1、 LEADERSHIP BEHAVIOR AND EMPLOYEE VOICE: IS THE DOOR REALLY OPEN? In todays hypercompetitive business environment, employee comments and suggestions intended to improve organizational functioning are critical to performance because,as Senge wrote,it is just not possible any longer to figure it out f
2、rom the top(1990:4； see also Morrison & Milliken, 2000). Edmondson(1999, 2003), for example, found that the willingness of all members to provide thoughts and ideas about critical work processes characterizes successful learning in various types of teams. Yet, despite this learning imperative, many
3、individuals do not work in environments where they perceive it as safe to speak up(Milliken, Morrison, & Hewlin, 2003； Ryan & Oestrich, 1998). This presents an unsettling state of affairs: Voice, which we define as the discretionary provision of information intended to improve organizational functio
4、ning to someone inside an organization with the perceived authority to act,even though such information may challenge and upset the status quo of the organization and its power holders, is critical to organizational well-being yet insufficiently provided by employees, who see the risks of speaking u
5、p as outweighing the benefits. Thus,it is important to better understand who speaks up with potentially valuable information and the organizational conditions that favor or inhibit such behavior. In this study, we seek to contribute to such understanding. Three broad lines of research have addressed
6、 this line of inquiry to varying degrees. The most systematic research to date has focused on individual differences in personality and demographic characteristics as correlates of voice (Crant, 2003； LePine & Van Dyne,2001). The stated or implicit reasoning in this line of work is that some individ
7、uals are simply more likely than others to go the extra milein regard to speaking up. A second line of research, based on Hirschmans (1970) seminal work defining exit, voice, and loyalty as the primary options facing employees who are dissatisfied with some aspect of organizational functioning, trea
8、ts employee attitudes as the primary determinant of upward voice(Rusbult, Farrell, Rogers, & Mainous,1988； Withey & Cooper,1989). Finally,a third research stream focuses on aspects of an organizational context that may affect employees willingness to speak up. An implicit assumption in this view is
9、that even the most proactive or satisfied employees are likely to read the wind as to whether it is safe and/or worthwhile to speak up in their particular context (Dutton, Ashford, ONeill, Hayes, & Wierba, 1997； Edmondson, 2003； Milliken et al.,2003). Seeking to further develop the contextual stream
10、, we focus on the role that specific leadership behaviors play in influencing employees decisions to voluntarily provide comments or suggestions intended to spark organizational improvement. Qualitative research has identified a number of leader behaviors or attributes including approachability (Mil
11、liken et al., 2003； Saunders, Sheppard, Knight, & Roth,1992), action taking (Edmondson, 2003； Ryan & Oestreich, 1998), and accessibility (Edmondson, 1999) that lead subordinates to conclude it is either safe or unsafe to speak up. However,the few quantitative studies that have assessed some aspect o
12、f leadership influence on voice have produced less conclusive results (e.g., Ashford et al.,1998). For example, Saunders and colleagues(1992) developed a measure,supervisor as voice manager, and found it to be positively related to the likelihood of voice in two samples, but Janssen, de Vries, and C
13、ozijnsen (1998),controlling for several individual differences, found that the supervisor as voice manager construct was not significantly related to subordinates reported likelihood to voice novel ideas. In sum, the literature presents a troubling discrepancy: Studies with the highest face validity
14、 suggest that leadership behaviors are an important contextual antecedent of voice,but survey research has failed to replicate such findings. The specific purpose of this study was therefore to address in consistent findings about leadership behavior as an influence on subordinates improvement-orien
15、ted voice. In a two-phase field study, we addressed the questions,Is leadership behavior related to subordinate voice? and, If so, why and for what types of employees? Our study extends the literature in a number of ways. First, few of the labels used to describe leadership findings in previous work
16、 correspond directly with constructs developed in the broader leadership literature. We drew upon well-established theory on leadership and power to develop predictions for how specific leader behaviors influence employee voice. Second, we controlled for many of the personality and employee attitudi
17、nal explanations found important in prior voice research but usually lacking in the leadership-focused studies. Third, we heeded calls for more precision in voice research (Van Dyne, Ang, & Botero,2003) by limiting our voice construct to verbal behavior that is improvement-oriented and directed to a
18、 specific target who holds power inside the organization in question. Beyond advancing understanding of leadership behavior as a predictor of voice, this study also examines psychological safety as an important mediating cognition linking leadership and voice and addresses how subordinate performanc
19、e level might moderate the impact of leadership behavior on voice. Collectively, this research extends understanding of the leadership-voice relationship and points to specific ways leaders can foster employee input. LITERATURE REVIEW AND HYPOTHESES The notion of voice stems from the idea that emplo
20、yees recognize some source of dissatisfaction or opportunity for improving their own and/or their organizations well-being (Hirschman, 1970). Speaking up in such situations can feel risky because they involve pointing out need for improvement in a program or policy to those who may have devised, be
21、responsible for, or feel personally attached to the status quo. Given this, along with the reality that voice cannot be coerced or readily designed into the in-role requirements of a job (Van Dyne & LePine, 1998), an initial motivation to peak up is likely to manifest in behavior only when the net perceived benefits outweigh potential costs. Perceived potential benefits of speaking up include getting the problem solved as well as formal (e.g., money or promotion) or informal (e.g., recognition or status) rewards that might
