The duality of autocratic and democratic leadership has been seen throughout history. How should people be led? How should people be governed? How should people be guided? Two views prevailed, based on opposing doctrines about human nature. For centuries, theologians and philosophers argued that human nature was cursed by original sin or that human nature was blessed with the inherent ability to find salvation. The former proponents of original sin advised that people were essentially bad; they had to be controlled, directed, and uplifted by authority. The proponents of inherent salvation set forth that people were essentially good and they must be given freedom in which to act, learn, grow, and overcome. Adam Smith, a moral philosopher as well as a political economist, saw both sides of human nature. He agreed with David Hume that self-interest required a distrust of human intentions and motives, but at the same time he also believed in the innate goodness of human-kind and the power of education to enlighten and liberate (Herman, 2001).
The framers of the U.S. Constitution recognized the benefits and the costs of an independent executive with authority and a democratic legislative body. They handled the dilemma posed by balancing the powers of the executive against those of a bicameral legislature: two senators from each of the 13 sovereign states and population-based representatives of each state’s male, white, property-qualified electorate. This arrangement recognized the contribution of the democratic majority vote of statewide and electoral district representatives to the commitment, interests, and satisfaction of the various constituencies. At the same time, an indirectly elected president could provide the need for order and for the execution and maintenance of federal laws, foreign affairs, and national defense. Much authority and power for direction and control were given to the president, but with safeguards. The president was commander in chief of the armed forces, but he had to face reelection and could be removed from office by the legislature. He could veto the legislative decisions of the representatives, but the veto could be overridden by the legislature.
The same dilemma confronts organizational leaders today. They must balance the advantages of a more democratic approach, which ordinarily contributes to the commitment, loyalty, involvement, and satisfaction of followers, with a more autocratic approach, which contributes to order, consistency, and the resolution of conflict. Whitty and Butts (1989) argue that authoritarian and elitist leadership sets the rules and has the power to control behavior. But more culturally accepted principles encourage democratic sharing of influence to obtain higher productivity, friendly interpersonal relations, and group cohesiveness.
Leaders and managers vary in how they deal with the dilemma of autocratic leadership at one extreme and democratic leadership at the other. These dichotomies are simplifications. Sweeney, Fiechtner, and Samores (1975) conducted a factor analysis of the leadership focus of 103 male employed part-time college students that corroborated the high degree of complexity to be found in examining autocratic and democratic leadership. The factors were: (1) authoritarian role preference, (2) authoritarian role pressure, (3) egalitarian role preference, (4) egalitarian role pressure, (5) balanced manager, (6) people-oriented manager, (7) assumed similarity between opposites, (8) contemptuous indulgence, (9) supportive values, (10) tolerance of people; and (11) organizational tolerance. Nevertheless, the same persons who engage in one type of autocratic behavior when, say, initiating structure, are also likely to be seen as facilitating work and persuading. The same persons who engage in one type of democratic behavior are likely to be seen as supportive, considerate, and people-oriented as well, but the empirical correlations are far from perfect (Edwards & Rode, 1986).
The autocratic-authoritarian cluster encompasses being arbitrary, controlling, power-oriented, coercive, punitive, and closed-minded. The cluster has often been described in pejorative terms. Stripped of negatives (emphasized by so many social scientists), it means taking full and sole responsibility for decisions and control of followers’ performance. Autocrats stress obedience, loyalty, and strict adherence to roles. They make and enforce the rules. They see that decisions are carried out (Smither, 1991). Powerful autocratic leaders throughout history have often been praised for their ability to develop reliable and devoted followers and to act as the principal authority figures in establishing and maintaining order.
Authoritarian sports coaches may be disliked by their players. But this dislike may evolve into respect, appreciation, and fondness. Players may hate autocratic coaches but play well under them (Bernath, 1991). A list of the toughest bosses in America is compiled regularly by Fortune magazine. They are described as merciless with people who tell them less than the truth. They are intensely persistent, imperious, unwilling to entertain ideas that don’t fit with their own. They get the job done, but the cost to themselves and others may be high (Flax, 1984).
The dark side of autocratic behavior is abusiveness, creating fear and distrust, using arbitrary and unconditional punishment, ignoring subordinates’ information and inputs to decisions, and relying exclusively on one’s own judgment (Smither, 1991). Henry Ford II was described by Lee Iacocca (1984), who worked directly under him, as the consummate autocrat in control of Ford Motor Company. Ford was bigoted and hypocritical. He would fire people without reason or for trivial reasons such as not wearing the right clothes. But autocrats can also be considerate. Pope John Paul II was called a spiritual dictator; he declared autocratically, “The Church is not a democracy and no one from below can decide the truth” (November 20, 1998). Yet repeatedly, he could show personal concern and sympathy for others whether or not they liked or agreed with him (Bane, 2005).
Abusive leaders engage in hostile verbal and nonverbal behaviors (Tepper, 2000). Lyndon Johnson, exemplifying the autocratic taskmaster, abused and exhausted his staff. He paid them the lowest salaries and worked them the longest hours. He played favorites and publicly humiliated his loyal aides (Reedy, 1982). He was terrorizing (though he could be tender at times), ruthless, impatient, petty, bullying, devious, brutal, suspicious, and crafty. Yet he could sometimes also be caring, patient, devoted to family and friends, compassionate, and generous (Valenti, 1975). Another autocratic leader was a terminated former president of Kellogg Foods, who was abrasive, abrupt, unwilling to listen, and inclined to manage without being questioned (Manz & Sims, 1991). In a review of the literature, Smither (1991) conceived of authoritarian managers in work organizations as taking full responsibility for decisions and subordinate performance. They believe their assignment is to make decisions and see that the decisions are carried out. They do not share power. They may be exploitative or benevolent in their relations with others (R. Likert, 1961a). Wallechinsky (2003) listed as the 10 worst living political dictators those who suppressed freedom and human rights taken for granted in democracies, executed or jailed political opponents, tortured prisoners, caused their citizens to starve, and used violence domestically and to interfere in other countries. Joseph Stalin was a model adopted by Saddam Hussein (Montefiore, 2004).
Although investigations use many terms whose meanings do not entirely overlap, correlations generally will be high among descriptions of various authoritarian ways of organizing to get things done. That is, the same leaders who are described as autocratic or authoritarian (Lewin & Lippitt, 1938) will also be described as directive (Bass & Barrett, 1981; Heller, 1969a), coercive and persuasive (Bass, 1960), concerned with production (Blake & Mouton, 1964), lone decision makers (Vroom & Yetton, 1974), initiators of structure (Fleishman, 1953c), production-centered (R. Likert, 1961a), goal emphasizers and work facilitators (Bowers & Seashore, 1966), task-oriented (Fied ler, 1967a), and concerned about performance (Misumi, 1985). Howard and Wellins (1994) listed five types of autocratic leader behavior in work settings: (1) The controller enforces a prescribed way of working. (2) The commander tells what to do and expects obedience. (3) The ruler considers decision making a privilege of management. (4) The judge evaluates subordinate performance and metes out rewards and punishments. (5) The guard protects turf and hoards resources.
The “bulls of the woods”—the early-twentieth-century shop foremen, exploitative autocrats, and punitive task-directed leaders—made the decisions for their groups and told workers what to do. They used physical force when necessary to cow their workers (Muszyk & Rei-mann, 1987). Such leaders discouraged subordinates from contributing to the decision process. They paid little or no attention to their subordinates’ needs. More modern and less punitive autocratic leaders still reserve decisions for themselves and remain more concerned about getting the job done than about the needs of their subordinates.
Autocratic leaders tend to initiate structure, provide information, determine what is to be done, issue the rules, promise rewards for compliance, and threaten punishments for disobedience. They use their power to obtain compliance with what they have decided. They depend on their knowledge of policies and regulations and their official rank to regulate the behavior of their subordinates. They use their technical knowledge to solve problems to gain their subordinates’ respect and willing compliance with their orders (Nelson, 1950). They are more often charismatic than consensual (Zaleznik, 1974). Autocratic leaders are ideologues who believe that people are basically lazy and must be driven and controlled by external rewards and punishments.
The Warrior Model of Leader. Nice (undated) has enumerated the characteristics of a subcategory of authoritarian leader. This model is of leadership related to conflict and to triumph over opposition. The flow of information is controlled. Results are more important than the methods used to achieve them. The warrior leader knows friend and foe. Battles are selected carefully and unnecessary fighting is avoided. Plans and preparations are made for future contingencies in a world seen as dangerous and hostile, in which few people can be trusted.
The democratic or egalitarian leadership cluster reflects concern about the followers in many different ways. Leadership is considerate (Fleishman, 1953c), democratic (Lewin & Lippitt, 1938), consultative and participative (Bass, 1976), consensual (Zaleznik, 1974), employee-centered (R. Likert, 1961a), concerned with people (Blake & Mouton, 1964), concerned with the maintenance of good working relations (Misumi, 1985), supportive and oriented toward facilitating interaction (Bowers & Seashore, 1966), relations-oriented (Fiedler, 1967a), oriented toward joint decision making (Heller, 1969a), and oriented toward group decision making (Vroom & Yetton, 1974). Democratic leaders are Theory Y ideologues (McGregor, 1960). They solicit advice, opinions, and information from their followers and share decision making with their followers. Democratic leaders use their power to set the constraints within which followers are encouraged to join in deciding what is to be done. Democratic leaders depend on their followers’ skills, as well as on their own interpersonal ability and knowledge of their followers’ individual needs, interests, and capabilities (Nelson, 1950). Democratic leaders believe that workers are internally motivated to do well, and seek autonomy and the opportunity to prove their worth. These leaders move decision making in organization hierarchies to lower levels, encourage questioning and ideas from below about better ways of doing things, are open to criticism, treat subordinates’ mistakes as learning opportunities, celebrate subordinates’ accomplishments, and promote subordinates’ ideas to higher authority (Howard & Wellins, 1994). But democratic leaders often need to do more to involve their followers in decision processes. Concerned with the corruption and shortcomings of democratic governance, Rost (1996) would reform it by changing the many passive followers in an organization or community into active collaborators who help their leaders to set the major policies of governance. According to Gill (1996), five principles govern all democratic societies: (1) personal responsibility, (2) empowering others to become leaders, (3) inclusiveness, (4) equality, and (5) full deliberation. Democratic leadership has internal conflicts and is often messy. It is not to be confused with laissez-faire leadership (discussed in Chapter 6), in which the leader abstains, withdraws, or abdicates responsibility and shows none of the concern seen in the authoritarian or democratic cluster.
Illustrating the meaningfulness of the clusters, Stanton (1960) reported higher scores for consideration and for a human relations orientation among managers of clerical employees of one firm that was known for more democratic supervision, compared with another firm that was known for more autocratic supervision. Conversely, the amount of initiation of structure was much higher in the autocratic firm than in the democratic firm.
Edwards and Rode (1986) asked 100 students, mainly in ROTC, to complete four different ways of measuring authoritarian and democratic leadership preferences. The Leader Behavior Opinion Questionnaire1 assessed how frequently they should initiate structure as leaders and how considerate they should be. Blake and Mouton’s (1978) Managerial Grid was used to describe a concern for people and a concern for production. Hersey and Blanchard’s (1973) LEAD-Self questionnaire (Leadership Effectiveness and Adaptability Description) generated authoritative telling and selling scores and democratic, participating, and delegating scores. Fiedler’s (1967a) Least Preferred Coworker (LPC) score was used to assess task and relations orientation.2 The intercorrelations among the authoritarian task measures were positive but rather low. The same was true for the democratic measures. However, a path model of the intercorrelations, adjusted for measurement errors by Jöreskog and Sorbos (1978), indicated that the latent structure underlying the observed correlations was a continuum stretching from authoritative behavior at one end to democratic behavior at the other.
This chapter and Chapters 18, 19, and 20 will look at concepts of the differences in leadership styles or patterns of leadership behavior and their effects: the overarching concept of and evidence about autocratic versus democratic leadership (this chapter), participative versus directive decision making (Chapter 18), relations versus task orientation (Chapter 19), and consideration versus initiation (Chapter 20). Laissez-faire leadership versus motivation to manage was discussed in Chapter 6. Autocratic-authoritarian versus democratic-egalitarian leadership is the most multifaceted issue. It refers to the way power is distributed, whose needs are met, and how decisions are made. Participative versus directive leadership refers primarily to how decisions are made. Relations-versus task-oriented leadership focuses on whose needs are met. Consideration versus initiation of structure is a behaviorally factor-derived dichotomy. Consideration and initiation refer to how decisions are made and to the structuring of tasks, goals, and role relationships. Laissez-faire leadership and motivation to manage refer to the extent to which leadership is either avoided or attempted. It should be kept in mind that styles of leadership are distinguished not only by behavioral differences but also by differences in cognition and intentions. To the degree that leaders are consistent in what they think they ought to do, they will have a consistent style of leadership (Wofford & Goodwin, 1998).
Why do we emerge with just two overarching clusters of leadership styles? Possibly because there are only two ways to change a follower’s behavior (apart from using drugs or physical force). The leader alters either the follower’s information, understanding, and ability to cope with the task at hand, or the follower’s motivation to deal with the task. When the leader has more relevant knowledge than the follower, authoritarian, task-focused direction can transfer the information quickly. Powerful leaders can arouse follower motivation. But in many situations, followers have at least as much information as the leader. Power may be shared. Follower motivation can be enhanced by involving them in decisions about handling the task, their concerns about the task and themselves.
Misumi (1985) argued that two fundamental dimensions—the performance leadership function (P) and the maintenance leadership function (M)—underlie all the others in the clusters. Thus, for instance, participative and democratic leadership are situationally specific manifestations of the underlying maintenance function, whose aim is to promote social integration and group stability. The manifestation of the maintenance function may appear as consideration if it is directed primarily toward satisfying subordinates’ needs.
Maclver (1947) and Bass (1960) noted that authoritarian leaders may depend on their power to coerce and their ability to persuade. An able leader successfully persuades others to follow him or her because they expect that following the leader’s suggestions will result in solving the problems the group faces. A powerful person successfully coerces others to follow him or her because the power of the leader’s position or the power of the leader as a person makes others expect that the leader will reward them for compliance or punish them for noncompliance. An able leader can indirectly reinforce the behavior of others. Such a leader can provide the cues that help them attain their goals. A powerful leader can directly reinforce the behavior of others by granting or denying them rewards or punishments (Bass, 1960).
These types of authoritarian leadership were described by F. C. Bartlett (1926). He observed that leaders in any complex social group maintain their success either because of the social prestige of their position and their personal capacity to impress and dominate or by virtue of their personal capacity to persuade their followers. Blau and Scott (1962) described the authoritarian supervisor as one who, among other things, uses power to be strict rather than lenient, to supervise closely, and to ensure adherence to procedures. Using one-minute sampling to observe the spontaneous play of children in nursery school, Parten (1932) found two types of leadership possible in the situation: persuasive leadership, which uses diplomatic, “artful” suggestion; and leadership through personal power, which uses brute force to dominate others. Zillig (1933) observed the same two types of leadership in the German classroom—leaders who dominate and leaders who direct and guide.
Colin Powell (1995, p. 332) described how he used his power and ability when he was a senior official in the Department of the Army to make Policy Review Group meetings more effective. He was delegated responsibility by his boss, Frank Carlucci, to chair the interagency discussions. Frustrated by “endless, pointless, time-wasters,” he used his legitimate power and perspicacity to structure the meetings highly. Everyone could contribute in advance to the agenda, but Powell controlled the final listing. No one else could change it. The meeting was to last only one hour. In the first 5 to 10 minutes, Powell reviewed its purposes and what had to be decided. Participants could present their positions without interruption for the next 20 minutes, followed by open discussion until the last 15 minutes, when Powell spent 5 minutes summarizing everyone’s positions, allowing the participants one minute each to critique the summary and reserving the last few minutes for himself to present the conclusions. Participants’ objections could be taken up with their superiors and referred by them to Carlucci.
The processes of democratic leadership usually require maturity and some education. Some leaders may be identified as democratic on the basis of their use of parliamentary procedures and majority decision making. Others may consult; strive for consensus; and pursue an open, trusting, follower-oriented relationship. Such consensus leadership has deep roots in the American national character (Cooley, 1956). According to Zaleznik (1974), the idealized image of the leader in America is that of brother rather than father. The leaders are first among equals whose sense of timing may be more important than their expertise. Leaders may be more dependent on the followers than the followers are on the leaders. Compromise, caring, and a sense of responsibility and attachment to followers characterize the consensus leader.
In the first experiments—using instructed camp counselors leading preadolescent boys and girls—to contrast the authoritarian and the democratic leader, the authoritarian leader dictated what was to be done and was unconcerned about the group members’ needs for autonomy and development. The democratic leader shared the decision making with followers and was concerned about their need to contribute to deciding what was to be done. The authoritarian leader was personal in praise or reproof of each follower; the democratic leader was factual and rational. The authoritarian leader emphasized his or her social distance from the followers; the democratic leader deemphasized social distance.
Following the earlier experiments with authoritarian and democratic leadership, subsequent studies tended to concentrate on one of four aspects of the distinction between authoritarian and democratic leadership: (1) whether the leader shared decision making, (2) whether the followers were of primary concern to the leader, (3) whether social distance was maintained, and (4) whether punishment and coercion were used.
It must be stressed that these are ideal types of authoritarian and democratic leadership. In practice, one is likely to find much variation and overlap. For instance, benevolent autocrats, although they are likely to be dictatorial, may also be considerably concerned about the needs of their followers. The democratic manager may encourage group decision making but may also emphasize getting the job done, as well as the needs of the group. The charismatic leader may rely on personal power and the followers’ desire for identification, but also may choose to support democratic efforts and the attainment of prosocial goals. Some researchers have taken greater care than others to define and “purify” the patterns of leadership behavior to be studied. But often it is difficult to determine whether the pattern of leadership behavior examined was democratic, participative, considerate, relations-oriented, or some combination of these characteristics.
Consistent with Kurt Lewin’s ideas (1939) about the behavioral dynamics in groups, two seminal experiments were conducted (Lewin & Lippitt, 1938; Lippitt, 1940a) to explore the effects of democratic and authoritarian atmospheres upon the behavior of group members. These studies were preceded by a report by Mowrer (1938) on how infractions of rules were reduced at the New Haven Children’s Center in 1937 when a democratic approach was offered to the children (aged 4 to 12) to deal with problems, in contrast to the authoritarian staff control that had been in place before.3 Reporting on the two experiments, Lippitt (1940) defined democratic and authoritarian leader behavior. By means of careful coaching and practice, the authoritarian leader was trained to (1) determine all policy for group members, (2) dictate the methods and stages of goal attainment one step at a time, (3) direct the actions and interactions of group members, and (4) praise the members in a personal manner. The democratic leader was trained to (1) encourage group members to determine their own policies, (2) give them perspective by explaining in advance the steps toward attaining the goals, (3) award them freedom to initiate their own tasks and interactions, and (4) praise them in an objective manner. The leaders were adult counselors. The group members were 10-year-old boys and girls, closely matched on several control variables. Two groups of five members each worked on hobby projects. The behavior of leaders and members was recorded by trained observers.
Both the leaders and the group members initiated more actions in the authoritarianled group. Members of the authoritarianled group had more submissive reactions to the leader, however, and treated him less as an equal than was the case in the democraticled group. In the authoritarianled group, members became progressively more submissive to the leader. Although they tended to respond to the leader rather than to initiate interaction with each other, they hesitated to approach him because to do so might further reduce their personal power and freedom of movement. The democratic form of leadership, on the other hand, tended to increase the freedom of action of group members. The members of the democratically led group exhibited less tension and hostility, and their subgroups were more cohesive and enduring than was the case for the authoritarianled group.4 Curfman (1939) replicated (1940a) Lippitt’s first experiment with two clubs of fifth-and sixth-graders, with similar results.5
Research and development by Rensis Likert (1961a, 1967, 1977b) flowed from a direct inference from the prototype experiments that the democratic approach was to be advocated in organizations to foster effectiveness and satisfaction. In this, Likert was joined by the entire human relations movement (see, for example, Argyris, 1957; McGregor, 1960). In counteraction, joining with the classical “scientific managers,” Miner (1968), among others, argued for an emphasis on the manager-leader as a task-oriented authority figure. Blake and Mouton (1964) pioneered the point of view that the best leaders and managers were highly concerned with both production and people and could integrate the two approaches. A fourth group of advocates, such as Fiedler (1967a), Vroom and Yetton (1974), and Hersey and Blanchard (1977), argued that which style was best depended on the situation.
As we shall see in this and succeeding chapters, there is some truth in all these positions. Evidence has accumulated that in specified circumstances, authoritarian direction may, in fact, result in heightened productivity, particularly in the short term; but overall, the democratic approach is likely to be more effective, particularly in the long run. And in general, under most conditions, working for a democratic supervisor will be more satisfying. This last conclusion is simplest to document.
Over 30 empirical laboratory experiments and 10 field studies on the subject were conducted between 1940 and 1975. Under democratic leadership, members’ satisfaction with the leaders was the rule rather than the exception. On the other hand, more positive correlations were found between the authoritative leadership style and productivity in the laboratory experiments. Fortunately, a more rigorous meta-analysis, including some of these findings as well as more recent reports, provided a greater opportunity to be more specific about these effects on productivity. In field studies, it could be concluded that a democratic style worked better. In the laboratory, the autocratic style resulted in more productivity.
Miller and Monge (1986) first conducted such a meta-analysis of short-term laboratory experiments of the effects on productivity and satisfaction of “employees” of contrived autocratic versus participative “supervisors” by Ivancevich (1974); Katzell, Miller, Rotter, and Venet (1970); McCurdy and Lambert (1952); and Shaw (1955). In these experiments, productivity was greater when the leader in the laboratory manipulation was more authoritarian than democratic. For the four short-term laboratory experiments in which the “supervisor” was trained to act autocratically or democratically, productivity was greater with authoritarian, directive supervision. The mean correlation between productivity and authoritarianism was .33.
However, Miller and Monge (1986) also completed a meta-analysis for 10 longer-term field studies. Consistent with what will be reviewed next, they found strong support in reports from 41 analyses. In such real-life settings, the mean correlation was .27 between democratic supervision and the productivity of the persons or groups supervised. The researchers also concluded that the more democratic supervisory styles correlated with the satisfaction of subordinates. The mean correlation was .34. Consistent with this conclusion is the massive survey evidence accumulated by R. Likert (1977b) and his colleagues. Despite the fact that in the short run, in brief laboratory experiments, authoritarian supervision pays off in more productivity than does democratic supervision, in the long run democratic approaches tend to generate bigger improvements in an organization’s productivity as well as in the satisfaction of its employees.
When It Pays to Be an Autocratic Leader. Jack Welch, CEO of General Electric, headed the 1984 Fortune list of the 10 toughest bosses in America (Flax, 1984). Yet he moved General Electric to the top of the list in earnings and successfully changed its line of products and competitive performance. Muczyk and Steel (1998) cite successful turnarounds by business executives. During organizational crises, despite the aversion to autocratic leadership, when organizations need to be turned around quickly, autocratic and directive leadership is needed. Sometimes unpopular decisions are required along with legitimate and reasonable goals and fair and respectful treatment of subordinates. Teachers, coaches, and religious leaders exert powerful influence over many followers who learn from them by being told what they should do (Smither, 1991). Not unexpectedly, the military tends to value the use of authority more favorably. In a survey of 30,735 U.S. Army superiors, peers, and subordinates of commissioned and noncommissioned officers, Penner, Malone, Coughlin, and Herz (1973) found that leaders who established a high level of discipline were likely to be rated much higher in their overall performance by their superiors. Consistent with this, Torrance (1959) reported that U.S. Air Force aircrews who were given feedback by highly authoritarian methods improved their performance more than those who were given feedback by less highly structured methods.
Supportive evidence was also found elsewhere. Hise (1968), studying simulated business groups, found that productivity was positively related to close rather than general supervision. M. E. Shaw (1955) obtained results indicating that the speed and accuracy of a group’s performance in a highly structured communications network were significantly higher under autocratic than under democratic leadership. In a study of groups in a formal organization, Shepherd and Weschler (1955) found that psychosocial distance between the leader and the followers was associated with fewer communication difficulties. Working closely together was related to greater difficulty in communication. Bergum and Lehr (1963) studied subordinates’ monitoring performance under different conditions of supervision. They found that the vigilance of subordinates could be maintained at fairly high levels under authoritarian conditions.
Whether authoritarian leadership pays off will depend on the extent to which the leader has more knowledge about what needs to be done as well as control of the necessary resources. This payoff can be routinely demonstrated in leadership training exercises. Authoritarian leadership can be productive if the trainee who is assigned the role of autocratic leader happens, by chance, to know more about the problem to be solved than does the person who is playing the role of subordinate (Shack-leton, Bass, & Allison, 1975). When autocratic leaders know the correct answer in such exercises, they can ensure highly accurate group outcomes better than can democratic leaders. But authoritative leaders with misinformation can lead their groups farther astray than can democratic leaders with the same degree of misinformation (Cammalleri, Hendrick, Pittman, et al., 1973). An increase in rules constraining managerial behavior was found salutary in Malaysia (Mansor & Ali, 1998).
When Autocratic Leadership Is Worse. The American public was led to believe that 11,500 unionized air controllers went on strike in August 1981 for unreasonable economic demands and because of peer pressure. However, a U.S. Department of Transportation task force discovered that they had struck because they wanted improvements in working conditions to alleviate job stress and because their managers put autocratic values and beliefs into practice (Bowers, 1983). In 1989, the newly formed union of air controllers made the same claim that morale was low because of autocratic management (Cushman, 1989). Ziller (1957) found group members to be least satisfied under autocratic leadership. In a study of personnel turnover, Ley (1966) found that the supervisor’s authoritarian behavior was the factor most frequently associated with subordinates quitting their jobs.
More deleterious effects are likely when autocratic leadership takes the form of abusive, punitive, and disciplinary actions. The only U.S. Navy captain to be relieved of duty “for cause” as commander of a nuclear submarine was relieved for abusive treatment of the crew, resulting in their “despondency” and low morale (Ricks, 1997). Richman, Flaherty, Rospenda, et al. (1992) found that abusive supervision of medical students and medical residents was related to their sense of distress and dissatisfaction. Ashforth (1997) reported that belittling of subordinates and noncontingent punishment by their supervisors were correlated with subordinates’ feelings of frustration, alienation, and helplessness. Tepper (2000) conducted follow-up surveys of 390 home telephone interviews of full-time employees of service, retail, manufacturing, and small-business organizations about their experience with abusive supervision. Abuse was infrequent on 15 items of behavior such as “Ridicules me” and “Puts me down in front of others” (1 = never uses this behavior; 5 = uses this behavior very often). The mean obtained was 1.38. Abusiveness correlated negatively with job satisfaction (–.35), organizational commitment (–.24, –.27), and positively with respondents’ emotional exhaustion.
Cultural norms dampen enthusiasm for using autocratic leadership. In earlier chapters, we observed that supervisors tend to avoid disciplinary, punitive action even when confronted with poor performance by subordinates. Thus, for instance, Maier and Danielson (1956) reported that even when disciplinary rules call for punishment, supervisors tend to avoid it. They perceive that punishment will reduce productivity, even when it is used to control the violation of rules. In fact, Keller and Szilagyi (1976) found that punitive leadership, rather than improving performance, primarily increased role ambiguity. Day and Hamblin (1964) found punitive and close supervision to be connected with reductions in productivity and group harmony. They studied 24 quartets who had to follow elaborate blueprints to assemble models of molecules using pegs, springs, and colored balls. In this complex task, the members were subjected to either punitive or nonpunitive supervision, as well as to close or general supervision. Punitive and close supervision, in contrast to nonpunitive and general supervision, tended to increase the subjects’ feelings of aggression toward their coworkers, as well as toward their supervisor, which significantly lowered their productivity.
Numerous other experiments point to the deleterious main and side effects of punitive supervision. French (1957) examined the effects of the “supervisor” who obtained compliance by using the power to fine paid participants who were working at a simple task—sorting IBM punch cards according to the total number of holes the cards contained. The assignment was supposedly part of a research project. Participants were fined for failure to maintain the standard of speed and accuracy. In comparison with participants who were rewarded with extra pay when they reached and maintained the standard, participants who were punished by fines for failure were more likely to show signs of resistance to maintaining production. After four working periods, punishment for failure became detrimental to speed and accuracy. The participants had a greater desire to leave the work and were more likely to want to do something else and to make suggestions for changing the work situation. Furthermore, they showed greater feelings of aggression, liked their supervisor less, and were less likely to accept the supervisor as competent to evaluate their work. In a similar experiment, Raven and French (1958b) levied fines when participants failed to conform to the demands of “supervisors.” These fines led the participants to resist by overconforming to the suggestions of their supervisors. That is, metaphorically, if participants were ordered against their will to “polish the silverware,” they rubbed so hard that they rubbed off all the silver.
Experiments by deCharms and Hamblin (1960) yielded similar results finding that punitive supervision resulted in increased tension in “employees” and lowered their productivity. These studies of the effects of punitive supervision are consistent with the general psychological findings that severe punishment tends to be disruptive and anxiety producing. Mild punishment, however, may provide, under certain circumstances, more appropriate attention to the failure to comply with authority and serve as a way to eliminate undesired acts. Thus Georgopoulos (1965) observed higher productivity to be associated with a pattern of supervisors’ disciplinary behavior that employees regarded as “just right” and “not too strict.” Mild aversive reinforcements and negative feedback have their uses.
When It Pays to Be a Democratic Leader. Smither (1991) listed four conditions in which democratic leadership works best in organizations: (1) it is visibly supported by higher authority; (2) members are well-educated and support the organization’s goals; (3) leaders have the skills to conduct meetings with the members; and (4) time can be afforded for trust to develop. Comrey, Pfiffner, and Beem (1952) studied employees at six levels of organization in the U.S. Forest Survey. Supervisors of more effective departments were described as sympathetic, democratic, social, and willing to share information. Similarly, Comrey, High, and Wilson (1955a, 1955b) studied supervisors and workers in an aircraft factory. Supervisors of effective groups were characterized by adequate authority, communication downward, and sympathy; these supervisors were not arbitrary, and their attitude toward employees was not hypercritical. Argyle, Gardner, and Ciofi (1958) also reported higher rates of productivity, reduced personnel turnover, and reduced absenteeism under democratic, nonpunitive supervision in work groups in seven British factories. Zweig (1966) found that democratic supervisors were rated by higher management as being more effective than supervisors exhibiting less democratic styles of behavior. Similarly, Hall and Donnell (1979) reported that the managers whose career advancement was the fastest, compared with those who advanced more slowly, were less likely to subscribe to Theory X beliefs6 that workers are lazy and immature, and need to be carefully monitored and controlled. Levy-Leboyer and Pineau (1981) obtained interview data from 151 French laboratory supervisors. The success of their laboratories was appraised subjectively and from research publication records. Leaders of more successful laboratories supervised less strictly, allowed for participation by technicians in decision making, and held more frequent meetings and evaluations.
In a federal agency, J. C. White (1972) observed that both the effective county office managers and the effective managers at headquarters used a more democratic style. The effectiveness of business supervisors and purchasing managers was also seen by H. C. White (1971a, 1972) to be associated with the extent to which they were more democratic and less autocratic. Hollman (1973) and Tanimoto (1977) found that the effectiveness of management by objectives, as seen by subordinates, was greater when practiced by a more democratic manager than by a more autocratic manager. Other studies have shown that democratic approaches in organizations have a favorable effect on the physical and mental health of the members (Caplan, Cobb, French, et al., 1975).
Ordinarily, satisfaction and morale are likely to be lower with autocratic supervision and higher with democratic supervision. Under autocratic leadership greater resentment, less loyalty, less commitment, less involvement, and less satisfaction are commonly found (Gould-ner, 1954). Such subjective reactions to the supervisors and poorer relations with them affect the willing compliance of subordinates with their supervisors’ initiatives (Barnard, 1938). In both the short term and the long term, subordinates generally will be more satisfied with democratic leaders.
Beam (1975) showed that enlisted personnel in the U.S. Navy had a strong desire to be treated democratically, regardless of the level of technology or physical activity in which they were involved. Mohr (1971) obtained similar results in 144 works groups from 13 local health departments, as did Pennings (1975) in 40 branch offices of a large U.S. brokerage firm. The subordinates’ satisfaction was strongly associated with democratic supervision in these as well as in many other large-scale field studies.
Baumgartel (1957) studied attitudes and motivations of scientists in governmental research-and-development labs under three leadership conditions (democratic, authoritarian, and laissez-faire). He found that scientists who worked under the democratic leadership held the most favorable attitudes and had a greater job motivation, whereas the least favorable attitudes were found among those who worked under the authoritarian leadership. Harnquist (1956) observed that group members tended to feel more satisfied under democratic than under autocratic leadership. Mullen (1965, 1966b) reported that the satisfaction of employees was associated with democratic supervision but saw no relationship between supervisory style and employees’ requests for transfers.
Mandell and Duckworth (1955) reported that the overall morale of 64 trade employees in civil service was high if the employees said that their “supervisor lets them know how they are doing.” In a series of surveys, R. Likert (1961a) found that employees of public utilities revealed higher job satisfaction when their supervisors were more “personal” than “institutional” in their dealings and more “downward” or employee-oriented, and when the supervisors trained the subordinates for better jobs. Supervisors of high-morale groups differed from supervisors of low-morale groups in that they reviewed their subordinates’ work more frequently, welcomed the discussion of mutual problems with subordinates, carried on group discussions, and kept subordinates posted on new infor-mation.
D’Angelo (1973) found that sales managers who believed themselves to be practicing a democratic, human relations style brought about more change in their subordinates in an organizational development program than did those who believed they were practicing either authoritarian or “human resources” leadership. H. H. Meyer (1968) studied two plants, one managed according to Theory Y and the other managed according to Theory X.7 Workers under the more democratic type of management (Theory Y) reported higher felt responsibility, risk, reward, warmth, and identity. Beehr and Gupta (1987) compared two manufacturing firms—one, formally democratic; the other, more traditional—that were similar in size and technology. They found that employees’ perceptions, attitudes, and behavior were more favorable in the democratic firm.
Hendrix and McNichols’s (1982) survey of 4,786 military and civilian personnel in the U.S. Air Force showed that managers who were described as self-enhancing, outspoken, and demanding and who used their authority as their primary means of influencing their subordinates were seen as contributing to a much less attractive organizational climate, to much less job satisfaction, and to much less perceived productivity than were other more democratic managers who tended to show concern for their subordinates and for group processes. Results were the same for three different work settings involving customer service, routine jobs, and unique jobs with a lot of autonomy.
Brollier (1984) surveyed 93 directors of departments of occupational therapy and 348 staff therapists. The directors’ democratic leadership had much more of a positive effect on the staff’s satisfaction than on the staff’s performance. A survey by Field (1984) of 295 human services professionals indicated that democratic leadership tended to promote greater job satisfaction among the staff, particularly satisfaction with salaries and career advancement.
When Democratic Leadership Is Worse or When Alternative Styles of Leadership Are Better. Studebaker went out of business in 1964, after competing against other auto manufacturers that practiced less industrial democracy and less committee management (Clemens & Meyer, 1987). Contingent reinforcement and transformational leadership may be more effective than the emphasis on either authority or democracy (Bass, 1998). Ziller (1957) observed greater problems among members who were led democratically. Neither autocratic nor democratic leadership was as effective as a type of reinforcement leadership that consistently approved the correct performance of members and made suggestions for improvements. Spector and Suttell (1956) demonstrated that this type of leader brought about better achievement by the group than did an authoritarian type who made decisions for the group or a democratic type who permitted maximum participation by members. D’Angelo (1973) reported that a human resources style, which involves “striving to continually expand the areas over which the manager’s subordinates have self-direction and self-control,” was associated with more effective work groups before and after an organizational development program for 103 sales managers and their 360 salesmen subor dinates. Autocratic or human relations styles were less effective. Howell (1985) showed that in contrast to task-structuring and considerate leaders, actors who were trained as charismatic leaders were most effective in enhancing productivity in the face of normative resistance among subordinates.
Anticipating future interest in transformational leadership, Litwin and Stringer (1966) organized 45 business students into three “firms.” The firms competed in the construction and marketing of “radar equipment” manufactured from Erector set materials. Three different business climates were created: (1) an authoritarian-structured business, with a strong emphasis on the careful definition of duties and the exercise of formal authority; (2) a democratic-friendly business, in which cooperative behavior, group loyalty, teamwork, freedom from punishment, and a loose informal structure were emphasized; and (3) an achieving (transformational) business, in which innovation was encouraged, competitive feedback was given, pride in the organization was fostered, a certain amount of risk taking was deemed desirable, and high personal goals were encouraged. The “president” of each company was a member of the research staff who adopted an authoritarian, democratic, or achieving leadership style. The achieving (transforming) style resulted in the greatest dollar volume, the most new products, and the most cost-saving innovations. The authoritarian style did succeed in producing the finished goods with the highest quality, primarily by never deviating from the specifications. Consistent with most other findings, students who were working with a democratic president were more satisfied with their jobs than were the students in the two other firms.
When There Are No Differences. Some studies have failed to find that democratic or autocratic supervision had any significant effect on satisfaction or productivity. Thus, with experimental groups of students, Hamblin, Miller, and Wiggins (1961) reported no relationship between authoritarian leadership and group morale. J. D. White (1963) examined whether the morale of boards of directors was affected by how they were led. Although the power of members was found to be higher on boards with democratic leaders than with autocratic leaders, morale was not related to the different styles of leadership. In the case of sports coaches, Browne and Mahoney (1984) concluded that it sometimes appears best for the coach to act in an authoritarian manner. At other times, it is best for the coach to be more democratic by talking to players with individualized consideration and by allowing them to participate in goal setting. Still other investigators found no significant differences in productivity between democratically and autocratically led groups. Thus Lyle (1961) and Spector and Suttell (1956) found no differences in the performance of groups under democratic and autocratic leadership. However, Lyle (1961) observed a tendency for democratic groups to work faster under restricted communication, whereas authoritarian groups worked faster under open communication. Results obtained by S. Adams (1952) indicated that bomber crews performed more effectively under medium than under high or low degrees of democratic leadership. Sales (1964) obtained no significant difference in the performance scores of groups with democratic and autocratic leaders. Johnson and Smith (1953) studied classes taught traditionally and under democratic leadership. They found no significant differences in achievement gains or in the students’ evaluations.
Mullen (1965, 1966b) also failed to find that group productivity was related to supervisory style. Likewise, T. A. Mahoney (1967) obtained no relationship between democratic supervision and measures of organizational effectiveness in a study of industrial organizations. Similarly, Swartz (1973) found that whether football coaches were autocratic, democratic, or laissez-faire was unrelated to their success in winning games. G. H. Graham (1969) conducted one class democratically according to Theory Y and another autocratically according to Theory X.8 The two groups did not differ in examination scores. The top quartile of students did better when led by the democratic instructor, but the lower 75% of students got better grades under the autocratic instructor.
Chapter 7 noted that authoritarian leadership works better with authoritarian followers and in authoritarian cultures. Other potential personal and interpersonal modifiers were also discussed. Chapters 25 through 29 will look at how situational conditions may affect the leader’s style, and Chapters 31 to 33 will examine how race, sex, and culture make a difference. Some of the findings will be introduced here that have demonstrated that the effects of authoritarian and democratic leadership depend on these antecedent conditions.
Taylor (1980) pointed out that immature, dependent, inexperienced subordinates are more likely to expect and accept authoritarian direction and that democratic leadership is likely to result in their more rapid development. Nisbett (1986) suggested that a more democratic leadership is desirable as the workforce becomes more educated and seeks greater participation, as business becomes more complex and requires a team of experts to deal with its problems, and as the use of high technology increases. The symphony conductor Daniel Barenboim (2001) felt that one could learn to live in a democratic society through experience in a symphony orchestra: “For when you do so, you know when to lead and when to follow. You leave space for others and at the same time you have no inhibitions about claiming a place for yourself.” The fall of communism in Eastern Europe gave Luthans and Riolli (1997) the opportunity to compare managers in Albania at a fully privatized company starting over, a company slowly converting to privatization, and a large old-line state-owned enterprise. Most managers in the privatized company preferred and practiced a democratic style of getting everyone involved in decisions. Most in the slowly converting company preferred and practiced a consulting style, getting input from others. Most in the state enterprise practiced benevolent autocracy, explaining their decisions, but preferred a more consultative style.
In examining the effects of followers’ expectations about leadership, we need to take into account cultural changes that have occurred in the past 60 years. Earlier studies are likely to show willingness to accept autocratic leadership by many followers; later studies are less likely to show this. Nevertheless, Smither (1989) noted that in the right context authoritarian leadership can be effective in accomplishing organizational goals. Tough leaders can be effective, according to anecdotal evidence. Autocratic leadership may be what is required if employees are poorly educated and are uninterested in responsibility or in the organization’s mission. Immediate productivity may be more important than employee satisfaction. Employees may have strong emotional ties to the leader and may not resent tough leadership.
According to French, Morrison, and Levinger (1960), autocratic leadership is likely to generate dissatisfaction and hostility in subordinates unless they see it as a legitimate part of the supervisor’s role. Thus Foa’s (1957) study of groups of Israeli workers under democratic and autocratic leadership found that groups with authoritarian and democratic expectations were about equally well satisfied with democratic leaders. But when leaders were autocratic, crews with authoritarian expectations were better satisfied than those with democratic expectations. Hemphill (1949b) noted that arbitrary inconsistency and reversal of opinion by the leader are more readily tolerated in groups that lack well-established rules and regulations, well-defined goal direction, and strong cohesiveness among the members. Vroom and Mann (1960) studied industrial work groups that varied in size and in style of supervision. In the small work groups, high rates of interaction occurred between workers and supervisors, and egalitarian leaders were preferred. In the large work groups, members interacted less frequently with each other and with their leaders and exhibited more positive attitudes toward authoritarian leaders.
Tepper, Duffy, and Shaw (2001) found that constructive resistance to abusive supervision was greater if subordinates were conscientious. But resistance was dysfunctional if the subordinates were also agreeable. Tepper, Duffy, and Hoobler (2002) reported that for 176 employees, abusive supervision made a difference in how their coworkers’ organizational citizenship behavior (OCB) affected these employees’ job satisfaction. When abusive supervision was low, coworker OCB contributed to employee job satisfaction. When abusive supervision was high, coworker OCB had a negative impact on employee satisfaction.
Calvin, Hoffmann, and Harden (1957) constructed experimental groups that differed according to their members’ scores on intelligence tests and the authoritarianism or egalitarianism of the leaders. The performance of groups composed of bright members did not differ under authoritarian and egalitarian leadership. But dull members in authoritarianled groups were more effective than were dull members in egalitarianled groups. McCurdy and Eber (1953) arranged for leaders to be coached in democratic and autocratic patterns of behavior. Authoritarian participants solved problems somewhat less speedily under democratic than under autocratic leadership.
Circularity. Kruglanski (1969) reported that managers tended to supervise less closely those subordinates whom they trusted. However, a circularity occurs. In an experiment, Strickland (1967) found that laboratory “supervisors” came to mistrust “subordinates” whom they were directed to monitor closely and increased their trust of “subordinates” whom they were not instructed to watch closely. In the same way, McFillen (1978) and McFillen and New (1978) failed to find any significant relationship between supervisory rewards or punishment and subordinates’ performance, but subordinates’ performance caused a difference in closeness of supervision. Low-performing subordinates were more closely supervised. McFillen and New (1979) demonstrated experimentally that not only is mistrust increased under close supervision, but the supervisor attributes more success to the closely supervised subordinate who succeeds than to the generally supervised subordinate who succeeds, and more failure to the closely supervised subordinate who fails.
Which leadership style works best depends on the task of the leader and the led. As Plato observed, there may be no time for a democratic vote on a ship in a storm. Sailors may not know as much about what to do as the captain. The captain’s individual judgment may be better than the sailors’ consensus. When the democratic management of Steve Jobs, which had promoted the creativity and innovation that launched Apple Computer, was no longer effective in competing with IBM’s entry into the personal computer market, Jobs was replaced by the tough John Scully, a top-down decision maker (Clemens & Meyer, 1987).
According to Rudin (1964), a punitive style of supervision leads to good performance on simple tasks and to poor performance on complex tasks. Becker and Baloff (1969) suggested that the optimum style of leadership is likely to depend on whether the task involves information processing or the generation of ideas. In a laboratory experiment involving complex mechanical tasks under close or general supervision and punitive or nonpunitive styles, Day and Hamblin (1964) found that close supervision produced a large increase in aggressive feelings toward the supervisor. Close supervision did not affect satisfaction with the task, but production was significantly lower. The punitive style also resulted in increased aggressive feelings toward the supervisor, but again without any effect on satisfaction with the task. As with close supervision, punitive supervision led to a decrease in production However, Patchen (1962) obtained a positive relationship between close supervision and better performance in manual work when there was strong group cohesiveness and when the supervisor was seen as rewarding rather than punitive. Thus although close supervision is usually considered part of a more autocratic style, its effects may depend more on whether it also includes a punitive component. (A benevolent autocrat would be more rewarding and less punitive.)
The phase in the task also makes a difference. R. C. Anderson’s (1959) survey of leadership in experimental groups disclosed that groups under authoritarian leaders required less time in the planning phases but were less efficient in the task-solution phases. Consistent with what was said earlier, democratic leadership resulted in greater satisfaction for members in both phases of problem solving. Subsequently, Doyle (1971) ascertained that egalitarian leadership was most effective in the analysis phase of group problem solving, but in the final synthesizing phase that required coordination, groups with powerful leaders were particularly effective.
As was seen in Chapter 16, managers’ behavior toward subordinates depends on how the managers’ superiors act toward the managers. Thus D. Katz (1951) found that highly productive groups had less close supervision from their supervisors, who in turn were less closely supervised by their superiors. Hunt, Osborn, and Larson (1975) observed that whether or not upper-level management was autocratic had an impact on what leadership style was most effective at lower levels. Morse (1953) studied employees and supervisors in an office situation. General supervision and delegation of authority facilitated the workers’ satisfaction with the work group. Job satisfaction or satisfaction with the company was not fostered unless the supervisors’ orientation toward employees was reinforced by higher management.
Bowers (1997) argued that top-down hierarchical command of business is out of date—that decision making is needed at all levels of a firm. At all levels, leaders and subordinates need to work out decisions together, and a democratic model should replace command and control systems. The outstanding application of democratic as opposed to autocratic processes in organizations was an extensive effort at the University of Michigan. Strong evidence was accumulated from over 500 studies completed between 1950 and 1977 that in the long run, democratic leadership pays off in both higher productivity and greater satisfaction for employees. Katz, Maccoby, and Morse (1950) began the line of investigation in the home office of a large insurance company. Twenty-four work groups were studied; half were high and half were low in productivity. Each highly productive unit was matched with another unit that was low in productivity. Differences in supervisory behavior between the high-and low-productivity units were assessed by means of interviews with supervisors and their subordinates. Highly productive supervisors were more frequently employee-centered rather than production-centered. They were more likely to exercise general rather than close supervision, and they were more likely to differentiate their roles from those of their subordinates in terms of the duties they performed.
Next, Katz, Maccoby, Gurin, and Floor (1951) studied railroad maintenance-of-way workers. Again, the more productive supervisors were found to be more employee centered and to exercise more general supervision than those whose units were less productive. However, no difference in role differentiation was found. The line of investigation was continued by Morse and Reimer (1956). They showed that although authoritarian methods contributed more to increased productivity in an insurance firm during the first year of an experimental effort to change, a sizable drop in performance followed in subsequent years because of the adverse impact of the authoritarian approach on human factors.
These studies led to the formulation of a rationale for organizational improvement. This rationale relied partly on democratizing the leadership patterns in the organization, predicated on the efficacy of democratic over autocratic processes.
Borrowing heavily from the original experimental concepts and results of Lewin and Lippitt (1938), R. Likert (1961a) conceived of four systems of interpersonal relationships in large organizations: (1) exploitative autocratic, (2) benevolent autocratic, (3) consultative, and (4) democratic. These systems varied as System 1 to System 4 on a number of criteria. Likert proposed and demonstrated that moving organizations away from Systems 1 and 2 and toward Systems 3 and 4 would result, given sufficient time for effects to take place, in increases in both the productivity and satisfaction of employees.
The Profile of Organizational Characteristics (POC) was constructed using 18 survey questions such as: “How much confidence and trust is shown in subordinates?” “How free do subordinates feel to talk to superiors about the job?” and “How often are subordinates’ ideas sought and used constructively?” An assessment was generated to indicate where the organization lay between System 1 and System 4. The correlations among the 18 scales of leader, peer, subordinate, and organizational behaviors ranged between .40 and .80. These correlations indicated that there was considerable consistency among the various questions. For leadership and influence in the autocratic Systems 1 and 2, the exploitative autocrat of System 1 emphasizes threats, fear, and punishment with some promise of reward. The benevolent autocrat of System 2 emphasizes more positive and less negative reinforcement. Top-down communication is stressed. Subordinates have little influence on goals and methods. Decisions and controls are centralized and are made person to person. The leadership and influence processes in the consultative System 3 and the democratic System 4 are such that supervisors and subordinates trust each other; and the supervisors are very supportive, very easy to talk to, and virtually always get subordinates’ ideas to try to make constructive use of them. There is an emphasis on economic and achievement motivation as well as on personal worth. The subordinates’ participation in goal setting is encouraged, along with bottom-up communication. Subordinates are influential in determining goals, tasks, and methods. Decisions and controls are decentralized. Virtually no confidence is shown by supervisors to subordinates in System 1 and a great deal of such confidence is shown in System 4.
R. Likert (1967) applied Bowers and Seashore’s (1966) four dimensions of leadership behavior to distinguish among autocratic and democratic leaders. He found that System 4 leaders were highest and System 1 leaders were lowest on the four dimensions of a leader’s behavior: (1) support—friendly, pays attention to what one is saying, listens to subordinates’ problems; (2) team building—encourages subordinates to work as a team and encourages the exchange of opinions and ideas; (3) goal emphasis—encourages best efforts and maintains high standards; and (4) helps with work—shows ways to do a better job; helps subordinates plan, organize, and schedule; and offers new ideas and solutions to problems.
Correlations of the respondents’ mean scores on the POC with the quality and quantity of organizational performance ranged from .30 to .60 (R. Likert, 1977a). In other words, in the more than 500 studies completed by 1997, positive associations generally were found between measures of the organizations’ performance. The modal system was System 3. Surveys before and after shifts upward from more autocratic to more democratic systems were reflected in improvements in organizational performance. For example, R. Likert (1975) reported that the shift from System 1 to System 4 between 1969 and 1970 at a General Motors plant resulted in substantial increases in direct labor efficiency in 1971 and 1972. Although indirect labor efficiency declined between 1970 and 1971, it increased sharply between 1971 and 1972. Particular effects cannot be attributed exclusively to changes in leadership style. However, given the pattern of large-scale, long-term changes reported by R. Likert (1977b), associated with changes in both leadership and other aspects of organizational development, it seems plausible to attribute many of these effects to the changes in leadership. By 1997, more than 500 studies had been completed in petroleum, automotive, pharmaceuticals, investment banking, insurance, delivery service, publishing, utilities, textiles, office equipment, packaging, paper making, and railroad companies. Research also had been conducted in governmental organizations, hospitals, schools, colleges, correctional institutions, military organizations, and voluntary organizations. Data were obtained from more than 20,000 managers at all hierarchical levels, and from more than 200,000 nonsupervisory employees.
These results need to be understood in terms of the differential impact of authoritarian and democratic systems on immediate, compared with long-term, labor costs. When belt-tightening was autocratically imposed in three continuous-processing plants, it produced an immediate reduction in costs in one organization of 600. However, this reduction occurred at the expense of the employees’ deteriorating motivation, dissatisfaction with company policy, and dissatisfaction with the leadership and was reflected in lowered quality and increased grievances, turnover, work stoppages, and failures to meet delivery dates. In one study, it took three or more years for these effects to show up clearly.9 The immediate savings of $250,000 actually produced losses of $450,000 in the longer term, owing to the fact that employees became more hostile, less motivated, and less individually productive as a result of the autocratic imposition (R. Likert, 1977b). Consistent with this finding, Dunnington, Sirota, and Klein (1963) noted that when engineered work standards were imposed on managers and supervisors at an IBM manufacturing plant, the employees resented the pressure, and the same adverse trends occurred in the measurements of the human consequences. Yet Dunnington, Sirota, and Klein did find exceptions. When particular supervisors were more democratic, employees showed much less resentment of the engineered work standards than did employees whose supervisors were more autocratic. The employees who worked for democratic supervisors felt less resentful because the democratic supervisors were more likely to try to do something if an employee complained that the work standard was unreasonable.
Contingency research, to be discussed in Chapter 25, supports the need for different organizational structures in different industries (see Lawrence & Lorsch, 1967a, 1967b; Woodward, 1965). Assembly plants (mass and batch production) have different organizational structures from oil refineries (continuous-processing). Nevertheless, Likert (1977b) noted that regardless of industry, the better-performing plant or department was likely to be closer to System 4 in its leadership processes, and the poorer-performing plant or department was likely to be closer to System 1 in its management system.10
Effects on Business Organizations. Results from 30 studies in 35 business firms involving some 260 sections, departments, or similar organizational units containing more than 50,000 employees were reported by R. Likert (1961a), Likert (1967), Likert and Likert (1976), and Likert and Fisher (1977). They demonstrated the efficacy of democratic as opposed to autocratic systems of management.
Nineteen of the studies, as summarized by Likert (1977b), were comparisons of more democratic organizations that were closer to System 4 in their leadership and management with organizations that were closer to the autocratic System 1. The differences in productivity and earnings favoring System 3 or 4 over System 1 or 2 ranged from 14% to 75%. In “before-after” studies, in which management shifted toward System 3, productivity and earnings improved from 15% to 40% one or two years after the shift. In two comparisons in which control groups were available, no such improvements were obtained for the control groups.
These improvements continued if the democratic shift was maintained. In departments of fewer than 200 employees, the improvement usually resulted in annual savings of $50,000 to $100,000. In a large plant of 6,000 employees, the annual saving was more than $5 million. Guest (1962a) observed similar results for productivity, quality, and safety when a new manager of the poorest-performing of six plants shifted the organization toward democratic Theory Y leadership.
For 15 business firms, Taylor and Bowers (1972) reported the relations between the measures obtained from the POC and various organizational outcomes 6 months prior to the survey to 18 months afterward. Correlations between having a democratic climate and efficiency reached as high as .80. But correlations were somewhat lower with reductions in absenteeism, minor injuries, ill health, and grievances.
Effects on Governmental Agencies. Heslin (1966) found that the high-producing units in a federal government agency engaged in automatic data processing were closer to Likert’s System 4 than to System 1 in their management, as seen by the employees, than were the low-producing units. Operational bureaus of the Department of State were seen as providing better budgeting, space, travel, and personnel services if their management (according to their own subordinates) was closer to System 4 than to System 1 (Warwick, 1975). Similarly, Likert (1977b) reported that city managers, when asked to compare the highest-producing unit with a matched lowest-producing unit they knew well, described the highest unit as between Systems 3 and 4 and the lowest unit as pursuing benevolent autocratic (System 2) management. Bruns and Shuman (1988) administered the POC to 298 Arizona police sergeants and 67 police lieutenants. The modal leadership system in departments of large and medium size was System 2, benevolent-autocratic. The modal leadership system in smaller departments was consultative, System 3.
Effects on Military Organizations. In data from 20 ships and 18 shore stations of the U.S. Navy, Bowers (1975) found a strong relationship between an individual’s intention to reenlist (which is a good predictor of actual reenlistment) and the extent to which the ship or station was closer to System 4 and farther from System 1. Likert (1977b) reported that among 14 U.S. Navy crews, the absence of mishaps and operational failures was associated with supervisory facilitation of work and team development. D. E. Johnson (1969) found that of 93 U.S. Air Force ROTC units, those that were judged by higher authority to be operating closer to System 4 than to System 1 were also evaluated as better-performing units.
Effects on Educational Institutions. Summarizing 40 studies in school systems, R. Likert (1977b) concluded—from school POC surveys of members of boards of education, superintendents, central staff, principals, department heads, teachers, students, and parents—that school systems closer to System 4, compared with those closer to System 1, exhibited better communications, cooperation, and coordination (Lepkowski, 1970). They were more flexible and innovative (Broman, 1974)11 and more effective overall (Ferris, 1965; Riedel, 1974). Their personnel felt a greater sense of self-actualization and satisfaction from their work (Wagstaff, 1970).12 Furthermore, they were judged as achieving superior educational results. They had better board-employee relations (R. C. Key, 1974) and union-management relations (Bernhardt, 1972; Haynes, 1972). Their students were more highly motivated and attained higher educational achievement for given IQ and socioeconomic levels (Belasco, 1973; A. K. Gibson, 1974). Their students had more favorable attitudes and were less likely to engage in disruptive behavior or acts of aggression against the schools (Cullers, Hughes, & McGreal, 1973; Morall, 1974).
For 12 studies of higher education, Likert (1977b) concluded that institutions whose administrations were closer to System 4 than to System 1 experienced more favorable outcomes. The faculty members were more satisfied with administrative decision making (A. B. Smith, 1971).13 There was less need for collective bargaining (Cline, 1974), more innovativeness (Bowers, 1976; Hanna, 1973) and commitment to college objectives (T. G. Fox, 1973; Laughlin, 1973), and more favorable student outcomes (Bowers, 1976; Gilbert, 1972).
Effects in Health Care Organizations. On the basis of data from 351 nurses in 55 patient teams in eight hospitals, Munson (reported by Likert, 1977b) found that the closer the head nurse was to System 4, the more generally satisfied the nurses were. H. C. White (1971b, 1971c, 1971d) obtained similar results for peers who described effective and ineffective supervisors they had known. System 4 was seen to be more effective than System 1 in three outpatient clinics (National Tuberculosis and Respiratory Disease Association, unpublished) and by Ketchel (1972), who studied the effectiveness of volunteer health planning in 17 Ohio counties.
Effects in Other Nonprofit Organizations. In three community-based reintegration centers for ex-convicts in Ohio, McGruder (1976) concluded that in comparison with the most autocratic center on Likert’s POC, the most democratic center was most effective, as measured by graduation rates and low rates of recidivism and reincarceration. carceration. Similarly, Marchant (1976) obtained data on Likert’s POC from the staffs of 22 research-oriented university libraries. The closer the libraries were to System 4 and the farther they were from System 1, staffs were more satisfied, and faculty evaluations of service were higher. Haggard (as reported by Likert, 1977b) found that as management of a YMCA shifted away from System 2 toward System 4, the number of people served by the YMCA increased from 11,064 to 23,794, and the budget increased (from $173,000 to $303,000.
Methodology. Despite the amount of support attesting to the efficacy of democratic leadership in the long term, disquieting ambiguities remain. Thus among the reasons why Miller and Monge (1986) excluded the results of 15 journal articles from their meta-analysis was their determination that the democratic approach had not been clearly measured or experimentally manipulated. They found additional methodological problems in seven other studies.
Differing Outcome Measures. Other difficulties arise, reducing confidence in the reported conclusions, owing to the extent to which differences occur from one study to the next in the definition and measurement of the satisfaction of followers and the productivity of groups. Some studies measure global satisfaction, whereas others measure satisfaction with leadership, the job, the group, or the organization. Some researchers count units of output or rates of performance as measures of productivity. Others use ratings of quantity or quality of output as productivity measures.
Circumstances. Leadership style may be a product of circumstances rather than of personal preferences. Scully, Sims, Olian, et al. (1994) found in 56 high-tech business firms that when financial performance was poor rather than better, the CEOs tended to be tougher, more directive, and more punitive toward their top managers. Southern Baptist preachers become more authoritarian in response to ambiguity in their role definition. Ingraham (1981) suggested that the ambiguity of the pastor’s role in Southern Baptist churches results in the development of an authoritarian self-image among pastors. When the pastors are unsuccessful in influencing their congregations, they either withdraw from their attempts to lead or try to become manipulative authority figures.
Effects on the Leader. The leadership style that is adopted may affect the leader as well as the follower. Kipnis, Schmidt, Price, and Stitt (1981) randomly assigned 113 business students to act as authoritarian or democratic leaders of five-person work groups that manufactured model airplanes. They found that those who acted as democratic leaders perceived their group members to be more internally motivated to work effectively than did those who acted as authoritarian leaders. As a consequence, the democratic leaders gave the members more favorable evaluations, although their productivity was not necessarily higher.
Confounding of Cause and Effect. Chapter 16 noted that followers condition a leader’s behavior; that is, the leader of a productive group can afford to be more considerate to his or her subordinates than can the leader of a poorly performing group. In addition, Mitchell, Larson, and Green (1977) showed that subordinates’ descriptions of a leader’s initiation and consideration are erroneously confounded with the perceived quality of his or her group’s success and morale. The error is compounded because the same source of information about the leader’s behavior provides the indicators of perceived group success and morale. Under these conditions, the correlations between democratic leadership on the one hand, and group success and morale on the other, are inflated above the true correlations.
In interpreting the findings, which comes first, the chicken or the egg? A concurrent study of leadership and its consequences may in reality be a study of leadership and its antecedent conditions. Leaders may be authoritarian because their groups are unproductive, or can afford to be democratic because their groups are productive. The longitudinal studies are few; the concurrent studies are many. Even the efforts to move organizations in a democratic direction suggest that results may not be immediately apparent. One or two years may elapse before a change in leadership style has measurable effects on organizational performance.
Reflection of Implicit Theories in Results. The issue of cause and effect is complicated by the extent to which correlations reflect the raters’ implicit theories of leadership. That is, a rater may assume that subordinates will be more productive under a particular pattern of supervision. They may make their report about the supervisor’s style of behavior as a consequence of how they see the group performing (Rush, Thomas, & Lord, 1977). In Mitchell, Larson, and Green’s (1977) study, knowledge that a group performed well caused increases in the rated consideration and initiating of structure of that group’s supervisor, whereas knowledge that a group performed poorly caused large decreases in the supervisor’s rated consideration and initiating of structure. These distortions in leadership ratings because of the knowledge of performance also occurred when raters of high-and low-performing leaders were exposed to identical and highly salient leadership behaviors (Lord, Binning, Rush, & Thomas, 1978).
A large cluster of styles can be included in democratic leadership and member-related behaviors: employee-oriented, considerate, concerned with the satisfaction of needs, maintenance-oriented, rewarding and nonpunitive, supportive, relations-oriented, open, close, informal, warm, System 3 or System 4, and people-centered. Included in autocratic leadership or work-related behaviors are opposites of those just mentioned: job-centered, structuring, task-oriented, punitive, closed, distant, formal, cold, System 1 or System 2, and work-centered.
The positive effects of democratic approaches are most apparent if based on the results of large-scale field surveys and lagged productivity measurements rather than on small-group laboratory experiments with immediate, concurrent effects. Generally, the patterns of behavior included in democratic leadership are more satisfying than those associated with autocratic leadership. But in the short term, productivity may be enhanced more by autocratic leadership than by democratic leadership. This is especially so when the democratic leadership ignores concern for the task and production goals. Nevertheless, in the long term, the positive effects of democratic leadership are evident, especially if the employees’ development, commitment, loyalty, and involvement are important to productivity. But numerous conditions, such as the authoritarianism of subordinates or the nature of the task, increase the utility of autocratic methods, particularly in the short run.
Looking at their components provides a fuller appreciation of the autocratic and democratic styles of leadership behavior. Leaning toward the autocratic are direction, task orientation, and initiation of structure. Leaning toward the democratic are participation, relations orientation, and consideration for followers. These components are covered in the chapters that immediately follow.