Contingency Theories of Leadership

By Hughes, R.L., Ginnett, R.C., Curphy, G.J.

Edited by Paul Ducham


In its earlier form (the vertical dyad linkage model), LMX was one of the simplest of the contingency models. Looking at our leader–follower– situation model, it is easy to see that LMX, even today, is largely about the process of relationship building between the leader and the follower. The situation has barely crept in, and only if we consider the desire to increase organizational effectiveness by maximizing the number of in-groups the leader might develop. From an application perspective, perhaps the biggest limitation of LMX is that it does not describe the specific behaviors that lead to high-quality relationship exchanges between the leader and the follower. Nonetheless, LMX, as opposed to some of the subsequent contingency models, continues to generate research into the present decade.


Like the other theories in this chapter, the normative decision model  was designed to improve some aspects of leadership effectiveness. In this case Vroom and Yetton explored how various leader, follower, and situational factors affect the degree of subordinates’ participation in the decision-making process and, in turn, group performance. To determine which situational and follower factors affect the level of participation and group performance, Vroom and Yetton first investigated the decision-making processes leaders use in group settings. They discovered a continuum of decision-making processes ranging from completely autocratic (labeled “AI”) to completely democratic, where all members of the group have equal participation (labeled “GII”). These processes are listed in Highlight 13.1.

Levels of Participation in the Normative Decision Model

Autocratic Processes
    AI: The leader solves the problem or makes the decision by himself or herself using the information available at the time.
    AII: The leader obtains any necessary information from followers, then decides on a solution to the problem herself. She may or may not tell followers the purpose of her questions or give information about the problem or decision she is working on. The input provided by them is clearly in response to her request for specific information. They do not play a role in the definition of the problem or in generating or evaluating alternative solutions.

Consultative Processes
  CI: The leader shares the problem with the relevant followers individually, getting their ideas and suggestions without bringing them together as a group. Then he makes a decision. This decision may or may not reflect the followers’ influence.
  CII: The leader shares the problem with her followers in a group meeting. In this meeting, she obtains their ideas and suggestions. Then she makes the decision, which may or may not reflect the followers’ influence.

Group Process
   GII: The leader shares the problem with his followers as a group. Together they generate and evaluate alternatives and attempt to reach agreement (consensus) on a solution. The leader’s role is much like that of a chairman, coordinating the discussion, keeping it focused on the problem, and making sure the critical issues are discussed. He can provide the group with information or ideas that he has, but he does not try to press them to adopt “his” solution. Moreover, leaders adopting this level of participation are willing to accept and implement any solution that has the support of the entire group.


After establishing a continuum of decision processes, Vroom and Yetton established criteria to evaluate the adequacy of the decisions made— criteria they believed would be credible to leaders and equally applicable across the five levels of participation. Although a wide variety of criteria could be used, Vroom and Yetton believed decision quality and decision acceptance were the two most important criteria for judging the adequacy of a decision.
    Decision quality means simply that if the decision has a rational or objectively determinable “better or worse” alternative, the leader should select the better alternative. Vroom and Yetton intended quality in their model to apply when the decision could result in an objectively or measurably better outcome for the group or organization. In the for-profit sector, this criterion can be assessed in several ways, but perhaps the easiest to understand is, “Would the decision show up on the balance sheet?” In this case, a high-quality (or, conversely, low-quality) decision would have a direct and measurable impact on the organization’s bottom line. In the public sector, we might determine if there was a quality component to a decision by asking, “Will one alternative have a greater cost saving than the other?” or “Does this decision improve services to the client?” Although it may seem that leaders should always choose the alternative with the highest decision quality, this is not always the case. Often leaders have equally good (or bad) alternatives. At other times, the issue in question is trivial, rendering the quality of the decision relatively unimportant.
       Decision acceptance implies that followers accept the decision as if it were their own and do not merely comply with the decision. Acceptance of the decision outcome by the followers may be critical, particularly if the followers will bear principal responsibility for implementing the decision. With such acceptance, there will be no need for superiors to monitor compliance, which can be a continuing and time-consuming activity (and virtually impossible in some circumstances, such as with a geographically dispersed sales staff).
       As with quality, acceptance of a decision is not always critical for implementation. For example, most organizations have an accounting form that employees use to obtain reimbursement for travel expenses. Suppose a company’s chief financial officer (CFO) has decided to change the format of the form for reimbursing travel expenses and has had the new forms printed and distributed throughout the company. Further, she has sent out a notice that, effective June 1, the old forms will no longer be accepted for reimbursement—only claims made using the new forms will be processed and paid. Assuming the new form has no gross errors, problems, or omissions, our CFO really has no concern with acceptance as defined here. If people want to be reimbursed for their travel expenses, they will use the new form. This decision, in essence, implements itself.
        On the other hand, leaders sometimes assume that they do not need to worry about acceptance because they have so much power over their followers that overt rejection of a decision is not likely to occur. A corporate CEO is not apt to see a junior accountant stand up and openly challenge the CEO’s decision to implement a new policy, even though the young accountant may not buy into the new policy at all. Because followers generally do not openly object to the decisions made by leaders with this much power, these leaders often mistakenly assume that their decisions have been accepted and will be fully implemented. This is a naive view of what really goes on in organizations. Just because the junior subordinate does not publicly voice his opposition does not mean he will rush to wholeheartedly implement the decision. In fact, the junior accountant has a lot more time to destructively undermine the policy than the CEO does to ensure that it is being carried out to the letter.


Having settled on quality and acceptance as the two principal criteria for effective decisions, Vroom and Yetton then developed a normative decision model. (A normative model is based on what ought to happen rather than describing what does happen.) They also developed a set of questions to protect quality and acceptance by eliminating decision processes that would be wrong or inappropriate. Generally, these questions concern the problem itself, the amount of pertinent information possessed by the leader and followers, and various situational factors.
      To make it easier for leaders to determine how much participation subordinates should have to optimize decision quality and acceptance, Vroom and Yetton incorporated these questions into a decision tree (see Figure 13.1). To use the decision tree, we start at the left by stating the problem and then proceed through the model from left to right. Every time a box is encountered, the question associated with that box must be answered with either a yes or a no response. Eventually all paths lead to a set of decision processes that, if used, will lead to a decision that protects both quality and acceptance.
      Having reached a set of feasible alternatives that meet the desirable criteria for quality and acceptance among followers, the leader may then wish to consider additional criteria. One practical consideration is the amount of time available (see Highlight 13.2). If time is critical, the leader should select the alternative in the feasible set that is farthest to the left, again noting that the feasible set is arranged from AI through GII. It generally takes less time to make and implement autocratic decisions than it does to make consultative or group decisions. Nevertheless, the first step is to protect quality and acceptance (by using the model). Only after arriving at an appropriate set of outcomes should leaders consider time in the decision-making process. This tenet is sometimes neglected in the workplace by leaders who overemphasize time as a criterion. Obviously there are some situations where time is absolutely critical, as in life-or-death emergencies. Certainly no one would have expected U.S. Airways Captain Chesley “Sully” Sullenberger to pull out his Vroom-Yetton decision model after his Airbus A320 struck a flock of geese and he found himself plummeting toward the Hudson River in what had become a very large glider. But too often leaders ask for a decision to be made as if the situation were an emergency when, in reality, they (the leaders, not the situation) are creating the time pressure. Despite such behavior, it is difficult to imagine a leader who would knowingly prefer a fast decision that lacks both quality and acceptance among the implementers to one that is of high quality and acceptable to followers but that takes more time.
      Another important consideration is follower development. Again, after quality and acceptance have been considered using the decision tree, and if the leader has determined that time is not a critical element, she may wish to follow a decision process more apt to allow followers to develop their own decision-making skills. This can be achieved by using the decision tree and then selecting the alternative within the feasible set that is farthest to the right. The arrangement of processes from AI to GII provides an increasing amount of follower development by moving from autocratic to group decisions.
     Finally, if neither time nor follower development is a concern and multiple options are available in the feasible set of alternatives, the leader may select a style that best meets his or her needs. This may be the process with which the leader is most comfortable (“I’m a CII kind of guy”), or it may be a process in which he or she would like to develop more skill.

figure 13.1

How Much Time Do I Have?

In a world of instant messages that require lightning-fast responses, Steven B. Sample, president of the University of Southern California, is touting the benefits of “artful procrastination.” In his course on leadership and his book The Contrarian’s Guide to Leadership, a key lesson is never make a decision today that can reasonably be put off to tomorrow:

      With respect to timing, almost all great leaders have understood that making quick decisions is typically counterproductive. I’m not talking about what to have for breakfast or what tie to wear today. President Harry Truman almost personified this concept. When anyone told him they needed a decision, the first thing he would ask is “How much time do I have—a week, 10 seconds, six months?” What he understood was that the nature of the decision that a leader makes depends to a large extent on how much time he has in which to make it. He also understood that delaying a decision as long as reasonably possible generally leads to the best decisions being made.

Other lessons from Sample include these:

• Think gray. Don’t form opinions if you don’t have to.
• Think free. Move several steps beyond traditional brainstorming.
• Listen first, talk later. And when you listen, do so artfully.
• You can’t copy your way to the top.


The Situational Leadership ® model has evolved over time. Its essential elements first appeared in 1969, with roots in the Ohio State studies, in which the two broad categories of leader behaviors, initiating structure and consideration, were initially identified. As Situational Leadership ® evolved, so did the labels (but not the content) for the two leadership behavior categories. Initiating structure changed to task behaviors, which were defined as the extent to which a leader spells out the responsibilities of an individual or group. Task behaviors include telling people what to do, how to do it, when to do it, and who is to do it. Similarly, consideration changed to relationship behaviors, or how much the leader engages in two-way communication. Relationship behaviors include listening, encouraging, facilitating, clarifying, explaining why a task is important, and giving support.
         When the behavior of actual leaders was studied, there was little evidence to show these two categories of leader behavior were consistently related to leadership success; the relative effectiveness of these two behavior dimensions often depended on the situation. Hersey’s Situational Leadership ® model explains why leadership effectiveness varies across these two behavior dimensions and situations. It arrays the two orthogonal dimensions as in the Ohio State studies and then divides each of them into high and low segments (see Figure 13.3). According to the model, depicting the two leadership dimensions this way is useful because certain combinations of task and relationship behaviors may be more effective in some situations than in others.
        For example, in some situations high levels of task but low levels of relationship behaviors are effective; in other situations, just the opposite is true. So far, however, we have not considered the key follower or Situational Characteristics with which these combinations of task and relationship behaviors are most effective. Hersey says these four combinations of task and relationship behaviors would increase leadership effectiveness if they were made contingent on the readiness level of the individual follower to perform a given task.

Figure 13.3


In Situational Leadership ® , follower readiness refers to a follower’s ability and willingness to accomplish a particular task. Readiness is not an assessment of an individual’s personality, traits, values, age, and so on. It’s not a personal characteristic, but rather how ready an individual is to perform a particular task. Any given follower could be low on readiness to perform one task but high on readiness to perform a different task. An experienced emergency room physician would be high in readiness on tasks like assessing a patient’s medical status, but could be relatively low on readiness for facilitating an interdepartmental team meeting to solve an ambiguous and complex problem like developing hospital practices to encourage collaboration across departments.


Now that the key contingency factor, follower readiness, has been identified, let us move on to another aspect of the figure—combining follower readiness levels with the four combinations of leader behaviors described earlier. The horizontal bar in Figure 13.3 depicts follower readiness as increasing from right to left (not in the direction we are used to seeing). There are four segments along this continuum, ranging from R1 (the lowest) to R4 (the highest). Along this continuum, however, the assessment of follower readiness can be fairly subjective. A follower who possesses high levels of readiness would clearly fall in the R4 category, just as a follower unable and unwilling (or too insecure) to perform a task would fall in R1.
     To complete the model, a curved line is added that represents the leadership behavior that will most likely be effective given a particular level of follower readiness. To apply the model, leaders should first assess the readiness level (R1–R4) of the follower relative to the task to be accomplished. Next a vertical line should be drawn from the center of the readiness level up to the point where it intersects with the curved line in Figure 13.3. The quadrant in which this intersection occurs represents the level of task and relationship behavior that has the best chance of producing successful outcomes. For example, imagine you are a fire chief and have under your command a search-and-rescue team. One of the team members is needed to rescue a backpacker who has fallen in the mountains, and you have selected a particular follower to accomplish the task. What leadership behavior should you exhibit? If this follower has both substantial training and experience in this type of rescue, you would assess his readiness level as R4. A vertical line from R4 would intersect the curved line in the quadrant where both low task and low relationship behaviors by the leader are most apt to be successful. As the leader, you should exhibit a low level of task and relationship behaviors and delegate this task to the follower. On the other hand, you may have a brand-new member of the fire department who still has to learn the ins and outs of firefighting. Because this particular follower has low task readiness (R1), the model maintains that the leader should use a high level of task and a low level of relationship behaviors when initially dealing with this follower.
      Hersey suggests one further step leaders may wish to consider. The model just described helps the leader select the most appropriate behavior given the current level of follower readiness. However, there may be cases when the leader would like to see followers increase their level of readiness for particular tasks by implementing a series of developmental interventions to help boost follower readiness levels. The process would begin by first assessing a follower’s current level of readiness and then determining the leader behavior that best suits that follower in that task. Instead of using the behavior prescribed by the model, however, the leader would select the next higher leadership behavior. Another way of thinking about this would be for the leader to select the behavior pattern that would fit the follower if that follower were one level higher in readiness. This intervention is designed to help followers in their development, hence its name (see Highlight 13.3).

A Developmental Intervention Using SLT

Dianne is a resident assistant in charge of a number of students in a university dorm. One particular sophomore, Michael, has volunteered to work on projects in the past but never seems to take the initiative to get started on his own. Michael seems to wait until Dianne gives him explicit direction, approval, and encouragement before he will get started. Michael can do a good job, but he seems to be unwilling to start without some convincing that it is all right, and unless Dianne makes explicit what steps are to be taken. Dianne has assessed Michael’s readiness level as R2, but she would like to see him develop, both in task readiness and in psychological maturity. The behavior most likely to fit Michael’s current readiness level is selling, or high task, high relationship. But Dianne has decided to implement a developmental intervention to help Michael raise his readiness level. Dianne can be most helpful in this intervention by moving up one level to participating, or low task, high relationship. By reducing the amount of task instructions and direction while encouraging Michael to lay out a plan on his own and supporting his steps in the right direction, Dianne is most apt to help Michael become an R3 follower. This does not mean the work will get done most efficiently, however. As we saw in the Vroom and Yetton model earlier, if part of the leader’s job is development of followers, then time may be a reasonable and necessary trade-off for short-term efficiency.


In Figure 13.4 we can see how the factors in Situational Leadership ® fit within the L-F-S framework. In comparison to the Vroom and Yetton model, there are fewer factors to consider in each of the three elements. The only situational consideration is knowledge of the task, and the only follower factor is readiness. On the other hand, the theory goes well beyond decision making, which was the sole domain of the normative decision model.
          Situational Leadership ® is usually appealing to students and practitioners because of its commonsense approach as well as the ease of understanding it. Unfortunately there is little published research to support the predictions of Situational Leadership ® in the workplace. A great deal of research has been done within organizations that have implemented Situational Leadership ® , but most of those findings are not available for public dissemination.
         In 2007 Blanchard modified the Situational Leadership ® prescriptions to specify more clearly the four definitions of follower developmental level and their four corresponding optimal styles of leadership. Although this revision of the model, perhaps as a result of much criticism concerning the lack of prescriptive specificity, does create a more discrete ypology of follower styles, recent research suggests that the original model is a better predictor of subordinate performance and attitudes than the revised version.
        Nevertheless, even with these shortcomings, Situational Leadership ® is a useful way to get leaders to think about how leadership effectiveness may depend somewhat on being flexible with different subordinates, not on acting the same way toward them all.

figure 13.4


To determine a leader’s general style or tendency, Fiedler developed an instrument called the least preferred co-worker (LPC) scale. The scale instructs a leader to think of the single individual with whom he has had the greatest difficulty working (that is, the least preferred co-worker) and then to describe that individual in terms of a series of bipolar adjectives (such as friendly–unfriendly, boring–interesting, and sincere–insincere). Those ratings are then converted into a numerical score.
          In thinking about such a procedure, many people assume that the score is determined primarily by the characteristics of whatever particular individual the leader happened to identify as his least preferred coworker. In the context of contingency theory, however, the score is thought to represent something about the leader, not the specific individual the leader evaluated.
         The current interpretation of these scores is that they identify a leader’s motivation hierarchy. Based on their LPC scores, leaders are categorized into two groups: low-LPC leaders and high-LPC leaders. In terms of their motivation hierarchy, low-LPC leaders are motivated primarily by the task, which means these leaders gain satisfaction primarily from task accomplishment. Thus their dominant behavioral tendencies are similar to the initiating structure behavior described in the Ohio State research or the task behavior of SLT. However, if tasks are being accomplished in an acceptable manner, low-LPC leaders will move to their secondary level of motivation, which is forming and maintaining relationships with followers. Thus low-LPC leaders will focus on improving their relationships with followers after they are assured that assigned tasks are being satisfactorily accomplished. If tasks are no longer being accomplished in an acceptable manner, however, low-LPC leaders will refocus their efforts on task accomplishment and persist with these efforts until task accomplishment is back on track.
      In terms of motivation hierarchy, high-LPC leaders are motivated primarily by relationships, which means these leaders are satisfied primarily by establishing and maintaining close interpersonal relationships. Thus their dominant behavioral tendencies are similar to the consideration behaviors described in the Ohio State research or the relationship behaviors in SLT. If high-LPC leaders have established good relationships with their followers, they will move to their secondary level of motivation, which is task accomplishment. As soon as leader–follower relations are jeopardized, however, high-LPC leaders will cease working on tasks and refocus their efforts on improving relationships with followers.
       You can think of the LPC scale as identifying two different sorts of leaders, with their respective motivational hierarchies depicted in Figure 13.5. Lower-level needs must be satisfied first. Low-LPC leaders will move “up” to satisfying relationship needs when they are assured the task is being satisfactorily accomplished. High-LPC leaders will move “up” to emphasizing task accomplishment when they have established good relationships with their followers.
       Because all tests have some level of imprecision, Fiedler suggested that the LPC scale cannot accurately identify the motivation hierarchy for individuals with intermediate scores. Research by Kennedy suggested an alternative view. Kennedy has shown that individuals within the intermediate range of LPC scale scores may more easily or readily switch between being task- or relationship-oriented leaders than those individuals with more extreme scale scores. They may be equally satisfied by working on the task or establishing relationships with followers.

figure 13.5


The other important variable in the contingency model is situational favorability, which is the amount of control the leader has over the followers. Presumably the more control a leader has over followers, the more favorable the situation is, at least from the leader’s perspective. Fiedler included three subelements in situation favorability. These were leader– member relations, task structure, and position power.
      Leader–member relations are the most powerful of the three subelements in determining overall situation favorability. They involve the extent to which relationships between the leader and followers are generally cooperative and friendly or antagonistic and difficult. Leaders who rate leader– member relations as high feel they have the support of their followers and can rely on their loyalty.
     Task structure is second in potency in determining overall situation favorability. Here the leader objectively determines task structure by assessing whether there are detailed descriptions of work products, standard operating procedures, or objective indicators of how well the task is being accomplished. The more one can answer these questions affirmatively, the higher the structure of the task.
     Position power is the weakest of the three elements of situational favorability. Leaders who have titles of authority or rank, the authority to administer rewards and punishments, and the legitimacy to conduct follower performance appraisals have greater position power than leaders who lack them.
         The relative weights of these three components, taken together, can be used to create a continuum of situational favorability. When using the contingency model, leaders are first asked to rate items that measure the strength of leader–member relations, the degree of task structure, and their level of position power. These ratings are then weighted and combined to determine an overall level of situational favorability facing the leader. Any particular situation’s favorability can be plotted on a continuum Fiedler divided into octants representing distinctly different levels of situational favorability. The relative weighting scheme for the subelements and how they make up each of the eight octants are shown in Figure 13.6.
       You can see that the octants of situational favorability range from 1 (highly favorable) to 8 (very unfavorable). The highest levels of situational favorability occur when leader–member relations are good, the task is structured, and position power is high. The lowest levels of situational favorability occur when there are high levels of leader–member conflict, the task is unstructured or unclear, and the leader does not have the power to reward or punish subordinates. Moreover, the relative weighting of the three subelements can easily be seen by their order of precedence in Figure 13.6, with leader–member relations appearing first, followed by task structure, and then position power. For example, because leader–member relations carry so much weight, it is impossible for leaders with good leader–member relations to have anything worse than moderate situational favorability, regardless of their task structure or position power. In other words, leaders with good leader–member relations will enjoy situational favorability no worse than octant 4; leaders with poor leader–member relations will face situational favorability no better than octant 5.

Figure 13.6


Before reviewing the empirical evidence, perhaps we can attain a clearer understanding of the contingency model by examining it through the L-F-S framework. As shown in Figure 13.8, task structure is a function of the situation, and LPC scores are a function of the leader. Because position power is not a characteristic of the leader but of the situation the leader finds himself or herself in, it is included in the situational circle. Leader–member relations are a joint function of the leader and the followers; thus they belong in the overlapping intersection of the leader and follower circles.
       As opposed to the dearth of evidence for Hersey and Blanchard’s situational theory, Fiedler and his fellow researchers have provided considerable evidence that the predictions of the model are empirically valid, particularly in laboratory settings. However, a review of the studies conducted in field settings yielded only mixed support for the model. Moreover, researchers have criticized the model for the uncertainties surrounding the meaning of LPC scores, the interpretation of situational favorability, and the relationships between LPC scores and situational favorability. Despite such questions, however, the contingency model has stimulated considerable research and is the most validated of all leadership theories.

FIgure 3.8


The four types of leader behavior in path–goal theory can be seen in Table 13.2. Like SLT, path–goal theory assumes that leaders not only may use varying styles with different subordinates but might also use differing styles with the same subordinates in different situations. Path–goal theory suggests that, depending on the followers and the situation, these different leader behaviors can increase followers’ acceptance of the leader, enhance their level of satisfaction, and raise their expectations that effort will result in effective performance, which in turn will lead to valued rewards (see Highlight 13.5).

TABLE 13.2 The Four Leader Behaviors of Path–Goal Theory

Directive leadership. These leader behaviors are similar to the task behaviors from SLT. They include telling the followers what they are expected to do, how to do it, when it is to be done, and how their work fits in with the work of others. This behavior would also include setting schedules, establishing norms, and providing expectations that followers will adhere to established procedure and regulations.

Supportive leadership. Supportive leadership behaviors include having courteous and friendly interactions, expressing genuine concern for the followers’ well-being and individual needs, and remaining open and approachable to followers. These behaviors, which are similar to the relationship behaviors in SLT, also are marked by attention to the competing demands of treating followers equally while recognizing status differentials between the leader and the followers.

Participative leadership. Participative leaders engage in the behaviors that mark the consultative and group behaviors described by Vroom and Yetton. As such, they tend to share work problems with followers; solicit their suggestions, concerns, and recommendations; and weigh these inputs in the decision-making process.

Achievement-oriented leadership. Leaders exhibiting these behaviors would be seen as both demanding and supporting in interactions with their followers. First they would set challenging goals for group and follower behavior, continually seek ways to improve performance en route, and expect the followers to always perform at their highest levels. But they would support these behaviors by exhibiting a high degree of ongoing confidence that subordinates can put forth the necessary effort; will achieve the desired results; and, even further, will assume even more responsibility in the future.

Shifting Behaviors at Caterpillar

James Despain was a leader with a very directive leadership style. He began his career at Caterpillar Inc. as a young man, sweeping the factory floor. He followed the lead of others of his generation—the 1950s were a time when leaders were the ultimate authority and words like participative and consultative were unheard of. Despain worked his way into supervisory positions and finally was named vice president of the track-type tractor division. Despain claims he “spent much of [his] career as a manager focusing on what employees were doing wrong.” He focused on the tasks at hand and little else. But in the early 1990s Despain had to face some hard facts: his $1.2 billion division was losing millions of dollars per year, his management team was getting hundreds of grievances from their employees, and morale at the Caterpillar plant was extremely low.
        Despain and his leadership group identified the need for a strategic plan to transform the working culture. Key to the plan was determining a strategy for dealing with employee attitudes and behavior. Despain and his transformation team identified nine behaviors or “common values” that they wanted every employee to emulate every day— trust, mutual respect, customer satisfaction, a sense of urgency, teamwork, empowerment, risk taking, continuous improvement, and commitment. Employee evaluations were based on the manifestation of these behaviors. Above and beyond those behaviors, top executives and management were expected to lead by example and commit themselves to practice 100 positive leadership traits. Statements such as “I will know every one of my employees by name . . . will recognize their accomplishments with praise . . . will trust my employees to do their work” became the new mantras for those in charge.
       Through this process, Despain came to understand that “the most important thing for employees in the workplace is to achieve self-worth.” The principal change he was striving to achieve was to make employees accountable for how their jobs got done; for workers that meant stretching a little more every day to achieve their full potential. For managers it meant shifting away from achieving traditional metrics and toward drawing desired behavior from workers. “And we found that the more we focused on behavior, the better the metrics got.” The result: Despain’s division cut its breakeven point in half within five years of launching the transformation.


Before getting into the research surrounding path–goal theory, you may wish to examine the theory using the L-F-S framework. As shown in Figure 13.11, the components of path–goal theory fit nicely into the L-F-S model. The four leader behaviors fit into the leader circle, the characteristics of the followers fit into the follower circle, and the task and the formal authority system fit into the situation circle. Of all the components of path–goal theory, the only “mismatch” with the L-F-S model deals with the primary work group. The norms, cohesiveness, size, and stage of development of groups are considered to be part of the follower function in the L-F-S model but are part of the situation function in path–goal theory. In that regard, we hasten to note we use the L-F-S framework primarily for heuristic purposes. Ultimately the concepts described in these five theories are sufficiently complex and ambiguous that there probably is no right answer in any single depiction.
         In terms of research, the path–goal theory has received only mixed support to date. Although many of these mixed findings may be due to the fact that the path–goal theory excludes many of the variables found to impact the leadership process, that may also be due to problems with the theory. Yukl maintained that most of these criticisms deal with the methodology used to study path–goal theory and the limitations of Expectancy Theory. Moreover, the path–goal theory assumes that the only way to increase performance is to increase followers’ motivation levels. The theory ignores the roles leaders play in selecting talented followers, building their skill levels through training, and redesigning their work.
       Nonetheless, path–goal theory is useful for illustrating two points. First, as noted by Yukl, “path–goal theory has already made a contribution to the study of leadership by providing a conceptual framework to guide researchers in identifying potentially relevant situational moderator variables.” Path–goal theory also illustrates that, as models become more complicated, they may be more useful to researchers and less appealing to practitioners. Our experience is that pragmatically oriented students and in-place leaders want to take something from a model that is understandable and can be applied in their work situation right away. This does not mean they prefer simplicity to validity—they generally appreciate the complexity of the leadership process. But neither do they want a model that is so complex as to be indecipherable.

figure 13.11