«J Bus Econ (2015) 85:793–815 DOI 10.1007/s11573-014-0753-1 ORIGINAL PAPER New insights on CEO charisma attribution in companies of different sizes ...»
We applied regression techniques to test Hypothesis 2 assuming that prior company performance leads to higher CEO charisma attribution as well the potential moderation of company size and ownership structure. We assessed the interrater reliability for the companies in which two or more raters assessed the charisma attribution and/or the performance measures by calculating rwg indices (James et al.
1984), which should be above 0.70 for acceptable aggregation. That was the case for all measures in our study.
Table 3 shows the means, standard deviations, and intercorrelations of the study variables. As expected, prior company performance in the previous year was positively related to CEO charisma attribution (r = 0.25, p \ 0.01). Furthermore, family ownership showed a negative relation to charisma attribution (r = 0.32, p \ 0.01).
Table 4 indicates that the effect between prior company performance and CEO charisma attribution is positive and signiﬁcant as shown in step 2 of the regression results (B = 0.13, p \ 0.05), even when including control variables. Prior company performance alone explains ﬁve percent additional variance in CEO charisma attributes, which indicates a moderate effect size (Cohen 1988). These results further supported Hypothesis 1 with ﬁeld study data. In an additional step (Model 3 in Table 4), we included the two interaction terms in the model. In this model the company size moderation was signiﬁcant (B = 0.13, p \ 0.05), while the family ownership moderation was not signiﬁcant (B = -0.04, ns), thus supporting Hypothesis 2 and rejecting Hypothesis 3. The company size moderation term explained three percent additional variance in CEO charisma attribution, indicating a relevant effect size for moderation analyses (McClelland and Judd 1993).
Following Aiken and West’s (1991) recommendation for testing interactions we
n = 69 participants. Reliability values are in parentheses * p \ 0.05 (two-sided) ** p \ 0.01 (two-sided) *** p \ 0.001 (two-sided) Table 4 Results of regression analysis: the effects of prior company performance on CEO charisma attribution
3.6 3.5 3.4 3.3 3.2 3.1
Fig. 1 The moderation effect of company size in the relationship between prior company performance and CEO charisma attribution also performed a graphical inspection of the results as well as simple slope test for the signiﬁcant size moderation term. Figure 1 illustrates the moderating effect of company size. As proposed in Hypothesis 2, we observe a strong positive effect between prior company performance and CEO charisma attribution in large companies, whereas we see an almost zero relationship in small companies. These results are further substantiated through simple slope testing, where the high company size slope was positive and signiﬁcant (b = 0.45, p \ 0.01), while the low company size slope was not signiﬁcant (b = -0.03, ns.). Hence, we validated Hypothesis 2 in a ﬁeld-setting design. Finally, we inspected the variance inﬂation factors (VIFs) for our analysis, which ranged between 1.06 and 5.02 indicating that multicollinearity was not a major threat to our analysis.
Our study contributes to the literature on charismatic leadership by investigating and clarifying the relationship between prior company performance and CEO charisma attribution through a mixed-method design in a sample of companies with different sizes and ownership structures. We applied a laboratory and ﬁeld study demonstrating that prior company performance is an antecedent of CEO charisma attribution. Consequently, in contrast with prior studies, our study design allows us to draw conclusions about causal effects and simultaneously determine important
808 B. Michaelis et al.
implications for real business settings. Thus, we answer Waldman et al.’s (2001) call for more systematic empirical research in the area of charismatic leadership.
Our results challenge prior ﬁeld studies reporting charisma’s effects on company performance by ﬁnding that reverse causality is also possible. Thus, prior studies might be strongly biased because they failed to control for knowledge of prior company performance when measuring charisma. In light of our results, prior company performance can be added as another factor leading subordinates to perceive that their leaders have charisma.
Furthermore, our study answers the call for more research on organizational-level boundary conditions in the relationship between prior company performance and CEO charisma attributions. Acknowledging previous criticisms (Agle et al. 2006), we conducted a ﬁeld study to test whether company size (i.e., number of employees) and ownership structure (i.e., shares held by family members) can strengthen or weaken the impact of prior company performance on CEO charisma attribution. We found evidence that (1) the size of the company moderates the positive relationship between prior company performance and CEO charisma attributions, but the relationship is signiﬁcant only in large companies, not in small companies; (2) company ownership structure does not strengthen or weaken the impact of prior company performance on CEO charisma attribution. Thereby, we provide several important theoretical and practical implications for scholars and executives on how to transfer and apply the concept of charismatic leadership to companies of different sizes and ownership structures.
5.1 Theoretical implications
From a theoretical perspective, our results reveal that prior company performance could be added to the attributional theory of charismatic leadership framework (Conger and Kanungo 1987). Prior company performance could be considered another crucial element inducing subordinates to attribute leaders with charisma, in addition to leaders’ statements and actions (Weber 1947). Top executives might use this ﬁnding for persuading their subordinates to see them as charismatic leaders.
Second, our study illuminates organizational-level boundary conditions in the relationship between prior company performance and CEO charisma attribution. We build on previous research (Finkelstein and Boyd 1998; Hambrick and Finkelstein
1987) suggesting that senior executives in small companies have more managerial discretion and hence communicate more directly with their employees, demonstrating that company size plays a crucial contextual role for the attribution of CEO charisma. In small companies, employees are likely to interact directly with the CEO who is ‘‘actually running the business’’ (Garten 2001). In larger, professionally managed ﬁrms, CEOs are more distant, interact only indirectly with most employees, and thus have less-pronounced ﬁrm-wide inﬂuence (Lubatkin et al.
Thus, in small companies, prior company performance seems to be less important for judging charisma, as most employees can use their direct personal interactions to assess charisma. In contrast, large corporations are more distant institutions; CEOs tend to communicate indirectly through middle management, being more concerned New insights on CEO charisma attribution 809 with public relations, strategy (e.g., mergers and acquisitions), and ﬁnancial resources rather than with daily operations. Because employees lack personal contact with the CEO, they cannot evaluate the CEO’s charismatic qualities through interpersonal interactions. Consequently, they will substitute prior company performance for attributing charisma. Conﬁrming our hypothesis, prior company performance in large corporations had a stronger positive inﬂuence on CEO charisma attribution than it had in small companies. Thereby, our study is the ﬁrst demonstrating that CEOs’ managerial discretion and the associated different interactions with employees shape perceptions about CEO charismatic qualities.
Third, our results support theoretical elaborations regarding the impact of uncertain and turbulent environments on leadership processes (Shamir and Howell 1999). Large companies have typically complex organizational structures and have more uncertain and turbulent environments, so that subordinates are more likely to over-attribute prior company performance as a basis for attributing CEOs with charisma. Employees in small companies, characterized by more stable and predictable environments, are less likely to use prior performance to measure CEO charisma.
Surprisingly, our hypothesized interaction effect between prior company performance and ownership structure was not supported. Hypotheses 3 might have been rejected because of our rather crude measurement of ownership structure with a dummy variable. We might have found stronger effects if we had used information regarding CEO relations to ownership structure; that is, whether the CEO was the founder or at least part of the founding family. Therefore, we encourage future research to replicate our study with a more precise measurement of company ownership structure.
Our speciﬁc measurement might also explain why our data show a strong negative main effect between ownership structure and charisma attributions.
Family-owned companies apparently provide an environment less conducive for attributing charisma to the CEO, whatever the company’s prior performance. This ﬁnding indicates that our argument about different CEO foci in privately and family owned companies (i.e., CEOs in publically-owned companies focus on short-term and in family-owned businesses on long-term performance) might not hold true.
Moreover, the CEO’s high stature in family-owned companies (Jayaraman et al.
2000) and the associated managerial discretion (Hambrick and Finkelstein 1987) do not necessarily lead to a positive image of the CEO. Although speculative, beyond the role of ownerships structure, other moderating variables such as the CEO’s leadership style and behavior, the company’s lifecycle stage, or the organizational culture might crucially inﬂuence the relationship between prior company performance and the attribution of CEO charisma. We therefore encourage researchers to investigate family-owned ﬁrms in more detail to discover if and when CEO charisma attributions emerge.
5.2 Practical implications
attributional and thus perceptual phenomenon. Therefore, charisma lies not within the leader, but rather within the relationship between the leader and subordinates (Klein and House 1995). To maximize their inﬂuence, leaders must strive for high performance outcomes, especially in large corporations where prior company performance is crucial for a CEO to receive high charisma attribution. In small companies, in contrast, prior company performance seems to be less relevant. CEOs should, therefore, rather focus on demonstrating their charismatic behavior through direct social interaction.
Second, our study is consistent with the ‘‘substitutes for leadership’’ approach and demonstrates that certain organizational characteristics such as size and ownership structure signiﬁcantly inﬂuence leadership processes (Kerr and Jermier 1978), which has important implications for leadership and management development (Gray and Mabey 2005; Michaelis et al. 2012). We show that certain contextual conditions can induce a ‘‘need for leadership’’ in organizational members, whereas others ‘‘reduce the need for leadership.’’ We, therefore, encourage CEOs to be aware of environmental conditions and pay close attention to the operational environment.
Third, our ﬁndings imply that organizational characteristics are likely to inﬂuence subordinates to perceive CEO charisma and leadership capabilities. We recommend that classical leadership programs should train CEOs and top executives to go beyond increasing company performance and to also analyze and understand organizational situations and characteristics.
5.3 Limitations and future research implications
Despite our study’s methodological strengths in using an experimental design combined with a ﬁeld study, we acknowledge some limitations. Although an experimental design was necessary to investigate the effects studied (Mook 1983), the artiﬁcial setting causes some limitations. In Study 1, participants were asked to pretend that they were real-life subordinates in an imaginary company, but most were university students. They were unlikely to have had the experience necessary to evaluate organizational outcomes in a leadership context. However, other studies have successfully used vignettes (e.g., De Cremer and van Knippenberg 2002, 2004), and the method provides some advantages, for example in terms of efﬁciency and economy. Furthermore, the limitations of experimental designs are not unique to our study. Some researchers even criticize the artiﬁcial design of experiments in general, positing that they threaten external validity (Campbell et al. 1963).
However, we countered some concerns by cross-validating our main result (i.e., the prior company performance/CEO charisma attribution relationship) with ﬁeld data in Study 2.
Although we gathered as much information as possible on the companies in our sample, we could not determine CEO tenure. As time passes, charisma is unstable, routinized, and traditionalized (Weber 1947) and declines (Yukl 1999). Thus, company performance might have the greatest impact on CEO charisma attributions earlier in a CEO’s tenure. To investigate the inﬂuence of CEO tenure on the performance–charisma relationship, CEO charisma attributions and company New insights on CEO charisma attribution 811 performance could be measured immediately after the CEO is appointed. This design could help researchers understand the impact of prior company performance on CEO charisma attributions while reducing CEO tenure effects.
Another limitation stems from the sample size. A larger sample in the experimental setting would have strengthened the generalizability of our results, although the additional ﬁeld study increased the real-world generalizability of our ﬁndings.