Major Trends and Future Directions
Lowe and Gardner (2000) reviewed the last 10 years of research published in the journal Leadership Quarterly and highlighted where the field has been and where it is heading. Other papers included in that same issue pointed to new directions in the leadership field, including several themes that we pursue in this chapter: strategic leadership, e-leadership,
^Wlective leadership, and leadership development. To the extent that the Leadership Quarterly contains a representative sampling of what has and perhaps will be published, some of the past trends in leadership research identified by Lowe and Gardner offer a basis for launching our discussion.
Lowe and Gardner (2000) reported nearly an equivalent emphasis in articles published on theory and research over the last decade in the Leadership Quarterly, i.e., 467c vs. 55%, respectively. This pattern points to a field in transition, as new theoretical perspectives entered the field in the 1980s and 1990s, shaping directions in research. Lowe and Gardner also reported that most articles were still published by American authors, supporting House and Aditya's (1997) claim that 98% of leadership research still originates in North America. However, some giant steps are being made to promote and include research from other cultures.
fc Lowe and Gardner (2000) identified several trends that have shaped the field of leadership studies during the 1990s. These included work on transformational leadership (Avolio, 1999; Bass, 1998; Bass & Avolio, 1993,1994) and neo-charis-matic leadership theories based on House's (1977) theory of charismatic leadership, as well as Burns (1978) and Bass (1985; see also Bass & Steidlmeier, 1999;Bryman, 1993;Con-ger & Hunt, 1999; Conger & Kanungo, 1987,1998; House & Shamir, 1993). Lowe and Gardner also pointed to renewed interest in cross-cultural leadership research, which received a tremendous boost from House and his associates' Global Leadership and Organizational Behavior Effectiveness study (GLOBE) project, as well as other cross-cultural research.
Another driving force behind the transformation of leadership research and theory during the 1990s involved the emphasis on levels of analysis in theory building, research design, and measurement (Dansereau, Alutto, & Yammarino,
Areas Requiring Further Exploration 279
1984; Dansereau, Yammarino, & Markham, 1995; Dansereau & Yammarino, 1998; Klein & House, I995j. A levels-of-analysis frame of reference provided a huge leap toward more sophisticated multilevel models of leadership that now include the context in which leadership is embedded. Today, work on levels of analysis has dramatically shaped the conceptualization of how leadership is defined, measured across research streams, and within context over time (Brass, 2001; House & Aditya, 1997; Zaccaro & Klimoski, 2001). The context now being examined includes a broad range of constructs such as the nature and level of change, the medium through which leaders and followers interact, the cultures in which each are embedded, the organization level, the type of work group or unit, and the network. Future research on leadership now needs to standardize approaches to examining individual, dyadic, group, and larger collective phenomena (Lowe & Gardner, 2000). We envision that all future research will allocate greater attention to defining the level at which leadership is investigated, as well as the levels at which various models are tested and hold across different contexts (Zaccaro & Banks, 2001).
House and Aditya (1997) noted an emerging trend over the previous decade regarding the focus on strategic leadership. However, there is still a thin base of research on what constitutes strategic leadership, as well as strategic change in organizations (Boal & Hoojberg, 2000; Lowe & Gardner, 2000). Moreover, much of the research in this area has been based on case analyses, cross-sectional designs, and small samples, and it lacks a strong and coherent theoretical base.
Ironically, we know very little about how leadership actually changes people, units, organizations, and larger collectives (Burns, personal communication, November 2000; Yukl, 1999). We need to examine how leadership affects fundamental change in individuals, groups, and organizations. How can we evaluate such change while also taking into consideration the context in which leadership is embedded? What criteria can we use for assessing change at the individual, group, organizational, and even community levels? For example, can we measure behavioral or attitudinal change at each of these respective levels? What does the performance domain look like when we cut across these levels and examine change over time? Surprisingly, we know very little about the leadership of CEOs across all types and sizes of organizations. Nonetheless, some authors have attributed between 20% and 40% of organizational effectiveness to executive leadership (Ireland & Hitt, 1999).
Prior research has not explored how shared or collective leadership associated with top management teams contributes to an organization's adaptability and effectiveness (Elron, 1997). However, several studies have produced a
280 Leadership Models, Methods, and Applications
positive relationship between strategic leadership and firm performance, but in most cases prior research has not examined how the context moderates or mediates strategic leadership performance (Finklestein & Hambrick, 1996). Demonstrating the importance of the context, Waldman, Ramirez, House, and Puranam (2001) showed that executives who were rated by followers as more charismatic had little impact on firm performance under stable environmental conditions, but under unstable or uncertain conditions charismatic leadership significantly predicted financial performance.
A major decision for organizations about leadership deals with its succession (Lauterbach, Vu, & Weisberg, 1999). Organizations that rely on internal networks enjoy smooth transitions and the continuation of strategy. Yet, selecting a new leader who comes from the same stock as the previous one may result in staying the course when radical change is required (Hambrick & Mason, 1984). The area of leadership and succession, particularly at the tops of organizations, deserves more attention in the literature.
A number of questions highlight the need for future research on the context in which leadership is observed and interpreted, including the following: Are the same leadership styles differentially effective as we move up the organizational hierarchy? How does the legacy left by a former leader affect his or her successor's ability to maintain and enhance organizational performance? How do strategic leadership styles vary in terms of their impacts across different sectors of the economy? Are some strategic leadership styles generic and equally effective across all sectors? As the world of work becomes more complex, can leadership at the top be more effectively shared; if so, what are optimal strategies for developing and deploying shared leadership? How will strategic leadership change after inserting advanced IT in organizations in which every employee is connected to the CEO?
Leadership and CreatingMeaning
Smircich and Morgan (1982) provided an alternative definition of leadership, which has not been thoroughly explored and may have even greater relevance as leaders lead at a distance from followers in today's global economy. They stated that "leadership is realized in the process whereby one or more individuals succeeds in attempting to frame and define the reality of others" (p. 258). Are there differences in how leaders create impressions and use impression-management strategies to influence followers to support their positions? How do different types of leaders use impression-management strategies to create the meaning that they want their followers to derive from a particular situation, and will the creation of meaning work differently across cultures? For example, W. L. Gardner and Avolio (1998)
argued that charismatic versus noncharismatic leaders use different impression-management styles to gain commitment and trust from followers. How do these differences extend across cultures?
Smircich and Morgan (1982) related the management of meaning to a number of important areas including the emergence of leadership: "They emerge as leaders because of their role in framing experience in a way that provides a viable basis for actions" (p. 258). Simply put, leaders define the situation in which followers find themselves, shaping their range of perceived and actual choices. Charismatic leaders frequently emerge in times of crisis because they offer a viable alternative interpretation to resolve the crisis while managing an impression of confidence and an ability to inspire followers to pursue a solution (Conger & Kanungo, 1998).
The management of meaning has direct relevance to studying strategic leadership. Such leaders usually do not have direct contact with all followers and therefore must manage the meaning of events at a distance—and in today's organizations through advanced IT. Strategic leaders have a mandate to define the current and future reality of their organizations (Ireland & Hitt, 1999). However, we know that many strategic leaders lose their mandate when their framing of reality does not make sense to followers, lacks credibility, or ultimately does not contribute to success. Indeed, according to Howard (2001), somewhere between 30% and 50% of CEOs are prematurely ousted from their jobs.
To make sense of each follower's future requires the leader to develop a relationship through which followers come to identify with the leader's vision (Shamir, House, & Arthur, 1993). According to Shamir and his associates, to energize followers, leaders must successfully link the follower's self-concept to the collective thinking or concept of the group. This creates a sense of alignment around the vision to move forward. Visions represent one of the highest forms of managing meaning (Awamleh & Gardner, 1999; Graen & Uhl-Bien, 1995) and could be examined in terms of their diffusion through organizations as being an outcome of successful leadership. Specifically, all other things being equal, the extent to which a vision is wired into each employee's thinking and behavior could be used as a criterion measure of leadership effectiveness in an organization.
In sum, Smircich and Morgan (1982) stated that "leadership as a phenomenon depends upon the existence of people who are prepared to surrender their ability to define their reality for others" (p. 270). The way leaders transform reality for followers is a fertile area for future research {Yukl, 1999), including differences in followers' willingness to surrender their interpretations of reality or to share in the responsibility of creating the interpretation of the future.
LeadershipSubstitutes
Whereas leaders are said to influence how followers derive meaning from events, another line of research argues that we attribute too much meaning to leaders as the central causes of events. Followers erroneously attribute events or performance outcomes to the influence of a leader when in fact it is due to the context (Kerr & Jermier, 1978). This may be particularly evident with charismatic leaders, who emerge during times of crisis {Beyer, 1999). Meindl, Ehrlich, and Dukerich (1985) argued that people use leadership as a way of explaining or interpreting what goes on in organizations when they do not fully understand the cause of events. In these cases, using a levels-of-analysis framework in the design of models and methods can facilitate a better understanding of what causes what over time.
Although the literature on leadership substitutes provides an interesting perspective on how the context can moderate
^pieimpact of leadership on follower perceptions and performance, the weight of evidence shows that substitutes for leadership do not substitute as predicted by Kerr and Jermier (1978). Podsakoff, Niehoff, MacKenzie, and Williams (1993) were unable to find sufficient evidence to support J. P. Howell, Dorfman, and Kerr's (1986) claim that leadership substitutes moderate the relationship between leader behaviors and various intermediate process and outcome measures. Podsakoff, MacKenzie, and Fetter (1993) examined effects of substitutes for leadership with a sample of professional employees. The authors concluded that a substantial portion of the variance in perceptions and performance was shared by both leadership behaviors and substitutes. All substitutes for leadership independently influenced at least one of the criterion measures, accounting for 30% to 40% of the variance in employee attitudes, 18% to 23% in employee role percep-
f tions, and 7% in performance. Podsakoff, MacKenzie, and Bommer (1995) concluded that "although the notion that subordinate, task, and organizational characteristics moderate the effects of a leader's behavior seems intuitively appealing, the weight of empirical evidence does not support it" (p. 381). However, these same substitutes added to leadership can augment the prediction of employee role perceptions, job attitudes, and performance. Podsakoff et al. reported that the total variance explained in employee attitudes and in role performance by leadership and its substitutes (e.g., task feed-hack, professional orientation, rewards outside of the leader's control, etc.) were 35% and 33%.
We now turn to addressing some of the issues just mentioned by focusing on the individual, later returning to the strategic context in which individual leadership is observed and developed.
Areas Requiring Further Exploration 281
Implicit LeadershipTheory
Calder (1977), as well as Mitchell (1979), argued that leadership was not directly observable. Observer ratings were based in part on attributions, thus introducing some degree of error or bias into all leadership ratings (Lord, Binning, Rush, & Thomas, 1978: Rush, Phillips, & Lord, 1981). Summarizing a long line of research on implicit notions of leadership, Lord and Maher (1991) defined leadership as a process perceived by others and then labeled leadership. Calder's (1977) work provided the basis for a cognitive revolution in leadership research that continues to emerge and have a significant influence on the field of leadership research and theory today. Much of this research is based on the work of Lord and his associates. Early experimental research by Rush, Thomas, and Lord (1977) demonstrated that college students exposed to the same experimental leadership conditions interpreted leadership behaviors differently. Phillips and Lord (1981) attributed those differences to a cognitive categorization process that uses contextual and behavioral cues to classify leadership behaviors. Each observer comes to the situation with a preexisting mental structure, which results in observed behaviors being encoded based on the rater's categorization process. Differences in category systems or implicit theories (e.g., between men and women leaders) can result in an encoding and recall of behavior that is different for men and women.
Lord, Fori, and DeVader (1984) suggest that perceptions of leadership are based on hierarchically organized categories, which are each represented by a prototype. Prototypes are formed based on experiences with individuals or events. A prototypical category might be the use of a political or military leader. Traits that people associate with prototypical categories become important facets of how perceivers construct their categories and prototypes (Mischel, 1973). Observed leadership behavior is then categorized more or less automatically based on prototypical matching between an observer's implicit theory of leadership and actual behavior.
Work on implicit leadership theory (ILT) clearly has implications for leadership research as it moves across different cultures. ILT also has implications for theory development as we examine how followers internalize leadership messages and identify with a leader's vision within and between cultures. Work on ILT has implications for leadership development as well. ILT also affects whether an individual sees himself or herself as a leader worthy of development.
In sum, the work by Lord and his associates has demonstrated the importance of viewing leadership as being in the eye of the beholder. This stream of research has implications for examining biases in leadership measurement, for developing new theories that capture how leaders manage the meaning
282 Leadership Models, Methods, and Applications
of events for followers, and the development of leadership itself. Other issues deserving closer scrutiny include the following: To what extent are changes in generational views of leadership being shaped by the media exposure on the topic? How can leaders build trust when they are so overexposed in the media? Does the generation now entering the workforce come with different expectations about what constitutes trusted leadership?