of the firm and may provide additional information to reconcile the debate of performance below aspiration level leading to risk seeking or risk averse behavior. In view of the fact that firms’ environmental conditions are of great influence to interpreting information and the firms’ behavior it can be suggested that environmental conditions are accountable for the
differences in the relation between performance feedback and risk behavior. Therefore, it is interesting to examine the moderating effect of the environmental conditions dynamism and munificence on the relationship between performance and risk behavior. With this, the study further responds to the call for greater consideration of moderators of the aspiration-strategic behavior relationship (Shinkle, 2012). Schimmer and Brauer (2012) already extended the study of Shinkle (2012) by focussing on the moderating influence of environmental conditions on the relationship between firm performance and strategic divergence-convergence. They find that dynamism makes it more difficult for managers to interpret past and present performance and narrows the managerial search scope. Munificence seems to enforce differentiation from the strategic groups. However, they only focus on financial performance behavior and the movements away or towards a firm’s current strategic group and not specific analyze the risk behavior of the firm. Additionally, I research if firms prefer one aspiration level to the other, to make R&D investments when confronted with environmental conditions to attend to the signals emanating from historical or social aspirations this study responds to the calls from Kim et al. (2015). They find that firms’ acquisition behavior vary significantly depending on whether historical or social aspirations are used. However, they do not include different conditions, which may impact the preference for one or other aspiration level. It may appear that decision makers experiencing environmental conditions may prefer one aspiration to another. To examine this relationship I aim to answer the following research question:
What is the moderating effect of environmental conditions on the relationship between performance below aspiration level and risk behavior, in terms of research and development (R&D) investments,
Environmental Conditions
Firms are affected by numerous uncertain and complex environmental factors outside the boundaries of the firm, called environmental conditions. Firms attempting to ignore environmental conditions, or refusing to respond to such conditions create trouble for themselves and diminish their competitive disadvantage. A firm has to adapt with the environment in order to survive and remain viable (Morgan, 2006; Duncan, 1972). A firm must always bow to the relationship with the environment since: ‘An organisation cannot evolve or develop in ways which merely reflect the goals, motives or needs of its members or of its leadership, since it must always bow to the constraints imposed on it by the nature of its relationship with the environment’ – Sadler and Barry (1970:58).
Moreover, it is found that different environmental conditions and different types of relationships with outside parties will, require different types of organizational structural accommodation to achieve a high level of performance (Child, 1972). To cope with environmental conditions, firms must make decisions. These organizational decisions are dependent on the ability to forecast the environment. Decision makers base their decision-making on their forecast, which also is implied in the behavior theory of the firm (Cyert & March, 1963).
Environmental conditions are presented in three dimensions; Munificence, Dynamism and Complexity (Aldrich, 1979 in Dess & Beard, 1984). First, environmental dynamism means that firms experience change that is difficult to predict; this increases uncertainty for the organizational members. Second, environmental munificence means the ability to which the environment can support sustained growth (Starbuck, 1973). Firms try to retain munificence in order to ensure a flow of resources that can provide slack, which provide possibilities for growth or diversification. Third, Child (1972) describes environmental complexity as “the heterogeneity of and range of an organization’s activities.” In other words, environmental complexity is the complexity of the activities within an organization. However, I will focus solely on dynamism and munificence. I
consider complexity more as an internal condition of the firm, since it explains the conditions (the activities) within the organization and not the conditions outside of the organization. Moreover, complexity is a condition chosen by the firm before, and may be part of the firm's strategy. In contrast, dynamism and munificence are conditions where the organization can or should react to and deal with.
Before, I indicated that the relationship between performance feedback and firms risk behavior is based on firm performance aspiration levels. In the following paragraphs, I will discuss how the environmental conditions dynamism and munificence, moderate this relationship. Additionally, I will discuss on which performance aspiration level the firms under environmental conditions mainly rely.
The Moderating Effect of Environmental Dynamism
I define environmental dynamism as an environment where there is rapid and discontinuous change in demand, competitors, technology, or regulation. That information is often inaccurate, unavailable or obsolete (Eisenhardt & Bourgeois 1998a;b) and changes among environmental elements are unpredictable (Child, 1972; Starbuck, 1973; Randolph & Dess, 1984; Castrogiovanni, 2002).
Cyert and March (1963) propose that low performance widens decision makers’ search scopes to seek more innovative solutions, which initiate R&D investments, to increase performance. Thus argue that performance below aspiration is positively related with risk seeking behavior. However, when the organizational environment is dynamic, decision makers experience more uncertainty and instability than in a stable environment (Duncan, 1972; Dess & Beard, 1984; Schimmer & Brauer, 2012). Uncertainty and instability leads to more and complex information, making it more difficult to be rational, process information, learn from past and present firm performance, and predict patterns and irregularities (Galbraith, 1974; Eisenhardt & Bourgeois, 1988a; 1988b; Eisenhardt, 1989; Gort, 1969; Simon, 1972; Duncan, 1972; Hrebiniak & Joyce,
1985; Schimmer & Brauer, 2012). Moreover, information uncertainty makes decision-making more complex and stressful (Tushman, 1979; Waldman et al., 2001). Consequently, when an environment is more dynamic, it becomes more difficult to seek for innovative solutions to increase firm performance.
Information is important for decision makers to identify and forecast patterns in order to support their possible choices. Decision makers in dynamic environments need to process more information in a dynamic environment than in a static environment (Duncan, 1972). Consequently, decision makers in dynamic environments are unable to engage in structured analysis and make a formal strategic planning (Eisenhardt & Bourgeois, 1988a). Dynamism not only makes it harder to engage a structured analysis, but also decreases the firms’ ability to anticipate with certainty on changes in the environment or respond to their competitors (Williamson, 1975; Lepak, Takeuchi & Snell, 2003). So, a dynamic environment, due to feelings of uncertainty and instability, makes it more difficult, complex and stressful for decision makers to perceive and react on performance aspirations. Environmental dynamism worsens the ability of decision makers in low performing firms to process information and have behavior flexibility. As a result, when environmental dynamism increases, it will be more difficult, complex and stressful for decision makers to make decisions about R&D expenditures. Moreover, because R&D expenditures are large and uncertain expenditures where the payoff is mostly only noticeable after a long time (Lim & McCann, 2014).
Dynamism not only increases uncertainty and information complexity, but also the anxiety to make changes and risky choices (Dess & Beard, 1984; Duncan, 1972; Schimmer & Brauer, 2012). Brown and Eisenhardt (1997) suggest that failures of changes are worse in highly dynamic environments. For that reason it is crucial to avoid mistakes in a more dynamic environment (Eisenhardt & Bourgeois, 1987, 1988a). The anxiety increases because decision makers have a strong need for security and are motivated to avoid bad outcomes. Anxiety restricts information
processing and reduces behavioral flexibility, which makes decision makers rigid and risk averse (Audia & Greve, 2006).
A traditional way to avoid failure and cope with uncertainty is to simply wait to see how the events unfold or to imitate others. Although in a dynamic environment it is hard to imitate others because competitors are changing rapidly (Eisenhardt & Bourgeois, 1988a, 1988b) and the environmental changes and discontinuities destroy competencies and capabilities (Tushman & Anderson, 1986; Leonard-Barton, 1992). Additionally, R&D investments are considered as an uncertain decision, since the return of the investment is hard to predict. As a result, it is difficult for firms to cope with uncertainty in the traditional way. This will lead to less investment in R&D, because these are uncertain decisions. As a result firms become homogenous (March & Simon, 1958; Duncan, 1972; Dess & Beard, 1984; Eisenhardt & Bourgeois, 1987).
Together, these effects suggest that environmental dynamism affects decision makers’ ability to search for better practices, find support for their decisions, and therefore their confidence to make risky decisions, such as R&D investments. So when environmental dynamism increases, decision makers are less likely to invest in R&D. Moreover, environmental dynamism may worsen the positive relationship between performance below aspiration and risk behavior, stated by the behavioral theory and prospect theory.
Concluding, environmental dynamism, due to uncertainty, makes it more difficult, complex and stressful for managers to interpret information and reduces behavior flexibility. Also, it will increase anxiety to take risk, since avoiding failure is more crucial. For this reason dynamism will have a negative influence on the statement of the behavioral and prospect theory. This indicates that firms threatened by low performance, and information uncertainty, firms will focus their attention toward avoiding failure by lessen their R&D investments, and as a result reduce their risk seeking behavior. When environmental dynamism increases, firms will be less likely to make R&D investments when having performance below aspirations. So, the environmental condition
dynamism will lessen the risk seeking behavior when firms are underperforming; therefore I state the following hypothesis:
Hypothesis 1: When performance is below the aspiration level, performance decreases lead to less risk taking among firms who operate in a dynamic environment than firms who operate in a less dynamic environment.
The Moderating Effect of Munificence
Munificence means the ability to which the environment can support firms with sustained growth and diversification (Starbuck, 1973; Dess & Beard, 1984). Firms that operate in munificent contexts are buffered from external threats, and thus, have more space available to accumulate slack resources (Finkelstein et al., 2009). Additionally, munificence reduces resource constraints and lowers rivalry-based mobility barriers, because market leaders less forcefully defend their market segments against competition (Miller & Chen, 2004; Schimmer & Brauer, 2012). This will increase the resource availability that gives firms more opportunities for improving their performance.
Managers of firms that perform below aspirations have broader search scopes to seek for more innovative solutions or diversification, because they feel the need to improve firm performance (Cyert & March, 1963; Bromiley, 1991; Shinkle, 2012). Munificence may enhance risk seeking when firms are underperforming. Both the behavioral theory of the firm and the prospect theory, as part of behavioral decision making theory, predict increased risk seeking behavior in the form of selection and adoption of innovative solutions and practices (Schimmer & Braurer, 2012). To develop these innovative solutions, firms are investing in R&D. Munificent environments will enhance these investments due to resource availability and lower rivalry-mobility barriers. First, resource availability gives R&D departments more possibilities to seek for innovative solutions or imitate innovative solutions of other firms. When there is great resource availability,
stakeholders are likely to exercise lenient supervision, as there is less dependency on resources. This facilitates risk-seeking managers to further invest in R&D investments to search for innovations (Schimmer & Brauer, 2012). Second, lower mobility barriers make diversification of the firm less costly and risky. Therefore underperforming firms hold an advantage under this condition. Munificence will enhance the argument of the behavioral theory that managers perceive performance decreases as a repairable gap. Not only because of the resource availability and the lower rivalry-mobility, but also because firms in munificent environments are better buffered from external threats (Finkelstein et al., 2009). Consequently, firms in high munificent environments have greater opportunities for profitable firm growth (Schimmer & Brauer, 2012) than firms in less munificent environments.
Aside from the fact that a more munificent environment will have greater resource availability, lower rivalry-based mobility barriers and more possibilities for growth and diversification, munificence also provides more certainty and predictability in the firm's environment (Dess & Beard, 1984; Georgakakis & Ruigrok, 2017). This will make it easier for decision makers to identify and forecast patterns in performance feedback in order to support their possible choices of R&D investments, than when there is low munificence. Strategists would be able to engage a structured analysis and make a formal strategic planning (Eisenhardt & Bourgeois, 1988a). A structured analysis and formal strategic planning will reduce the uncertainty of the R&D investment, and as a result decision makers will be more likely to invest in R&D.
I suspect that munificence strengthens the risk seeking behavior of managers in low performing firms. Managers of low performing firms in a more munificent context may look extra at the opportunities and are more prone to make investments in R&D than in less munificent environments. As there is increased resource availability in munificent environments, which stimulates risk seeking and lower mobility barriers, making diversification less costly. Moreover a munificent environment provides more certainty and predictability, which makes it easier to find
support for their R&D investments. As a result, munificence may increase risk the seeking behavior, thus investments in R&D, when firms are underperforming. Therefore I state the following hypothesis:
Hypothesis 2: When performance is below the aspiration level, performance decreases lead to more risk taking among firms who operate in a munificent environment than firms who operate in a less munificent environment.
Aspiration Level Preferences under Environmental Conditions
Previous research on performance aspirations assumes that historical and social aspirations work in parallel and influence risk behavior in a similar manner. Recent research shows that firms’ risk behavior varies significantly depending on the aspiration levels used (Kim et al., 2015). Firms risk behavior varies between the use of the social aspiration and historical aspiration to evaluate performance. Kim et al. (2015) posit that historical and social aspiration levels induce different interpretations as these are filtered through dissimilar cognitive and organizational processes. However, this study overlooked under what conditions managers have a preference for one to the other aspiration level when making risky decisions.
Historical and social aspirations give rise to different risk behaviors, because respective underlying benchmarks and processes are distinct and interpreted differently by managers (Kim et al., 2015). This can be explained by the fact that individuals, and thus managers, are bounded rational. As described before, bounded rationality means that managers cannot use all information available in performance evaluation. As result, managers simply process the aspiration that reflect organizational goals and serve as a benchmark (Cyert & March, 1963 in Kim et al., 2015).
Kim et al. (2015) found that firms’ acquisition behavior varies significantly depending on whether historical or social aspirations are used. Additionally, The study of Schimmer and Brauer
(2012) found that social aspirations motivate risk and change within firms. Considering that risk behavior may be different based on the aspirations, it is reasonable to assume that firms who operate in a dynamic or munificent environment are likely to attend more to a specific aspiration level. Managers possibly, due to bounded rationality, interpret the aspirations that can serve as a benchmark to find a solution for the low performance. It may be that decision makers who operate in a dynamic or munificent environment rely more on social aspiration levels. Given that decision makers in search for better practices scan the business environment in close proximity to firms’ current positioning (Cyert & March, 1963; Kacperczyk et al., 2015). It possibly will be that firms prefer one aspiration level to another, or that one aspiration level shows a stronger relationship with R&D intensity of the firm. Firms frequently set their own performance goals by observing the performance of a reference group (Fiegenbaum & Thomas, 1995). Social aspirations help decision makers how they should perform when the firm operates in a high dynamic or munificent environment. Historical aspirations may be less important to decision makers than recent results of their competitors.
As indicated before, in a dynamic environment there is discontinuous change in demand, competitors, technology, or regulation (Eisenhardt & Bourgeois 1998a;b). Additionally, dynamism can destroy firms’ capabilities and competencies (Tushman & Anderson, 1986; Leonard-Barton, 1992). The social aspiration is a useful performance benchmark to learn how others perform, but also how they achieved the observed performance (Kim et al., 2015). In order to keep notice of the dynamic environment and it competition, it is expected that firms in high dynamic environments rely more on social aspirations than on historical aspirations when making decisions.
In contrast with high dynamism, when there is low dynamism, the environment is more stable. In a stable environment, firms shift their focus from product market differentiation to production efficiency (Miles, Snow & Sharfman, 1993). Firms will make decisions based on their past performance in order to increase efficiency, and due to bounded rationality they will rely more
on historical aspirations than on social aspirations. Considering this, it is expected that underperforming firms shift their focus from historical aspirations to social aspirations as dynamism moves from low to high. I state the following hypothesis:
Hypothesis 3a: When environmental dynamism increases, from low dynamism to high dynamism, firms who perform below their aspiration level change their focus from historical aspiration to social aspiration.
A high munificent environment might stimulate firms to look more at their opportunities within the environment when they are underperforming. A munificent environment can support sustained growth or diversification from competitors (Starbuck, 1973; Dess & Beard, 1984). In order to notice these opportunities firms must analyze the current performance of their competitors. Social aspirations help decision makers assess how they should perform in order to increase their performance. Additionally, firms taking on similar strategic actions become useful reference points for managers to evaluate the effectiveness of their own strategies (Fiegenbaum, Hart & Schendel, 1996; Kim et al., 2015). Thus it is expected that firms in high munificent environments rely their decisions more on their social aspirations than firms in less munificent environments. When there is less munificence, there is less resource availability and higher mobility barriers (Schimmer & Brauer, 2012). This makes it more difficult for firms to differentiate from other firms and managers will shift their focus on operational efficiencies. To increase their performance, firms will focus on historical aspirations to examine the areas of improvements and make their decisions. Therefore, I hypothesize:
Hypothesis 3b: When environmental munificence increases, from low munificence to high munificence, firms who perform below their aspiration level change their focus from historical aspiration to social aspiration.
METHOD
This chapter explains the research approach of my thesis. First, the sampling strategy will be discussed. Then, I will explain the dependent, independent, moderating, and control variables. The last section shows details the data.
Sampling Strategy
This study uses secondary panel data gathered from Wharton CRSP – COMPUSTAT Merged database, which collects operational and financial data for all publicly traded U.S. Companies. From this collection I choose to focus on manufacturing, as this can be compared with prior research (e.g., Chen & Miller, 2007; Chen, 2008; Lim & McCann, 2014) and maintains comparable industry demographics. The dataset consists of manufacturing firms with Standard Industrial Classification (SIC) codes ranging from 2000 to 3999. This will prevent having confounding results due to major differences in industry activities and will allow comparisons of the results with prior research. Including different industries within the manufacturing industry could show weak results, because it is possible that more innovative or technological industries have greater correlation with R&D intensity than less innovative or technological manufacturing industries. This could offset any association between performance feedback and risk behavior (Chen & Miller, 2007). For this reason, my sample only consists of the most represented industry of the dataset (22.5 %), SIC 36, which is the Electronic & other Electrical Equipment & Components except Computer Equipment industry, from now on called Electronic Hardware Industry. Additionally, I want to exclude small industries within the Electronic Hardware Industry, with less than five firms to avoid possible biases in our estimates of industry average R&D intensity. However, there are no small industries within Electronic Hardware Industry, the to exclude. To examine the risk behavior, by R&D intensity, over time I use data from 1979 to 2014.
Dependent variable
The debate in the literature is focuses on firms being risk averse or risk seeking when performance falls below aspirations. Consistent with a large number of prior studies, I use R&D intensity as the proxy for Risk behavior. R&D intensity is computed by firms R&D expenditures in millions of U.S dollars divided by sales, (Cohen & Levinthal, 1989; Chen & Miller, 2007). Following the behavioral theory of the firm (Cyert & March, 1963), I focus on firms with R&D intensity less than or equal to 1.0 (Chen & Miller, 2007; Lim & McCann, 2014). I reason, that firms with R&D expenses greater than sales are start-ups or firms with enormous investments from their parent company in order to find or change their strategy. My theoretical arguments applied to firms that engage in ongoing production and sales activities. The resulting sample has mean R&D intensity at 0.11 and a standard deviation of 0.13, which is similar to Chen & Miller’s (2007) and Lim & McCann (2014) results.
Independent variable
The independent variable Performance below aspiration level is measured using one traditional accounting measure of returns; return on assets (ROA). ROA is the main accounting-based proxy for firm profitability within the manufacturing industry and is consistent with prior studies (e.g., Bromiley, 1991; Greve, 2003; Lepak et al., 2003; Miller & Chen, 2004; Audia & Greve, 2006; Chen & Miller 2007; Desai, 2008; Iyer & Miller, 2008; Schimmer & Brauer, 2012; Lim & McCann, 2014). Given that performance measures are evaluated against aspiration levels based on historical and social aspiration levels (Cyert & March, 1963) and I want to investigate on which aspiration level the risk behavior relies the most (Kim et al., 2015), I include two alternate measures for aspiration levels. One related to the firm’s own prior performance, the historical aspiration level, and the other related to the competitors in the same industry, the social aspiration level. Historical aspiration levels are determined by the recent history of performance of the organization, and social
aspiration levels are determined by the performance of peers. The historical aspiration level is generated by taking the average ROA of the firm’s past three years, which are the past values of the performance variable (Greve, 1998; Levinthal & March 1981; Audia & Greve, 2006). When the data is not available for three years, I use the last two or only the previous year. The social aspiration level is computed by taking the median ROA of all companies within each industry at the four-digit SIC level, which is consistent with Lim and McCann (2014). Actual firm performance was measured at year t-1, and each aspiration level is proxied at t-2. I compute the firm’s performance feedback as the difference between the firm’s actual performance and the aspiration level (social or historical), following recent research of Lim and McCann (2014). To indicate performance below and above aspiration, I adopt the method from Lim and McCann (2014). Firms with performance feedback above 0 are indicated as performance levels above aspirations, and below 0 as performance levels below aspiration (Desai, 2008). I split both the variables historical and social performance feedback in performance below aspiration and performance above aspiration.
Moderating variables
The underlying assumption of this study, as described before, is that the environmental conditions of a firm is the dominant focus or frame of reference for most corporate decisions. This assumption is supported in the literature (Keats & Hitt, 1988; Morgan, 2006). The selection of indicators of the variable environmental dynamism is based on the paper of Dess and Beard (1984) and the data is obtained by using NBER-CES Manufacturing Industry Database. Keats and Hitt (1988) have also used this method.
Environmental Dynamism is measured as regressing time against industry sales for the five
divided by the mean sales value to obtain the value from dynamism (Lepak et al., 2003). This is consistent with prior studies (e.g. Keats & Hitt, 1988; Dean & Snell, 1996).
The selection of the indicators of the variable munificence is based on the paper of Dess and Beard (1984). This method is consistent with the study of Schimmer and Brauer (2012). Munificence is measured by the regression slope coefficient divided by the mean value (Boyd, 1995; Castrogiovanni, 2002).
Control variables
Control variables are entered to rule out alternative explanations for the results. First I control for
firm size. Controlling for size reduces the possibility that experience and age effects are due to
unobserved heterogeneity in each companies scale of operations, this might increase firm risk taking and thus risk behavior (Wright et al., 2007; Desai, 2008; Lim & McCann, 2014). I measure firm
size, consistent with prior literature, as the number of employees (e.g. Hitt et al., 1991; Audia &
Greve, 2006; Lim & McCann, 2014).
Second, I also control for firm slack, because slack levels may influence firm risk behavior (e.g., Singh, 1986; Lim & McCann, 2014). Consistent with prior literature working capital-to-sales ratio and current ratio were chosen as proxies for firm slack (Bourgeois, 1981; Singh, 1986; Chen & Miller 2007; Lim & McCann, 2014).
Third, I control for distance from bankruptcy, as this may have a negative influence on risk behavior (March & Shapira, 1987). I have measured the distance from bankruptcy, consistent with Lim & McCann (2012; p. 270), with Altman’s (1983) Z-score computed based on Chen and Miller’s (2007) formula: “(1.2 x working capital divided by total assets) + (1.4 retained earnings divided by total assets) + (3.3 x income before interest expense and taxes divided by total assets) + (0.6 x market value of equity divided by total liability) + (1.0 x sales divided by total assets).” Lastly, I control for industry R&D growth, because industry prospects could influence the
firms’ investments decision. This variable is computed as the percentage change in industry sales from t-1 to t (Chen & Miller, 2007).
Data analysis
For the analysis I use panel data to test the hypotheses. Before I make the analysis, I check the data. I use Breusch-Pagan (BP) tests for all the models to check for heteroskedasticity in the data. All the results are consistent in rejecting the null hypothesis of homoskedasticity (Breusch & Pagan (1979). Therefore, the statistical evidence implies that heteroskedasticy is present and thus indicates that a simple OLS estimation is not appropriate.
Further, I perform Hausman’s specifications test to decide between performing a fixed effects model or a random effects model. The results of the Hausman test are consistent in rejecting the null hypothesis for all models. Consequently, I have used fixed panel regression models and a t-test to t-test the hypotheses. I have use STATA statistical package to run the within (fixed effects) models.
Hypotheses 1 and 2 suggest that different environmental conditions affect the relationship between performance aspiration levels and risk behavior. To test this effect for shifts in risk behavior based on the two aspiration levels. I create different models for social and historical aspirations as performance feedback. First, I test the direct effect of performance feedback on risk behavior. Second to test the moderation affect of the environmental conditions, I use a model that includes the interaction of environmental conditions. To test the interaction effect, I first perform a
t-test to determine if the two aspirations are significantly different from each other when
experiencing environmental conditions. To further analyse the interaction effect I create interaction variables of the dynamism and munificence variables with performance feedback variables. I have mean-centred the variables, before creating the interaction, following Lim et al. (2014).
RESULTS
This chapter shows the results of the research. First, the descriptive statistics are given to provide an overview of the data. Then, multiple hierarchical multiple regressions are performed to test the first two hypotheses formulated in the theoretical framework of this paper. At last, a t-test and interaction figures are performed to test hypotheses 3a and 3b.
Descriptive Statistics and Correlations
Table 1 shows descriptive statistics and a bivariate correlation matrix for all study variables. The sample consists of 11987 observations from 1979 to 2014 in the Electronic Hardware Industry (SIC 36). I could not detect any coefficient instability when variables were added individually and hierarchically (Greenberg & Parks, 1997; Kennedy, 2003; Kim et al, 2015). To further assess the potential model estimation issues that might be introduced by multicollinearity, I calculated variance inflation factors (VIFs). The VIFs for all variables in all models are below 10, a common rule of thumb used to detect multicollinearity problems (Hardin, 1996), and range from 1.14 to 1.16. So, I did not find evidence of multicollinearity in my models.
Overall the control variables show negative relations with R&D intensity, except for Slack. Slack shows, as expected a positive correlation with R. This is consistent with the study of Chen and Miller (2007). The other control variables are not consistent with the studies of Chen and Miller (2007) and Audia and Greve (2006). The control variable Distance to Bankruptcy shows the strongest correlation with R&D intensity of all the variables, and is negatively correlated with Risk Behavior (r=-0.26). Industry R&D growth shows a negative, but weak, correlation (r=-0.08). Firm Size is negatively correlated with Risk Behavior (r=-0.23). It is interesting to see that Firm Size is negatively correlated and Slack is positively correlated, since Audia and Greve (2006) argue that larger firms are risk seeking because they have large stocks of resources, which implies slack.
R&D intensity shows a positive correlation with Dynamism (r=0.21) and a negative correlation with Munificence (r=-0.10). As expected there is a positive correlation between Industry R&D growth and Munificence (r=0.51), since munificent environments provide opportunities for firm growth. In contrast, Industry R&D growth shows negative relationship with Dynamism (r=-0.27). Considering the negative correlation between Munificence and Dynamism (r=-0.45) it is possible that when there is one of the two, the other is reduced.
R&D intensity shows a positive correlation with Performance above historical aspirations (r =0.06). Performance above social aspirations shows a negative correlation (r=-0.16). Furthermore, Performance below historical aspirations shows a negative relation (r=-0.17) with R&D intensity, as well as Performance below social aspirations (r=-0.23). Dynamism is positively correlated with Performance above historical aspirations (r=0.08) and with Performance above social aspiration (r=0.13). Further, Dynamism is negatively correlated with Performance below historical aspirations (r=-0.08) and Performance below social aspirations (r=-0.05). Munificence does not exhibit significant correlations with either performance below historical and social aspirations. However, Munificence shows significant negative correlations with Performance above historical aspirations (r=-0.08), and Performance above social aspirations (r=-0.06). Overall the relationships between the environmental conditions and performance aspirations are very weak. For all correlations please go to Table 1.
Regression Models
Table 2 reports the results of the fixed regression analysis. I use separated models for performance feedback for Historical [models 2, 3 & 4] and Social aspirations [model 5, 6 & 7]. Each feedback shows a model with the direct effect of performance feedback and R&D intensity and two models with the interactions of the environmental conditions. First the interaction with Dynamism and second the interaction with Munificence. For each of the models in Table 2, the incremental
improvement in R2 beyond the baseline model is around 1% to 2%. As the variables in the baseline model are not orthogonal to the variables added in the other models, I cannot unambiguously assign the proportion of variance explained to each variable. For the nonorthogonal variables, the proportion of variance explained depends on the order in which variables are added to the multivariate model (Desai, 2008).
When interpreting the coefficients for performance feedback below aspirations, I consider that a positive coefficient for performance below aspirations means that the further the past performance drops below the aspiration level, the lower R&D intensity. A negative coefficient means that the further the past performance falls below their aspirations, the higher the R&D intensity (Chen & Miller, 2007; Iyer & Miller, 2008; Lim & McCann, 2014).
In table 2, model 1 shows that the control variables Industry R&D growth, Distance to Bankruptcy and Slack have a significant effect on R&D intensity. As expected according to Chen & Miller (2007) Slack shows a positive effect on R&D intensity, which implies that firm’s intensity in R&D increases with the firms’ slack resources. Distance from Bankruptcy (Altman’s Z) is negatively associated with R&D intensity. In other words, firms increase their risk seeking behavior as they approach bankruptcy. The other control variable, Industry R&D growth is negatively associated with R&D intensity. As Industry R&D growth rates rise, firms appear to favor to invest in R&D, so become more risk seeking.
Regarding the effect of performance feedback on R&D intensity, model 2 shows that the coefficient for historical performance below aspiration is negative and significant (β = -0.049, p < .01). Model 5 shows that the coefficient for social performance below aspiration is negative and significant (β =0.067, p < .01). This suggests that performance below aspirations is positively related with R&D intensity. This means that firm increase risk taking behavior as performance falls further below aspiration.
