Performance Feedback and organizational Risk-taking Behavior:
Moderating Effects of Dynamism and Munificence.
Abstract
To survive in the current dynamic and competitive world, adaptation is key. In order to adapt, organizations must learn. One of the ways organizations learn is through performance
feedback. A better understanding of the concept would be helpful for organizations. However, current empirical evidence shows contrary results, regarding organizational risk-taking following performance feedback. In some cases, performance below aspiration level leads to an increase in organizational risk taking, whereas in other cases it leads to a decrease in organizational risk taking. In order to help explain the contradiction in the literature, this study aimed to look at contextual industry factors that influenced the relationship. We conducted quantitative research by using panel data from the COMPUSTAT database. Munificence and dynamism were tested as moderators of the relationship between performance below aspiration level and organizational risk taking. Regression analysis showed that both were negative moderators on the relationship between performance below aspiration level and firm performance. Suggesting that in situations of industry munificence and industry dynamism organizations tend to be more willing to take risk when performance falls below the aspiration level.
Key words: Dynamism; Munificence; Performance feedback; aspiration; organizational risk-taking behavior; behavioral theory of the firm
Student: Dylan Veerman (10252800)
MAc. Business Administration Strategy
University of Amsterdam, Faculty of Economics and Business Supervisor: B. Silveira Barbosa Correia Lima
University of Amsterdam, Amsterdam Business School
Statement of originality
This document is written by student Dylan Veerman, 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.
Table of contents
Abstract ... 1 Statement of originality ... 2 Table of contents ... 3 Introduction ... 4 Theoretical framework ... 9Literature gap and research question ... 11
Moderating effects of dynamism ... 13
Moderating effect of munificence ... 16
Method ... 20
Data and sampling strategy ... 20
Dependent variable ... 20 Independent variable ... 21 Moderating variables ... 21 Control Variables ... 22 Statistical method ... 23 Results ... 25
Descriptive statistics and correlation analysis ... 25
Regression analysis ... 25
Discussion ... 33
Key findings ... 33
Theoretical implications ... 33
Managerial implications ... 37
Limitations and future research ... 37
Conclusion ... 38
References ... 39
Introduction
In the current competitive and dynamic environment, organizational learning, which enables the ability to adapt to change, is key to organizational survival (e.g. Audia & Greve, 2006; Greve, 2003a; Greve, 2003b; Levitt & March, 1988). Organizational learning is achieved under different circumstances. One of the ways organizations learn is through performance feedback (e.g. Audia & Greve, 2006; Fiol & Lyles, 1985; Desai, 2008). The main theory supporting the idea of organizational learning through performance feedback is the behavioral theory of the firm.
According to the behavioral theory of the firm, organizations set certain performance goals based on an aspiration level. The aspiration level is set by looking at own historical firm performance and by looking at the performance of similar firms, which often operate in the same industry (Cyert & March, 1963). Performance feedback theory states that organizations will initiate search processes when performance falls short of the goals or aspirations that management has set for the firm (Cyert and March, 1963; Schimmer & Brauer, 2012). These search processes in some cases will lead to organizational change in order to close the performance gap that has developed. When performance is above the aspiration level, organizations tend to show contrary behavior. They tend to proceed with their current strategy, processes and routines and therefore problemistic search will reduce (Bromiley, 1991; Greve, 2003a).
Further research on performance feedback led to a contradiction in the literature. As researchers studied the effects of performance below aspiration on the risk-taking behavior of organizations, they failed to provide consisting evidence. Some studies found that when performance is below the aspiration level, organizations will take more risks in order to overcome the performance gap (Bromiley, 1991; Greve, 1998; Greve, 2003a; Desai, 2008). This is in line with the behavioral theory of the firm (Cyert & March, 1963). Other studies
found contrary results, suggesting that when performance falls below the aspiration level organizations become more risk averse. This is in line with research on organizational decline, and threat rigidity theory (Staw et al., 1981; Wiseman & Bromiley, 1996; Cameron, 1983). March & Shapira (1992), offered a theoretical framework which helped explain the contradiction in the literature. The theory states that when performance falls below the aspiration level the performance gap can either be seen as a repairable gap or one that is a threat to firm survival (Audia and Greve, 2006; March & Shapira, 1992). When the gap is perceived as a repairable gap, managers will be focused on the aspiration level. Organizations will start looking for a solution to the performance gap and will take more risks to overcome it. They believe the risk could yield them the improvement in performance required to exceed the aspiration level again (Desai, 2008; March & Shapira, 1992; Miller & Chen, 2004; Ocasio 1995). Typical solutions that organizations tend to look for are adopting new routines or strategies, increased spending in research and development and engaging in acquisitions (e.g. Audia & Greve, 2006; Miller & Chen, 2004; Chen & Miller, 2007; Bromiey, 1991). When the gap is perceived as a threat to organizational survival, managers switch their attention from the aspiration level towards the survival level. In order to reduce the probability of
organizational failure, managers will prefer low-risk options or are unable to consider riskier alternatives, which results in organizations to become more risk-averse (Desai, 2008; Audia & Greve, 2006; Staw et al., 1981). Typical behavior of managers that focus on the survival point are centralization of decision-making, allocation of slack resources in order to cover operating expenses, avoiding risky investments and risky activities (Desai, 2008; Audia & Greve, 2006).
In addition to this, several studies have looked for contingency factors that determine whether performance feedback leads to either risk-taking or risk-aversion. The idea behind this is that organizations react to or interpret performance feedback differently when they face
different organizational or environmental conditions. Recent research emphasized the importance of looking into moderating factors if we are to gain a better understanding of the relationship between performance feedback and organizational risk behavior. Existing empirical evidence shows examples of organizational factors that influence the relationship: firm size (Greve, 2011; Audia & Greve, 2006), firm age (Desai, 2008), slack resources (Chattopadhyay et al., 2001; Singh, 1986), distance to bankruptcy (Desai, 2008; March & Shapira, 1992), and stock option grants (Lim & McCann, 2014). Although, organizations environment is one of its major contingencies, studies that use industry contingency factors as moderator on the relationship between performance feedback and organizational risk taking are lacking (Desai, 2008; Audia & Greve, 2006; Shinkle, 2012). The most common way of conceptualizing an organization environment is by using the three underlying dimension as proposed by Aldich (1979). The three contextual industry characteristics are dynamism, munificence and complexity (Aldrich, 1979; Dess & Beard, 1984). Recently, Schimmer and Brauer (2012) made a good contribution to the literature by looking at the moderating effect of industry dynamism and munificence on the relationship between performance feedback and strategic repositioning. Though, in order to gain a better understanding of when
organizations become either risk-taking or risk-averse following performance feedback, more research should be done on external industry factors that could possible moderate the
relationship.
As earlier management research has shown, the environment influences organizations structures, processes, strategies and performance (Goll & Rasheed, 2004). In order to
reconcile the debate, we will look at industry munificence and dynamism as possible
moderators of the relationship between performance below aspiration level and organizational risk taking. We elaborate on Schimmer and Brauer (2012), and examine the moderating effect of dynamism and munificence on the relationship between performance below aspiration
level and organizational risk-taking. Where Schimmer and Brauer (2012), looked at strategic repositioning, we will look at R&D intensity as a proxy for organizational risk-taking. We build our theoretical model around the shifting-focus theory of March & Shapira (1992). We propose that industry dynamism as a measure of environmental uncertainty moderates the relationship between performance below aspiration level and firm performance. Therefore, when operating in a dynamic industry, managers are more likely to see the performance gap as a threat to organizational survival. We also investigate whether munificence, as a measure of available resources in the industry, moderates the relationship between performance below aspiration level and firm performance. Therefore, when operating in a munificent industry, managers are more likely to see the performance gap as a repairable gap. We conduct random-effects panel regression analysis in order to test the moderating effects of munificence and dynamism over time. Panel data used was derived from the Compustat database in the period of 1980 to 2005. The sample we used consist of manufacturing organizations with a SIC-code ranging from 2000-3999.
The results on dynamism show that managers operating in a dynamic environment are more willing to take risks when performance falls below the historical aspiration, compared to when they operate in a less dynamic environment. In dynamic industries we expected
managers to be anxious, stressed and restrictive in terms of information processing, and therefore to become risk-averse as a consequence. Therefore, the results are contrary to what we expected. Results of the research show that when organizations operate in a munificent environment managers are willing to take more risk following performance below the social aspiration level, compared to when they operate in a less munificent environment. This supports our expectations and shows a negative moderating effect.
Our research contributed to the behavioral theory of the firm and performance feedback theory, as we answer the call to look at moderators that influence the relationship
between performance feedback and organizational risk taking. In order to support our assumptions with a theoretical model, we combined multiple theories. Our models suggest that industry munificence and dynamism influence whether organizations become risk-averse or more risk-taking when their performance falls below the aspiration level. Therefore, this study contributes to the literature by showing that industry factors moderate firms’ response to performance feedback. By doing that, we gain a better understanding of the conditions under which organizations respond differently to performance feedback. A deeper
understanding of industry effects on firm risk-taking and organizational learning from performance feedback is gained, which helps explain the contradiction in the current literature. Furthermore, our findings empirically support and elaborate the theoretical assumptions of previous research.
The paper will follow up with a theoretical framework which will provide insight into the previous research done on the research topic and includes the hypothesis of this paper. Then, the methodology and design of the research will be briefly explained, after which the results will be presented. Finally, there will be a discussion section and a conclusion.
Theoretical framework
The behavioral theory of the firm (Cyert & March, 1963) has been foundational for much recent research on organizational stability and change. The key concepts presented by Cyert and March (1963) are bounded rationality, problemistic search, the dominant coalition, standard operating procedures, and slack search (Cyert & March, 1963; March & Simon, 1958; Simon, 1955). Due to bounded rationality, decision makers try to simplify performance evaluations into either a success or a failure, instead of a continuous measure of performance (March, 1988; Greve, 1998). Organizations evaluate performance against a target (aspiration level) and perception of success or failure relative to the target induces organizations to make changes in their behavior (Shinkle, 2012; Chen & Miller, 2007). The aspiration level is the point which delineates what is determined a success and what a failure, and thus the starting point of doubt and conflict in decision making (Schneider, 1992; Lopes, 1987; Greve, 1998). During this study and throughout the literature, aspiration level is seen as: “the smallest outcome that would be deemed satisfactory by the decision maker” (Schneider, 1992, p. 1053). Aspiration levels are set by looking at both the performance of competitors and past performance of the organization (Cyert & March, 1963; Levinthal & March, 1981; Greve, 1998). Therefore, performance feedback theory states that organizations learn from their experience by making the probability of changes depend on their history (Greve, 1998; Levit & March, 1988; Cyert & March, 1963). So, organizations learn from performance feedback by using either a social or historical aspiration level. Following that, they evaluate their performance and take actions based on how they performed relative to the set aspirations (Cyert & March, Greve, 2008; Jordan & Audia, 2012; Hoang, 2016).
The behavioral theory of the firm (Cyert & March, 1963) states that organizations use their performance relative to an aspiration level to determine their behavior. This includes both organizational decision making when performance is above and beyond a certain
aspiration level (Desai, 2008; Shinkle, 2012; Chen & Miller, 2007; Cyert & March, 1963). Reactions regarding performance relative to aspiration levels are contrary when it comes to performance below and above the aspiration level (Schimmer & Brauer, 2012). When organizations are facing a performance shortfall, decision makers will start searching for a solution to overcome it. They feel the need to overcome the performance gap with bolder actions (Audia & Greve, 2006; Cyert & March, 1992; Schimmer & Brauer, 2012). This searching for a solution is often referred to as ‘problemistic search’ (Cyert & March 1963, Greve, 1998, Sitkin ,1992; Desai, 2008). Prior research has provided empirical evidence of this behavior. For example, research by Greve (2003b) found evidence that research and development expenses are increased when low performance causes "problemistic search". Other research has shown that, in response to poor performance, firms are more likely to imitate others’ strategy or change their strategy in order to improve the performance (Park, 2007; Bromiley, 1991; Fiegenbaum et al., 1996; Shinkle, 2012; Schimmer & Braurer, 2012). In line with this, firms performing far below their aspiration level have shown to focus more on exploring their environment and thus learning from their competitors’ practices rather than relying on their own knowledge and experience (Baum & Dahlin, 2007). This exploring and changing in strategy are assumed to involve greater risk (Bromiley, 1991; Schimmer & Brauer, 2012). Behavioral theory of the firm therefore argues that organizations and managers will be more inclined to risky behavior when performance falls below the aspiration level (Cyert & March, 1963; Schimmer & Brauer, 2012; Greve, 2003b; Audia & Greve, 2006; Desai, 2008; Miller & Chen, 2004; Haong, 2016).
Where managers will try to overcome performance gaps, and thereby increase risk taking, they will become more risk averse when the performance exceeds the aspirations (Schimmer & Brauer, 2012). When firms perform well, managers tend to repeat actions or stick to strategies which they think have led to the current success of the organization, even
though a causal link might not exist (Schimmer & Brauer, 2012; Levitt & March, 1988; Miller & Chen, 1994, 1996; Milliken & Lant, 1991). Managers strive to increase performance even further, though they are not willing to take more risks in order to achieve this (Haong, 2016; Iyer & Miller, 2008; Chen & Miller, 2004). Behavioral theory of the firm therefore argues that when performance is above the aspiration level, organizations will become more risk averse (Cyert & March, 1986, Kahneman & Tversky, 1979; Staw et al., 1981).
Literature gap and research question
Although there is a lot of research that states that performance below aspiration level will lead to increased risk-taking, researchers failed to provide consisting evidence. More specifically, multiple studies showed that, under certain conditions, organizations tend to become risk-averse when their performance falls below the aspiration level (Staw et al., 1981; Wiseman & Bromiley, 1996; Cameron, 1983; Audia & Greve, 2006; Miller & Chen, 2004). The theory which supports these findings is threat rigidity theory. This suggest that
performance below aspiration leads to risk-aversion, rather than increased risk-taking (Staw et al, 1981; Desai, 2008). Following poor performance, individuals restrict information
processing and revert to responses deemed successful in the past. Moreover, organizations tend to centralize decision making, forego long-term planning, divert slack resources to cover operating expenses, and curtail risky investments or activities (Desai, 2008; Audia & Greve, 2006). In order to reduce the probability of organizational failure, managers either choose low-risk options or are unable to consider riskier alternatives. This leads to the problemistic search being constrained and the organization becoming more risk-averse (Desai, 2008; Audia & Greve, 2006).
Following the conflicting results, several perspectives have been developed that help explain the contradiction in the literature. March and Shapira (1992) developed a shifting focus model which could help explain whether decision makers will become riskier or more
risk-averse following performance shortfall. They state that decision makers shift their attention between aspiration level and a survival point at which an organization fails due to extremely low performance (Desai, 2008). When performance falls below aspiration level and decision makers’ focus changes towards the survival point, they will start to become more risk averse. If the focus is on survival, decision makers will believe that risky actions with uncertain outcomes could lead to performance below the organization’s threshold for failure. When the focus is upward on exceeding the aspiration level in the future, organizations are likely to take more risks. They believe the risk could yield them the improvement in performance that is required to exceed the aspiration level again (Desai, 2008; March & Shapira, 1992; Miller & Chen, 2004; Ocasio 1995). In this model, therefore, the impact of performance shortfalls on risk taking depends on whether a manager’s focus is on aspirations or survival points (March & Shapira 1992, Miller & Chen 2004, Ocasio 1995). Various determinants of attention focus have been proposed. For example, extremely low performance like a major catastrophe or the threat of bankruptcy appears to restrict risk-taking, as it shifts manager’s attention towards organizational survival (March & Shapira, 1992, Miller & Chen, 2004, Sitkin, 1992).
Other factors suggested to influence the relationship are certain internal contingency factors (Osiyevskyy et al., 2015). Firm size (Greve, 2011; Audia & Greve, 2006), survival reference point (Iyer & Miller, 2008; Shimizu, 2007), slack resources (Chattopadhyay et al., 2001; Singh, 1986), distance to bankruptcy (Desai, 2008), are amongst the factors having an effect on whether firms will be more likely to take risks, or whether they become more risk-averse. Despite the research done on relationship between performance below aspiration level and risk-taking, there is still no consensus in the literature. Where researchers call for more research into moderating variables that influence the relation, they seem to overlook
Where previous research showed that an organizations environment is one of its major contingencies, studies that use industry contingency factors as moderator on the relationship between performance feedback and organizational risk taking are lacking (Desai, 2008; Audia & Greve, 2006; Shinkle, 2012). Therefore, in order to gain a better understanding of how organizations react to or interpret performance feedback, we should study external contingency factors that might influence whether they become either taking or risk-averse following performance feedback.
Moderating effects of dynamism
The reasoning behind this view is that organizations tend to react to environmental changes in order to keep an optimal fit. It is important that organizations adapt properly to the environmental changes, as it is argued that the effectiveness of an organization is influenced by the degree of fit between the organization and its environment (Doty, Glick & Huber, 1993; Miles & Snow, 1978; Chattopadhyay et al., 2001). Alignment between the organization and its environment will maintain the competitiveness and survival of the firm over the long run (Hambrick, 1983; Fiol & Lyles, 1985). Proposed by the resource dependency literature and argued by several researchers, external industry context therefore influences
organizational behavior significantly. A considerable amount of research has shown the impact that the environment has on organizations structures, processes, strategies and
performance (Goll & Rasheed, 2004; Özsomer et al., 1997; Lawrence & Lorsch, 1967; Chen et al, 2017; Keats and Hitt, 1988; Palmer and Wiseman, 1999; Walters et al., 2010; Aldrich, 1979; Dess and Beard, 1984). Thus, there is much evidence that supports the idea that organizations react differently when facing different environmental conditions (Duncan, 1972; Downey, Hellrigel & Slocum, 1975; Lawrence & Lorsch, 1967; Yasai-Ardekani, 1989. To measure and conceptualize organizations industry environment, Aldrich (1979) and Dess and Beard (1984) conceptualized three dimension of an industry. Although there is little
consensus about a scheme of industry contextual dimensions, two often-used contextual industry characteristics are dynamism and munificence (Keats and Hitt, 1988; Palmer and Wiseman, 1999; Walters et al., 2010; Schimmer & Brauer, 2012; Yasai-Ardekani, 1989).
Industry dynamism is a measure that is often used to indicate uncertainty and instability of an industry (Dess & Beard, 1984; Duncan, 1972; Schimmer & Brauer, 2012). More specifically, industry dynamism tries to capture to what extend an environment is characterized by: “change in technologies, variations in customer preferences, and
fluctuations in product demand or supply of materials” (Jansen et al., 2006, p. 1664; Dess and Beard, 1984; Chan et al., 2016). Dynamic environments are perceived as highly uncertain and often considered to be very risky. Higher uncertainty makes it more difficult for managers to interpret and predict regularities and patterns (Eisenhardt and Bourgeois, 1988; Chen et al., 2017; Duncan, 1973). It is harder to respond to a lack of fit with the environment and it can lead to high volatility in performance (Chen et al, 2017). As a result, managers tend to suffer from greater information processing burdens, they experience higher levels of stress and have more feelings of anxiety (Tushman, 1979; Waldman et al., 2001). Managers have the feeling that wrong decisions could result in severe trouble and possibly put the survival of the organization at risk (Waldman et al., 2001).
When operating in less a dynamic environment, decision-makers can rely on pre-established rules and routines, when responding to performance feedback (Duncan, 1973). Decision-makers’ information procession capabilities tend to be more effective (Duncan, 1973). In a stable environment, interpretation and predicting regularities and patterns becomes easier (Duncan, 1973). Lower levels of uncertain will lead to less stress, as managers can rely on the pre-established rules, routines, strategies. Research by Zahra (1996) suggests that when organizations operate in a more stable environment they tend to modify existing products and processes and are less likely to use external sources in order to keep up with the rapid changes
(Miles and Snow 1978; Zahra, 1996). Organizations operating in a stable environment tend to be “inclined to exploit existing products, services and markets” (Jansen et al, 2006, p. 1664). This suggests that managers have confidence in their way of working and should therefore be less anxious when their performance falls below their aspiration level. Because there is less uncertainty, managers should not have the feeling that changes in performances are a threat to the survival of the firm. As a result, they are more likely to see the performance shortfall as a repairable gap. Therefore, we suggest that when performance falls below the aspiration level and there are low levels of dynamism, organizations are willing to take more risks in order to overcome the gap.
Therefore, we suggest that industry dynamism is likely to affect managers’ decision making. Because of the uncertainty, anxiety, stress and decreased confidence they switch to a more risk averse strategy following performance shortfalls (Schimmer & Brauer, 2012; Waldman et al., 2001). Managers will switch their focus towards the survival point instead of the aspiration point (March & Shapira, 1992) and therefore will take less risk compared to managers that operate in a less dynamic environment. Managers operating in a less dynamic environment could still be focused on the aspiration level and will therefore be inclined to take more risks. Therefore, we propose a positive moderating effect of dynamism on the relationship between performance below aspiration level and organizational risk taking.
Hypothesis 1: When performance is below the aspiration level, decreases lead to less risk taking among firms that operate in a dynamic environment than among firms that operate in a less dynamic environment.
Moderating effect of munificence
Next to environmental uncertainty, the scarcity-munificence component of the environment may also be an important determinant of organizational actions in the environment (Staw and Szwajkowski, 1975). “Munificence has been defined as the abundance and availability of external resources that can support organizational growth” (Chen et al., 2017; p. 126). In a munificent industry, resource dependencies are limited. Furthermore, there are more opportunities for organizational growth and profits (Dess & Beard, 1984; Duncan, 1972; Staw & Szwajkowski, 1975; Schimmer & Brauer, 2012). It is argued that munificence can increase organizations’ confidence and managers’ confidence (Barney, 1991; Chen et al., 2017). Next to that, the growth opportunities in a munificent industry are not uniformly distributed, which therefore implies that the industry is
characterized as a ‘multi-peaked growth-landscape’ (Levinthal, 1997; Schumpeter, 1939; Chen et al., 2017). Gavetti and Levinthal (2000) argued that a multi peaked growth-landscape can best be searched with a broader search strategy. Under these conditions, firms that have performance shortfalls below their aspiration level, are therefore more likely to take risks and have a broader search scope, will have an advantage (Schimmer & Brauer, 2012; Cyert & March, 1968; Brittain & Freeman, 1980; Miller & Chen, 1994). Miller and Chen (1994) argue that munificence also lowers mobility barriers of an industry, as market leaders are less determined to protect their market segments against possible new entrants and existing competitors(Schimmer & Brauer, 2012). Therefore, when performance falls below the aspiration level, managers see the performance gap as a gap that can be overcome, and their focus will therefore be on the aspiration level (March & Shapira, 1992). As explained before, when the focus is on the aspiration level managers are more likely to increase risk-taking following performance shortfalls (March & Shapira, 1992.
Yasai-Ardekani (1989) states that organizations react differently in times of scarcity than in times of munificence. Where organizations’ reactions to uncertainty in times of munificence are lower formality, higher levels of structural complexity, and greater decentralization of decision making (Duncan, 1972; Downey, Hellrigel & Slocum, 1975; Lawrence & Lorsch, 1967; Yasai-Ardekani, 1989), typical responses to scarcity are
formalization of procedures, centralization of decision making and restriction of channels of communication (Cameron & Zammuto, 1983; Levin et al., 1978; Zammuto, 1983; Yasai-Ardekami, 1989). The reactions to environmental scarcity are similar to organizational responses to crises and environmentally induced threats (Staw, Sandelands & Dutton, 1981). Responses to crises as suggested by researchers include a tendency to increase the
formalization of procedures and standardization. Furthermore, adherence to previously established decision rules is emphasized and problem solving becomes more rigid than before. Next to the amount of people involved in the decision-making process, which tends to decrease, decision making tends to shift to higher hierarchical levels in times of crisis
(Billings, Milburn, & Schaalman, 1980; Hedberg, Nystrom, & Starbuck, 1976; Smart & Vertinsky, 1977; Yasai-Ardekami, 1989). Other research by Schimmer and Brauer (2012) argues that organizations’ choice to diverge or converge from their strategic group after performance had fallen below their aspiration level was influenced by external factors such as munificence. In munificent times, organizations were more likely to diverge from their
current strategy. In a sense we could say organizations are more likely to take risks in times of munificence.
Building on the behavioral theory of the firm (Cyert & March, 1963), the shifting focus model (March & Shapira, 1992) and the research described previously (Schimmer & Brauer, 2012; Chen et al., 2017), I expect organizations to increase risk-taking in times of munificence when their performance falls below the aspiration level. As mentioned earlier,
managers could have increased confidence, there are more opportunities for growth and it is easier for organizations to get their hands on resources in order to overcome the performance gap (Schimmer & Brauer, 2012; Chen et al., 2017; Yesai-Ardekami, 1989; Dess & Beard, 1984). It results in managers seeing the performance gap as a gap that could be overcome, and their focus will therefore be on the aspiration level (March & Shapira, 1992). Furthermore, earlier research suggests that they naturally want to overcome the performance gap, as they attribute more weight to losses than to gains (Cyert & March, 1963; Kahneman & Tversky, 1979). Combining this will lead to increased risk taking when performance falls below the aspiration level.
Building on the threat rigidity theory (Staw et al., 1981), the shifting focus model (March & Shapira, 1992) and previous research (Schimmer & Brauer, 2012; Greve, 2003; Sitkin 1992; Yasai-Ardekani, 1989; Desai, 2008), I expect that in times of scarcity managers tend to categorize the environmental conditions as a threat. As suggested by Staw et al., (1981), the reactions to environmental scarcity are similar to organizational responses to crises and environmentally induced threats. It could induce fear and managers could perceive the performance shortfall in combination with the scarce environment as a threat to
organizational survival. Therefore, when performance falls below the aspiration level managers might switch their focus from increasing the performance to surviving the adversity, resulting in them becoming more risk averse (March & Shapira, 1992).
I expect organizations to increase risk-taking when performance falls below the aspiration when the industry is munificent. As mentioned earlier, managers could have
increased confidence, there are more opportunities for growth and it is easier for organizations to get their hands on resources in order to overcome the performance gap. Contrary to that, I expect managers to become more risk-averse in times of industry scarcity. They perceive the environment as a threat and the performance fallbacks will more likely be seen as
unrepairable gaps or as threats to organizational survival. I expect organizational risk taking following performance below aspiration level to increase in times of industry munificence and decrease in times of scarcity.
Hypothesis 2: When performance is below the aspiration level, decreases lead to less risk-taking among firms that operate in a scarce environment than among firms that operate in a munificent environment.
Method
Data and sampling strategy
The panel data was derived from the Standard and Poor’s Compustat database. Compustat collects operational and financial information for all publicly traded U.S.
companies (Chen & Miller, 2007). In order to compare the results with other previous studies on performance feedback (e.g. Chen & Miller, 2007; Chen, 2008; Iyer & Miller, 2008), we selected manufacturing companies with SIC codes ranging from 2000 to 3999. Another reason for restricting the industries is to prevent confounding results due to major differences in industry activities (Iyer & Miller, 2008; Brush, 1996; Davis & Thomas; 1993). The total sample includes 41,334 firm-year observations. The major two-digit industry categories were chemicals and allied products (17.92%), industrial machinery and equipment (13.91%), electronic and other electric equipment (17.48%), instruments and related products (13.71%). In this research the control and independent variables were lagged by one year in order to capture the effect on the dependent variable. In order to examine organizational risk-taking behavior over time, we used panel data from 1980 to 2005.
Dependent variable
Organizational risk-taking. In line with performance feedback literature, R&D
intensity (RDI) is used as a proxy for organizational risk-taking (e.g. Chen & Miller, 2007; Lim & McCann, 2014). R&D intensity is calculated as the R&D expenditures divided by sales. Firms with a R&D intensity higher than 1, where R&D expenditures exceed sales, make decisions based on different criteria then what we hypothesize (Chen & Miller, 2007). Firms with R&D Intensity smaller than -1, make decisions based on different criteria then what we hypothesize. Therefore, organizations with R&D Intensity smaller than -1 and bigger than 1 were removed from the sample (similar to Chen & Miller, 2007; Lim & McCann, 2014). The final sample has a mean of .14 and standard deviation of .09. This is in line but a little higher
than Chen and Miller (2007) and Lim and McCann (2014). The reason for this is because they do not remove firms with R&D intensity smaller than -1.
Independent variable
Performance below aspiration level. For this study, return on assets (ROA) is used as
a measure of performance. The performance variable is lagged one-year relative to the dependent variable (Similar to Chen & Miller, 2007; Latham & Braun, 2008). As for aspirations we chose to run two different models with different aspiration proxies: one for own firm past performance (historical) and one for industry median past performance (social). Performance measures are evaluated relative to aspiration levels determined by the
organization's recent performance history or the performance of comparable others (Cyert and March 1963, Greve 1998). The historical aspiration level was calculated as the average ROA of the firm’s past three years. The social aspiration level was calculated as the mean ROA of all companies within each industry (defined at the SIC4 level). Actual performance was measured at year t – 1, and each aspiration level was proxied at year t – 2. With this information the firm’s performance feedback was computed. The historical performance feedback was calculated by the firm’s ROA – firm’s average ROA of last three years. The social performance feedback was constructed as firm’s ROA – the mean ROA of all companies within each industry (defined at the SIC4 level) (similar to Hoang, 2016). To examine performance below the aspiration level, we constructed spline variables by splitting ROA into two variables based on the aspiration level (Greve 1998; Chen & Miller, 2007; Desai, 2008). In order to prevent model misspecification, we also include performance above aspiration level, although it will not be included in the hypothesis (Greve 1998; Chen & Miller, 2007; Desai, 2008).
Moderating variables
1995, p. 312). Similar to Boyd (1995), dynamism was operationalized using a standardized measure of the volatility of industry sales growth rate over a 5-year period. The standard error of regression slope coefficient was divided by the mean value. The coefficients are based on a regression of time against the value of shipments. Estimates for any given year are based on the five preceding years, for example the dynamism estimate for 2000 is based on data for 1995-2000. Industries are defined using 4-digit SIC-codes (Boyd, 1995). The data on
dynamism was obtained from the Compustat database and ranged from 1980 to 2005. Higher values of dynamism suggest higher volatility and instability in an industry.
Munificence. Munificence measures “the abundance of resources in an industry”
(Boyd, 1995, p. 312). Similar to Boyd (1995) and other studies (e.g. Goll & Rasheed, 1997), munificence-scarcity was operationalized using a standardized measure of industry sales growth over a 5-year period. The data on munificence was obtained from the Compustat database and ranged from 1980 to 2005. The regression model is the same as used for
dynamism (Boyd, 1995). Higher values of munificence suggest that there are more resources available in an industry.
Control Variables
Firm size. is a commonly used control measure, which will also be included in our
models (e.g. Coad & Rao, 2010; Osiyevski et al., 2014; Greve, 2003; Chen & Miller, 2007). Firm size is said to be directly related to R&D intensity (Baysingen & Hoskisson, 1989), and is argued to influence the directionality of organizational actions (Chattopadhyay et al., 2001). Firm size can be measured with several different measures (i.e. log employees, log revenues, cash, market value,). In tis study we use the logarithm of the number of employees (similar to Chen & Miller, 2007; Osiyevski, 2014; Gaba & Joseph, 2013)
Financial slack. Prior research on slack resources showed the impact of slack
we control for the effects of slack. In line with previous research (e.g. Chen & Miller, 2007; Bourgeois, 1981; Singh, 1986), we used the current ratio as measure of organizational slack. The current ratio is calculated by dividing the current assets by the current liabilities.
Distance to bankruptcy. Organizations that are on the verge of bankruptcy are more
likely to become risk-averse when their performance falls below the aspiration level.
Therefore, and in line with previous research we controlled for distance to bankruptcy (March & Shapira, 1992; Desai, 2008; Chen & Miller, 2007; Iyer & Miller, 2008; Lim & McCann, 2014). For distance to bankruptcy we used Altman’s (1983) Z score. Altman’s Z is defined as (1.2 x working capital divided by total assets) + (1.4 x 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). A lower Z value means a higher likelihood of bankruptcy (similar to Chen & Miller, 2007; Iyer & Miller, 2008).
Statistical method
For our empirical analysis, we took into account the cross-section and time-series nature of the data. The nature of the dataset leads us to either fixed- or random-effects panel regression analysis. In order to check whether we should use fixed- or random-effects models we ran an Hausman test on our control model with independent variables (model 2). Results of the Hausman test rejected the null hypothesis (χ2 (1) = 857.02, p < 0.01), therefore
suggesting that we should use a fixed-effect regression model (Hausman, 1978). In addition to this we ran tests checking for heteroscedasticity and auto correlation. In order to test for heteroscedasticity, we ran a modified Wald test for groupwise heteroscedasticity in fixed effect regression model. Results of this test suggests the presence of heteroscedasticity (χ2 (4122) = 3.7e+41, p < 0.01). Results of the Results of the Woolridge test for auto correlation in panel data rejected the null hypothesis (F(1,3261) = 93252.53, p < 0.01). This indicates the
presence of serial correlation (Woolridge, 2010; Drukker, 2003). In order to control for heteroscedasticity and auto correlation, we used robust standard errors.
Results
Descriptive statistics and correlation analysis
Table 1 provides means, standard deviations, and correlations for all the variables used in this study. Similar to Chen & Miller (2007), we removed firms with a Research and
Development Intensity (RDI) greater than one and smaller than minus one. We removed these firms because they make decisions based on different criteria then what we hypothesize. As we calculated ROA by dividing profit by the total assets, we also removed companies with total assets that are smaller than one, as we do not want to divide by numbers smaller than 1. The final sample consisted of 41,334 observations from 1980 to 2005. The major two-digit industry categories were chemicals and allied products (17.92%), industrial machinery and equipment (13.91%), electronic and other electric equipment (17.48%), instruments and related products (13.71%).
Table 1 also shows the correlations between the different variables used in this study. The control variable firm size (r = -0.32) showed a negative correlation with the dependent variable R&D intensity. Slack resources (r = 0.25) and distance to bankruptcy (r = 0.10) showed a positive correlation with R&D intensity.
The correlation between the independent variable performance below aspiration level and the dependent variable R&D intensity show for both social aspirations (r = -0.27) and historical aspirations (r = -0.19) a negative correlation.
The correlation between the moderating variables dynamism (r = 0.09) and
munificence (r = 0.06) with dependent variable R&D intensity show very weak but positive correlations. Please check table 1 for the full set of correlations.
Regression analysis
Table 2 shows the fixed-effect panel regression results for dependent variable Research and Development Intensity (RDI). Table 2 consists of 4 models. Every model is
made up for both historical and social aspirations. Model 1 is the control model and consists of all the control variables and moderator variables. The independent variables are added in Model 2. Model 3 and 4 include the interaction terms showing the moderating effects for dynamism (model 3) and munificence (model 4).
Model 1, representing the control model shows the direct effects of the control variables and moderating variables on the dependent variable R&D intensity. The results show that distance to bankruptcy has no significant direct effect on R&D intensity (b = 0.002, p > 0.05). Slack resources (b = .003, p < 0.001) and dynamism (b = .373, p < 0.001) have a positive and significant effect on R&D intensity. This suggest that when slack resources and dynamism increase organizations tend to invest more in R&D intensity. Firm size (b = -.012, p < 0.001) and munificence (b = -.023, p < 0.05) showed a negative and significant
relationship with R&D intensity. This suggest that when munificence and firm size decrease, organizations R&D intensity is likely to increase and when munificence and firm size
increase, organizations tend to lower their R&D intensity. The explained variance of the model is also given by table 2 (R2 = .111).
Model 2, represents the model which next to the control and moderating variables, includes the independent variables. It is important to note that when interpreting performance below and above aspiration one must be cautious. When performance is below the aspiration level, a positive coefficient means that firms will lower their R&D intensity following a further decrease in performance below the aspiration level. A negative coefficient of Performance below aspiration, means that organizations, following a further drop in
performance below aspiration level, will increase their R&D intensity (Chen & Miller, 2007; Iyer & Miller, 2008; Lim & McCann, 2014; Hoang, 2016). Model 2 shows the main effect of performance below aspiration level on R&D intensity to be significant and negative for both social (b = -0.060, p < 0.01) and historical aspirations (b = -.051, p < 0.05). As explained
before, these results suggest that R&D intensity increases as performance falls further below the aspiration level.
Hypothesis 1 predicted a positive moderating effect of dynamism on the relationship between performance below aspiration level and R&D intensity. Therefore, organizations in a dynamic industry are less likely to increase their R&D intensity as performance falls further below the aspiration level. Model 3 shows the moderating effects of dynamism on the
relationship between performance below aspiration level and R&D intensity. The results show a negative but non significant interaction effect in the social model (b = -.450, p > 0.05). This suggests that dynamism has no influence on the relationship between performance below aspirations level and R&D intensity, if managers act upon social aspirations. For the historical model it shows a strong and negative significant (b = -1.378, p < 0.001) moderating effect of dynamism on the relationship between performance below aspiration level and R&D
intensity. This suggests that managers in dynamic environments are more likely to take risk when performance falls below their historical aspiration level, compared to managers that operate in an industry with less dynamism. This is contrary to what we predicted and therefore hypothesis 2 is rejected. The explained variance of the historical model slightly decreased compared to the control model. (R2 = .118).
In order to gain a better understanding of the moderating effect we plotted the simple slope effects of historical dynamism. In order to compute the simple slopes, weplotted the relationship between performance below aspiration level and R&D intensity for different values of dynamism. The range of dynamism was determined by subtracting and adding one SD from the mean. This gave us the simple slope effects ranging from different measures of dynamism ranging from .004 to .029. The results showed that for the lowest values of dynamism (0.004 and 0.009) the moderating effect was not significant (p > 0.05). For performance below aspiration level, we used the maximum and minimum values reported in
the model. Namely, -22 and 0. As for performance below aspiration level, we only have negative measures, ranging from 0 to -22. Figure 1 shows the interaction plot of the moderating effect of dynamism on the relationship between historical performance below aspiration level and R&D intensity. The results show that for both high and low values of dynamism there is a negative relationship between historical performance below aspiration level and R&D intensity. Therefore, when performance falls further below the aspiration level, R&D will increase more. When industry dynamism is high the negative relationship is stronger. R&D This implies that dynamism has a negative moderating effect on the
relationship between historical performance below aspiration level and R&D intensity. Therefore, in times of dynamism organizations are more willing to take risk when their performance falls below the historical aspiration level, compared to when they operate in a less dynamic, stable environment.
Hypothesis 2 predicted a negative moderating effect of munificence on the
relationship between PBA and R&D intensity. Therefore, organizations that operate in a more munificent industry are more likely to increase their RDI when performance falls below their aspiration level. Model 4 shows the results for both the social and historical aspiration level. The social model shows a negative and significant (b = -.312, p < 0.01) moderating effect. This suggests that, managers facing a munificent environment are more likely to increase R&D intensity following performance below their social aspiration level, compared to managers operating in a low munificent environment. This is line with hypothesis 2 and therefore we should accept the hypothesis for social aspirations. The explained variance of the model increased slightly compared to the control model (R2 = .136). For the historical model, results are also negative but not significant (b = -.104, p > 0.05). This suggests that there is no moderating effect of munificence on the relationship between performance below aspiration level and R&D intensity, if managers react to performance shortfalls compared to their
historical aspirations.
In order to gain a better understanding of the moderating effect we plotted the simple slope effects of social munificence. In order to compute the simple slopes, we plotted the relationship between performance below aspiration level and R&D intensity for different values of munificence. The range of munificence was determined by subtracting and adding one SD from the mean. This gave us the simple slope effects ranging from different measures of munificence ranging from (-0.01) to (0.12). The results showed that for the lowest values of munificence (-.01 and .01) the moderating effect was not significant (p > 0.05). For performance below aspiration level, we used the maximum and minimum values reported in the model. Namely, -22 and 0. We only have negative measures of performance below aspiration, as we changed values above zero to zero as we made spline variable for performance feedback. Figure 2, shows the simple slopes of the relationship. When interpreting the results, we can see that under conditions of industry munificence, the relationship between performance below aspiration level and R&D intensity is negative but stronger. Therefore, the results show a negative moderating effects of munificence on the relationship. In times of less munificence, or scarcity, managers also tend to increase their R&D intensity when their performance falls further below the aspiration level, though this effect is weaker if you compare it to high munificent environment.
Tab le 1 . D es cr ip tiv e stati sti cs th e fo r p an el d ata Var iab le N Me an S.D. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 1. R& D In te ns ity 41,334 0. 09 0. 14 - 2. Fir m s ize 40,419 6. 84 2. 15 -0. 32* - 3. Sl ac k re so ur ce s 40,891 3. 31 3. 72 0. 25* -0. 31* - 4. D is tan ce to b an kr up tc y 40,535 5. 95 18. 40 0. 10* -0. 14* 0. 59* - 5. D yn am is m 40,341 0. 02 0. 01 0. 09* -0. 02* 0. 03* -0. 01* - 6. Mu ni fic en ce 40,345 0. 06 0. 07 0. 06* -0. 11* 0. 04* 0. 04* -0. 19* - 7. PA A So ci al 41,261 0. 12 0. 15 -0. 03* 0. 01* 0. 11* 0. 12* -0. 02* 0. 07* - 8. PA A Hi sto ric al 37,560 0. 06 0. 18 0. 12* -0. 18* 0. 03* 0. 05* 0. 03* -0. 01 0. 10* - 9. PB A So ci al 41,261 -0. 07 0. 26 -0. 27* 0. 23* 0. 06* 0. 09* 0. 02* -0. 00 0. 20* -0. 01 - 10. PB A Hi sto ric al 37,560 -0. 07 0. 22 -0. 19* 0. 15* 0. 03* 0. 11* -0. 05* -0. 01 0. 16* 0. 09* 0. 87* - Co rr el ati on s w ith a * ne xt to it ar e si gn ifi can t at th e p < 0. 05 le ve l.
Tab le 2 . F ix ed e ffe cts p an el re gr es si on s fo r R& D in te ns ity : r es ul ts fo r th e fu ll sam pl e. Mo de l 1 Mo de l 2 Mo de l 3 Mo de l 4 So ci al Hi sto ric al So ci al Hi sto ric al So ci al Hi sto ric al Coe ff SE Coe ff SE Coe ff SE Coe ff SE Coe ff SE Coe ff SE Coe ff SE Co ns tan t .1 56 *** (.012) .1 40 *** (.013) .1 60 *** (.013) .1 47 *** (.012) .1 42 *** (.013) .1 52 *** (.013) .1 47 *** (.013) Fi rm S ize -.0 12 *** (.001) -.0 09 *** (.002) -.0 13 *** (.002) -.0 10 *** (.002) -.0 10 *** (.002) -.0 10 *** (.002) -.0 10 *** (.002) Sl ac k re so ur ce s .0 03 *** (.000) .0 03 *** (.000) .0 02 *** (.001) .0 02 *** (.001) .0 02 ** (.001) .0 02 *** (.001) .0 02 ** (.001) D is tan ce to b an ktr up tc y c -.0 23 (.016) -.0 01 (.013) -.0 03 (.017) -.0 17 (.018) -.0 11 (.019) -.0 19 (.018) -.0 12 (.019) D yn am is m .3 73 *** (.057) .3 51 *** (.054) .3 47 *** (.056) .2 24 *** (.061) .1 56 * (.063) Mu ni fic en ce -.0 23 * (.012) -.0 24 * (.011) -.029* (.012) -.0 21 * (.011) -.0 15 (.012) Pe rf or m an ce b el ow as pi rati on -.0 60 ** (.022) -.0 51 * (.021) -.0 15 (.013) -.0 01 (.009) -.0 08 (.010) -.0 15 (.012) Pe rf or m an ce ab ov e as pi rati on -.0 97 *** (.009) -.0 38 *** (.011) -.0 34 *** (.006) -.0 16 * (.008) -.0 33 *** (.006) -.0 18 * (.008) Pe rf or m an ce b el ow as pi rati on x D yn am is m -.4 50 (.458) -1 .3 78 *** (.371) Pe rf or m an ce b el ow as pi rati on x Mu ni fic en ce -.3 12 ** (.105) -.1 04 .(127) Mo de l F 26 .3 2*** 39 .1 7*** 17 .4 5*** 14 .9 4*** 12 .1 6*** 16 .1 2*** 7. 27 *** R² .1 31 .154 0. 122 .140 .1 18 .1 36 .1 09 N 38705 38641 35245 35123 31757 35127 31761 a Si gn ifi can ce le ve ls : †p <0 .1 *p <0 .0 5, **p <0 .0 1, ***p <0 .0 01 ; b E sti m ate d co ef fic ie nts an d as so ci ate d ro bu st stan dar d er ro rs (i n par en th es es ) ar e re po rte d; c Co ef fic ie nts an d stan dar d er ro rs ar e m ul tip lie d by 1 0² .
Figure 1. Plot of moderating effect of dynamism on the relationship between historical performance below aspiration level and R&D intensity.
Figure 2. Plot of moderating effect of munificence on the relationship between social performance below aspiration level and R&D intensity.
Low performance below aspiration
High performance below aspiration R& D In te n si ty Low Dynamism High Dynamism
Low Performance below
aspiration High Performance below aspiration
R& D In te n si ty Low Munificence High Munificence
Discussion
Key findingsPerformance below aspiration level is significant and negatively related in our main model. This implies that organizations tend to take more risk when performance falls below the aspiration level. This is in line with the performance feedback theory and the problemistic search arguments. The regression models show that, when industry dynamism is high,
organizations tend to be willing to more risk when performance falls below the historical aspiration level, compared to when they operate in a stable environment. Our models did not find an interaction effect of dynamism when organizations look at the social aspiration level. The regression models further showed that organizations tend to take more risks when their performance falls below the social aspiration level when the values of munificence are higher. Our models did not support an interaction effect of munificence when performance falls below the historical aspiration level. While munificence had a significant impact on the relationship, it contributed to only a small part of the variance in R&D intensity. Most of the variance was explained by the control variables size, financial slack and how close firms were to bankruptcy, which are firm effects.
Theoretical implications
Looking at the theoretical implications, there is support for both the main effect and the moderation variable munificence. The main effect shows a negative relationship between performance below aspiration level and R&D intensity. Both historical and social aspirations show the same effects on the dependent variable R&D intensity. The negative relationship between performance below aspiration level and R&D intensity is in line with earlier research. Earlier research suggested that organizations will search for solutions and thereby take risks when performance is poor. In the literature, this search for solutions is referred to as problemistic search (Cyert and March 1963, Greve 1998, Sitkin 1992; Desai, 2008). We also
included performance above aspiration level in the analysis, though we did not draw up any hypothesis for this variable. Analysis showed, however, that in line with previous research and behavioral theory of the firm, there is a negative relationship between performance above aspiration level and organizational risk-taking.
Behavioral theory of the firm, as mentioned before, states that poor performance indicates that an organization’s routines and operating practices are not well suited for its current environment (Boeker & Goodstein, 1991; Desai, 2008). This often results in decision-makers searching for solutions and taking risks in order to remedy the performance shortfall (Cyert & March 1963, Greve 1998, Sitkin 1992; Desai, 2008). The results of this study are thus in line with prospect theory (Kahneman & Tversky, 1979) which argues that
organizations performing below their aspiration level will be more likely to take risks and therefore focus on change (Bromiley, 1991; Greve, 1998; Greve, 2003a).
Drawing on the behavioral theory of the firm, we looked at contingency factors that might affect the relationship between firm performance below aspiration level and its risk behavior. In line with earlier research done regarding the effects of the firm’s industry environment, we looked at the effect of munificence and dynamism. The results showed that although only a small part of the variance is explained by the interaction, the moderation effect of munificence is significant. Research by Schimmer and Baum (2012) suggested that the influence of performance aspirations on strategic repositioning is contingent on industry factors like munificence and dynamism. Research by Staw and Szwajkowski (1975) found that, when organizations face a less munificent environment, they are more likely to engage in illegal activities and irresponsible actions. Strategic repositioning and illegal activities can be seen as risk activities. This study thus contributes to the literature by showing similar
moderating effects of munificence-scarcity on the relationship between performance below aspiration level and organizational risk-taking.
According to Chattopadhyay et al. (2001), executives’ perceptions in particular determine the course of action an organization makes following environmental changes. Executives appear to filter and interpret incoming information and the decisions they make are dependent on these interpretations (Chattopadhyay et al., 2001; Hambrick & Mason, 1984; Starbuck & Milliken, 1988; Thomas et al., 1993). They seem to categorize the environmental changes into being either a threat or an opportunity (Chattopadhyay et al., 2001; Dutton & Jackson, 1987, Fredericksen, 1985; Jackson & Dutton, 1988). Staw et al. (1981) showed that reactions to environmental scarcity are similar to organizations’ responses to crises and environmentally induced threats. Therefore, a possible explanation why
managers become more risk-averse rather than risk taking in a situation where there is
scarcity and performance below aspiration level could be that managers change their attention from increasing performance to surviving the adversity. Threat-rigidity theory states that when organizations shift their attention towards surviving the adversity as a result of the perceived threat to organizational survival, they tend to become more risk-averse (Staw et al., 1981). Our study supports this theory, by showing that, in times of scarcity, managers are less likely to take risks when their performance falls below their set aspiration level. In a more general sense, the results support the idea that organizations’ environments and external factors have an effect on a firm’s performance feedback behavior.
The results of our analysis on dynamism showed results that were not in line with our expectations. We proposed that managers working for organizations that operate in a dynamic environment are more likely to perceive performance shortfalls as a threat to the survival of the firm, and will therefore switch their focus to the survival point. Results showed that when performance is below the aspiration level, organizations operating in a dynamic environment are more likely to take risk and increase their R&D intensity. Existing literature does provide a possible expalanation for the behavior. Resaerch on dynamism suggests that organizations
that operate in a dynamic environment should focus on exploration. Learning by doing and the development of effective routines becomes superstitious (Lave & March 1975) as the environment is volatile and unpredictable. The success factors of the past have little meaning in the future (Sorenson, 2003; Farjoun & Levin, 2011). Managers’ ability to accurately predict the future state of the environment will decrease as a result of industry dynamism.
Organizations could in turn focus on experimentation, rely on structural adjustments and get used to temporary competitive advantages (Beinhocker, 2006; Farjoun & Levin, 2011). It could be that firms operating in dynamic environments are used to performance shortfalls and therefore put greater emphasis on experimentation. It might be that managers operating in a dynamic environment are used to the volatility in performance, and therefore rely on their ability to overcome the performance gap (Chen et al, 2017). This would be in line with literature about behavioral decision theory and prospect theory (Cyert & March, 1963; Kahneman & Tversky, 1979).
Overall, we could say that the model used for the moderating effect of munificence worked out the way we expected. Previous research, the existing literature and theories used to explain the effect are in line with our findings. With the results we contributed to strategic management literature by showing that contextual industry characteristics have significant effects on organizations’ decision-making. Industry characteristics munificence and dynamism seem to have a considerable impact on the relationship between performance feedback and organizational risk-taking. With this, we contribute by explaining part of the contradiction between research on organizational decline (Staw et al., 1981; Wiseman & Bromiley, 1996) and behavioral decision theory (Cyert & March, 1963; Kahneman & Tversky, 1979). Also, we add empirical evidence to the contingency theory of firm, which dictates that both internal and external factors influence decision making for organizations (Donaldson, 2001).
Managerial implications
The results of this study have several implications for managers. First, the results show the impact of contextual industry factors on managers’ decision making. Organizations should keep this in mind when formulating their strategy and while making risky decisions. As our results suggest it seems that when organizations operate in a dynamic environment and they react to performance feedback based on historical aspirations, they more inclined to take risks compared to when operating in a less dynamic environment. As research suggests, when operating in a dynamic environment, organizations should best be changing continuously in order to stay up with the environment. Therefore, increasing R&D intensity following performance below the aspiration level could be the right way to go. Formulating a strategy that is in line with managers’ overall response to performance below aspiration level could help get the support needed for proper execution. As our results suggest it seems that when organizations operate in a munificent environment and they react to performance feedback based on social aspirations, they more inclined to take risks compared to when operating in a less munificent environment. It therefore seems that managers are more risk-averse because of scarcity. Our study contributed, as we gained a better understanding of what drives risk-taking behavior. Thereby, we increase the capability to design proper strategies regarding organizational risk-taking.
Limitations and future research
We did not focus on a particular industry during the research. We drew a sample from the complete Compustat database. What other research could do is focus on specific industries to see whether there are differences between industries. This might help explain the
differences in response to performance feedback as existent in prior studies. Furthermore, we did not take organizational complexity into account during our research. Existing literature does in many cases incorporate this industry characteristic together with dynamism and
munificence. Future research could look at the effects of complexity on the relationship between performance feedback and organizational risk taking. With regard to our dynamism variable, Farjoun and Levin (2011) proposed a different way to measure dynamism, which also incorporates complexity. It would be interesting to see if the results of this study will be similar when tested with the new measure for dynamism. Next to the different industry characteristics, different measures of risk-taking could be taken into account and compared to each other in order to check whether methodological issues arise with the measures.
Conclusion
In response to researchers’ call for further studying moderating factors on the relationship between performance feedback and organizational decision making, we
contributed by looking at the effects of munificence and dynamism. Our results showed two significant negative moderating effects for munificence and dynamism on the relationship between performance below the aspiration level and organizational risk taking. In our theoretical framework we included literature on resource dependency, contingency theory, threat rigidity theory, behavioral theory of the firm and prospect theory, thereby combining multiple fields of research. Lastly, we made some recommendations for future research in order to further explore the interesting field of performance feedback theory.
References
Aiken, M., & Hage, J. (1968). Organizational interdependence and intra-organizational structure. American sociological review, 912-93.
Aldrich H., (1979). Organizations and Environments. Stanford University Press: Stanford, CA.
Altman, E. I. (1983). Corporate distress: A complete guide to predicting, avoiding, and dealing with bankruptcy. New York.
Ansoff, I. H., (1979). Strategic management. London: Macmillan.
Ansoff, I. H., (1987). The emerging paradigm of strategic behavior. Strategic Management Journal, 8, 501-515.
Antonelli, C. (1989). A failure-inducement model of research and development expenditure: Italian evidence from the early 1980s. Journal of Economic Behavior & Organization, 12(2), 159-180.
Argote, L., & D. Epple. (1990). Learning curves in manufacturing. Science. 247(4945) 920– 924.
Argote, L., & Greve, H. R. (2007). A behavioral theory of the firm-40 years and counting: Introduction and impact. Organization Science, 18(3), 337-349.
Audia, P. G., & Greve, H. R. 2006. Less likely to fail: Low performance, firm size, and factory expansion in the shipbuilding industry. Management Science, 52, 83-94.
Baird, I. S., & Thomas, H. (1985). Toward a contingency model of strategic risk taking. Academy of management Review, 10(2), 230-243.
Baum, J. A., & Dahlin, K. B. (2007). Aspiration performance and railroads’ patterns of learning from train wrecks and crashes. Organization Science, 18(3), 368-385.
Baum, J. A. C., Rowley, T. J., Shiplov, A. W., & Chuang, Y. (2005). Dancing with strangers: Aspiration performance and the search for underwriting syndicate partners. Administrative Science Quarterly, 50, 536-575.
Barlevy, G. (2005). Why don’t recessions encourage more R&D spending? Chicago Fed Letter, 220, 1-4.
Baysinger, B., & Hoskisson, R. E. (1989). Diversification strategy and R&D intensity in multiproduct firms. Academy of Management Journal, 32(2), 310-332.
Beinhocker, E. D. (2006). The origin of wealth: Evolution, complexity, and the radical remaking of economics. Harvard Business Press.
Billings, R. S., Milburn, T. W., & Schaalman, M. L. (1980). A model of crisis perception: A theoretical and empirical analysis. Administrative Science Quarterly, 300-316.
