The Interaction Between Leader Power and Legitimacy on
Risk-Taking and the Mediating Role of Risk Perception
Msc Human Resource Management University of Groningen Faculty of Economics and Business
Risk-Taking and the Mediating Role of Risk Perception
ABSTRACT
Risk-taking has many negative consequences, such as decreased firm performance and decreased firm value. Therefore, it is important to understand why and when leaders engage in risk-taking behavior. Building on the approach inhibition theory of power, we examine the role of leader power and legitimacy of that power for leader’s risk-taking behavior. The approach inhibition theory suggests that power is associated with increased risk-taking and initial experimental research has demonstrated this relationship. The present research aims to replicate these findings in an organizational field setting. Moreover, we aim to extend former research by incorporating the moderating role of legitimacy and the mediating role of risk perception for the relationship between leader power and risk-taking. We aim to show that leader power is associated with increased risk-taking, but only when power is perceived as legitimate. Results suggest that leader power is associated with increased risk-taking, although this effect did not reach significance. Moreover, results show that there is no significant interaction between leader power and legitimacy on risk-taking or on risk perception. More research is necessary to determine whether the relationship between leader power and risk-taking exists in practice, and which variables influence this relationship.
THEORY ... 8
Leader Power and Risk-Taking ... 8
The Moderating Role of Legitimacy ... 9
The Mediating Role of Risk Perception ... 11
METHOD ... 13 Data Collection ... 13 Measurement ... 13 RESULTS ... 16 Assumptions ... 16 Descriptive Statistics ... 16 Hypotheses Testing ... 18 DISCUSSION ... 24 Implications ... 24
Limitations and Further Research ... 25
Conclusion ... 27
REFERENCES ... 28
INTRODUCTION
Risk-taking by leaders has many negative consequences for both society and organizations. Many people assume, for instance, that the financial crisis of 2008 was partially caused by leaders taking excessive risks (Nelson & Katzenstein, 2013; Larcker, Ormazabel, Tayan, & Taylor, 2014). Specifically, the crisis of 2008 has created awareness for the catastrophic effects of excessive risk-taking for the society. Pessimistic expectations about the future, negative home equity, and high unemployment levels are just a few of the negative effects of the financial crisis of 2008 (Hurd & Rohwedder, 2010). Furthermore, risk-taking has negative consequences for organizations. Research has shown, for instance, that risk-taking is associated with reduced firm value (Guo, Jalal, & Khaksari, 2015), and poor firm performance (Bromiley, 1991; Naldi, Nordqvist, Sjöberg, & Wiklund, 2007). Considering these negative effects of risk-taking it is important to understand why and when leaders engage in risk-taking behavior.
To date, however, most research with regard to risk-taking has focused on identifying the many negative consequences of risk-taking, whereas little research has examined
antecedents of leaders’ risk-taking behavior (Boyer, 2006). Building on the approach inhibition theory of power (Keltner, Gruenfeld, & Anderson, 2003), the present research examines the roles of leader power and legitimacy of such power for leaders’ risk-taking behavior.
finding in organizational field setting and show that leader power is associated with increased risk-taking.
According to the approach inhibition theory, however, power does not always lead to increased approach behavior (Keltner et al., 2003). Specifically, legitimacy moderates the relationship between power and approach. Keltner et al. (2003) argue that the relationship between power and approach behavior is broken when power is perceived as illegitimate. If powerful people perceive their power as legitimate they do not observe threat and are therefore free to act. However, a threat to the legitimacy of powerful individuals leads to instability, which reduces the freedom of these illegitimate powerful individuals to act and thereby activates inhibition (Keltner et al., 2003). Therefore, the approach inhibition theory suggests that legitimacy moderates the relationship between power and risk-taking. To date, first studies have found that legitimacy moderates the relationship between power and
approach (Lammers, Galinsky, Gordijn, & Ottens, 2008; Smit, Jost, & Vijay, 2008). Lammers et al. (2008) argue that illegitimate powerful people might not act, because they are afraid to lose their power and they focus on avoiding losses.
Based on the approach inhibition theory, we expect that people’s focus changes when they perceive their power as illegitimate (Keltner et al. 2003; Lammers et al. 2008). If powerful people are afraid to lose their power, they are no longer focused on rewards but rather on threats and avoiding losses. As mentioned above the approach inhibition theory suggests that powerful people take more risks since they focus on rewards and benefits (Anderson & Galinsky, 2006). If, however, powerful people perceive their power as illegitimate, they focus on the potential losses of risks rather than benefits (Lammers et al., 2008), and therefore they are less likely to take risks (Anderson & Galinsky, 2006). Hence, we expect that power only leads to more risk-taking when power is perceived as legitimate (see Figure 1). The second aim of the present research is to show that legitimacy moderates the relationship between leader power and risk-taking.
Moreover, the present research proposes that risk perception mediates the interaction between leader power and legitimacy on risk-taking. More specifically, Anderson & Galinsky (2006) suggests that power leads to a more optimistic risk-perception and therefore increases risk-taking. Furthermore, they argue that power promotes optimism and lowers one’s
awareness of the downside of risks, which increases risky behavior. Power reduces the
third aim of the present research is to show that the interaction between leader power and legitimacy on risk-taking is mediated by risk-perception.
The present research offers two important theoretical contributions to the existing literature on power and risk-taking. First, this study examines the relationship between power and risk-taking in an organizational field setting. To date, the relationship between power and risk-taking has not been examined in the field. While informative, former studies about the relationship between power and risk-taking are based on experiments with students who choose to participate in exchange for either money or course credits (Anderson & Galinsky, 2006; Ronay & Hippel, 2010). Therefore, the generalizability of previous studies is low (Al-Ubaydli & List, 2013). Consequently, the purpose of the current study is to explore the relationship between power and risk-taking, by replicating the findings of former research outside of the laboratory, within an organizational field setting. This helps to assess whether actual power-holders would engage in more risk-taking behavior, instead of using a
manipulation of power.
Second, beyond replication, we seek to extend prior research by examining risk perception as a mediator of the interaction between leader power and legitimacy on risk-taking. The existing literature on risk-taking suggests that the relationship between power and risk-taking is more complex than the effect that power increases risk-taking (Ronay & Hippel, 2010; Maner, Gailliot, Butz, & Peruche, 2007). The present research offers new information regarding the antecedents of risk-taking behavior, which contributes to a better understanding about why and when leaders engage in risk-taking behavior. Specifically, the present research includes both perceived legitimacy of power and risk-perception to explain why and when power leads to more risk-taking. In doing so, the present research will add new information about the influence of both legitimacy and risk-perception on the relationship between power and risk-taking.
excessive risk-taking by leaders organizations should carefully consider who they give power and how much.
THEORY Leader Power and Risk-Taking
In the existing literature on power different types of power and applicable definitions can be found (French, & Raven, 1959; Goldhamer, & Shils, 1939). The most conforming definition of power in the existing relevant literature defines power as ‘the capacity to influence others via the control of resources and the ability to administer rewards and punishments’ (Anderson, & Galinsky, 2006: 512). Important in this definition of power by Anderson and Galinsky (2006), is that power implies control over resources. This is important since control of resources enables powerful people to act. Moreover, leader power is the ability of a leader to influence followers to do things they normally would not do (Atwater & Yammarino, 1996).
According to the approach inhibition theory of power people with power behave different than people without power (Anderson & Berdahl, 2002; Keltner et al, 2003; oskowitz, 2004). Specifically, this theory proposes that high power is associated with increased approach behavior, whereas low power is associated with increased inhibition behavior. The approach inhibition theory argues that power is associated with more access to resources, such as financial resources, physical comforts, and social resources (Kelner et al. 2003; Smith & Bargh, 2008). These resources enable powerful people to act and lead to increased levels of approach behavior, such as looking for rewards and opportunities
(Galinsky et al., 2003; Ferguson, Ormiston, & Moon, 2010). Powerless people, on the other hand, have less access to resources and tend to show inhibition behavior, such as attention to threats. Furthermore, having power reduces the awareness of constraints and enables people to act without interference from others (Keltner et al., 2003; Anderson & Berdahl, 2002; Whitson, Liljenquist, Galinsky, Magee, Gruenfeld, & Cadena, 2013).
To date, first experimental research has shown that power increases risk-taking behavior. Anderson and Galinsky (2006) for instance, showed that power increases actual risk-taking behavior and preferences for riskier options. In one of their experiments
participants were randomly assigned to a high power condition, a neutral condition or a low power condition (Anderson & Galinsky, 2006). Then all participants were asked to choose between two alternatives, in which one was more risky than the other. Results showed that participants assigned to high power conditions were drawn by the potential rewards inherent in the riskier option, and preferred the riskier alternative more compared to participants assigned to neutral and low power conditions.
Similarly, Inesi (2010) studied the effect of power for loss aversion with four
experiments. In these experiments participants were randomly assigned to a high power, a low power, or a neutral condition. Then all participants were presented with a choice scenario. Results showed that participants assigned to the high power condition were less loss averse and thus less risk averse regarding their choice patterns compared to participants assigned to neutral and low power conditions. Inesi (2010) suggests that powerful people perceive negative outcomes as less painful and therefore engage in more risk-taking.
Based on the approach inhibition theory and former research about the relationship between power and taking, we expect that leader power is associated with increased risk-taking. In accordance, the present research aims to replicate these empirical findings in an organizational field setting and hypothesizes the following:
Hypothesis 1: Leader power is associated with increased risk-taking.
The Moderating Role of Legitimacy
2008). In contrast, Smith, Jost, and Vijay (2008) suggest that power is perceived as
illegitimate when no explanation for power is given or when the assigned level of power is based on nepotism. In this research legitimacy means that leaders believe that they have the right to expect that employees follow their orders because they are the boss and because they have information that their employees do not have (Frost, & Stahelski’s, 1988).
The approach inhibition theory gives several explanations why legitimacy moderates the relationship between power and approach (Keltner et al., 2003; Lammers et al. 2008). The theory proposes that perceived illegitimacy implies instability in the power relationship. Specifically, powerful people feel threatened and are afraid to lose their power, which activates inhibition behavior (Kelner et al. (2003). Moreover, the relationship between a leader and followers changes from cooperation to resistance when power is perceived as illegitimate. In a situation of illegitimacy, powerful people may not want to use their power undeserved. In contrast, illegitimate powerless may try to change the situation and use
approach behavior. Finally, illegitimate power influences the emotions of the leader. Powerful people who perceive their power as illegitimate may feel fearful, they are afraid to lose their power. On the other hand, illegitimate powerless may feel angry, and show more approach behavior. Except for the suggestion that these emotion influence approach or inhibition tendencies, they also influence risk perceptions (Lerner & Keltner, 2001). According to Lerner and Keltner (2001) fearful people tend to be more risk-averse, while angry people tend to take more risks. In a situation of illegitimacy, powerful people may feel fearful and
therefore take fewer risks. Therefore, we suggest that legitimacy influences the relationship between leader power and risk-taking.
To date, first research has shown that legitimacy moderates the relationship between power and approach behavior (Lammers et al., 2008; Smith, Jost, & Vijay, 2008). Lammers et al. (2008) and Smith, Jost, and Vijay (2008), showed that power leads to more approach behavior when this power is perceived as legitimate. However, when power is perceived as illegitimate, the powerless people show as much approach as the powerful, or even more.
In four experiments, Lammers et al. (2008) randomly assigned participants to one of the following four conditions: legitimate powerful, legitimate powerless, illegitimate powerful or illegitimate powerless. Then participants received different instructions depending on their condition. Results showed that illegitimate powerless engaged in more approach behavior than illegitimate powerful. Results even indicate that under the condition of illegitimacy, the difference between powerful and powerless regarding risk preferences disappears.
Similarly, Smith, Jost, and Vijay (2008) randomly assigned participants to a high- or low power role. Then participants were told the role assignments were either based on
legitimate or illegitimate reasons, or they were given no explanation at all. The results showed that participants who were given an illegitimate reason engaged in the same amount of
approach behavior regardless of their power role.
Overall, research shows that illegitimacy changes the effect of power on approach (Lammers et al., 2008; Keltner et al., 2003; Smith, Jost, & Vijay, 2008), and this suggests that illegitimacy changes the effect of power on risk-taking. We expect that people’s focus
changes when they perceive their power as illegitimate. Illegitimate powerful people are afraid to lose their power, and are therefore focused on threats and avoiding losses. This change in focus makes it less likely that illegitimate powerful people engage in risk-taking behavior. Therefore, we expect that power is associated with increased risk-taking but only when leaders perceive their power as legitimate. The present research hypothesizes the following:
Hypothesis 2: Leader power is associated with increased risk-taking when power is perceived as legitimate, but not when power is perceived as illegitimate.
The Mediating Role of Risk Perception
Finally, the present study proposes that risk perception mediates the interaction between leader power and legitimacy of that power on risk-taking. Consistent with the approach inhibition theory, Hecht (2013) suggests that power increases approach behaviors which are associated with optimism. More specifically, power is associated with a more optimistic risk perception as power changes one’s focus from potential losses to potential payoffs. Furthermore, research indicates that an optimistic risk-perception encourages
powerful people to take more risks (Hecht, 2013). If, however, powerful people perceive their power as illegitimate they are afraid to lose their power and they focus on potential losses (Lammers et al., 2008), which leads to a less optimistic risk perception and ultimately less risk-taking.
Risk perception is defined as a subjective evaluation of the probability of a certain outcome and its consequences (Sjöberg, Moen, & Rundmo, 2004). So risk perception includes both an evaluation of the probabilities and the consequences of a certain outcome. An
lower probability of negative consequences and are therefore more optimistic in their risk perceptions (Anderson & Galinsky, 2006).
To date, first experimental research has shown that risk perception mediates the relationship between power and risk-taking. Anderson and Galinsky (2006) studied whether powerful people engage in more risk-taking because they have a more optimistic risk perception. Participants were randomly assigned to either a high or low power condition. Then all participants were presented with a certain scenario which involved a risky situation. The participants had to indicate the likelihood of actually engaging in this risky behavior (i.e., engaging in unprotected sex) and they had to indicate their perceptions of the risky behavior. Results showed that participants assigned to the high power condition perceived the risky behavior as less dangerous and therefore engaged in more actual risk-taking behavior compared to participants assigned to low power conditions. Based on five experiments Anderson and Galinsky (2006) argue that power leads to a more optimistic risk perception and this increases people’s tendency to engage in risk-taking behavior.
Overall, research indicates that powerful people have a more optimistic risk perception and therefore take more risks (Anderson & Galinsky, 2006). However, as mentioned above illegitimacy would break the relationship between power and risk perception. Therefore, the present research hypothesizes the following:
METHOD Data Collection
The data collection method that was used in this study was a questionnaire. The questionnaire was conducted in an organizational field setting. The data was collected by several students to be able to gather as much data as possible in a short amount of time. Therefore, the questionnaire consists of several parts of which only a part is used to contribute to this study. The questions that contribute to this study are some general questions about the respondent, such as age and gender. The other questions that contribute to this study are directly related to the main variables leader power, legitimacy, risk perception, and risk-taking. In total, we contacted 110 leaders, of which 83 completed the questionnaire (a
response rate of 75.5%). Three participants did not answer the risk-taking and risk perception questions and were therefore excluded from the analysis. Therefore, there are 80 participants included in the analyses. Of these 80 participants 57.5 % is male and 42.5 % is female, with a mean age of 43.64 (SD = 11.11). These individuals are all leaders who work in a wide range of organizations from a variety of sectors (e.g., transport, government, education) and who hold different functions. Their average organizational tenure was 14.24 years (SD = 10.36) and their average job tenure was 7.31 years (SD = 7.85). The majority of participants are Dutch and they have permanent employment contracts. Of the 80 participants, 6.3 % highest level of education achieved is lower general secondary education, 13.8 % achieved higher general secondary education, 2.5 % achieved pre-university education, 41.3 % achieved higher vocational education, and 36.3 % achieved university. Furthermore, their incomes range from 0 - 10.000 Euro (3.8 %) to 80.000 Euro or more (20 %).
Measurement
The measurement items of all variables can be found in Appendix A. Moreover, unless indicated otherwise all items were measured using a seven-point rating scale, ranging from 1 (strongly disagree) to 7 (strongly agree).
Alpha of .80, which indicates a high reliability. Therefore these four items were combined to form one power measure, called composite power.
Legitimacy was measured with three slightly adjusted items taken from Frost and Stahelski’s (1988). These items include, ‘I expect that my employees follow my orders’, ‘I expect my employees to do what I say, because I am the boss’ and, ‘I expect my employees to do what I say, because I possess more information’. The reliability analysis showed a
Cronbach’s Alpha of .74, which indicates a high reliability.
Legitimacy was also measured using a measurement from Lammers et al. (2010). The item that was used is, ‘Given my qualities, I deserve my current position in the organization’. Including this item in the reliability analysis showed a Cronbach’s Alpha of .70, which was just high enough to be able to combine all four items. Therefore, these four items were combined to form one legitimacy measure, which is used in the analysis.
Risk-taking was measured with eight items taken from Weber, Blais and Betz’s (2002) Domain-specific Risk-attitude Scale (DOSPERT scale). Financial risk-taking is one of the five domains included in the DOSPERT scale. Risk-taking was measured by asking participants to indicate their likelihood of engaging in each activity or behavior. Sample items included, ‘Investing 10% of your annual income in a moderate growth diversified fund’ and, ‘Betting a day’s income at the horse races’. These items were measured using a seven-point Likert scale, ranging from 1 (very unlikely) to 7 (very likely). The reliability analysis showed a Cronbach’s Alpha of .74, which indicates a high reliability. Therefore, these eight items were combined to form one risk-taking measure called financial risk-taking.
Risk-taking was also measured using Mandrik and Bao’s (2005) 6-item measure. Sample items included, ‘I do not feel comfortable about taking chances’ and, ‘Before I make a decision, I like to be absolutely sure how things will turn out’. The reliability analysis showed a Cronbach’s Alpha of .76. Therefore, these six items were combined to form one risk-taking measure called general risk-taking.
The reliability analysis of both of the above measures of risk-taking combined showed a Cronbach’s Alpha of .67. As this is considerably lower than the Cronbach’s Alpha’s of the separate measures we decided to test our hypotheses based on both of the above measures separately.
seven-point Likert scale, ranging from 1 (not at all risky) to 7 (extremely risky). The
reliability analysis showed a Cronbach’s Alpha of .86, which indicates a very high reliability. Therefore, these eight items were combined to form one risk perception measure.
RESULTS Assumptions
To be able to test the hypothesis it is necessary to check the assumptions which need to be fulfilled for a regression analysis. First of all we checked for missing values. For the 80 participant that are included in the hypothesis testing there are no missing values, except for two participants who did not fill in their age. Second, both the independent and dependent variables are measured at a continuous level, and scatterplots indicate a linear relationship between the independent variables and the dependent variable. Furthermore, we check for independence of observation by using the Durbin-Watson statistic. The value of the Durban Watson test indicates that the independent and dependent variables are independent of each other (D = 1.93). These assumptions were easily met, the following two assumptions, however, are a bit more complicated.
The first assumption that was not directly met was the assumption regarding outliers in the data. For financial risk-taking there are two extreme outliers with more than three standard deviations from the mean. As these outliers have a big impact on the results, we perform the regression analyses two times, ones without the outliers and ones including the outliers. The top part of tables 2 to 4 shows results of the analyses without the outliers and the bottom part of tables 2 to 4 shows results including the outliers. The outliers are excluded in the
correlation table.
Second, we looked at the normality of the dependent variables. According to the Shapiro-Wilk test general risk-taking is approximately normally distributed. On the other hand, the results of the Shapiro-Wilk tests are significant for financial risk-taking (W = .86, p < .01) and for risk perception (W = .88, p < .01), which indicates that this data is not normally distributed. However, the kurtosis and skewness values of both variables indicate that there is no reason to assume non-normality (West, Finch, & Curran, 1995). The skewness of financial risk-taking is .63 (SE = .27) and the kurtosis is -.95 (SE = .54). The skewness of risk
perception is -1.4 (SE = .27) and the kurtosis is 2.31 (SE = .54). Descriptive Statistics
variables indicate that the participants score relatively high on power, which makes sense while all participants are leaders.
Moreover, Table 1 includes the correlations between all study variables. An interesting finding is that there is no significant correlation between general risk-taking and financial risk-taking, R = .15, p = .21. This suggests that general risk-taking and financial risk-taking are not related, which is peculiar since both variables should measure risk-taking. Results show that there is a significant positive relationship between the composite power measure and general risk-taking, R = .25, p < .05. This indicates that an increase in power is related to an increase in general risk-taking. However, these results do not take into account the effects of control variables. Moreover, there is no significant relationship between the remaining leader power variables and financial risk-taking or general risk-taking. This implies that how leader power and risk-taking are measured influences the results. Therefore, the separate measurements of both leader power and risk-taking are included in the testing of the hypotheses.
TABLE 1
Means, Standard Deviations, and Correlations
*p < .05 **p < .01
Hypotheses Testing
To test the first and second hypothesis, six moderated regression analyses are conducted (using standardized predictor variables). We start with the results regarding the first hypothesis that leader power is associated with increased risk-taking. Results of the regression analysis regarding financial risk-taking were not significant for subjective power, B = -.10, t = -1.65, p = .10 (see table 2). Table 3 shows that there is no significant relationship between control over outcomes and financial risk-taking when excluding the outliers of financial risk-taking, B = .07, t = 1.08, p = .28. However, there is a significant positive relationship between control over outcomes and financial risk-taking when including the outliers, B = .16, t = 2.05, p = .04. Lastly, table 4 indicates that there is no significant
relationship between composite power and financial risk-taking, B = .01, t = .09 p = .93. Since the results show only one significant positive relationship (when including all outliers), this indicates that there is no sufficient support for the first hypotheses. Moreover, table 5, 6, and 7 depict results of the regression analysis regarding general risk-taking. Results show that there is no significant relationship between power and general risk-taking using the subjective (B = .12, t = 1.06, p = .29), control over outcomes (B = -.03, t = -.27 p = .79), and composite power measure (B = .17, t = 1.37, p = .17). The control variables gender and income are included in
these analyses; however they do not have a significant influence. Results show that there is no significant support for the first hypothesis. Therefore, hypothesis one is rejected. However, the overall results do suggest that leader power might be positively associated with risk-taking. Although most beta values are not significant, most are in the right direction and suggest a positive relationship between power and risk-taking.
Moreover, Tables 2 to 7 depict results regarding the second hypothesis that leader power is associated with increased risk-taking when power is perceived as legitimate, but not when power is perceived as illegitimate. Results show that there are no significant interactions between power and legitimacy on financial risk-taking using the subjective (B = .05, t = .61, p = .55), control over outcomes (B = .04, t = .65, p = .52), and composite power measure (B = -.08, t = -1.30, p = .20). This implies that there is no support for the second hypothesis. Furthermore, results indicate that there are no significant interactions between power and legitimacy on general risk-taking using the subjective (B = .10, t = .51, p = .51), control over outcomes (B = .09, t = .76, p = .45), and composite power measure (B = .05, t = .47, p = .64). The control variables gender and income are included in these analyses; however they do not have a significant influence. Overall, results indicate that there is no significant interaction between leader power and legitimacy on risk-taking. Therefore, hypothesis two is rejected.
TABLE 2:
Regression analysis Subjective Power and Financial Risk-taking Model 1 Model 2 B SE B SE Subjective Power -.10 (.07) -.11 (.07) Legitimacy .04 (.06) .02 (.07) Subjective Power x Legitimacy .05 (.09) Change R square .04 .01 B SE B SE Subjective Power -.09 (.08) -.09 (.08) Legitimacy .03 (.08) .03 (.09) Subjective Power x Legitimacy .01 (.11) Change R square .02 .00
TABLE 3
Regression analysis Control over Outcomes and Financial Risk-taking Model 1 Model 2
B SE B SE
Control over Outcomes .07 (.07) .08 (.07)
Legitimacy .04 (.06) .04 (.06)
Control over Outcomes x Legitimacy
.04 (.07)
Change R square .02 .01
B SE B SE
Control over Outcomes .16* (.08) .17* (.08)
Legitimacy .03 (.08) .03 (.08)
Control over Outcomes x Legitimacy
.04 (.08)
Change R square .05 .00
Notes. The top part of the table excludes the outliers and the bottom part includes the outliers. *p < .05
TABLE 4
Regression analysis Composite Power and Financial Risk-taking Model 1 Model 2 B SE B SE Composite Power .01 (.07) .00 (.07) Legitimacy .04 (.07) .04 (.07) Composite Power x Legitimacy -.08 (.06) Change R square .04 .01 B SE B SE Composite Power .03 (.08) .02 (.08) Legitimacy .03 (.08) .04 (.08) Composite Power x Legitimacy -.13 (.08) Change R square .00 .03
TABLE 5
Regression analysis Subjective Power and General Risk-taking Model 1 Model 2 Model 3
B SE B SE B SE Gender -.17 (.12) -.16 (.12) -.14 (.12) Income .20 (.12) .16 (.12) .19 (.13) Subjective Power .12 (.11) .12 (.11) Legitimacy -.08 (.12) -.09 (.12) Subjective Power x Legitimacy .10 (.16) Change R square .09* .02 .01 *p < .05 TABLE 6
Regression analysis Control over Outcomes and General Risk-taking Model 1 Model 2 Model 3
B SE B SE B SE
Gender -.17 (.12) -.19 (.12) -.19 (.12)
Income .20 (.12) .19 (.12) .20 (.13)
Control over Outcomes -.03 (.12) -.02 (.12)
Legitimacy -.08 (.12) -.09 (.12)
Control over Outcomes x Legitimacy
.09 (.12)
Change R square .09* .01 .01
TABLE 7
Regression analysis Composite Power and General Risk-taking Model 1 Model 2 Model 3
B SE B SE B SE Gender -.17 (.12) -.13 (.13) -.12 (.13) Income .20 (.12) .16 (.12) .16 (.12) Composite Power .17 (.12) .17 (.12) Legitimacy -.11 (.12) -.11 (.12) Composite Power x Legitimacy .05 (.11) Change R square .09* .03 .00 *p < .05
TABLE 8
Regression analysis Subjective Power and Risk Perception Model 1 Model 2 Model 3
B SE B SE B SE Age .35* (.14) .37* (.14) .37* (.15) Subjective Power .01 (.14) .01 (.14) Legitimacy .13 (.14) .12 (.14) Subjective Power x Legitimacy .03 (.18) Change R square .08* .01 .00 *p < .05 TABLE 9
Regression analysis Control over Outcomes and Risk Perception Model 1 Model 2 Model 3
B SE B SE B SE
Age .35* (.14) .38** (.14) .37** (.14)
Control over Outcomes -.14 (.14) -.10 (.14)
Legitimacy .13 (.14) .13 (.14)
Control over Outcomes x Legitimacy
.16 (.14)
Change R square .08* .02 .02
*p < .05 **P < .01
TABLE 10
Regression analysis Composite Power and Risk Perception Model 1 Model 2 Model 3
DISCUSSION Implications
The present research examined the roles of leader power, the legitimacy of that power, and risk perception for leaders’ risk-taking in organizational work settings. The aim of this research was to find more evidence about the relationship between leader power and risk-taking, and to add information about how the legitimacy of power and risk-perception influence this relationship. The present research hypothesized that leader power is associated with increased risk-taking. However, even though we used several measures of power and risk-taking, we only found one significant relationship between leader power and risk-taking. This significant relationship was only found when including all outliers. Research suggests that there are strong arguments for the removal of outliers, since they tend to significantly influence results and the accuracy of the analyses (Osborne & Overbay, 2004). When removing the outliers, we found no significant relationship between leader power and risk-taking. Moreover, the present research included the role of legitimacy as a boundary condition for understanding the relationship between leader power and risk-taking. However, we found no significant interactions between leader power and legitimacy on risk-taking. Furthermore, we found no significant interactions between leader power and legitimacy on risk perception.
Although we found no significant relationship between leader power and risk-taking, the results do suggest that power is positively associated with risk-taking. As expected, most beta values are positive which suggest that an increase in leader power is associated with an increase in risk-taking. Specifically, results indicate a trend in the hypothesized direction. However, it might be that the sample is too small to indicate significance. As the results already suggest a positive relationship, a bigger sample size would help to find a more accurate p-value (Johnson, 1999) and possible acceptance of the hypothesis. Acceptance of the hypothesis would suggest that power leads to increased risk-taking. However, our results only provide a hint of evidence to suggest such an association between power and risk-taking, as the results did not reach significance.
There are several explanations why our results deviate from what we expected. First, there is a clear difference between our research and former research. Former research used experiments to explain the relationship between power and risk-taking and the mediating effect of risk-perception while the current research is conducted in an organizational field setting (Anderson & Galinsky, 2006; Ronay & Hippel, 2010). It might be that the
high-power group instead of using natural high-power differences. This enables researchers to compare manipulated high and low power individuals, but it lacks an equal power group and in reality power differences can be much smaller. Moreover, actual power holders are influenced by situational variables and possible boundary conditions (not included in the experiments) which might lead to less risk-taking (Moskowitz, 2004). Therefore, manipulating power in experiments might not lead to the same risk-taking behavior as the risk-taking behavior of actual power-holders. In practice this means that actual power-holders might not always engage in more risk-taking, which leads to an insignificant relationship between leader power and risk-taking. Moreover, the manipulation of power used in experiments might not
influence an individual’s risk perception in the same way as with actual power holders. This is consistent with Hovland (1959), who argue that differences in results can be caused by
differences in the research design itself using either an experimental or survey method. Second, other research can explain why our results deviate from what we expected. The prospect theory suggests that some powerless individuals take more risks since they are in a so called domain of loss and they have less to lose, while some powerful people are more risk averse since they are in a so called domain of gains and they have a lot to lose
(McDermott, 1998, 2004; Maner et al., 2007). This implies that power is sometimes
associated with less risk-taking. Moreover, research suggests that the influence of power on risk-taking is depending on variables both within the person and the situation (Maner, Gailliot, Butz, & Peruche, 2007). In practice this means that the impact of power on risk-taking may depend on several variables, and that power is not always associated with increased taking. On the other hand, our present findings with regard to financial risk-taking suggest that people are in general risk averse regardless of their power status, which is consistent with Tetlock (2002). Tetlock (2002) suggest that both powerful and powerless people are afraid to fail and are therefore risk averse, which indicates that power does not influence risk-taking behavior.
Limitations and Further Research
outliers as it had a considerable effect on the results. Both the small sample size and the presence of outliers have made it more complicated to find significant results.
Furthermore, this study includes some measurement problems. Specifically, the correlations depicted in Table 1 indicate that the two measurements of risk-taking are not associated, which is striking since they should both measure risk-taking. However, this can be explained because one of the measures uses a domain-specific measurement of risk-taking (i.e., financial taking) while the other measurement is more general (i.e., general risk-taking). It might be that individuals’ financial risk-taking behavior and individuals’ general risk-taking behavior is too different to be associated. The results indicate that individuals take less financial risks compared to general risks. This suggests that there is a different
relationship between leader power and financial risk-taking and leader power and general risk-taking. Moreover, it might be that leader power is only associated with a certain type of risk-taking behavior.
Future research should replicate this research with a few adjustments. First, further research should include a bigger sample to increase the generalizability and to increase the chance of finding significant results. Moreover, future research should include more or different measurements of both power and risk-taking. Including different measures would help to gain a better understanding of the relationship between leader power and risk-taking as our results suggest that the chosen measurement has a big influence on the results. Lejuez et al. (2002) argue that using a behavioral measure of risk-taking (i.e., BART) in addition to a self-assessment improves the overall assessment of a wide range of actual risk-taking behaviors. Research suggests that the BART is a valid and useful laboratory instrument to measure actual risk-taking behavior (Lejuez et al, 2002; Lejuez, Aklin, Daughters, Zvolensky, Kahler, & Gwadz, 2007). These adjustments would eliminate the limitations of the current research with regard to the small sample size and the measurement problems.
Moreover, further research may include other potential moderators for the relationship between leader power and risk-taking. For example, testosterone could moderate the
personal variables such as power motivation can moderate the relationship between power and risk-taking. Further research should examine these moderators, as these moderators might help to explain how power differences influence risk-taking behavior.
Conclusion
In conclusion, the present study suggests that leader power is associated with increased risk-taking behavior, although this effect did not reach significance. There is not enough support to actually demonstrate a positive relationship between leader power and risk-taking. Moreover, the present findings have found no support with regard to the moderating role of legitimacy or the mediating role of risk-perception. However, it might be that such a relationship does not exist in practice or there are other variables that influence this
REFERENCES
Al-Ubaydli, O., & List, J. A. 2013. On the Generalizability of Experimental Results in Economics: With a response to Camerer. Working paper no. 19666, National Bureau of Economic Research, Cambridge, MA.
Anderson, C., & Berdahl, J. L. 2002. The experience of power: Examining the effects of power on approach and inhibition tendencies. Journal of Personality and Social Psychology, 83(6): 1362-1377.
Anderson, C., & Galinsky, A. D. 2006. Power, optimism, and risk-taking. European Journal of Social Psychology, 36: 511-536.
Atwater, L. E., & Yammarino, F. J. 1996. Bases of power in relation to leader behavior: A field investigation. Journal of Business and Psychology, 11(1): 197-217.
Boyer, T. W. 2006. The development of risk-taking: A multi-perspective review. Developmental Review, 26(3): 291-345.
Bromiley, P. 1991. Testing a causal model of corporate risk taking and performance. Management Journal, 34(1): 37-59.
Byrnes, J. P., Miller, D. C., & Schafer, W. D. 1999. Gender difference in risk taking: a meta-analysis. Psychological Bulletin, 125(3): 367-383.
Ferguson, A. J., Ormistion, M. E., & Moon, H. 2010. From approach to inhibition: The influence of power on responses to poor performers. Journal of Applied Psychology, 95(2): 305-320.
French, J., & Raven, B. 1959. The Bases of Social Power. In Cartwright, D. Studies in Social Power. Ann Arbor, Michigan: Institute for Social Research.
Frost, D., & Stahelsk, A. 1988. The systematic measurement of French and Raven’s bases of social power in workgroups. Journal of Applied Social Psychology, 18(5): 375-389. Galinsky, A. D., Gruenfeld, D. H., & Magee, J. C. 2003. From power to action. Journal of
Personality and Social Psychology, 85: 453–466.
Goldhamer, H., & Shils, E. A. 1939. Types of power and status. American Journal of Sociology, 45: 171–182.
Guo, L., Jalal, A., & Khaksari, S. 2015. Bank executive compensation structure, risk taking and the financial crisis. Review of Quantitative Finance and Accounting, 45: 609-639.
Hovland, C. I. 1959. Reconciling conflicting results derived from experimental and survey studies of attitude change. American Psychologist, 14(1): 8-17.
Hurd, M. D., & Rohwedder, S. 2010. Effects of the financial crisis and great recession on American households. Working Paper No. 16407, National Bureau of Economic Research, Cambridge, MA.
Inesi, M. E. 2010. Power and loss aversion. Organizational Behavior and Human Decision Processes, 112(1): 58-69.
Keltner, D., Gruenfeld, D. H., & Anderson, C. 2003. Power, approach, and inhibition. Psychological Review, 110(2): 265-284.
Lammers, J., Galinsky, A. D., Gordijn, E. H., & Otten, S. 2008. Illegitimacy moderates the effects of power on approach. Psychological Science, 19(6): 558-564.
Lammers, J., Galinsky, A. D., Gordijn, E. H., & Otten, S. 2012. Power increases social distance. Social Psychological and Personality Science, 3(3): 282-290.
Lammers, J., Stapel, D. A., Galinsky, A. D. 2010. Power increases hypocrisy moralizing in reasoning, immorality in behavior. Psychological Science, 21(5): 737-744.
Lammers, J., Stoker, J. I., & Stapel, D. A. 2010. Power and behavioral approach orientation in existing power relations and the mediating effect of income. European Journal of
Social Psychology, 40: 543-551.
Larcker, D. F., Ormazabal, G., Tayan, B., & Taylor, D. J. 2014. Follow the Money: Compensation, Risk, and the Financial Crisis. Working paper no. 14-34, Stanford University Graduate School of Business, Stanford, CA.
Lee, A. S., & Baskerville, R. L. 2003. Generalizing Generalizability in Information Systems Research. Information Systems Research, 14(3): 221–243.
Lejuez, C. W., Aklin, W. Daughters, S., Zvolensky, M., Kahler, C., & Gwadz, M. 2007. Reliability and validity of the youth version of the Balloon Analogue Risk Task (BART-Y) in the assessment of risk-taking behavior among inner-city adolescents. Journal of Clinical Child & Adolescent Psychology, 36(1): 106-111.
Lejuez, C. W., Read, J. P., Kahler, C. W., Richards, J. B., Ramsey, S. E., Stuart, G. L., Strong, D. R., & Brown, R. A. 2002. Evaluation of a behavioral measure of risk taking: The Balloon Analogue Risk Task (BART). Journal of Experimental Psychology, 8(2): 75-84.
Lerner, J.S., & Keltner, D. 2001. Fear, anger, and risk. Journal of Personality and Social Psychology, 81: 146–159.
Management Science, 36(4): 422-435.
Mandrik, C. A., & Bao, Y. 2005. Exploring the concept and measurement of general risk aversion. Advances in Consumer Research, 32: 531-539.
Maner, J. K., Gailliot, M. T., Butz, D. A., & Peruche, B. M. 2007. Power, risk, and the status quo: Does power promote riskier or more conservative decision-making? Personality & Social Psychology Bulletin, 33: 451–462.
McClelland, D. C., & Watson, R. I. 1973. Power motivation and risk-taking behavior. Journal of Personality, 41(1): 121–139.
McDermott, R. 1998. Risk-taking in international politics: Prospect theory in American foreign policy. Ann Arbor: The University of Michigan Press.
McDermott, R. 2004. Prospect theory in political science: gains and losses from the first decade. Political Psychology, 25(2): 289-312.
Moskowitz, D. S. 2004. Does elevated power lead to approach and reduced power to inhibition? Comment on Keltner, Gruenfeld, and Anderson (2003). Psychological Review, 111(3): 808-818.
Naldi, L., Nordqvist, M., Sjöberg, K., & Wiklund, J. 2007. Entrepreneurial orientation, risk taking, and performance in family firms. Family business review, 20(1): 33-47. Nelson, S. C., & Katzenstein, P. J. 2013. Uncertainty, risk, and the financial crisis of 2008.
International Organization, 68(2): 361-392.
Osborne, J. W., & Overbay, A. 2004. The power of outliers (and why researchers should always check for them). Research & Evaluation, 9(6): 1-12.
Ronay, R., & Hippel, W. 2010. Power, testosterone, and risk-taking. Journal of Behavioral Decision Making, 23(5): 473-482.
Sjöberg, L., Moen, B. E., & Rundmo, T. 2004. Explaining risk perception. An evaluation of the psychometric paradigm in risk perception research. Rotunde, 84: 1-39.
Slovic, P., Peter, E., Finucane, M. L., & MacGregor, D. G. 2005. Affect, risk, and decision making. Health Psychology, 24(4): 35-40.
Smith, P. K., & Bargh, J. A. 2008. Nonconscious effects of power on basic approach and avoidance tendencies. Social Cognition, 26(1): 1-24.
Smith, P. K., Jost, J. T., & Vijay, R. 2008. Legitimacy Crisis? Behavioral approach and inhibition when power differences are left unexplained. Social Justice Research, 21(3): 358-376.
Tetlock, P. E. 2002. Social-functionalist metaphors for judgment and choice: The politician, theologian, and prosecutor. Psychological Review, 109: 451-472.
Turner, C., & McClure, R. 2003. Age and gender differences in risk-taking behaviour as an explanation for high incidence of motor vehicle crashes as a driver in young males. Injury Control and Safety Promotion, 10(3): 123-130.
Weber, E. U., Blais, A. R., & Betz, N. 2002. A domain-specific risk-attitude scale: Measuring risk perceptions and risk behaviors. Journal of Behavioral Decision Making, 15: 263-290.
APPENDIX Leader Power
Control over outcomes
What is your current management position? Subjective power
Indicate whether you feel to be at the bottom or the top of the power hierarchy of your organization?
Composite power
1. How much power do you have? 2. How much influence do you have? 3. How much status do you have? 4. How much authority do you have?
Legitimacy
I expect that my employees follow my orders
I expect my employees to do what I say, because I am the boss
I expect my employees to do what I say, because I possess more information Given my qualities, I deserve my current position in the organization
Risk-taking and Risk Perception
Items measuring risk-taking and risk perception 1. Betting a day’s income at the horse races.
2. Investing 10% of your annual income in a moderate growth diversified fund. 3. Betting a day’s income at a high-stake poker game.
4. Investing 5% of your annual income in a very speculative stock.
5. Betting a day’s income on the outcome of a sporting event (e.g. baseball, soccer, or football)
6. Investing 10% of your annual income in a new business venture.
Items measuring risk-taking
1. I do not feel comfortable about taking chances. 2. I prefer situations that have foreseeable outcomes.
3. Before I make a decision, I like to be absolutely sure how things will turn out. 4. I avoid situations that have uncertain outcomes.
5. I feel comfortable improvising in new situations.
