Psychologists who have advanced normative theories of management have typically advocated highly participative processes for making decisions. The principal basis for such prescriptions is the motivational benefit that results from a leader involving group members in decision making. In spite of this advocacy, reviews of the literature suggest a much more mixed picture of the consequences of participation.
One way of reconciling the inconsistent evidence is to attempt to identify the moderating variables that regulate these different effects. Such moderating variables could then be incorporated into a contingency theory to guide managers in selecting the degree of participation appropriate to each situation. In the early 1970s, Victor Vroom, working with a graduate student, Philip Yetton, formulated a normative model of leadership style that had that objective. Expressed as a decision tree, the model distinguished five degrees of participation and eight situational factors believed to interact with participation in determining its effectiveness. The Vroom-Yetton model inspired many studies aimed at determining its validity as well as its usefulness in leadership training. The validity data, summarized 15 years later by Vroom and Jago, showed that the incidence of successful decisions was about twice as high when the decision process used was consistent with the model as when it was inconsistent. Clearly the model had promise, but the research suggested that there was much room for improvement.
In the three decades since its original publication, the Vroom-Yetton model has been substantially revised, first by Vroom and Arthur Jago in 1988 and in 2000 by Vroom. Its current structure is shown as a balance scale in Figure 1.
The five decision processes have undergone significant modification from the Vroom-Yetton model, as have the eight situational factors, which have been expanded to 11. The factors at the left-hand side drive the recommended process toward the more autocratic end of the spectrum, whereas those at the right favor a more participative approach. Finally, the two in the center, decision significance and likelihood of disagreement, interact with those at the left or right to determine the sensitivity of the scale. For example, when a highly significant decision is combined with factors at the left, the recommended process is shifted further toward autocratic methods. When it is combined with factors at the right, it will shift further toward participation.
Of course, the scale is only a metaphor for the actual model, which is driven by a set of equations. To use the model, a manager, faced with a specific problem to solve or decision to make, is asked for judgments (typically on a five-point scale) concerning each of the 11 factors. These judgments are entered into four equations that estimate the effects of each of the five processes on the quality of the decision, its likely implementation, the time consumed in making it, and the developmental benefits resulting from the process. Finally these four consequences (quality, implementation, time, and development) receive differential weights corresponding to the manager’s judgments of their importance in that problem.
Using the model sounds complicated but can be accomplished in less than one minute using a computer program called Expert System. Once the judgments are entered, the manager sees not a single recommended process (as in the Vroom-Yetton model) but a bar graph showing the relative estimated effectiveness of each of the five processes.
Scientific Implications of Normative Theories
It has been said that a theory should be evaluated not only in terms of its validity but also in terms of the questions it raises and the quality of research it stimulates. Jago has recently compiled a list of more than 100 studies in scientific journals and more than 40 doctoral dissertations dealing directly with the Vroom-Yetton-Jago models. The models have also stimulated the development of a novel measure of leadership style that has proven useful both in research and in leadership development. The measure uses a set of 30 real or realistic cases, each depicting a manager faced with a decision to make that would affect his or her team. For each case, the manager chooses from the five alternative decision processes the one that he or she would select. The cases are not selected randomly, but rather on the basis of a multi-factorial experimental design in which eight principal situational factors are varied independently of one another. This property makes it possible to systematically determine how managers change their intended behavior as elements of the situation are changed.
With the advent of the Internet, it is now possible for a manager to view and respond to the cases online in one of several available languages. As an inducement to enter their choices, managers can choose two groups from a list, varying in organizational level, nationality, and industry, with whom they would like to be compared. Finally, the manager downloads a 12-page individualized report comparing his or her style with the model and with the chosen comparison groups.
The data obtained from this measure has taught us a lot about the correlates of leadership style including the influence of nationality, gender, functional specialty, and hierarchical position. Managers do vary their behavior over situations in a manner not unlike that shown in the previous figure. However, they differ from one another in two respects. The most obvious is a preference for one side of the scale or another. This is similar to what is meant by describing mangers as autocratic or participative. But it should not be thought of as a general trait, because it accounts for only 10% of the total variance in behavior. Using the metaphor of the balance scale, it can be thought of as an extra weight added to one side of the scale or the other.
The other respect in which managers differ is the specific situational factors that govern their choices among the five styles. Although the pattern shown in the figure is a reasonable approximation of that of a manager choosing the modal response on each of the 30 cases, each individual manager displays a different pattern, ignoring factors or sometimes responding to them in a manner opposite to that prescribed by the normative model.
Practical Implications of Normative Theories
Apart from the impact that the Vroom-Yetton-Jago models and derivative tools have had in the science of leadership, it is safe to say that it has had an even greater impact on the practice of leadership. About 200,000 managers around the world have now been trained in the models. Invariably, such training has included feedback on responses to sets of cases showing managers how they compare with their peers, with occupants of positions to which they aspire, and with the model. In effect, managers can compare their model of decision making and leadership with those of other groups and with the normative model.
- Vroom, V. H. (2000). Decision making and the leadership process. Organizational Dynamics, 28(4), 82-94.
- Vroom, V. H. (2003). Educating managers in decision making and leadership. Management Decision, 41(10), 968-978.
- Vroom, V. H., & Jago, A. (1988). The new leadership: Managing participation in organizations. Englewood Cliffs, NJ: Prentice Hall.
- Vroom, V H., & Yetton, P. W. (1973). Leadership and decision making. Pittsburgh, PA: University of Pittsburgh Press.