Out of this ferment emerged two distinct approaches to understanding leadership in organizations that have coexisted for more than a century, traveling more or less side by side, with only occasional nods to one another. One track, which we label quantitative- analytic, emphasizes testing hypotheses with quantitative data to develop leadership theory. The work is typically published as articles in scholarly journals (an overview of eras in the evolution of this strand appears in Exhibit 17.1). A second track, qualitative-holistic, relies
330 Reframing Organizations
WEBC17 05/26/2017 2:40:14 Page 331
Exhibit 17.1. A Short History of Quantitative-Analytic Leadership Research.
Leadership Theory Examples Central Idea Current Status
Trait theory: how Galton, 1869; Terman, Leaders possess distinctive Fell out of favor in the 1950s when are leaders 1904; Kirkpatrick and personal characteristics reviewers found weak empirical support, different? Locke, 1991; Zaccaro, (intelligence, self-con!dence, but has returned to favor in recent
2007 integrity, extraversion, and so on). decades.
Leadership style Lewin, Lippitt, and White, Leadership depends on style Mixed evidence stimulated move toward theory: how do 1939; Likert, 1961; (democratic vs. autocratic, task- contingency theories, which often include leaders act? Fleishman and Harris, oriented vs. people-oriented, etc.). leader style variables.
Contingency Fiedler, 1967; Lawrence Effective leadership depends on No single contingency view has found theory: how do and Lorsch, 1967; Evans, the characteristics of followers consistent empirical support or wide circumstances 1970; House, 1971, 1996 and context: what works in one acceptance, but most modern leadership affect leadership? situation may not work in research incorporates the idea that
another. leadership depends on circumstances.
Leader-member Dansereau, Graen, and Leadership is rooted in the quality Advocates of LMX theory have been exchange (LMX) Haga, 1975; Graen and of the relationships between actively conducting research since the theory: what Uhl-Bien, 2008 leaders and individual followers. 1970s; many LMX propositions have happens in the empirical support, but the approach is leader-follower criticized for complexity and viewing relationship? leadership too narrowly.
Transformational Burns, 1978; Bass, 1985; Transformational (or charismatic) Evidence suggests transformational leadership Conger and Kanungo, leaders use inspiration, idealized leadership makes a difference, but more theory: how do 1998 in”uence, and the like to research is needed on when and how it leaders transform generate followers’ trust and works best. followers? willingness to go above and
Reframing Leadership 331
WEBC17 05/26/2017 2:40:14 Page 332
on case studies and interviews with practitioners to develop ideas and theory about how leadership works in practice. Such work is often published in books aimed at an audience that includes both practitioners and scholars. We will survey quantitative and then qualitative work before trying to capture the current state of the !eld.
Quantitative-Analytic Research Since the early twentieth century, quantitative research has moved through several eras, gradually evolving from simpler to more complex views of leadership. The initial research, “owing from the “great man” theory of leadership (Carlyle, 1841), focused on !nding the distinctive traits that made leaders different from everyone else. Around 1950, multiple reviews (Stogdill, 1948; Gibb, 1947; Jenkins, 1947) concluded that there was little consist- ency in leadership traits across people and circumstances. That gave rise to two lines of research in the 1950s and subsequent decades: one on leadership style and another on situational contingencies. Style research focused particularly on the difference between task- oriented and people-oriented leaders. The results suggested that leaders who focused on people generated higher morale although not necessarily higher productivity, and that the most effective leaders were good at dealing with both tasks and people (Fleishman and Harris, 1962).
Contingency theorists examined characteristics of situations that interacted with leader behavior. One in”uential line, for example, found that task-oriented leaders did best in situations that were either highly favorable or highly unfavorable for the leader, while people-oriented leaders did best in situations in the middle (Fiedler, 1964, 1967).
Another contingency theory, Hersey and Blanchard’s situational leadership model (1969, 1977), had less research support (Hambleton and Gumpert, 1982; Graeff, 1983; Blank, Weitzel, and Green, 1990) but became more popular with practitioners because it is more intuitive and offers clearer practical guidance to practitioners. The model incorporates its own version of the distinction between task and people, using a two-by-two table to develop four different leadership styles (see Exhibit 17.2). Hersey and Blanchard argued that each style was appropriate for a different level of subordinate “readiness,” which they de!ned in terms of how able and willing subordinates were to do the work. If subordinates are neither willing nor able, then the leader should tell them how to do the job. If they want to do the job but lack skill, then the leader should sell or coach to build capacity. When subordinates are able but unwilling or insecure, then the leader should use a participative style to build motivation. If they are both able and willing, the leader should delegate and get out of the way.
332 Reframing Organizations
WEBC17 05/26/2017 2:40:14 Page 333
Exhibit 17.2. Situational Leadership Model.
High Relationship, Low Task: High Relationship, High Task:
Participate Sell (or Coach)
Use when followers are “able” but Use when followers are “unable” but “unwilling” or “insecure.” “willing” or “motivated.”
Low Relationship, Low Task: Low Relationship, High Task:
Use when followers are “able” and Use when followers are “unable” and “willing” or “motivated.” “unwilling” or “insecure.”
Hersey and Blanchard’s model continues to be popular for leadership training but has been criticized for lack of research support and for generating self-ful!lling prophesies. If, for example, managers give unwilling and unable subordinates high direction and low support, what would cause their motivation to improve? The manager of a computer design team told us ruefully, “I treated my group with a ‘telling’ management style and found that in fact they became both less able and less willing.”
The 1970s spawned a new line of research: leader-member exchange theory (LMX). LMX research began with the insight that leaders create different relationships with different followers, and, in particular, they create in-groups and out-groups by interacting with some subordinates in a more personal way while focusing strictly on task with others (Dansereau, Graen, and Haga, 1975; Graen and Uhl-Bien, 2008). One practical implication from this research is that leaders can get better results by creating strong relationships with all, not just some, of their subordinates (Graen and Uhl-Bien, 2008, p. 225).
A major new strand that emerged in the 1980s emphasized a distinction between transactional and transforming leadership (Burns, 1978). Transactional leadership involves practical, give-and-take exchanges, such as pay for performance. Transforming leaders, on the other hand, “champion and inspire followers . . . to rise above narrow interests and work together for transcending goals” (Burns, 2003, p. 26). Over the next two decades, research on transformational, or charismatic, leadership became a dominant research strand, producing a number of studies con!rming that transformational leaders had a more powerful impact than those who relied only on transactional approaches (Shamir, House, and Arthur, 1993; Conger and Kanungo, 1998).
Reframing Leadership 333
WEBC17 05/26/2017 2:40:14 Page 334
Qualitative-Holistic Leadership Studies The quantitative research tradition has both strengths and limits. Over more than a century, scholars have tested hypotheses, discarded ideas that don’t work, and gradually built theory that !ts the data. But the work has often simpli!ed the complexities of leadership by treating only a few variables at a time and by treating leadership as equivalent to what happens between managers and their subordinates. Qualitative research on real-world practice has viewed leadership in more nuanced and holistic ways, often developing ideas decades before they make their way into quantitative studies. Mary Parker Follett (1896, 1918, 1941), for example, was well ahead of her time in exploring distributed leadership, charisma, and the importance of the human element. Many of the major themes in Follett’s work were extended by two of the most in”uential management thinkers of the early twentieth century: Elton Mayo and Chester Barnard. Mayo, often viewed as the founder of the “human relations” school of management, conducted the famous studies that gave rise to the “Hawthorn effect” and promoted the idea, viewed as radical at the time, that human and social factors mattered as much as technical and economic ones (Mayo, 1933).
Chester Barnard, a telephone executive, was a practitioner rather than an academic, but he wrote one of the most in”uential management books of the midtwentieth century, The Functions of the Executive (Barnard, 1938). Barnard argued that the task of leadership is to balance technical and human factors to achieve cooperation among the many groups and individuals within an organization. Organizations rarely survive inde!nitely, he noted, because it is so challenging to solve two central issues: achieving goals while satisfying the needs of those who do the work.