Leadership Both for $20.00

shwnnjlotbug

advances_in_authentic_and_ethical_leadership.pdf

Home >Human Resource Management homework help >Leadership Both for $20.00

CHAPTER 8

A CRITICAL ASSESSMENT OF RESEARCH ON EFFECTIVE

LEADERSHIP BEHAVIOR Gary Yukl

University at Albany, SUNY

John W. Michel Loyola University Maryland

ABSTRACT

Much of the research on effective leadership over the past half century has involved studies of leader behavior. Progress in learning about effective lead- ership has been limited by a narrow focus on one or two broadly-defined behaviors or styles of leadership, by confounding observable behaviors with leader traits and values, by a lack of attention to situational variables, and by the frequent use of weak research methods. The weaknesses in much of the research on charismatic and transformational leadership, and in the earlier research on task-oriented and relations-oriented behavior, are also limiting progress in the recent research on ethical, servant, and authentic leader- ship. The research on effects of leader behavior is briefly reviewed, the limita- tions are explained, and ways to improve future research are suggested.

Co py ri gh t © 2 01 4. I nf or ma ti on A ge P ub li sh in g. A ll r ig ht s re se rv ed . Ma y no t be r ep ro du ce d in a ny f or m wi th ou t pe rm is si on f ro m th e pu bl is he r, e xc ep t fa ir u se s pe rm it te d

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

EBSCO Publishing : eBook Collection (EBSCOhost) - printed on 1/2/2017 2:50 PM via TRIDENT UNIVERSITY AN: 826526 ; Schriesheim, Chester, Neider, Linda L..; Advances in Authentic and Ethical Leadership Account: s3642728

210  G. YUKL and J. W. MICHEL

INTRODUCTION

For decades, scholars have sought to understand what it takes to be an ef- fective organizational leader. Much of the theory and research on effective leadership involves the effects of leadership behavior on outcomes such as subordinate attitudes and performance. It has been common practice to define and measure leadership behavior in terms of broadly defined constructs or “metacategories.” The focus on one or two metacategories in most studies on leadership behavior has weakened results and made them more difficult to interpret. Other limitations in much of the behavior re- search include confounding of behavior with leader traits and values, reli- ance on weak research methods, and insufficient attention to situational variables and explanatory processes. In this chapter, we will briefly review what was found in decades of research on prominent behavior metacat- egories. Then, the weaknesses that have limited progress in the behavior research are described, and suggestions are presented for improving future research on effective leadership.

RESEARCH ON BEHAVIOR METACATEGORIES

In the early period of leadership theory and research from 1955 to 1980, the dominant metacategories were task-oriented behavior and relations-ori- ented behavior. From the early 1980s to the current time, much attention was devoted to research on charismatic and transformational leadership. In the past decade, there has been growing interest in ethical leadership, ser- vant leadership, and authentic leadership. Each type of leadership behavior is described, along with findings in the research on it.

Task-Oriented and Relations-Oriented Behavior

The relations-oriented metacategory includes behaviors that are primar- ily intended to improve interpersonal relations between the leader and a subordinate or among subordinates in a group. The task-oriented meta- category includes behaviors that are primarily intended to improve task performance by an individual or group. The labels used for the two meta- categories varied for different scholars; examples include Consideration and Initiating Structure (Fleishman, 1953; Halpin & Winter, 1957), Sup- portive and Instrumental Leadership (House, 1971), Employee-centered and Production-centered leadership (Likert, 1961), and Task-oriented and Relations-oriented leadership (Yukl, 1971). The specific component

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

A Critical Assessment of Research on Effective Leadership Behavior  211

behaviors for each metacategory also vary somewhat for different scholars and measures.

Examples of component behaviors for the relations-oriented metacat- egory include doing personal favors for a subordinate, listening to a sub- ordinate’s problems, defending a subordinate, and treating a subordinate as an equal. Later research also identified other component behaviors such as providing praise and recognition for subordinate achievements and contributions, and facilitating the development of skills relevant for a subordinate’s career success. Some versions of the relations-oriented metacategory include giving subordinates influence over leader decisions that affect them.

Examples of component behaviors for the task-oriented metacategory include assigning tasks to subordinates, clarifying role requirements, set- ting goals for individual or group performance, monitoring performance, and resolving problems that disrupt the work. In research on leadership in teams, additional task behaviors include planning, organizing, and co- ordinating team activities. Task-oriented and relations-oriented behavior can also occur in interactions with people outside of a leader’s work unit, but these behaviors were seldom examined closely in the early research, and they are also part of a different metacategory called external behavior (Yukl, 2012).

Many studies were conducted to determine how the two metacategories are related to leadership effectiveness (Bass, 2008). A meta-analysis of the results from the survey research using behavior description questionnaires found that both types of behavior are related to follower satisfaction, moti- vation, and job performance ( Judge, Piccolo, & Ilies, 2004). The strongest and most consistent finding is a positive correlation between relations-ori- ented behavior and subordinate satisfaction with the leader or leader-mem- ber relations. Results were weakest for studies with independent measures of subordinate or work-unit performance.

Unmeasured situational differences within and between studies are one reason for lack of stronger, more consistent results in research on task and relations metacategories. After the importance of the leadership situation was recognized, several contingency theories were proposed to explain how situational variables can enhance or limit the effects of a leader’s task-ori- ented and relations-oriented behaviors. The contingency theories based on behavior metacategories include Path-goal Theory (House, 1971), Leader- ship Substitutes Theory (Kerr & Jermier, 1978), and Situational Leadership Theory (Hersey & Blanchard, 1977). Many studies were conducted to test these contingency theories, but little support was found for them (Podsa- koff, MacKenzie, Ahearne, & Bommer, 1995; Thompson & Vecchio, 2009; Wofford & Liska, 1993).

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

212  G. YUKL and J. W. MICHEL

Charismatic and Transformational Leadership

The lack of progress in research on effects of task-oriented and relations- oriented behaviors encouraged leadership scholars to examine other types of leadership behavior, and much of the subsequent research involved test- ing theories of charismatic and transformational leadership. Charismatic leadership theories attempt to explain how leaders influence followers on an emotional and ideological level (Conger, 1989; Conger & Kanungo, 1998; House, 1977; Shamir, House, & Arthur, 1993). Conger and Kanungo (1998) proposed that charismatic behavior includes articulating an inno- vative strategic vision, taking personal risks and making self sacrifices to attain the vision, and identifying threats and opportunities in the external environment. The behaviors emphasized by House (1977) and Shamir et al. (1993) include articulating an appealing vision, emphasizing ideological aspects of work, communicating high performance expectations, express- ing confidence in subordinates, showing self-confidence, modeling exem- plary behavior, and doing things to increase identification with the team or organization. Some versions of the theory emphasize the importance of situational variables for attributions of charisma to a leader, but few empiri- cal studies were designed to assess the effects of situational variables.

Transformational leadership theories describe how some leaders influ- ence subordinate task motivation (Bass, 1985). The component behaviors vary somewhat for different scholars and even for the same scholars at dif- ferent points in time. A majority of the studies on transformational leader- ship have defined it in terms of four component behaviors identified by Bass and Avolio (1990). Idealized influence involves setting an example of task commitment and making self-sacrifices that benefit followers or the work unit. Intellectual stimulation involves encouraging others to view problems in a new way and find creative solutions. Individualized consideration in- volves providing support, encouragement, and coaching. Inspirational mo- tivation involves articulating an appealing vision and attempting to inspire commitment to the mission of the team or organization. Another measure of transformational leadership (Podsakoff, Mackenzie, Moorman, & Fet- ter, 1990) included six component behaviors: articulating an appealing vision, modeling appropriate behavior, providing individualized support, providing intellectual stimulation, fostering acceptance of group goals, and communicating high performance expectations. These behaviors were as- sumed to be relevant for all leaders, and few studies on transformational leadership have included situational variables.

Many leadership scholars regard charismatic and transformational lead- ership as equivalent constructs and use similar measures for them. Meta- analyses of this research find positive correlations with outcomes such as subordinate satisfaction, organizational commitment, job performance,

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

A Critical Assessment of Research on Effective Leadership Behavior  213

and work-unit performance (DeGroot, Kiker, & Cross, 2000; Judge & Pic- colo, 2004; Lowe, Kroeck, & Sivasubramaniam, 1996; Wang, Oh, Cour- tright, & Colbert, 2011). However, the results in studies with independent measures of leadership effectiveness were much weaker than studies with same-source measures, and some studies failed to support the theories (van Knipenberg& Sitkin, 2013). For example, research using case stud- ies of chief executives found that charismatic leadership was not required for effective organizational performance (e.g., Bennis & Nanus, 1985), and sometimes it resulted in weaker performance or a failed organization (e.g., Finkelstein, 2003; O’Connor, Mumford, Clifton, Gessner, & Connelly, 1995). In a longitudinal study of CEOs, ratings of charismatic leadership were correlated significantly with a company’s past financial performance but did not predict future performance (Angle, Nagarajan, Sonnenfeld, & Srinivasan, 2006).

Studies on transformational leadership sometimes include another be- havior metacategory called transactional leadership, which includes using reward contingencies to motivate subordinates, monitoring their perfor- mance, and taking corrective action when poor performance is found. Some measures of transactional leadership include examples of relations- oriented behavior (providing praise and recognition) and task-oriented be- havior (active monitoring). Research on the effects of transactional behav- ior suggests that it can have a positive effect on subordinate performance in some situations, but negative effects can also occur (Lowe et al., 1996). Bass (1985) proposed that effective leaders use a combination of transfor- mational and transactional leadership, but few studies have examined how the two metacategories jointly influence independent measures of work- unit performance.

Ethical Leadership

Ethical leadership has been defined in many different ways, and the con- struct usually includes a combination of values and behaviors (Brown & Treviño, 2006; Brown, Treviño, & Harrison, 2005). One key attribute of ethical leadership behavior involves acting in a way that is as consistent with widely accepted ethical standards. Behaviors commonly regarded as mor- ally correct include treating people fairly, providing accurate information and honest answers to questions, keeping promises and commitments, ob- serving the same rules and standards applied to others, and acknowledging responsibility for mistakes while also seeking to correct them.

Another type of ethical leadership behavior involves attempts to influ- ence the ethical behavior of others (Treviño, Brown, & Hartman, 2003). Examples include leader statements about the importance of ethics;

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

214  G. YUKL and J. W. MICHEL

dissemination of ethical guidelines for members of the organization; mod- eling ethical behavior to set an example for others, including ethical be- havior in the assessment of performance; and criticizing or punishing un- ethical behavior. A limitation of this aspect of ethical leadership involves questions about the extent to which it is appropriate for leaders to set moral standards for others in the organization or to make subjective judgments about the morality of subordinates.

Research on the consequences of ethical leadership is still limited, but several studies have found more employee satisfaction, task commitment, organizational citizenship behaviors, willingness to report problems, and ethical behavior (e.g., Kacmar, Bachrach, Harris, & Zivnuska, 2011; Mayer, Aquino, Greenbaum, & Kuenzi, 2012; Mayer, Kuenzi, Greenbaum, Bardes, & Salvador, 2009; Piccolo, Greenbaum, Den Hartog, & Folger, 2010; Wa- lumbwa, Morrison, & Christensen, 2012; Yukl, Mahsud, Hassan, & Prussia, 2013). However, most measures of ethical leadership include a diverse set of leader traits, values, and behaviors, and most studies only examined how the composite score on the measure was related to subordinate attitudes and behavior. How ethical leadership is related to objective measures of performance has not been closely examined, and some case studies found evidence that short-term company profits were increased by unethical prac- tices (e.g., Sims & Brinkman, 2003).

Servant Leadership

Servant leadership includes nurturing, defending, and empowering fol- lowers (Greenleaf, 1977), which are examples of relations-oriented behav- ior. Servant leadership also includes aspects of ethical leadership. Servant leaders must listen to followers, learn about their needs and aspirations, and be willing to share in their pain and frustration. Service includes nur- turing, defending, and empowering followers. Trust is established by being completely honest and open, keeping actions consistent with values, and demonstrating trust in followers. The servant leader must stand for what is good and right, even when it is not in the financial interest of the organiza- tion. Social injustice and inequality should be opposed whenever possible. Different questionnaires have been developed to measure servant leader- ship (Barbuto & Wheeler, 2006; Dennis & Bocarnea, 2005; Liden, Wayne, Zhao, & Henderson, 2008; van Dierendonck & Nuijten, 2011), but the best way to define and measure this construct has not been resolved. Most mea- sures require respondents to make difficult judgments about a leader’s in- tegrity, authenticity, and stewardship.

Research on the consequences of servant leadership is still limited, but several studies found positive outcomes such as more subordinate

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

A Critical Assessment of Research on Effective Leadership Behavior  215

commitment, self-efficacy, and organizational citizenship behavior (e.g., Eh- rhart, 2004; Liden et al., 2008; Mayer, Bardes, & Piccolo, 2008; Neubert, Kacmar, Carlson, Chonko, & Roberts, 2008; van Dierendonck, 2011; Wa- lumbwa, Hartnell, & Oke, 2010). However, the research seldom included independent measures of work-unit performance, and the objective of serving followers is sometimes inconsistent with the objective of improving performance. How leaders can resolve tradeoffs in benefits for different stakeholders remains an unresolved question.

Authentic Leadership

Authentic leadership theories have been proposed by several scholars (Avolio, Gardner, Walumbwa, Luthans, & May, 2004; George, 2003; Ilies, Morgeson, & Nahrgang, 2005). The definition varies somewhat across dif- ferent versions, but they all emphasize the importance of leader integrity. Authentic leaders have positive core values (e.g., honesty, kindness, fair- ness, accountability, and optimism) that motivate them to do what is right and fair for followers. These leaders create a special type of relationship that includes high mutual trust, transparency, shared objectives, and em- phasis on follower welfare and development. Core component behaviors include keeping leader actions consistent with espoused and actual values, articulating an appealing vision, modeling appropriate behaviors, and ex- pressing optimism and encouragement when there are problems in accom- plishing task objectives. With regard to other leadership behaviors, there is less agreement among the different versions of the theory. The measures of authentic leadership include leader traits and values such as self-awareness and an internalized moral perspective in addition to observable behaviors. As with ethical and servant leadership, it is not clear to what extent each attribute is necessary for effective leadership or is only an ideal that any leader should strive to attain (Caza & Jackson, 2011).

The amount of research on authentic leadership is still limited, but a recent review by Gardner, Cogliser, Davis, and Dickens (2011) showed that it was related to follower job satisfaction and organizational commitment. However, few studies included objective measures of performance or iden- tified the independent effects and relative importance of the values and behaviors that define authentic leadership.

LIMITATIONS OF THE BEHAVIOR RESEARCH

Several conceptual and methodological limitations in much of the behavior research have made it more difficult to find strong, consistent results that

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

216  G. YUKL and J. W. MICHEL

explain how leaders influence individual, group, or organizational perfor- mance. The limitations include varying content for a metacategory, overlap among metacategories, unique and joint effects of component behaviors, confounding of behavior with traits and values, weak survey studies, failure to examine curvilinear relationships, failure to examine lagged effects, lack of attention to explanatory processes, lack of multiple independent crite- ria, and lack of attention to situational variables.

Varying Content for a Behavior Metacategory

Most of the behavior metacategories used in the leadership research lack a clear definition and adequate criteria for identification of relevant com- ponent behaviors. One basis for grouping specific behaviors into a metacat- egory is that they have the same objective. The task-oriented and relations- oriented metacategories are examples of a taxonomy based on leadership objectives. A limitation of this approach is that some specific types of leader behavior can be used to achieve more than one objective, and attempts to create measures of mutually exclusive metacategories may result in the deletion of these effective behaviors. Another limitation is ambiguity about what objectives should be considered when developing a behavior taxono- my. Behavior metacategories with important objectives such as facilitating change and influencing outsiders were not included in the early behavior research. The problems created by behaviors with multiple objectives and exclusion of relevant behaviors can be minimized by using accurate mea- sures of specific behaviors likely to influence all important outcomes for the type of leader studied and by focusing attention on these relationships rather than on results for the metacategories. Unfortunately, this type of leadership study is very rare.

The component behaviors for a metacategory are not the same for dif- ferent versions of a leadership theory, when the measures are developed by different researchers, or when the definition of the metacategory changes over time as more is learned about it. For example, the early definition of relations-oriented behavior did not explicitly include some of the com- ponent behaviors found in more recent measures of this metacategory (e.g., providing praise and recognition, increasing member confidence, en- couraging cooperation among subordinates, empowering subordinates). It is more difficult to compare results from different studies or to interpret results from a meta-analysis of many studies when the same component behaviors are not used in each study and only the composite score for a metacategory is used in the data analyses.

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

A Critical Assessment of Research on Effective Leadership Behavior  217

Overlap Among Metacategories

Conceptual overlap among supposedly different metacategories is a related problem in leader behavior research (DeRue, Nahrgang, Well- man, & Humphrey, 2011). Sometimes the same component behavior is included in different metacategories. For example, providing praise and recognition has been included in some measures of relations-oriented be- havior, transformational leadership, and transactional leadership. When the metacategories included in a study have some of the same content and only composite scores for the metacategories are used in the analy- ses, it is more difficult to interpret the results. A related problem occurs when the metacategory in a study includes some component behaviors from unmeasured metacategories. For example, measures of transforma- tional leadership include some relations-oriented behaviors (e.g., support- ing and developing subordinates) and some change-oriented behaviors (e.g., articulating an appealing vision, encouraging innovative thinking). Research that examined results for a broad range of specific behaviors found that the effects attributed to transformational leadership can be ac- counted for primarily by relations-oriented and change-oriented behaviors (Michel, Lyons, & Cho, 2011).

Unique and Joint Effects of Component Behaviors

Most leader behavior studies use only the composite score on a metacate- gory to assess the effects on outcomes rather than examining the unique ef- fects of specific component behaviors. Broadly defined categories of leader behavior have limited utility for understanding how leaders can influence work-unit performance. The component behaviors are not equally relevant for influencing performance; they have different relationships with medi- ating variables, and they may be affected in different ways by the context. For example, clarifying goals and problem solving are both task-oriented behaviors. However, clarifying goals helps to ensure that subordinates know what to do, how to do it, and the expected results; whereas, problem solving is used to deal with disruptions of normal operations and member behav- ior that is unsafe or illegal. Supporting and developing are both relations- oriented behaviors. However, supporting is used to show positive regard, build cooperative relationships, and help people deal with stressful situa- tions; whereas developing is used to increase subordinate skills and con- fidence. Sometimes effective leadership involves using a combination of complementary behaviors from the same metacategory or from different metacategories (Piccolo et al., 2012). The relatively small number of studies

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

218  G. YUKL and J. W. MICHEL

that examine the unique and joint effects of specific behaviors is one likely reason for the lack of stronger results in the behavior research.

Confounding Behavior With Values and Traits

Many leadership studies are attempts to test a theoretical conception of an ideal leader (e.g., authentic, ethical, and servant leadership), or a theory about universally effective forms of leader influence (e.g., transformational leadership). The theories and measures used to test them usually include personality traits and values as well as a mix of diverse behaviors. Behaviors are different from values or personality traits. Most leader behaviors can be directly observed, but values and traits are usually inferred from behavior or measured with some type of personality test or self-report questionnaire. Traits and values are relatively stable characteristics for adults, whereas behaviors can be changed with training and development interventions. Leader traits influence leader behaviors, but there is not a simple, direct causal relationship. When these different types of constructs are included in the same scale and results are analyzed for a composite score, their ef- fects will be confounded and the results difficult to interpret (Yukl, 2012). A better research strategy is to treat other types of leader attributes as a separate set of variables that can influence behavior or moderate its effects (DeRue et al., 2011).

Weak Survey Studies

The dominant method in the research on effects of leader behavior has been a survey study with a leader behavior questionnaire used by subordi- nates to retrospectively rate how much or how often a leader has used each type of behavior (Gardner, Lowe. Moss, Mahoney, & Cogliser, 2010). The high correlations commonly found among supposedly different behaviors in these studies suggest that the ratings of leader behavior are influenced by common response biases, implicit theories about effective leadership, and general satisfaction with the leader. When the same respondents (usually subordinates of each leader) provide the data for both the leader behavior and the outcomes, the correlations are likely to be inflated by respondent biases and attributions. These problems are not adequately assessed by the statistical analyses many researchers use to claim that common method vari- ance is not a problem.

Other research methods can be used to measure leadership behavior, including coding of behaviors in audio or video recordings, diaries, in- terviews, critical incidents, case studies, biographies, and documents for

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

A Critical Assessment of Research on Effective Leadership Behavior  219

prominent leaders (e.g., presidential speeches). Leader behavior can be manipulated in scenario studies, laboratory simulations, and field experi- ments. All research methods have limitations, and the best strategy is to use multiple methods that are relevant and feasible for the research ques- tion. Consistent findings from different methods enhance confidence in the findings (Bryman, 2004).

Curvilinear Relationships

Another limitation in most of the behavior research is to focus on linear relationships between leader behavior and outcomes when curvilinear re- lationships are more plausible. In some cases, the effects of a behavior are best described by an inverted U-shaped curve in which benefits increase up to an optimal point, after which they decline. In some other cases, the benefits from a behavior will increase until reaching an optimal point, after which more use of the behavior will not result in any additional increases in benefits. Even when doing more of a behavior does not reduce the ben- efits or have negative side effects, spending more time than necessary on a behavior means that the leader is losing the opportunity to use more ben- eficial types of behavior (Yukl, 2012). Several studies have found curvilinear relationships between leader behavior and outcomes (Fleishman & Harris, 1962; Kaplan & Kaiser, 2006; Pierce & Aguinis, in press; Stouten, van Dijke, Mayer, De Cremer, & Euwema, 2013). However, merely testing for curvilin- ear relationships is not enough, because it is difficult to detect them unless there are accurate measures of leader behavior and effectiveness.

Lagged Effects

Most leader behavior studies examine events that occur during a time interval that is too short for the independent variables to have an effect on the mediator or dependent variables. Cross-sectional research is unable to determine how leaders develop relationships with subordinates, improve motivation and optimism, build effective teams, and lead change initia- tives over weeks, months, or years. The effects for mediators, such as sub- ordinate skill and task commitment, usually occur sooner than effects on performance outcomes that depend on the mediators. The lagged effect is longer for some types of leader behaviors than for others, and some ef- fects are negative before they become positive. For example, introducing a major change often results in a temporary decline in performance before it results in significant improvements. Sometimes the beneficial effects of a leader’s action or decision are only temporary and will eventually vanish

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

220  G. YUKL and J. W. MICHEL

unless the leader continues to use appropriate behaviors. Reciprocal cau- sality may also occur, as when leader behavior is adjusted after receiving feedback about the initial effects of the behavior. A longitudinal study with repeated measures of behavior and outcomes is necessary to detect these complex relationships. Relatively few longitudinal studies have been con- ducted on the effects of leadership behavior, and some did not include accurate measures and an appropriate time interval.

Explanatory Processes

A major limitation in most of the leader behavior research is insufficient attention to mediating variables and causal processes that can explain lead- er influence on the type of outcomes of primary interest to the researcher (e.g., the performance of an individual subordinate, team, or organization). Prior research on organizational behavior has already identified common performance determinants for an individual, group, or organization, and they can be used to identify specific leadership behaviors likely to be effective in a given situation. Unfortunately, most leader behavior studies do not mea- sure any mediating variables, and they are seldom used to select relevant, spe- cific behaviors to include in a study. In studies that include only one or two of the relevant mediators, confounding with unmeasured mediators makes it difficult to interpret the results. Studies that examine mediators only for metacategories fail to identify differential mediating effects for the specific component behaviors and are less likely to find strong, clear results.

Multiple Criteria and Tradeoffs

Independent measures of leadership effectiveness are more useful for iden- tifying the effects of leader behavior than ratings provided by the same persons who describe the behavior, but same-source measures are much more com- mon in the leader behavior research. Relatively few studies examine the ef- fects of leader behaviors on a wide range of relevant performance outcomes (Gardner et al., 2010; Kaiser, Hogan, & Craig, 2008). Studies that measure only one or two outcomes are unable to detect unintended negative consequences for outcomes that are important but not measured. Examining a wide range of outcomes is especially important for ethical and servant leadership, because decisions or actions carried out to benefit subordinates in some way may harm them in other ways or at a later time. Moreover, when the primary objective is to enhance or protect subordinate welfare, a leader’s decisions and actions are more likely to have adverse consequences for other stakeholders and may reduce organizational performance.

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

A Critical Assessment of Research on Effective Leadership Behavior  221

Situational Variables

Many situational variables can enhance or limit the effects of leader be- havior on outcomes such as subordinate commitment and group perfor- mance (Howell, Dorfman, & Kerr, 1986; James & Brett, 1984; Yukl, 2012). Examples of situational variables include the nature of the work performed by the leader’s group (e.g., task complexity, structure and novelty, skill re- quirements, quality requirements); dependence on others for resources, information, approvals, and assistance; frequency and seriousness of dis- ruptions in work-unit operations (e.g., equipment failures, supply shortag- es, natural disasters, terrorist attacks, strikes or sabotage); leader authority, position power, and autonomy; and the amount of change and uncertainty in the external environment for the organization (e.g., new technology, in- tense competition, social-economic change).

Deficiencies in any of the performance determinants for an individual, group, or organization represent another type of situational variable, be- cause a leader can improve performance by eliminating these deficiencies (Yukl, 2013). For example, if subordinates lack essential skills and experi- ence to perform an important new task, the leader can provide coaching, arrange for them to get more training, provide personal coaching, hire more talented employees, or hire consultants to help with the work.

An important determinant of effectiveness for many leaders is the ability to adapt their behavior to fit changing conditions in their current position or different conditions in a new position (Yukl & Mahsud, 2010). Leaders with behavioral flexibility are able to use a wide range of specific behaviors, and they use behaviors that are relevant for the situation (Hart & Quinn, 1993; Hooijberg, 1996; Yukl & Mahsud, 2010).

The effects of situational variables are not the same for all the specific component behaviors in a metacategory, and the differential effects will not be found by examining only results for composite scores on behavior metacategories. To understand how the situation can enhance or constrain the effects of specific leader behaviors, it is necessary to have accurate mea- sures and adequate variance for the situational variables and analyses that examine how they jointly moderate the effects of specific leader behaviors. This type of research is difficult, and it is seldom conducted.

SUGGESTED IMPROVEMENTS IN FUTURE BEHAVIOR RESEARCH

To make faster progress in learning about effective leadership, it is desirable to improve the way behavior constructs are defined and methods used to study leader behavior. Both subjects are discussed in this section of the chapter.

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

222  G. YUKL and J. W. MICHEL

Behavior Constructs and Taxonomies

Despite the limitations of metacategories, they have benefits that justi- fy using them for theory development and research. Organizing specific behaviors with a common purpose into metacategories can facilitate the development of general theories of effective leadership and make them more parsimonious and easier to understand. The metacategories should be meaningful in terms of important leadership roles, functions, and objec- tives. Each metacategory should include distinct and observable compo- nent behaviors. A wide range of behaviors should be examined, including some that are not adequately represented in any of the popular metacat- egories. Specific behaviors that have multiple objectives should not be ex- cluded. An example of a hierarchical taxonomy was recently proposed by Yukl (2012) to integrate much of the previous behavior research. The tax- onomy has four behavior metacategories: task-oriented, relations-oriented, change-oriented, and external behavior. Some of the change-oriented and external behaviors have been overlooked in much of the research on effec- tive leadership.

The objective for change-oriented behavior is to encourage and facilitate collective learning, innovation, and changes that will improve the effective- ness of the leader’s work unit (Ekvall & Arvonen, 1991; Yukl, 1999; Yukl, Gordon, & Taber, 2002). Examples of specific change behaviors include ad- vocating why change is necessary, communicating a clear and compelling vi- sion of the benefits to be gained, proposing major changes or new initiatives, taking personal risks to get proposed changes approved, planning how to implement changes, encouraging innovation and flexibility, and encourag- ing collective learning (Yukl, 2012). The change-oriented behaviors did not get much attention in the early leadership research, and only a few of the component behaviors are represented in measures of charismatic and trans- formational leadership. Research on change-oriented behavior as a distinct metacategory is still limited, but a positive relationship has been found with outcomes such as job attitudes, managerial effectiveness, and independent ratings of leader performance (Barling, Weber, & Kelloway, 1996; Gil, Rico, Alcover, & Barrasa, 2005; Kim & Yukl, 1995; Michel et al., 2011).

The objective for external (or “boundary spanning”) behavior is to rep- resent the work unit, promote and defend work-unit interests, and get im- portant information, resources, and political support for it. Examples of specific external behaviors include networking to build and maintain favor- able relationships, scanning the external environment to identify threats and opportunities, negotiating agreements with other units or outsiders, and lobbying for resources and assistance from bosses or peers (Yukl, 2012). Only a small number of leadership studies have included external behavior, and one reason may be that subordinates—who are often the source of

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

A Critical Assessment of Research on Effective Leadership Behavior  223

information about leader behavior—are seldom able to directly observe a leader’s external behavior. Several studies have found a positive relation- ship between external behaviors and indicators of leadership effectiveness (e.g., Ancona & Caldwell, 1992; Druskat & Wheeler, 2003; Grinyer, Mayes, & McKiernan, 1990; Kim & Yukl, 1995; Kotter, 1982).

Ethical aspects of leader behavior also deserve more attention in future research on effective leadership. The research should determine if it is use- ful to treat ethical leader behavior as a distinct metacategory with unique component behaviors. In research on ethical, servant, and authentic lead- ership, it is essential to identify any specific behaviors not already repre- sented in other metacategories and to differentiate these behaviors from other types of constructs such as leader values and personality traits. Fu- ture research should examine the separate and joint effects of these unique behaviors and other specific relations-oriented behaviors that are used to benefit subordinates (Yukl et al., 2013). Studies should be conducted to ex- amine how leader traits and values influence a leader’s choice of behavior and moderate its effects. Finally, more attention should be paid to leader intentions and interpretation of the situation when studying the determi- nants and consequences of leader behavior.

Research Methods

Survey studies with convenience samples and same-source measures should not continue to be the primary method for studying the effects of leadership behavior. Whenever possible, leadership researchers should use stronger methods like field experiments and lab experiments with simu- lated groups to assess the effects of leadership behavior. Because each type of research method has limitations, it is desirable to use more than one method whenever possible.

When survey studies are conducted, the accuracy of behavior description questionnaires should be improved. Behavior ratings should be provided by respondents who are trained to recognize the behaviors and have ample opportunity to observe them. Whenever possible, descriptions of actual in- cidents involving the behaviors should also be obtained and used to verify the accuracy of the behavior ratings and to help understand the behaviors. Instead of convenience samples, researchers should find samples that are appropriate for the research objectives.

In future research on effective leadership, multiple outcomes should be measured independently at appropriate times, and the outcomes should include some objective measures of performance for the leader’s work unit. Relevant mediating processes should be assessed as well as situational vari- ables likely to enhance or limit the effects of leader behavior. Longitudinal

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

224  G. YUKL and J. W. MICHEL

research designs should be used to assess lagged and temporary effects, and the timing of measures should be consistent with the underlying causal processes in the model being tested.

Credible rival hypotheses and alternative explanations should be iden- tified and investigated whenever feasible. Analyses should be conducted for specific behaviors as well as for metacategories, and competing models should be compared to determine the relative benefits of the two types of constructs for predicting indicators of effective leadership. When alterna- tive methods of data analysis are feasible but they do not provide consistent results, the discrepancies should be reported rather than selecting the one with the most favorable results. Any serious limitations in the sample, mea- sures, or analyses should be clearly identified in the research report and suggestions made for avoiding them in future research.

CONCLUSIONS

The purpose of this review was to provide an overview of research on leadership behavior during the past half century and evaluate progress in learning about effective leadership in organizations. The conclusion is that despite some important findings, several limitations have made progress much slower than it should have been. Progress was limited by a focus on a one or two vague metacategories, confounding of observable behavior with other constructs, a lack of attention to the situation and explanatory processes, and infrequent use of strong research methods.

A hierarchical taxonomy of behavior metacategories can be useful for integrating the complex results found for the effects of specific, observable behaviors into a comprehensive and parsimonious model of causal rela- tionships. However, it is important to remember that behavior constructs are conceptual tools to facilitate theory development and research, and there is no magic answer to the question of how to describe effective lead- ership behavior. As more is learned about effective leadership, taxonomies for describing leadership behavior will continue to evolve and improve. It is likely that some unique component behaviors will be necessary to describe how leaders influence an individual subordinate, a team, or an organiza- tion. It may also be necessary to identify some unique behaviors for differ- ent types of leaders than the managers and administrators studied in most of the leadership literature (e.g., elected political leaders, leaders of social groups, religious leaders, coaches of sports teams), and for leadership in specific contexts such as conducting meetings, making decisions, or lead- ing a multicultural team.

Faster progress in future research will also require the use of stronger research methods, including experiments in which leaders are trained how

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

A Critical Assessment of Research on Effective Leadership Behavior  225

to use effective behaviors. Survey studies can be useful, but only if appro- priate samples are obtained, behavior is measured accurately, and relevant outcomes are measured independently. More longitudinal studies should be conducted to assess lagged effects, temporary effects, and reciprocal causality. Whenever possible, leader behavior studies should also include measures of relevant situational variables and mediating processes.

It will not be possible to make faster progress in efforts to understand ef- fective leadership until limitations in much of the behavior research during the past half century are acknowledged and better studies are conducted. Improvements in leadership research will provide many benefits, including the development of better theories and more useful practical applications. A good understanding of effective leadership is an important way to im- prove the performance of groups and organizations, and it is a subject that deserves more intensive and systematic investigation by leadership scholars.

REFERENCES

Ancona, D. G., & Caldwell, D. F. (1992). Bridging the boundary: External activity and performance in organizational teams. Administrative Science Quarterly, 37, 634–665.

Angle, B. R., Nagarajan, J. N., Sonnenfeld, J. A., & Srinivasan, D. (2006). Does CEO charisma matter? An empirical analysis of the relationships among organiza- tional performance, environmental uncertainty, and top management team perceptions of CEO charisma. Academy of Management Journal, 49(1), 161–174.

Avolio, B. J., Gardner, W. L., Walumbwa, F. O., Luthans, F., & May, D. R. (2004). Unlocking the mask: A look at the process by which authentic leaders impact follower attitudes and behaviors. Leadership Quarterly, 15, 801–823.

Barbuto, J. E., & Wheeler, D. W. (2006). Scale development and construct clarifi- cation of servant leadership. Group & Organization Management</emphasis>, 31(3), 300–326.

Barling, J., Weber, T., & Kelloway, E. K. (1996). Effects of transformational leader- ship training on attitudinal and financial outcomes: A field experiment. Jour- nal of Applied Psychology, 81, 827–832.

Bass, B. M. (1985). Leadership and performance beyond expectations. New York, NY: Free Press.

Bass, B. M. (2008). Handbook of leadership: A survey of theory and research. New York, NY: Free Press.

Bass, B. M., & Avolio, B. J. (1990). The implications of transactional and transfor- mational leadership for individual, team, and organizational development. In W. Pasmore & R.W. Woodman (Eds.), Research in organizational change and development (Vol. 4, pp. 231–272). Greenwich, CT: JAI.

Bennis, W. G., & Nanus, B. (1985). Leaders: The strategies for taking charge. New York, NY: Harper & Row.

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

226  G. YUKL and J. W. MICHEL

Brown, M. E., & Treviño, L. K. (2006). Ethical leadership: A review and future direc- tions. Leadership Quarterly, 17, 595–616.

Brown, M. E., Treviño, L. K., & Harrison, D. A. (2005). Ethical leadership: A social learning perspective for construct development and testing. Organizational Behavior and Human Decision Processes, 97, 117–134.

Bryman, A. (2004). Qualitative research on leadership: A critical but appreciative view. Leadership Quarterly, 15, 729–769.

Caza, A., & Jackson, B. (2011). Authentic leadership. In A. Bryman, D. Collinson, K. Grint, B. Jackson, & M. Uhl-Bien (Eds.), The Sage handbook of leadership (pp. 352–364).London, UK: Sage.

Conger, J. A. (1989). The charismatic leader: Behind the mystique of exceptional leader- ship. San Francisco, CA: Jossey-Bass.

Conger, J. A., & Kanungo, R. (1998). Charismatic leadership in organizations. Thou- sand Oaks, CA: Sage.

DeGroot, T., Kiker, D. S., & Cross, T. C. (2000). A meta-analysis to review organiza- tional outcomes related to charismatic leadership. Canadian Journal of Admin- istrative Science, 17, 356–371.

Dennis, R. S., & Bocarnea, M. (2005). Development of the servant leadership as- sessment instrument. Leadership and Organization Development Journal, 26(8), 600–615.

DeRue, D. S., Nahrgang, J. D., Wellman, N., & Humphrey, S. E. (2011). Trait and behavioral theories of leadership: An integration and meta-analytic test of their relative validity. Personnel Psychology, 64, 7–52.

Druskat, V. U., & Wheeler, J. V. (2003). Managing from the boundary: The effective leadership of self-managed work teams. Academy of Management Journal, 27, 435–457.

Ehrhart, M. G. (2004). Leadership and procedural justice climate as antecedents of unit-level organizational citizenship behavior. Personnel Psychology, 57, 61–94.

Ekvall, G., & Arvonen, J. (1991). Change-centered leadership: An extension of the two-dimensional model. Scandinavian Journal of Management, 7, 17–26.

Finkelstein, S. (2003). Why smart executives fail. New York, NY: Portfolio. Fleishman, E. A. (1953). The description of supervisory behavior. Personnel Psychol-

ogy, 37, 1–6. Fleishman, E. A., & Harris, E. F. (1962). Patterns of leadership behavior related to

employee grievances and turnover. Personnel Psychology, 15, 43–56. Gardner, W. L., Cogliser, C. C., Davis, K. M., & Dickens, M. P. (2011). Authentic lead-

ership: A review of the literature and research agenda. Leadership Quarterly, 22, 1120–1145.

Gardner, W. L., Lowe. K. B., Moss, T. W., Mahoney, K. T., & Cogliser, C. C. (2010). Scholarly leadership of the study of leadership: A review of The Leadership Quarterly’s second decade, 2000–2009. Leadership Quarterly, 21, 922–958.

George, B. (2003). Authentic leadership: Rediscovering the secrets to creating lasting value. San Francisco, CA: Addison-Wesley.

Gil, R., Rico, R., Alcover, C., & Barrasa, A. (2005). Change-oriented leadership, sat- isfaction, and performance in work groups. Journal of Managerial Psychology, 20, 312–328.

Greenleaf, R. K. (1977). Servant leadership. New York, NY: Paulist.

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

A Critical Assessment of Research on Effective Leadership Behavior  227

Grinyer, P. H., Mayes, D., & McKiernan, P. (1990). The sharpbenders: Achieving a sustained improvement in performance. Long Range Planning, 23, 116–125.

Halpin, A. W., & Winter, B. J. (1957). A factorial study of the leader behavior descrip- tions. In R. M. Stogdill & A. E. Coons (Eds.), Leader behavior: Its description and measurement. Columbus: Bureau of Business Research, Ohio State University.

Hart, L. S., & Quinn, E. R. (1993). Roles executives play: CEOs, behavioral complex- ity, and firm performance. Human Relations, 46(5), 543–575.

Hersey, P., & Blanchard, K. H. (1977). The management of organizational behavior (3rd ed.). Englewood Cliffs, NJ: Prentice-Hall.

Hooijberg, R. (1996). A multidimensional approach toward leadership: An ex- tension of the concept of behavioral complexity. Human Relations, 49(7), 917–947.

House, R. J. (1971). A path-goal theory of leader effectiveness. Administrative Science Quarterly, 16, 321–339.

House, R. J. (1977). A 1976 theory of charismatic leadership. In J. G. Hung & L. L. Larson (Eds.), Leadership: The cutting edge (pp. 189–207). Carbondale: South- ern Illinois University Press.

Howell, J. P., Dorfman, P. W., & Kerr, S. (1986). Moderator variables in leadership research. Academy of Management Review, 11, 82–102.

Ilies, R., Morgeson, F. P., & Nahrgang, J. D. (2005). Authentic leadership and eu- daemonic well-being: Understanding leader-follower outcomes. Leadership Quarterly, 16(3), 373–394.

James, L. R., & Brett, J. M. (1984). Mediators, moderators, and tests for mediation. Journal of Applied Psychology, 69, 307–321.

Judge, T. A., & Piccolo, R. F. (2004). Transformational and transactional leadership: A meta-analytic test of their relative validity. Journal of Applied Psychology, 89, 36–51.

Judge, T. A., Piccolo, R. F., & Ilies, R. (2004). The forgotten ones? The validity of consideration and initiating structure in leadership research. Journal of Ap- plied Psychology, 89, 36–51.

Kacmar, K. M., Bachrach, D. G., Harris, K. J., & Zivnuska, S. (2011). Fostering good citizenship through ethical leadership: Exploring the moderating role of gen- der and organizational politics. Journal of Applied Psychology, 96(3), 633–642.

Kaiser, R. B., Hogan, R., & Craig, S. B. (2008). Leadership and the fate of organiza- tions. American Psychologist, 63(2), 96–110.

Kaplan, R. E., & Kaiser, R. B. (2006). The versatile leader: Make the most of your strengths without overdoing it. San Francisco, CA: Pfeiffer.

Kerr, S., & Jermier, J. M. (1978). Substitutes for leadership: Their meaning and measurement. Organizational Behavior and Human Performance, 22, 375–403.

Kim, H., & Yukl, G. (1995). Relationships of self-reported and subordinate-reported leadership behaviors to managerial effectiveness and advancement. Leader- ship Quarterly, 6, 361–377.

Kotter, J. P. (1982). The general manager. New York, NY: Free Press. Liden, R. C., Wayne, S. J., Zhao, H., & Henderson, D. (2008). Servant leadership:

Development of a multidimensional measure and multi-level assessment. Leadership Quarterly, 19, 161–177.

Likert, R. (1961). New patterns of management. New York, NY: McGraw-Hill.

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

228  G. YUKL and J. W. MICHEL

Lowe, K. B., Kroeck, K. G., & Sivasubramaniam, N. (1996). Effectiveness of corre- lates of transformational and transactional leadership: A meta-analytic review of the MLQ literature. Leadership Quarterly, 7, 385–442.

Mayer, D. M., Aquino, K., Greenbaum, R. L., & Kuenzi, M. (2012). Who displays eth- ical leadership, and why does it matter? An examination of antecedents and consequences of ethical leadership. Academy of Management Journal, 55(1), 151–171.

Mayer, D. M., Bardes, M., & Piccolo, R. F. (2008). Do servant-leaders help satisfy fol- lower needs? An organizational justice perspective. European Journal of Work and Organizational Psychology, 17, 180–197.

Mayer, D. M., Kuenzi, M., Greenbaum, R., Bardes, M., & Salvador, R. (2009). How low does ethical leadership flow? Test of a trickle-down model. Organizational Behavior and Human Decision Processes, 108, 1–13.

Michel, J. W., Lyons, B. D., & Cho, J. (2011). Is the full-range model of leadership really a full-range model of effective leader behavior? Journal of Leadership and Organizational Studies, 18, 493–507.

Neubert, M. J., Kacmar, K. M., Carlson, D. S., Chonko, L. B., & Roberts, J. A. (2008). Regulatory focus as a mediator of the influence of initiating structure and servant leadership on employee behavior. Journal of Applied Psychology, 93, 1220–1233.

O’Connor, J., Mumford, M. D., Clifton, T. C., Gessner, T. L., & Connelly, M. S. (1995). Charismatic leaders and destructiveness: A historiometric study. Lead- ership Quarterly, 6, 529–555.

Piccolo, R. F., Bono, J. E., Heinitz, K., Rowold, J., Duehr, E., & Judge, T. A. (2012). The relative impact of complementary leader behaviors: Which matter most? Leadership Quarterly, 23, 567–581.

Piccolo, R. F., Greenbaum, R., Den Hartog, D. N., & Folger, R. (2010). The relation- ship between ethical leadership and core job characteristics. Journal of Organi- zational Behavior, 31, 259–278.

Pierce, J. R., & Aguinis, H. (in press). The too-much-of-a-good-thing effect in man- agement. Journal of Management.

Podsakoff, P. M., MacKenzie, S. B., Ahearn, M., & Bommer, W. H. (1995). Searching for a needle in a haystack: Trying to identify the illusive moderators of leader- ship behaviors. Journal of Management, 21, 423–470.

Podsakoff, P. M., Mackenzie, S. B., Moorman, R. H., & Fetter, R. (1990). Trans- formational leader behaviors and their effects on followers’ trust in leader, satisfaction, and organizational citizenship behaviors. Leadership Quarterly, 1, 107–142.

Shamir, B., House, R. J., & Arthur, M. B. (1993). The motivational effects of char- ismatic leadership: A self-concept based theory. Organization Science, 4, 1–17.

Sims, R. R., & Brinkman, J. (2003). Enron ethics (or culture matters more than codes). Journal of Business Ethics, 45, 243–256.

Stouten, J., van Dijke, M., Mayer, D. M., De Cremer, D., & Euwema, M. C. (2013). Can a leader be seen as too ethical? The curvilinear effects of ethical leader- ship. Leadership Quarterly, 24, 680–695.

Thompson, G., & Vecchio, R. P. (2009). Situational leadership theory: A test of three versions. Leadership Quarterly, 20, 837–848.

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .

A Critical Assessment of Research on Effective Leadership Behavior  229

Treviño, L. T., Brown, M., & Hartman, L. P. (2003). A qualitative investigation of perceived ethical leadership: Perceptions from inside and outside the execu- tive suite. Human Relations, 55, 5–37.

van Dierendonck, D. (2011). Servant leadership: A review and synthesis. Journal of Management, 37, 1228–1261.

van Dierendonck, D., & Nuijten, I. (2011). The servant leadership survey: Develop- ment and validation of a multidimensional measure. Journal of Business & Psychology, 26, 249–267.

van Knipenberg, D., & Sitkin, S. B. (2013). A critical assessment of charismatic- transformational leadership research: Back to the drawing board? Academy of Management Annals, 7, 1–60.

Walumbwa, F. O., Hartnell, C. A., & Oke, A. (2010). Servant leadership, procedural justice climate, service climate, employee attitudes, and organizational citi- zenship behavior: A cross-level investigation. Journal of Applied Psychology, 95, 517–529.

Walumbwa, F. O., Morrison, E. W., & Christensen, A. L. (2012). Ethical leadership and group in-role performance: The mediating roles of group conscientious- ness and group voice. Leadership Quarterly, 23, 953–964.

Wang, G., Oh, I. S., Courtright, S. H., & Colbert, A. E. (2011). Transformational leadership and performance across criteria and levels: A meta-analytic review of 25 years of research. Group & Organization Management, 36, 223–270.

Wofford, J. C., & Liska, L. Z. (1993). Path-goal theories of leadership: A meta-analy- sis. Journal of Management, 19, 858–876.

Yukl, G. (1971). Toward a behavioral theory of leadership. Organizational Behavior and Human Performance, 6, 414–440.

Yukl, G. (1999). An evaluative essay on current conceptions of effective leader- ship. European Journal of Work and Organizational Psychology, 8, 33–48.

Yukl, G. (2012, November). Effective leadership behaviors: What we know and what questions need more attention? Academy of Management Perspectives, 66–85.

Yukl, G. (2013). Leadership in organizations (8th ed.). Englewood Cliffs, NJ: Pearson. Yukl, G., Gordon, A., & Taber, T. (2002). A hierarchical taxonomy of leader behav-

ior: Integrating a half century of behavior research. Journal of Leadership and Organizational Studies, 9, 15–32.

Yukl, G., & Mahsud, R. (2010). Why flexible, adaptive leadership is important. Con- sulting Psychology Journal, 62(2), 81–93.

Yukl, G., Mahsud, R., Hassan, S., & Prussia, G. E. (2013). An improved measure of ethical leadership. Journal of Leadership & Organizational Studies, 23, 38–48.

un de r U. S. o r ap pl ic ab le c op yr ig ht l aw .