Effectiveness of a moral and benevolent leader: Probing the
interactions of the dimensions of paternalistic leadership
Chun-Pai Niu, An-Chih Wang and Bor-Shuian Cheng Department of Psychology, National Taiwan University, Taipei, Taiwan
Paternalistic leadership has three dimensions: authoritarianism, benevolence and morality. Although it is impor-tant to understand how these dimensions interact to impact leadership effectiveness, previous studies have failed to identify consistent interaction effects of these dimensions, probably because of the high intercorrelations among the three dimensions. By manipulating the three dimensions independently in an experimental study (N= 265 Taiwanese employees), we found that: (i) benevolence and morality increased subordinates’ deference to supervisor and work motivation, although authoritarianism was unrelated to these outcomes; and (ii) benevo-lence and morality interacted to affect the same employee outcomes. Specifically, benevolent and moral leaders elicited more favourable employee outcomes than leaders exhibiting other leadership styles.
Key words: authoritarianism, benevolence, morality, paternalistic leadership.
Introduction
Based on their extensive review of the literature of overseas Chinese entrepreneurship, Farh and Cheng (2000) identi-fied paternalistic leadership as a leadership style that ‘combined strong discipline and authority with fatherly benevolence and moral integrity’ (p. 84). That is, Chinese paternalistic leaders demonstrate three features: (i) authori-tarianism: they use control strategies that assert strong authority over subordinates, apply control tactics to main-tain power status and highlight strict discipline; (ii) benevo-lence: they express holistic and individualized concern for subordinates’ well-being; and (iii) morality: they empha-size the pursuit of superior moral character, such as self-lessness and being a model person.
Paternalistic leadership is not only found in the Chinese context but also widely practiced in many non-Western regions, such as the Asia–Pacific, Middle East and Latin America (Uhl-Bien, Tierney, Graen, & Wakayabashi, 1990; Aycan et al., 2000; Martinez, 2003, 2005; Farh, Cheng, Chou, & Chu, 2006). Although Weber (1947) argued that paternalistic control in organizations would eventually be replaced by rational-legal power structures, as shown in prior research, paternalistic leadership is still effective in many contemporary non-Western organizations (Uhl-Bien et al., 1990; Aycan et al., 2000; Farh & Cheng, 2000; Mar-tinez, 2003, 2005; Aycan, 2006; Farh et al., 2006; Farh, Liang, Chou, & Cheng, 2008).
Farh and Cheng (2000) postulate that in Chinese contexts, the three dimensions of paternalistic leadership would inter-act to affect employee outcomes. Specifically, high benevo-lence is hypothesized to strengthen the positive linkage between authoritarianism and subordinate effectiveness, and leader’s morality is expected to further reinforce the joint effect of high authoritarianism and benevolence on employee effectiveness. However, several correlational studies (Cheng, Huang, & Chou, 2002a; Cheng, Chou, Huang, Farh, & Peng, 2003; Cheng, Chou, Wu, Huang, & Farh, 2004) that have examined the interactive effects of the three dimensions of paternalistic leadership on employee outcomes have failed to find the expected three-way inter-action. Nonetheless, this line of research has reported a consistent interaction of authoritarianism and benevolence on subordinates’ work attitudes (Cheng et al., 2004; Farh et al., 2006); when leader benevolence is high, leader authoritarianism has either positive or no relationship with subordinates’ attitudes. When leader benevolence is low, authoritarianism is negatively related to subordinates’ attitudes. Other two-way interactions were either non-significant or inconsistent across studies. For example, although the joint effect of benevolence and morality on employee outcomes is quite intuitive, it was not found in previous studies. In contrast, the interaction between authoritarianism and morality on subordinates’ outcomes observed in some past studies (Cheng et al., 2003, 2004) was difficult to interpret. Specifically, it is hard to understand why high leader morality would attenuate the positive asso-ciation between authoritarianism and employee outcomes.
We attribute these seemingly confusing findings to the correlational design used in previous studies, in which the three dimensions of paternalistic leadership were measured in a survey. In all previous studies, there was a highly
Correspondence: Bor-Shuian Cheng, Department of Psychology,
National Taiwan University, No. 1, Sec. 4, Roosevelt Road, Taipei 10617, Taiwan. Email: chengbor@ntu.edu.tw
Received 14 February 2008; accepted 27 February 2008.
positive relationship between benevolence and morality, and highly negative relationships between benevolence and authoritarianism and between morality and authoritarian-ism (Cheng et al., 2002a, 2003, 2004; Cheng, Shieh, & Chou, 2002b; Farh et al., 2006). Sometimes, the correlation coefficients, especially those between benevolence and morality, were greater than 0.60. Because of the high levels of intercorrelation among authoritarianism, benevolence and morality, distribution of leadership styles was not even in the samples. For example, following Smith and Foti (1998) and Lord and Maher (1991), Farh et al. (2008) used the theoretical mid-point (3.5) of the six-point Likert scale of the paternalistic leadership measures to dichotomize authoritarianism, benevolence and morality to identify eight (i.e. 2¥ 2 ¥ 2) different types of leaders. In two inde-pendent samples of Chinese leaders, a very high percentage of participants were low on authoritarianism, high on benevolence and high on morality. For example, in a sample of school principals, more than 60% of the partici-pants were low on authoritarianism, high on benevolence and high on morality, and almost 90% of them were rated as moral leaders. In another sample of managers, low authoritarianism/high benevolence/high morality leaders made up half of the sample, and 88% of the managers were rated as moral leaders. The overrepresentation of moral leaders in these studies may also reflect a self-selection bias. Moral and benevolent leaders, by virtue of their level of agreeableness, are particularly likely to agree to partici-pate in these studies. In short, given the high intercorrela-tions among the three paternalistic leadership dimensions, the interactions between the three leadership dimensions cannot be reliably determined.
To address this issue, in the present study, we conducted an experiment in which we independently manipulated the three paternalistic leadership dimensions by having the par-ticipants report their deference to the leader and work moti-vation in response to a leader who exhibited one of the eight leadership styles. This method was widely adopted in research regarding leadership decision-making (Zhou & Martocchio, 2001), leadership effectiveness (Deluga & Souza, 1991; Knippenberg & Knippenberg, 2005) and sub-ordinate perception (Trempe, Rigny, & Haccoun, 1985). Via our experimental design, we attempt to better investi-gate the interactive effects of leader’s authoritarianism, benevolence and morality.
Theory and hypotheses
Main effects of the dimensions of paternalistic leadership
Farh and Cheng (2000) contend that paternalistic leader-ship has deep roots in traditional Chinese culture. An
important aspect of Chinese culture is Confucian politics in social organization, which highlights the centrality of the vertical relationship between superiors and followers. It endorses the leader’s dominance, concern for the followers and moral self-cultivation, as well as the followers’ submis-sion to authority, obligations to reciprocate favours, and acceptance of the leader’s moral teachings. This cultural tradition has also impacted the construction of the roles of the leaders and the subordinates in the contemporary Chinese organizations. When both leaders and subordinates willingly accept and respect the paternalistic prescriptions in their respective roles, paternalistic leadership is often accompanied by positive employee responses, which, in turn, predict an increase in subordinates’ work motivation. Specifically, authoritarianism is expected to promote sub-ordinate compliance and dependence, leaders’ benevolence is expected to increase subordinates’ gratitude and recipro-cation, whereas leaders’ morality is expected to promote subordinates’ respect and identification; all three paternal-istic leadership dimensions are expected to increase sub-ordinates’ work motivation (Farh & Cheng, 2000). To elaborate, leaders practicing authoritarian leadership would promote dependency and compliance by imposing strict discipline and high performance standards on the subordi-nates while providing firm guidance and instructions to the subordinates. Leaders practicing benevolent leadership would enhance reciprocity by helping subordinates when they encounter difficulties and personal emergencies, expressing interest in the subordinates’ welfare even out-side the work settings, and rewarding subordinates who display appropriate or desirable behaviours. Finally, leaders practicing moral leadership would increase identification by being a role model for the subordinates; these leaders would keep promises, be fair to all subordinates and would not take advantage of the subordinates.
Consistent with these ideas, there is good evidence from previous studies for the positive relationship between leader benevolence and employee outcomes, and that between leader morality and employee outcomes (Cheng et al., 2002a, b, 2003, 2004; Farh et al. 2006). However, there is little evidence for the positive relationship between authoritarianism and employee outcomes. Contrarily, au-thoritarianism was found to be negatively related to satis-faction with supervisor, job performance, intention to stay (Cheng et al., 2002a), organizational citizenship behaviour (Cheng et al., 2002b) and organizational commitment (Farh et al., 2006). To explain these unexpected results, Farh et al. (2008) speculated that globalization and the fairness values that accompany it might have weakened the rel-evance of some control strategies used by authoritarian leaders to assert their authority in traditional Chinese societies (e.g. concealing and manipulating critical information, ignoring subordinates’ suggestions and underestimating subordinates’ contributions; King, 1979;
Yang, 1998; Liu, 2003). Thus, in the present study, we restrict authoritarianism to those control strategies that are less affected by globalization (e.g. strict discipline, in-sistence on high performance standards, intolerance of low performance and reputation building), and hypothesize a positive relationship between authoritarianism and employee outcomes. In short, based on Farh and Cheng’s (2000) model, we make the following hypotheses. Hypothesis 1: Leader authoritarianism increases
subordi-nates’ deference to the supervisor and work motivation.
Hypothesis 2: Leader benevolence increases subordinates’ deference to the supervisor and work motivation.
Hypothesis 3: Leader morality increases subordinates’ deference to the supervisor and work motivation.
Interactive effects of the dimensions of paternalistic leadership
Although Farh and Cheng’s (2000) model of paternalistic leadership suggests some potential interactive effects of the three dimensions of paternalistic leadership on employee outcomes, it does not provide a compelling argument for such effects. In addition, previous studies examining the interactive effects of authoritarianism, benevolence and morality on employee outcomes (e.g. Cheng et al., 2004) have not provided a strong theoretical rationale for these effects. To fill this void, we propose that although authori-tarian leaders set up strict discipline and high performance standards for their subordinates, subordinates may not comply with their leaders’ expectations unless the subordi-nates are motivated by reciprocation concerns or look up to the leaders as role models. Accordingly, leader authoritari-anism should be most effective when it is accompanied by either reciprocation-promoting benevolence or internalization-promoting morality. Thus, we make the fol-lowing hypotheses.
Hypothesis 4: Leader authoritarianism and benevolence in-teract to affect subordinates’ deference to the supervisor and work motivation in such a way that these outcomes are high only when both authoritarianism and benevolence are present.
Hypothesis 5: Leader authoritarianism and morality inter-act to affect subordinates’ deference to the supervisor and work motivation in such a way that these outcomes are high only when both authoritarianism and morality are present.
Alternatively, subordinates may also be motivated to reciprocate a benevolent leader’s favours by displaying positive subordinate outcomes. They are particularly likely to do so when they perceive that the leader’s benevolent behaviours arise from the leader’s moral, altruistic inten-tions. Thus, the positive effects of leader benevolence on subordinates’ outcomes should be magnified at the pres-ence of leader morality. Accordingly, we make the follow-ing hypothesis.
Hypothesis 6: Leader benevolence and morality interact to affect subordinates’ deference to the super-visor and work motivation in such a way that these outcomes are high only when both benevolence and morality are present.
Method
Leadership description generation and the pretest
To manipulate the three dimensions of paternalistic leader-ship, we created descriptions of eight leaders (high vs low authoritarianism by high vs low benevolence by high vs low morality) based on Farh and Cheng’s (2000) definition and Farh et al.’s (2008) suggested revision of the concept of authoritarianism. For example, the description for the high benevolence/high authoritarianism/high morality leader is as follows:
Mr Chen, an executive, takes care of his subordinates. He would accommodate, support and actively find ways to help his subordinates when they encounter difficulties or are in need, either at work or in private. He is a serious person who makes few compromises. He would reprimand his subordi-nates when they fail to meet his expectations. He keeps his word and is serious about his conduct and work perfor-mance. He would never break any rules or allow misbehav-iour to leave a stain on his honour. He is a model for his subordinates.
The description of the low benevolence/low authoritarianism/low morality leader is as follows:
Mr Chen, an executive, does not care about his subordinates at all. He feels indifferent to the difficulties they encounter or their needs, either at work or in private. He believes that his subordinates are responsible for those problems and would not extend a helpful hand. He is easy to get along with and would share his thoughts and other information at work with his subordinates. Even when they fail to produce satisfactory results, he would not hold them up to a high standard or reprimand them. He uses company resources for per-sonal purposes and takes advantage of his subordinates.
Sometimes he says one thing and acts differently, steals credit from others’ achievements, sheds responsibilities and violates regulations.
As a manipulation check, we had 80 college students rate the relevance of the eight leader descriptions to the three paternalistic leadership dimensions. Having learned the definitions of authoritarianism, benevolence and morality, these pretest participants rated each of the eight leader descriptions’ relevance to the three dimensions on a six-point Likert scale, ranging from 1 (totally irrelevant) to 6 (totally relevant). To control for order effects, four counter-balanced sequences (ABCDEFGH, GHABCDEF, EFGHABCD, CDEFGHAB) were created following a reduced Latin square design, and the participants were ran-domly assigned to one of the sequences. An authoritarian-ism (high or low)¥ benevolence (high or low) ¥ morality (high or low) ¥ order (ABCDEFGH, GHABCDEF, EFGHABCD, CDEFGHAB) anova was carried out on the ratings along each leadership dimension. No order effects were found (for authoritarianism, F1,76= 2.05, p> 0.05; for benevolence, F1,76= 0.58, p> 0.05; for morality, F1,76= 0.69, p> 0.05). The manipulations had the hypoth-esized main effect on the respective dimension of pater-nalistic leadership. As shown in Table 1, the high authori-tarianism leaders were perceived as more authoritarian than the low authoritarianism leaders, F1,76= 516.80, p< 0.01, h2
p= 0.87; the high benevolence leaders were viewed as
more benevolent than the low benevolence ones, F1,76= 668.69, p< 0.01, h2p= 0.90; and the high morality
leaders were seen as more moral than the low morality leaders, F1,76= 1188.48, p< 0.01, h2p= 0.94).
Participants and procedures
In the main study, we used a between-subjects design to minimize potential carryover effects. Participants were
employees from different private organizations in Taiwan. These organizations were in the industries of high-tech manufacturing, traditional manufacturing, retailing, com-mercial banking, insurance, educational service and other service industries. Twenty personnel officers from these organizations helped us deliver the survey to the employ-ees in their organizations. A total of 304 randomly chosen employees were contacted. Employees who consented to participate in the study were randomly assigned to respond to one of the eight leader descriptions, with the constraint that each experimental group (N= 38 in each group) would have a similar demographic composition. Participants were asked to carefully read the leader description, respond to the questions following it (see below), and return the completed survey in a sealed enve-lope to the authors through the personnel officers or by mail.
We received 265 completed and useable questionnaires (mortality rate= 13%). The final sample consisted of 54% female participants. Participants were relatively young (74% were under 35 years) and well educated (84% had Bachelor’s degrees or above). Most participants were sub-ordinates (only 25% identified themselves as supervisors). The average corporate tenure of the sample was 7.51 years. The eight experimental groups did not differ in the dis-tributions of gender (c2= 7.72, df = 7, p> 0.05), age (c2= 37.34, df = 28, p> 0.05), educational background (c2= 11.60, df = 14, p> 0.05), years of tenure (c2= 50.57, df= 28, p> 0.05), and position rank (c2= 8.08, df = 7, p> 0.05).
Dependent measures
Deference to supervisor. We adopted Cheng et al.’s (2004) 18-item scale to measure deference to supervisor. This scale measures three deference dimensions: identification Table 1 Means and standard deviations of leadership ratings in the pretest
Scenario (leadership type)
Authoritarianism (A) rating Benevolence (B) rating Morality (M) rating
M SD M SD M SD
High A, High B, High M 4.70a 1.00 4.95ab 0.84 5.46a 0.67
High A, High B, Low M 4.61a 0.88 4.74b 0.79 1.51b 0.57
High A, Low B, High M 4.95a 0.81 1.94cd 0.74 5.19a 0.86
High A, Low B, Low M 4.70a 1.05 1.76d 0.80 1.48b 0.66
Low A, High B, High M 2.00b 0.93 5.34a 0.57 5.45a 0.73
Low A, High B, Low M 2.04b 0.99 5.01ab 0.85 1.73b 0.84
Low A, Low B, High M 2.25b 1.12 2.38c 0.97 5.28a 0.81
Low A, Low B, Low M 2.25b 1.15 2.23cd 1.04 1.53b 0.59
F1,76 516.80** 668.69** 1188.48**
H2 0.87 0.90 0.94
**p< 0.01. Within each column, means having different superscripts were significantly different at the 0.05 family-wise error rate; two means sharing a common superscript were not significantly different.
(seven items), compliance (five items) and gratitude (six items). We dropped two identification items because of their low correlations with the other five identification items. On a six-point Likert scale ranging from 1 (never) to 6 (always), participants reported the extent to which they would display the deference behaviour depicted in the item if they were Mr Chen’s subordinates. The identification scale (a = 0.90) includes items such as ‘very much admire his manner and behaviour’ and ‘tell my colleagues or friends about his merits’. The compliance scale (a = 0.89) includes items such as ‘completely obey his instructions’ and ‘comply with his decisions even if I do not agree with him’. The gratitude scale (a = 0.87) includes items such as ‘repay him for his kindness when I get the opportunity’ and ‘appreciate his kindness’. We randomly assigned items of each scale to three composites and used these nine com-posites as indicators to perform a confirmatory factor analy-sis. Results showed that a second-order factor model with a higher-order latent factor that explains the covariance among identification, compliance and gratitude fitted our data well (c2= 48.56, df = 24, NFI = 0.97, NNFI = 0.98, CFI= 0.99, IFI = 0.99, RMSEA = 0.06, beta coefficients for paths from the second-order factor to three dimensions ranged from 0.81 to 0.93). We thus averaged the scores of the three scales to form an aggregate measure of deference to supervisor (a = 0.94).
Work motivation. Two items were created to measure par-ticipants’ work motivation: ‘If Mr Chen were your super-visor, would you be willing to work hard under his style of leadership?’ and ‘If you were a subordinate of Mr Chen and under his style of leadership, would you be willing to work hard?’. Participants were asked to report a score between 0 (very unwilling to work hard) and 100 (very willing to work hard) for each item. We then averaged the two scores to form a measure of the participants’ work motivation (a = 0.90).
Frequency of experiencing the given leadership style. To ensure that the participants perceived the eight leadership styles to be equally likely to occur in real life, we asked participants to answer two questions about the likelihood of encountering the leadership style depicted in the survey in their actual work experience. These two questions were: ‘How often have you ever worked under the leadership style of Mr Chen in your past experience?’ and ‘How often have you heard of a supervisor like Mr Chen in your past experience?’. The scale for these two items ranged from 0 (never) to 100 (always). We averaged these two items to measure the actual frequency of experiencing the given leadership style (a = 0.86).
Results
Frequency of experiencing the given leadership style
The eight leadership styles were reported to be similarly likely to occur in the participants’ work experiences (the reported frequencies of the eight leadership styles ranged from 33.20 to 46.44); the main effect of leadership styles on reported frequency was not significant, (F7,254= 1.23, p> 0.05).
Deference to supervisor
For each of the two dependent measures, we carried out the authoritarianism (high or low) by benevolence (high or low) by morality (high or low) ancova controlling for the participants’ demographics (gender, age, educational back-ground, years of tenure and position rank). None of the demographic variables were related to the dependent vari-ables and hence would not be discussed further.
Contrary to Hypothesis 1, which predicts positive effects of authoritarian leadership on subordinate’s deference to the supervisor, the main effect of authoritarianism was not significant (F1,237= 1.72, p> 0.05). Consistent with Hypotheses 2 and 3, the main effects of benevolence and morality were significant. For benevolence, F1,237= 60.16, p< 0.01, h2
p= 0.20, Cohen’s d = 1.02: the high
benevo-lence leaders (M= 3.61) received higher levels of deference than did the low benevolence leaders (M= 2.82). For morality, F1,237= 51.13, p< 0.01, h2p= 0.18; Cohen’s
d= 0.92: the high morality leaders (M = 3.57) received higher levels of deference than did the low morality leaders (M= 2.86).
In addition, there was support for Hypothesis 6. The interaction of benevolence and morality was significant, F1,237= 5.39, p< 0.05, h2p= 0.02. As shown in Table 2, post
hoc t-tests at the 0.05 family-wise error rate revealed that the high benevolence/high morality leaders received higher levels of deference (M= 4.10) than did the other three types of leader. In addition, the high benevolence/low morality leaders (M= 3.15) and the low benevolence/high morality leaders (M= 3.07) received higher levels of deference than did the low benevolence/low morality leaders (M= 2.56).
Contrary to Hypotheses 4 and 5, the interaction of authoritarianism and benevolence (F1,237= 1.36, p> 0.05) and that of authoritarianism and morality were both insig-nificant (F1,237= 0.01, p> 0.05).
Work motivation
Similar results were obtained for work motivation. Again, Hypothesis 1 was not supported: The main effect of authoritarianism was not significant, F1,237= 0.71, p> 0.05.
Consistent with Hypotheses 2 and 3, the main effects of benevolence and morality were significant; for benevo-lence, F1,237= 25.87, p< 0.01, h2p= 0.10, Cohen’s d = 0.67;
for morality, F1,237= 20.47, p< 0.01, h2p= 0.08, Cohen’s
d= 0.58. Work motivation was higher when leaders were high (vs low) on benevolence (M= 68.78 for the high benevolence leaders and 53.73 for the low benevolence leaders) and high (vs low) on morality (M= 67.85 for the high morality leaders and 54.74 for the low morality leaders).
Again, there was support for Hypothesis 6. The interac-tion of benevolence and morality was significant, F1,237= 6.57, p< 0.05, h2p= 0.03. As shown in Table 2, post
hoc t-tests at the 0.05 family-wise error rate indicated that work motivation was higher when the leaders were high on benevolence and morality (M= 78.97) than when the leaders displayed the other three leadership styles (M ranged from 50.07 to 59.34).
Again, contrary to Hypotheses 4 and 5, the interaction of authoritarianism and benevolence (F1,237= 0.28, p> 0.05) and that of authoritarianism and morality (F1,237= 1.39, p> 0.05) were not significant.
Discussion
Consistent with prior research, our results supported Farh and Cheng’s (2000) predictions that benevolence and morality would contribute to deference to the supervisor and work motivation. These results suggest that leader benevolence and morality can enhance leadership effective-ness in the Chinese context. Aside from the main effects of leader benevolence and morality, we showed that the co-presence of the leader benevolence and morality pro-duces the highest levels of deference to the leader and work motivation, supporting the idea that workers motivated to reciprocate the leader’s benevolence are particularly likely to do so when they perceive the leader’s benevolent actions
are reflections of the leader’s moral virtues (vs pretentious actions motivated by instrumental motives).
Examining the interactive effects of the dimensions of paternalistic leadership on subordinate outcomes has been identified as an important future research direction in the field of management and organizations (Farh et al., 2008; Pellegrini & Scandura, 2008). Since Farh and Cheng (2000) proposed the three-dimensional model of paternalistic lead-ership, a series of correlational studies have been conducted to investigate the interactive effects of the three dimensions (Cheng et al., 2002a, b, 2003, 2004; Farh et al., 2006). Perhaps because of the sampling problems in the correla-tional studies discussed in the Introduction, these past studies have yielded mixed results. In the present study, by manipulating the three dimensions independently in an experiment, we discerned a reliable and meaningful inter-active effect of leader benevolence and morality on some employee outcomes. The joint effect of benevolence and morality on deference to supervisor and work motivation found in the present study has important implications for management practices. Our results suggest that compared to ‘indulgent’ leaders, who primarily influence subordinates through acting nicely, or ‘ideological’ leaders, who mainly rely on personal virtues and social modelling as the influ-ence vehicles, individuals playing a role of ‘selfless bene-factors’ are most effective supervisors in the Chinese context.
Unlike past studies which have found a negative associa-tion between leader authoritarianism and employee out-comes (e.g. Cheng et al., 2002a, b; Farh et al., 2006), in the present study, leader authoritarianism did not impact def-erence to supervisor and work motivation negatively, prob-ably because, in the present study, we had excluded from authoritarian leadership control strategies that conflict with fairness values in contemporary Chinese societies. Despite this, our result does not support the hypothesis that leader authoritarianism would increase deference to the supervisor and work motivation. The lack of leader authoritarianism’s Table 2 Interaction effects of benevolence and morality on deference to supervisor and work motivation
Benevolence Morality N
Deference to supervisor Work motivation
M SD M SD High High 63 4.10a 0.67 78.97a 15.22 High Low 68 3.15b 0.78 59.34b 23.72 Low High 67 3.07b 0.84 57.39b 24.31 Low Low 67 2.56c 0.78 50.07b 26.23 F3,261 43.88** 18.74** h2 p 0.34 0.18
**p< 0.01. Within each column, means having different superscripts were significantly different at the 0.05 family-wise error rate; two means sharing a common superscript were not significantly different.
effects should not be taken to imply that leader authoritari-anism is irrelevant to paternalistic leadership. Instead, the evidence to date suggests that paternalistic leadership can impact employee behaviours via two routes. First, benevo-lent leadership promotes intrinsic work motivation and leader identification by increasing the obligation to recip-rocate the leader’s kindness, and this effect is particularly strong when the leader is also a moral leader. Second, authoritarian leadership increases extrinsic work motiva-tion and compliance through setting high standards and providing directive instructions to the employees. The effect of authoritarian leadership on extrinsic work motiva-tion and compliance should be particularly pronounced when the authoritarian leader has strong control over orga-nizational resources (Farh et al., 2006). Resource domi-nance refers to control over the budget, equipment and the information necessary for task completion, as well as the position power of determining subordinates’ rewards (e.g. bonus, wage increase and promotion). When leaders have sole control over these resources, subordinates tend to accept leaders’ power status and comply with leaders’ instructions (Blau, 1964). Accordingly, Farh and Cheng’s (2000) prediction of the positive relationship between leader authoritarianism and subordinates’ outcome may be found when the leader is high on resource dominance. Future research should examine this possibility.
References
Aycan, Z. (2006). Paternalism: Towards conceptual refinement and operationalization. In: K. S. Yang, K. K. Huang & U. Kim, eds. Scientific Advances in Indigenous Psychologies:
Empiri-cal, PhilosophiEmpiri-cal, and Cultural Contributions, pp. 445–466.
London: Sage.
Blau, P. M. (1964). Exchange and Power in Social Life. New York: John Wiley & Sons.
Aycan, Z., Kanungo, R. N., Mendonca, M., et al. (2000). Impact of culture on human resource management practices: a 10-country comparison. Applied Psychology: An International
Review, 49, 192–221.
Cheng, B. S., Chou, L. F., Huang, M. P., Farh, J. L. & Peng, S. (2003). A triad model of paternalistic leadership: Evidence from business organization in Mainland China. Indigenous
Psy-chological Research in Chinese Societies, 20, 209–252 (in
Chinese).
Cheng, B. S., Chou, L. F., Wu, T. Y., Huang, M. P. & Farh, J. L. (2004). Paternalistic leadership and subordinates responses: Establishing a leadership model in Chinese organizations.
Asian Journal of Social Psychology, 7, 89–117.
Cheng, B. S., Huang, M. P. & Chou, L. F. (2002a). Paternalistic leadership and its effectiveness: Evidence from Chinese orga-nization teams. Journal of Psychology in Chinese Societies, 3, 85–112 (in Chinese).
Cheng, B. S., Shieh, P. Y. & Chou, L. F. (2002b). The principal’s leadership, leader-member exchange quality, and the teacher’s
extra-role behavior: The effects of transformational and pater-nalistic leadership. Indigenous Psychological Research in
Chinese Societies, 17, 105–161 (in Chinese).
Deluga, R. J. & Souza, J. (1991). The effects of transformational and transactional leadership styles on the influencing behaviour of subordinate police officers. Journal of Occupational
Psy-chology, 64, 49–55.
Farh, J. L. & Cheng, B. S. (2000). A cultural analysis of paternal-istic leadership in Chinese organizations. In: J. T. Li, A. S. Tsui, & E. Weldon, eds. Management and Organizations in the
Chinese Context, pp. 85–127. London: Macmillan.
Farh, J. L., Cheng, B. S., Chou, L. F. & Chu, X. P. (2006). Authority and benevolence: employees’ responses to paternal-istic leadership in China. In: A. S. Tsui, Y. Bian, & L. Cheng, eds. China’s Domestic Private Firms: Multidisciplinary
Per-spectives on Management and Performance, pp. 230–260. New
York: Sharpe.
Farh, J. L., Liang, J., Chou, L. F. & Cheng, B. S. (2008). Pater-nalistic leadership in Chinese organizations: Research progress and future research directions. In: C. C. Chen & Y. T. Lee, eds.
Business Leadership in China: Philosophies, Theories, and Practices, pp. 171–205. Cambridge: Cambridge University
Press.
King, A. Y. C. (1979). From Tradition to Modernity. Taipei, Taiwan: China Times Publishing Co. (in Chinese).
Knippenberg, B. V. & Knippenberg, D. V. (2005). Leader self-sacrifice and leadership effectiveness: The moderating role of leader prototypicality. Journal of Applied Psychology, 90, 25–37.
Liu, S. (2003). Culture within culture: Unity and diversity of two generations of employees in state-owned enterprises. Human
Relations, 56, 387–417.
Lord, R. G. & Maher, K. J. (1991). Leadership and Information
Processing: Linking Perceptions and Performance. Boston:
Unwin Hyman.
Martinez, P. G. (2003). Paternalism as a positive form of leader-subordinate exchange: Evidence from Mexico. Journal of
Iberoamerican Academy of Management, 1, 227–242.
Martinez, P. G. (2005). Paternalism as a positive form of leader-ship in the Latin American context: Leader benevolence, decision-making control and human resource management practices. In: M. Elvira & A. Davila, eds. Managing Human
Resource in Latin America: An Agenda for International Leaders, pp. 75–93. Oxford: Routledge Publishers.
Pellegrini, E. K. & Scandura, T. A. (2008). Paternalistic leader-ship: A review and agenda for future research. Journal of
Man-agement, 34, 566–593.
Smith, J. A. & Foti, R. J. (1998). A pattern approach to the study of leader emergence. Leadership Quarterly, 9, 147–161. Trempe, J., Rigny, A. J. & Haccoun, R. R. (1985). Subordinate
satisfaction with male and female managers: Role of perceived supervisory influence. Journal of Applied Psychology, 70, 44–47.
Uhl-Bien, M., Tierney, P., Graen, G. & Wakayabashi, M. (1990). Company paternalism and the hidden investment process: Identification of the ‘right type’ for line managers in leading Japanese organizations. Group and Organization Studies, 15, 414–430.
Weber, M. (1947). The Theory of Social and Economic
Organi-zation (translated by A. M. Henderson & T. Parsons). New
York: The Free Press.
Yang, K. S. (1998). Chinese responses to modernization: A psy-chological analysis. Asian Journal of Social Psychology, 1, 75–97.
Zhou, J. & Martocchio, J. J. (2001). Chinese and American man-agers’ compensation award decisions: A comparative policy-capturing study. Personnel Psychology, 54, 115–145.
