The distribution of power and the leadership associated with it have been a main theme in both political history and the history of management-worker relations. Autocracies and oligarchies have represented concentrations of power; democracies have represented a wider distribution of power. In the former, the individual autocrat with a patronizing staff of subordinates dictated to a relatively powerless membership. In the latter, power was dispersed voluntarily or as legally mandated. The past century has witnessed in the developed world a substantial reduction in the range between the powerful and the powerless. The differences in power between women and men, parents and children, teachers and students, and supervisors and supervisees have all been reduced. Apart from some Asian, African, and Latin American countries, autocracies, dictatorships, feudalistic, and tribal societies have been replaced by more democratic forms of government, business, and other institutions. Nevertheless, in many organizations, communities, and countries, large differences in political, economic, and social power are the rule rather than the exception. In informal groups, power differences are likely to evolve. In initially leaderless groups, power accrues to the active and more knowledgeable members.
Mulder (1976) focused attention on such differences in power in organizations and the tendency of those who are striving for more power to reduce them. Hofstede (1980, 1997) unearthed a “power distance” factor in a large-scale survey of employees’ attitudes in 50 countries. A large power distance was indicated, for example, in countries in which employees expressed fear of disagreeing with their supervisors.
Power and Leadership Style. Hofstede suggested that with a large power distance between leaders and followers, or supervisors and employees, the followers and employees became either dependent or counterdependent. When differences in power are great, more autocratic leadership and more coercion are likely. Large differences within a group, organization, or society—particularly in coercive power, and to a lesser extent in personal, reward, and legitimate power—may guarantee public success and acceptance of the leadership. But resistance, at least private resistance, will bring about unintended consequences, such as resentment and apathy. On the other hand, the maintenance by leaders and supervisors of a small power distance between themselves and their followers and employees encourages consultation and participation. Many behavioral science theorists1 have observed that the sharing, leveling, and equalization of power among members of a group increases the members’ participation and results in their full commitment and their acceptance, both public and private, of the leader.
Elitism versus Populism. Even within a society’s democratic order, there are wide differences in power. Populism competes with elitism. Populists believe that all citizens share equally in the opportunity to exert influence through political activity. Elitists believe that power should be delegated to those who are committed to the rules of decision making and are highly knowledgeable and skilled in analysis, negotiation, persuasion, and manipulation. Populism calls for a high degree of citizen participation; elitism calls for a relatively passive, un-involved citizenry, who, from time to time, “in a single elemental choice,” elect active officials who will be responsible for the stability, efficiency, and authority of the state (Summers, 1987). Populism within the work organization translates into shop-floor democracy, in which all employees share consultation and decision making with their immediate superiors. Elitism within the organization translates into the election of experienced and knowledgeable union officials and council representatives, or heavy dependence on informal leaders from within the workforce.
Populism within the small group or team translates into shared decision making by the group members; elitism translates into elected heads whose power to decide is unchallenged by the members, who remain passive following the election. In elitist groups, most of the power resides in the leader; in populist groups, it is widely shared and equalized among the members and the leader. With such equalization of power, the members become as influential as the leader, and their effects on each other become important to the process of interpersonal influence. Thelen (1954) noted the power and control of the group over its members. Group agreements “have teeth in them.” The determination of what is possible and at what cost or with what reward is under the control of the group. Leadership occurs in changing the agreements and working out new ones to fit the continual diagnosis of realities in the group problem situation.
According to Tannenbaum and Massarik (1950), shared decision making increases the likelihood that workers will accept the goals desired by the management. This results in their greater satisfaction and increased efforts to move toward the selected goals. Likewise, McGregor (1944) argued that the effectiveness of workers will be increased when they are given opportunities to participate in finding solutions to problems, to discuss actions that affect them, and to assume responsibilities when they are ready to do so. If members of a group participate in setting goals, they will be more motivated to achieve the goals. If they have the capabilities, they can be delegated more responsibility or freedom to act without a review by a higher authority (Learned, Ulrich, & Booz, 1951).
When all members are equal in status and power, more communication of feelings will occur (Bovard, 1952). Differences in status and power in a group result in the inhibition of the ventilation of feelings by members who are lower in the status and power hierarchy. Such inhibition can be detrimental to the group’s functioning. Thibaut and Coules (1952) found that the communication of hostility toward instigators will reduce the residual hostility toward them. Again, power differences can offset the influencing effects of participation. Bass, Flint, and Pryer (1957b) and Bass (1963) ascertained that for individuals to influence a group’s decisions, they had to participate and to attempt to lead. Nonetheless, such active participation did not necessarily influence the group’s decision when the members of a small group varied widely in power.
Caveats. This does not mean that all communications are easier and all outcomes are better when members are equal in status and power. Although some believe that power sharing and participation are a universal panacea for promoting change, productivity, and satisfaction, experimental researchers such as Locke, Latham, and Erez (1987) have pointed out that considerable qualification is needed. For instance, if the leader is clear and supportive, the members’ acceptance of hard group goals can be obtained without the power sharing implied in group discussions and decision making. More competition from members and persistence by the leader in the face of rejection are likely to be required of the would-be leader of a group in which all members are equal in status and power. Thus Bovard (1951a) observed that in groups where there is little status differentiation among members and all have similar power to influence each other, verbal interaction is higher and more influencing is attempted than when a single person of higher status directs the activities. It may be that although communication seems easier when members are equal in power, it really is more difficult. Shepherd and Weschler (1955) studied work groups that differed in internal status stratification. The members of these groups felt they had fewer communication difficulties, but they actually experienced more such difficulties.
Formation of Coalitions. The distribution of power in a small group is dynamic. Two or more members can combine their power to increase their joint power and influence the course of events. Lawler (1975) found that expectations of support from others were an important cognitive determinant of the formation of a coalition against inequitable and threatening leaders. How reward power is distributed in the triad has been the focus of experiments regarding the bargaining process. Mills (1953) assembled triads in which one member was given high reward power; one, medium reward power; and one, low reward power. A strong tendency was found for two of the members to form a coalition that acted in opposition to the third member. Vinacke and Arkoff (1957) confirmed hypotheses advanced by Caplow (1956) that: (1) when all members of a triad are equal in power, all possible combinations will be formed; (2) when two equal members who are stronger than the third member combine, they will form a coalition in opposition to the third member; (3) when one member is weaker than the two equal members, he or she will form a coalition with one of them; and (4) when one member is stronger than the other two combined, no coalition will ensue. Turk and Turk (1962) confirmed the fourth of these hypotheses. But Kelley and Arrowood (1960) found that the results reported by Vinacke and Arkoff depended on the clarity of the game and on whether a player was assured of a certain return whether or not he or she joined a coalition.2
Coalitions are also organized as a tactic to increase one’s upward influence. Kipnis and Schmidt reported that respondents exerted more upward influence on their boss by obtaining the support of their coworkers and subordinates to back up their request, or they arranged a formal meeting with the boss at which they made their request. Such coalitions were used at about the same frequency to influence coworkers, subordinates, and the boss.
The stability of the distribution of power begins early. In an observational study of kindergarten children, Hanf-mann (1935) noted a stable power structure that was similar to a pecking order. The same child would continue to give orders to the same second child, who did the same to the same third child. Gellert’s (1961) observations of pairs of preschool children during three play periods found that in a significant proportion of these dyads, the same child maintained the same position of dominance or submission during the three periods. This again suggests that a pecking order was at work.
There are two opposing arguments about how power should be distributed in organizations. In the structuralist view, organizations are mechanistic rather than organic (T. Burns & Stalker, 1961). It is inherent in the nature of organizations with systems of command and control that power is distributed unevenly. As organizations develop, power accrues in varying amounts among the units and supervisors. Some units (and, therefore, their supervisors) obtain more power than others in influencing the allocation of the organization’s scarce resources (Pfeffer & Salancik, 1978). The units and supervisors with more power have control over strategic contingencies and the activities of other departments that are critical to their functioning; this makes it less likely that there will be any way of substituting for the more powerful units’ and supervisors’ activities. Also, information is provided to the less powerful units and supervisors to decrease uncertainties. Thus more powerful units and supervisors are in the center rather than at the periphery of information and work flow (Hickson, Hinings, et al., 1971). Support for these propositions was found in 20 subunits of five breweries by Hinings, Hickson, Pennings, et al. (1974). Similar results were obtained for 62 vice presidents, deans, and directors of six universities (Saunders & Scamell, 1982).
Power holders want to retain their power, so the power distribution in mechanistic organizations remains stable over time unless there are shifts in technological and environmental demands or the power holders are replaced (House, 1991). But temporary downward shifts in power become necessary as leaders’ positions enlarge. Increasingly, leaders need to rely on subordinate staff as intermediaries, deputies, and surrogates. Kings had to share their power in distant colonies with viceroys. Whyte (1953) and Simon (1957) argued for structure, power differences, and status differentiation. They contended that hierarchical organization is the natural, biological, and sociological solution for ensuring a group’s survival. R. B. Cattell (1953) hypothesized that status differentiation and the accompanying power differences promote speed in decision making and the faster attainment of goals.
Human Relations Perspective. The opposing human relations perspective sets forth that increasing the equalization of power in organizations was essential to the organization’s health. Reducing power differences was a basic tenet of modern organizational development theory and practice (Bennis, 1965) and the human potential movement (Maslow, 1965). The organic organization was favored. It had its roots in those aspects of democratic and egalitarian values that stress the importance of informal trusting relationships rather than a formal structure in which relationships depend on the authority of position and role requirements. Practical applications include unstructured leadership, job enlargement, leadership sharing, participative management, consultative supervision, joint consultation, workers’ councils, and other forms of industrial democracy. Patchen (1970) administered questionnaires to 90 employees in plants operated by the Tennessee Valley Authority and found that the degree of control workers had over their jobs was significantly correlated with their general interest in their jobs, concern with innovations in their work, and pride in their work. These findings were supported by communicationnet studies (McCurdy & Eber, 1953).
Argyris (1957) and McGregor (1960) were convinced that individuals, motivated by their basic desire for autonomy and self-actualization, were frustrated by organizational structure, specialization of work, and philosophies of management that assumed that workers would remain complacent and unproductive unless they were subjected to controls hampering unwanted initiatives from them. When workers were less encumbered by controls, their first-line supervisors also benefited. Atwater (1996) found that in organic organizations, more power was bestowed on first-line supervisors relative to the power at higher levels of management, as reported by 280 employees from 45 traditional or organic organizations.
Stability in Organizations. House (1984) deduced that the distribution of power in organizations would tend to become stable over time and would change only if there were major shifts in environmental or technological demands on the organization. But even here, changes in the power distribution would come slowly because those in power are reluctant to part with any of it. In fact, given the tendency of the powerful to gain more power, demands for change in the organization may bring about a greater concentration of power rather than equalization.
Implications for Organizational Design. The two views have been the basis of opposing prescriptions for how organizations should be designed. Structuralism emphasizes designing organizations with power differences; humanism emphasizes designing organizations with more power equalization. Legitimate differences in power are manifested in an organization in how much the positions of members are structured—that is, prescribed and bound by rules and regulations or the dictates and policies of a higher authority. An infantry squad is highly structured; an initially leaderless discussion group is not. In the infantry squad, there is a clear legitimate hierarchy of power from the squad leader down to the new private; there is no such legitimate hierarchy in the leaderless discussion group. Equalization of power is fostered by humanism in its support for informality, trust, learning, and the sharing of information.
Structuralism as enunciated by such classical management theorists as Taylor (1911), Fayol (1916), Follett (1918), and Davis (1942) depends on external constraints to gain the compliance and commitment of followers. The human relations viewpoint, through the equalization of power, expected to gain even more commitment as well as compliance through the individual’s awareness of and insight into effective interpersonal relations and trust.
Structuralism called for someone to be responsible for supervising all essential activities. The hierarchical pyramid of supervisor-subordinate relationships was mandated. Power sharing proposed that any team member could take on leadership responsibilities when he or she saw the need to do so. The autonomous team with no formally appointed leader was the extreme example.
Structuralism avoids the duplication or overlapping of responsibilities. However, overlapping, cross-training, rotation, and other forms of sharing responsibilities to increase the reliability of the system are encouraged by human relationists when informal relationships are developed among workers who are committed to see that their group prospers. The simplification of jobs is fundamental to a structuralist approach. According to this perspective, no individual should be responsible for a wide assortment of unrelated acts. In contrast, human relations theorists call for the enlargement and enrichment of jobs; they believe that subordinates who receive bigger jobs with greater responsibilities become closer in power to superiors.
The structuralist position calls for subordinates to receive clear, written job specifications and role assignments. The human relations position calls for making goals clear and allowing subordinates, commensurate with their training and experience, to decide how to reach the goals. Both structuralists and human relation-ists agree that authority should be delegated so that decisions take place as close as possible to the action. For structuralists, the decision maker is the superior in the situation; for human relationists, the decision is shared or delegated to the subordinates as much as possible.
Structuralists emphasize the chain of command. Human relationists argue that organizations require much communicating, reporting, proposing, influencing, complying, and deciding in complex vertical, diagonal, and horizontal paths that are not shown on the formal organizational charts. Again, participants need some sense of the equalization of power for this flow to occur effectively.
Neither the structuralists nor the human relationists have a monopoly on the most effective design (Bass & Ryterband, 1979). Situational elements and personal preferences affect which approach pays off. Young professional workers, compared with older unskilled employees, are likely to call for less structure and more equalization of power. Small-business managers, secure in their personal relations with their parents and attachment to their spouses, will be more comfortable in decentralized organizations with less structure; less secure counterparts who prefer distance in their personal relationships will want centralization of power (Johnston, 2000). Some pupils thrive in schools with a great deal of freedom to choose what and how to learn, but others do best in a highly structured environment. Juvenile delinquents who adjust most readily to institutionalization may have more difficulty avoiding antisocial, criminal behavior once they are back on the streets. The optimum differentiation of power is likely to lie between the extreme structuralist and extreme human relations positions, depending on the situation. In a pluralistic society, the only thing one can safely say is that wide differences in performance, attitudes, and response to hierarchy are likely to depend on task, time, place, immediate needs, and such.
The distribution of power provides the potential for leadership and influence over events and people. It shows itself and its effects in how it is distributed in communities, in formal organizations, and in smaller groups. In small communities, citizens may have the power to directly participate in decisions, such as at town meetings. In cities, they may have the power to participate indirectly through attendance at civic meetings, consultation with officials, and keeping themselves informed. They may reveal their power in elections, formal petitions, protests, initiatives, and referenda. Or they may be powerless and manipulable (Couto, 1992). Central to civic improvement in many U.S. metropolitan areas has been citizens’ power to directly participate in decisions regarding innovation and change. In Portland, Oregon, citizen participation became expected as mayors, governors, and legislators introduced land laws, new parklands, light rail, and a revived active downtown to provide a high quality of life. In San Antonio, civic organizations pushed their elected officials into major changes, such as on the riverfront. In many cities such as Denver, business and government leaders shared power and responsibilities in turning the economy around. Catholics were mobilized by a bishop in Cleveland to deal with urban sprawl and its effects on the poor of the inner city. Neighborhood associations, commonplace in cities and suburban communities, deal with local as well as communitywide issues (Pierce & Johnson, 1997).
The earliest research on power was concerned with identifying individuals who had the power to influence their local communities. Lynd and Lynd (1929) pointed the way in their study Middletown. F. Hunter (1953) used interviews in a large city to obtain a list of 175 persons who were mentioned as wielding influence in the community, with reasons for the belief that they had such power to influence. A panel of experts reduced the list to 40 persons who were regarded as constituting a monolithic power structure in the city. However, Dahl (1961) maintained that the nominating technique hid important details that could be uncovered by a study of community issues. He concentrated on identifying those individuals who had played key roles in promoting or blocking issues. His analysis of the data suggested that there was a pluralist power structure in the city. Using both Hunter’s and Dahl’s methods, Freeman, Fararo, Bloomberg, and Sunshine (1963) and Presthus (1964) found that the two methods identified a common core of power figures and that Dahl’s method revealed additional subgroups of individuals who could influence different issues.
The status and power structures of many communities were studied.3 Consistently, people who were identified as having power also tended to rate high in social and economic status. Educational level, as such, was less important (Alford & Scoble, 1968). However, the acceleration of the importance of new technology in everyday life has given professionals increased power in their communities. That is, as problems of health, economics, safety, and so on come to the fore, the expert power of professionals takes on increasing importance in the community.
When Hunter and Fritz (1985) studied the power structures in four Chicago suburbs, they found that the smaller communities contained less complex structures. This finding fit with V. Williams’s (1965) conclusion that smaller communities may be able to function effectively without the differentiation of roles and individuals in specialization, power, and authority. Similarly, Nix, Dressel, and Bates (1977) described the dispersion of power in small rural communities, drawing on 189 interviews and the reconstructed history of a country in the southeastern United States since 1900. They noted that the initial bossism was replaced by informal cliques, organized pluralism, bifactionalism, multifactionalism, and amorphous leadership. In the Chicago suburbs studied by Hunter and Fritz, the informal status structures were less complex in the richer suburbs than in the poorer ones.
Equal status and equal power are more commonly found when familiarity, homogeneity, and the potential to interact are high. Caplow and Forman (1950) observed that when neighborhoods contained families who were homogeneous in duration of their residence, their interests, and their type of dwelling, no status differentiation occurred. Form (1945) reported similar findings for Greenbelt, Maryland, when it was a newly established community for federal white-collar employees of similar occupations, age, nativity, and housing. Similarly, Munch (1945) found that the homogeneous inhabitants of isolated Tristan da Cunha in the South Atlantic maintained their community without status differentiation or institutionalized government.
Interlocking Office Holding. Perrucci and Pilisak (1970) compiled a list of 434 organizations in a community with a population of 50,000. Of 1,677 executives in these organizations, 1,368 held a position in only one organization. Twenty-six executives who occupied positions in four or more organizations were compared with 26 executives who occupied a position in only one organization. The multiorganizational leaders were regarded by both groups as being more powerful and as having more influence on actual and theoretical issues. These leaders were named more often as social and business friends and were found to constitute a powerful network of influence relationships. The concentration of power was seen in the city boss, studied in detail by Banfield and Wilson (1963), Moos and Koslin (1951), and Zink (1930). Mayor Richard Daley of Chicago epitomized the use of social exchange relationships with other bloc leaders and centers of community power to build the powerful political machine that claimed to “make Chicago work.” His son, as mayor, continued the tradition of relationships.
Power reveals itself in formal organizations in the extent to which it is associated with status and authority—the formal authority system—and in the beliefs shared by the members in the organization’s system of ideology. An organization’s system of authority involves the way legitimate power is distributed and enacted in goals, rewards, sanctions, and the division of labor. It is based on the legitimate power of the different roles and positions, and the status and authority of the position holders. The organization’s system of expertise is the way information and knowledge are distributed among its positions and position holders. It is more informally applied to solving the complex problems of the organization (Mintzberg, 1983). The interlocking of systems was illustrated by Majchrzak (1987), who found that rates of unauthorized absences in units in 20 Marine Corps companies were significantly reduced, in contrast to 20 control companies, if policies were clarified and communicated effectively to all levels involved and there was a hierarchical consistency in agreement and understanding of the policies and requirements at the different levels.
The systems offset each other. Although the authority system would dictate otherwise, the system of expertise may result in managers being less influential than their subordinate staff experts. B. Walter (1966) traced the transmission of influence in two municipal organizations. Superiors generally were found no more influential than their staff subordinates in making decisions; in fact, the subordinates were more influential when making novel decisions.
The political system operates outside the system of formal authority to benefit some constituents at the expense of others or at the expense of the organization (Mintzberg, 1983). Although the political system is likely to conflict with the other systems, it can also serve to mobilize power when needed for the acceptance and implementation of new strategies. The interacting systems of power can be observed in community settings as well.
Status and Power. One of the most consistent findings in social science is the general tendency for those higher in status in an organization to wield more power to influence those who are lower in status. Thus, at Walt Disney World leaders are expected to encourage their staffs of “cast members” to connect with visitors emotionally by “owning” their roles and monitoring their performances and the visitors’ reactions. Among many studies of the relationship, Blankenship and Miles (1968) found hierarchical position to be a more important determinant of managers’ decision-making behavior than the size of a company or the number of subordinates that the manager supervised. Bass (1960) argued that by creating greater differences in power and status between the leader and followers, one can increase the proportion of success in acts of leadership. This outcome was observed by Hemphill, Seigel, and Westie (1951) in a study of 100 groups that varied in status differentiation. Such leadership was also likely to be more dominating, directive, and coercive and more likely to define and structure the work for the membership. Again, as was expected, a related study of 212 aircrews by Rush (undated) found that commanders were much less considerate of their subordinates when the crews were highly stratified in status.
House (1988a) concluded that when social stratification (status differentiation) exists and the holders of authority feel strongly about their status, they will have fewer inhibitions against being coercive. The greater the social stratification, the more likely they are to be coercive in order to obtain compliance from those at lower levels. Nevertheless, high-status powerful persons may fail to be successful leaders over the powerless. They may become distrusted (Tenbrunsel & Messick, 2001), particularly if they usurp power or are coercive.
Effects of the Total Amount of an Organization’s Power and Control. A. S. Tannenbaum (1956a, 1956b, 1968) proposed that the total power and control in an organization varies rather than remaining fixed. Power and control can be expanded by increasing authority—legitimate power—at one or two levels or at all levels of the organization. An organization tends to gain most in the total amount of its control of its members when authority is expanded all the way down the line particularly to the lowest levels. In studies of two industrial plants (1956a) and four labor unions (1956b), Tannenbaum found that organizations of the same type could differ in the amount of authority exercised by executives at different levels and in the total amount of organizational control.
Smith and Tannenbaum (1963) studied 200 units of large firms. They found that the total amount of organizational control was positively and significantly related to the members’ loyalty, morale, and judged effectiveness but not to objective measures of effectiveness. Nevertheless, leader-member agreement on the ideal amount of control and the actual total amount of control over members was related to rates of effectiveness. On the other hand, when Smith and Tannenbaum (1965) examined the effects of the total amount of control by the members of women’s organizations, the total amount of control failed to increase the coordination of leadership, although such leadership contributed to the group’s effectiveness.
In a study of insurance agencies, Bowers (1964b) reported that total organizational control was related to the members’ satisfaction and performance but not to the volume of business, the growth of business, or the turnover of personnel. Furthermore, the amount of power at different levels of an organization was not associated with effectiveness. Smith and Ari (1964) observed that a consensus among members of work groups and among superiors and subordinates was related to the total amount of organizational control but not to the amount of control of individuals at the work level.
Ivancevich (1970) studied the effects of the total amount of organizational control and members’ satisfaction in insurance agencies. The agents’ satisfaction with status, autonomy, and growth was correlated about .30 with the manager’s control over the agency, the agents’ control over the agency, and the total amount of organizational control. Ivancevich found that agents were equally satisfied whether they or the manager exercised control. In another study of insurance agencies, Bach-man, Smith, and Slesinger (1966) found that the agents’ satisfaction was highly related to the total amount of control, control by the manager, and control by the agents.
In addition to the effect of differences in the total amount of control exercised by members, the total amount of control is higher when exercised by members in all echelons of the hierarchy rather than at just one or two levels. The total amount of control, but not control in the various echelons, is related to productivity and to members’ satisfaction. Furthermore, in some organizations members tend to feel more satisfied when the top-level leader is charismatic and exercises strong control over organizational activities (House, 1988).
Although this point has not been examined, the reason why high total control increases satisfaction and productivity may be that goals and expectations are much clearer for members and conflict is much lower than when organizational anarchy exists. Nonetheless, many of the older studies of satisfaction with control need replication, because organizations have become flatter and workers and managers expect more from participation and democratic leadership.
The distribution of power is ordinarily inferred from observing variance among members of a group or organization in the success of their efforts to influence each other. This inference leads to a tautology. Observed differences in influence indicate the distribution of power; differential power begets differential influence. The equalization of power is seen in shared leadership. Power differences are also inferred from differences in status and esteem. The general has more legitimate power than the colonel. The admired parent will have more referent power in the family than an estranged one.
Measuring the responsibility, authority, and delegation associated with a position is a more direct way to understand an incumbent’s power.4 To obtain a direct measure of power differentiation independent of observed influence and status, Bass and Valenzi (1974) asked subordinates and managers to describe the power that managers had over manager-subordinate relations and the power the subordinates had over these relations. Questionnaire items dealing with the managers’ power included the extent to which the managers had the power to override or veto any decisions the subordinates made; grant or deny promotion or salary increases to subordinates; reverse the priorities of subordinates; control the size of subordinates’ budgets; and get the support of a higher authority for what the managers wanted to do. The subordinates’ power included bringing outside pressure to support what they want; doing the opposite of what the manager wants done; maintaining final control over their own plans, assignments, and targets; ignoring the manager and submitting their own requests to a higher authority; and nominating or electing the manager. Bass and Valenzi obtained alpha coefficients of .60 for the reliability of both management’s and subordinates’ power. They found that on scores ranging from one to nine, superiors’ power was typically above six whereas the subordinates’ score was below four.
Dependency. Kotter (1978) took a different route to study power differences in organizations. His interviews focused on whom the incumbent in a particular management position felt dependent. Kotter also assessed the incumbent’s efforts to obtain cooperation, compliance, and deference from others in the system. Integrating these concerns about power and dependence with a concern for the organization’s goals was seen as the key to effective performance and was consistent, as was noted in Chapter 7, with McClelland and Burnham’s (1976) demonstration that effective managers have a strong need for power but are oriented toward organizational goals rather than self-aggrandizement.
Power of Departments. For their study of the distribution of power among departments at universities, Saunders and Scamell (1982) modeled their assessments after Hinings, Hickson, Pennings, and Schneck (1974). They were able to obtain valid measures of three determinants of how departments differed in power. From data obtained in interviews and questionnaires, they indexed the department’s centrality in the flow of work and information, the difficulty of substituting an alternative department for the department’s activities, and how much ability the department had to cope with uncertainties. Using small-space analysis, Shapira (1976) showed that the differences between the power of managers and subordinates, when combined with the information that each had, could predict, as theoretically expected, which style of leadership would be most often displayed by the managers. Managers with more power and information than their subordinates were most directive; managers with more power but less information were most consultative; managers with more information but less power were most negotiative; managers with less power and information were most delegative.
Autonomy. The distribution of power can be measured indirectly by how much is delegated to the less powerful and by how much autonomy and freedom they have to choose how to operate in the work setting. Objective characteristics of the situation may constrain autonomy. It also is a subjective experience, such as being one’s own boss. Sinha and Viswesvaran (1998) conducted a meta-analysis of 99 samples of autonomy from 56 published studies. The variety of scales and measurements registered a mean reliability of .69 for the managers’ perceived autonomy and .73 for nonmanagers. Meanings of autonomy included being allowed to try and fail without fear of reprisals; freedom from constant evaluation; freedom from close supervision; sense of ownership of the work; discretion in scheduling the work; determination of what needed to be done; freedom to make decisions without checking with a supervisor; taking part in decisions affecting the work situation; exercising personal judgment; having an opportunity to express ideas, and being treated as an equal by the supervisor.
Power sharing with all members of a group does not necessarily mean increased initiative and freedom for members. On the contrary, powerful groups can constrain and influence their members more strongly than any individual leader with power can. Mere membership in a group makes a difference. Deutsch and Gerard (1954) found that individuals in a group were more influenced by observed standards than were those who were not in a group. Katz and Lazarsfeld (1955) concluded that people are more likely to change their opinions if their relatives or friends are undergoing a similar change. The changes are also more likely to persist if supported by such group affiliations. In the same way, individuals are more influenced by the mass media when they are listening in groups than when they are alone.
Socialization processes in organizations are a familiar phenomenon. For example, after liberal students enter the world of work, they tend to adopt more conservative attitudes involving the norms of the firms they join. S. Lieberman (1954) found that workers who were promoted to foremen tended to shift their attitudes in favor of the management as they became members of management. Workers with the same attitudes as those who became foremen but who subsequently were elected as union shop stewards, shifted their attitudes in the union’s direction after 12 months as stewards.
Many early investigators demonstrated the tendency of members of a group to shift toward the attitudes and behavior of the group.5 Subsequently, Bass (1957b), Hare (1953), and McKeachie (1954) found in various experiments that group discussions were more effective than reading or listening to arguments in increasing group members’ agreement with one another, both publicly and privately. Generally, such group discussion leads to improvement in the members’ understanding, commitment, and decision making. But the reverse is also true. Janis (1972) saw groupthink as the reason why highly competent members made extremely poor collective decisions, such as in the Bay of Pigs fiasco shortly after Fidel Castro seized power in Cuba. Highly cohesive, highly intelligent top officials of the Kennedy administration assumed a unanimous view when it did not exist. Conformity occurred out of a sense of loyalty to the group.
The signs of conformity in workers’ performance as a consequence of the work group’s power over its members were well known when pay-for-performance was commonly established by the number of pieces a worker produced in an hour or day. The fear of “rate busting” and that management would lower the price per piece paid to workers resulted in distributions of outputs of workers that peaked at an arbitrary limit set by the group. The variance in output was not a normal probability distribution; the distribution was truncated at the upper end.6
Subtle Influences. In a series of well-known experiments, Asch (1952) showed that conformity to the opinion of a simulated group can be induced even when the group decision defies the senses. A fair proportion of duped participants will agree that the clearly shorter of two lines is longer if they see all the other persons (all confederates) in the same situation stating that the shorter line is the longer one. Subtler influences of group power were measured by A. L. Hoffman (1956) and Horwitz (1954). Hoffman confronted participants with group norms that alternately agreed and disagreed with the participants’ responses. The psychogalvanic skin responses of participants suggested that tension was reduced when they believed they had conformed to the group instead of being in disagreement with others. Horwitz (1954) found that individual members were under the greatest tension according to their greater recall of interrupted tasks (Zeigarnik effect), when they voted not to continue a task that the group had voted to continue.
The effects of conformity persist even after the group disbands. When partners are separated after experiencing a judging situation together, they continue to influence each other’s judgments when asked to judge alone (E. Cohen, 1956). Bovard (1948) found that the tendency to be affected in this way persisted for at least 28 days.
Speaking to government employees, Robert Reich, the secretary of labor, declared:
To provide quality service is through fundamentally changing the relationship between management and labor, by pushing responsibility downward and giving people on the front line the power they need to do their jobs.… It’s not giving up authority. It’s joining in partnership to give everybody more authority (including me). Reich & Copening, 1994, pp. 35, 36.
Group Discussion. The use of discussion in groups is an approach to increasing power sharing in the decision-making process and to increasing the influence of the group decision on individual members. K. Lewin (1943) conducted a seminal experiment to demonstrate how group discussion influenced housewives’ choice of meats for their families’ meals. Group discussion will produce more change in members than will the arguments of individuals; it has been found superior to lecture techniques in reducing biases in the merit-rating tendencies of supervisors, prejudice, hostile attitudes, alcoholism, and emotional disturbances in children. Group discussion has also proved superior to lectures in reaching solutions to community problems and modifying food habits (Lichtenberg & Deutsch, 1954). In addition, Radke and Klisurich (1947) observed that new mothers who engaged in discussions under the leadership of a dietitian much more readily adopted desired behavioral patterns that coincided with the usual recommended procedures than did a control group who received individual instruction. Similarly, Levine and Butler (1952) significantly reduced the halo effect in merit ratings by foremen by permitting them to discuss and make decisions regarding more realistic evaluations. Torrance and staff (1955) found that B-29 aircrews were more likely to reach a state of efficiency that enabled them to go into combat if they had been observed to participate in decision making earlier in survival school.
Coch and French (1948), J. R. P. French (1950), and French and Zander (1949) reported the results of an experiment in a clothing manufacturing plant. The factory had experienced a considerable turnover of labor after each change in operating methods. The experimenters studied four groups of employees. For the control group, a change was merely announced. For the three experimental groups, the workers were given an opportunity to discuss the change, to offer suggestions, and to agree on the necessity for the change. Productivity increased and turnover of personnel decreased in the experimental groups but not in the control group.
The same plant studied by Coch and French and a second plant were studied in 1962, 1964, and 1969 by Seashore and Bowers (1963, 1970) and by Marrow, Bowers, and Seashore (1968). Managers and workers were given group training in joint problem solving. Although the employees’ satisfaction did not change to a marked degree, their commitment to the task improved, productivity norms were raised, and productivity increased. Ronken and Lawrence (1952) studied a firm that was experiencing difficulties in communication. When small groups and committees were formed to discuss and analyze the problem and to suggest solutions, communications became freer and operations were conducted more smoothly.
Lawrence and Smith (1955) compared groups of workers who set their own production goals with other groups who merely discussed production problems. Groups who set their own goals showed significantly greater increases in productivity. Lawler and Hackman (1969) reported experimental results using group power to develop a pay-incentive plan for maintenance crews. Three groups developed their own pay-incentive plans to reward good attendance on the job. The plans were then imposed on two other groups. One type of control group was given a lecture about job attendance and the other received no experimental treatment. Only those in the groups who developed their own plans increased their attendance.
Sharma (1955) surveyed 568 teachers in 20 school systems and found that their satisfaction was related to the extent to which they reported that they were involved in decision making as individuals or as groups.
Zander and Gyr (1955) examined employees’ attitudes toward a merit rating system. They found that a significant change in attitudes required monthly feedback, and that consultation and discussion were somewhat more effective than mere explanations of the system. In addition, under both conditions, changes in attitudes occurred only when subordinates described the supervisor as sincere, when the supervisor knew the plan and the issues, when the supervisor’s and the subordinates’ opinions were in agreement, and when the supervisor was skilled as a chairperson.7
The more power the work group has over its members, the fewer differences will be seen in the members’ output. Thus Seashore (1954) found less variance in productivity among different members of the more cohesive of 228 factory groups. The more members expect to be rewarded and to avoid punishment through membership in the group, the more they value the group and—as in Asch’s experiments described above—the more the group will have the power to produce conformity (Deutsch & Gerard, 1954). Gerard (1954) noted that the greater the attraction of the members to a group, the more their opinions coalesced. Gorden (1952) observed more conformity in members who were identified most with the group. Similarly, Newcomb (1943) observed that the extent to which Bennington College students wanted to “belong” on campus determined the extent to which they shifted their social attitudes in the direction of the liberal attitudes prevailing on campus. Rasmussen and Zander (1954) found that teachers’ levels of aspiration conformed more to the ideal of their group the more they were attracted to the group—that is, the more they saw the group as potentially rewarding.
Visibility and Salience of the Group. Groups will exert more power over their members when the members’ behavior is visible than when it is unobservable (Deutsch & Gerard, 1954). The effects of conformity can be increased by giving the group more power and by making it more important to its members. For example, Dickinson (1937) observed that the effectiveness of a group’s wage incentives depended on mutual policing. Members who work for group rewards are more likely to keep up with each other in groups that are sufficiently small to permit effective mutual observation. Grossack (1954b) found that members of such groups, rewarded as groups, demanded more uniformity from each other than did individuals who were competing with others for rewards.
A significant factor in a group’s power over its members is whether the members are clear about the modal or majority opinion of the other members. The clearer the members are about what they must conform to, the more likely they are to conform. In judging intelligence from photographs, participants tended to shift their judgments toward the group decision if they obtained knowledge of it (S. C. Goldberg, 1954). Gorden (1952) observed that when expressing their public opinion, members of a cooperative living project tended to be influenced by what they perceived to be the group’s opinion. Bennett (1955) found more compliance with a request for volunteers among those who perceived most others complying. Pennington, Haravey, and Bass (1958) obtained the greatest objective increase in agreement among members in groups in which members made their initial opinions public in discussions or when a group decision was announced rather than kept secret; agreement was less when either group decision or group discussion was absent, and it was least when both were absent. Similarly, in studies of autokinetic judgment, N. Walter (undated) found that agreement among participants decreased when they were no longer given information that was ostensibly based on the judgments of typical students at a prestigious school.
As was noted earlier, differences in power are established in formal organizational structures. Some positions and roles are assigned more power than others, and power increases as people move from the bottom to the top of the organization. Management-worker hierarchies and political bureaucracies replace traditional power structures based on family, class, tribe, wealth, strength, and age. Even in a country with long-standing democratic traditions, preferences for formal power structures vary. Some citizens prefer highly structured settings, and others prefer highly unstructured ones.
In some situations, everyone wants total structuring. Automobile drivers on through streets want assurance that drivers coming from side streets will halt at stop signs and red lights, and that oncoming drivers will stay on their side of the white dividing lines. Conversely, the hostess at a social gathering may be faulted for trying to structure interactions. In work and educational settings, there is much divergence of opinion about structure. Thus Heron (1942) observed that if employees do not have access to information about rules and procedures and must learn by trial and error, they are dissatisfied and afraid to take the initiative because of possible infractions and penalties. Seeman (1950) found that a majority of schoolteachers preferred a group-oriented style of leadership yet also exhibited a substantial preference for a type of leadership that called for the differentiation of power.
Bradshaw (1970) obtained ratings of the ideal as well as the actual behavior of leaders in a professional organization. The rank order of the means for both observed and desired behavior was as follows: (1) structuring of expectations, (2) tolerance of freedom, (3) consideration, and (4) emphasis on production. The observed tolerance of freedom was close to ideal, but more structure was desired than was provided by supervisors. More autonomy and self-actualization were desired than realized. Bradshaw concluded that deficiencies in structure were associated with failure to satisfy higher-order needs.
Wispe and Lloyd (1955) found that low producers preferred a more structured group whereas high producers favored a less structured group. The high producers perceived their superiors as less threatening than did low producers. Lowin and Craig (1968) studied supervisors in experimental groups and found that supervisors advocated more structure and closer supervision of subordinates who were incompetent. This was consistent with what has been noted earlier about the effect of subordinates’ incompetence on their supervisor’s punitiveness.
If any sizable group is to reach a common objective, some degree of structure in its internal role relations is required. Individuals need to be able to predict each other’s behavior. Unless informal means are possible, the reliability of individual performances will be obtained only by distributing authority, power, and responsibility to holders of the various positions in the group through some structure. Otherwise, the group is likely to remain highly ineffective. Lucas (1965), who studied the effects of feedback on problem solving in groups, found that a group’s effectiveness was related to the degree of its structure. Gerard (1957) observed that leaders tend to perform more effectively when they have a wider scope of freedom but that followers are more effective in a somewhat more highly structured situation. Gross, Martin, and Darley (1953) found that groups with strong formal leaders were more productive and more cohesive than groups with weak informal leaders. Sexton (1967) studied 170 line workers in jobs whose structure varied widely, according to both the supervisors and the workers, and reported that the degree of job structure was positively and significantly related to the employees’ satisfaction of needs. Hickson, Pugh, and Pheysey (1969) studied a variety of organizations in En gland and found that the structuring of activities, role specialization, and functional specialization were positively and significantly related to the size and productivity of the groups. Similarly, in a study of manufacturing firms, Pheysey, Payne, and Pugh (1971) noted that formality and an orientation toward rules were related to the employees’ satisfaction with their promotions and with fellow workers, and to greater involvement by the managers with the group. Bass and Valenzi (1974) also obtained a positive association between the perceived effectiveness of work groups and the extent to which relations between managers and subordinates in such groups were structured.
Hobhouse, Wheeler, and Ginsberg (1930) rated the level and complexity of development of some 200 societies. They found that efficient government, which resulted from a strong and stable leadership position or council of leaders, was positively correlated with high cultural achievement. Two opposing factors were found to re-duce the power of the leader. In less well-developed societies, the frequent lack of a permanent government deprived the leaders of any effective structure for exercising power except in sporadic undertakings. In more highly developed societies, the presence of a stable governing council limited the powers of the primary leader.
Fox, Lorge, et al. (1953) compared the effectiveness of six-to-eight-man discussions by U.S. Air Force officers with groups of 12 or 13. Although smaller groups ordinarily would be more effective, here the larger groups yielded higher-quality decisions. The larger groups overcame their disadvantage by organizing themselves to solve the problem so as to make maximum use of the larger number of participants. F. L. Bates (1953) observed that the performance of four medium-size bomber wings was better when there was a greater use of authority and a greater frequency of production plans, orders, and instructions sanctioned by authority. Again, if additional members are not redundant in contributing to a group’s decision and if efficient computerization can be arranged to process, establish priorities, distribute, and pool their individual contributions, groups of 35 with computers may be able to become as effective in utilizing the informational resources of their individual members as can groups of six lacking computerization and information-processing technology (see Chapter 29).
Bureaucracy and the Distribution of Power. The theory of bureaucracy is concerned with problems involved in the structure of power relations in formal organizations. Weber (1924/1947) observed that bureaucracy is characterized by the continuous organization of official functions bound by rules, specified spheres of competence, the hierarchical ordering of authority relations, and impersonality. Responsibility and authority reside in the office, rather than in the person. Bureaucratic structure does not necessarily limit all autonomy for the organization’s managers, although as expected, Engel (1970) found that managers’ autonomy was greater in organizations with moderate bureaucratic structures than in those with extreme ones. Numerous case studies8 have suggested that when organizations become overly concerned with rules and formalities, they tend to lose touch with the external demands made on them and to become insensitive to the internal problems that they generate. (These defects are not confined only to bureaucracies.) Dissatisfaction and ineffectiveness are unintended effects. Although it is the rational purpose of a bureaucracy to use specialized talent effectively, bureaucratic organizations and institutions often emphasize formalisms and legalisms to such a degree that the perpetuation of the system takes precedence over whatever function it was intended to perform. In agreement, Burns (1978, pp. 295–296) declared:
Bureaucracy is the world of explicitly formulated goals, rules, procedures, … specialization, and expertise, with the roles of individuals minutely specified and differentiated … organized by purpose, process, clientele, or place. [It] prizes consistency, predictability, stability, and efficiency … more than creativity and principle. Roles and duties are prescribed less by superiors … than by tradition, formal examinations, and technical qualifications. Careers and job security are protected by tenure, pensions, union rules, professional standards, and appeal procedures. The structure … approaches the … elimination of personalized relationships and … reciprocity, response to wants, needs, and values. …
Through its methodical allotment of tasks, its mediating and harmonizing and “adjustment” procedures, its stress on organizational ethos, goals, and authority, bureaucracy assumes consensus, discounts and discredits clash and controversy, which are seen as threats to organizational stability. Bureaucracy discourages the kind of power that is generated by the tapping of motivational bases among employees and the marshaling of personal—as opposed to organizational—resources. Bureaucracy pursues goals that may as easily become separated from a hierarchy of original purposes and values as from human needs. And bureaucracy, far from directing social change or serving as a factor in historical causation, consciously or not helps buttress the status quo.
Thus the status and power differentiation of a bureaucracy, initially organized to meet objectives expeditiously and efficiently, often experiences hardening of the arteries when it fails to adapt to changes in outside conditions and its own personnel.
But a complete lack of power differentiation generates problems that are equally damaging to the group’s and organization’s performance. To illustrate, Kingsley (1967) described a group of therapists that fell apart and disbanded after the collapse of its structured power relations when the discussion leader left. Members were unable to maintain interactive relationships in the absence of the power structure provided by the leader. Similarly, over a period of one year, L. M. Smith (1967) observed an experimental elementary school whose administration, scheduling, curriculum, teaching, and discipline were highly unstructured and in which authority resided in the pupils. The teachers engaged in a ceaseless struggle for power, and there were continuous confusion, delay, noise, and frustration. All but two of the teachers decided to leave at the end of the year. J. W. Gardner (1998, p. 13) summed up the utility to organizations of sharing power to maintain their vitality. There needs to be
the willingness of many people scattered throughout the organization to take the initiative [in] identifying problems and solving them. Without that, the organization becomes another of those sodden, inert, non-adaptive bureaucracies that are the bane of corporate and governmental life—rigid, unimaginative, and totally unequipped to deal with a swiftly changing environment.
The past century saw the rise and acceleration of a movement to change the distribution of power by delegating decision making to lower levels of organizational managers and employees closer to the need for action (Kanter, 1977, 1993). Power sharing took on a life of its own as empowerment at all organizational levels became a popular strategy in the 1980s (Block, 1986; Burke, 1986). Empowerment gave organizational members the responsibility and authority to carry out their jobs so that psychologically they felt they “owned” their jobs. It aligned their personal mission to the organization as a whole (Potter, 1994). It was adopted by service industry firms such as American Airlines, Marriot, American Express, and Federal Express. It meant encouraging front-line employees to exercise initiative and imagination. It also implied giving overall direction with considerable autonomy to employees, and latitude to carry out their assignments. It could be limited to deciding on the best way to complete an assigned task. It could involve identifying and solving problems and dealing with larger organizational issues (Ford & Fottler, 1995). It also came to mean empowering individuals at all levels in the organization, although in a study of hospital personnel perceived empowerment was higher among those higher in rank and tenure (Koberg, Boss, Senjem, et al., 1999). “Organizations must move away from policies that encourage employees to ‘leave their brains at the door’ to those that nurture, develop, and more directly reward employees’ intellectual capital” (Avolio, 1997, p. 2). But Wall, Cordery, and Clegg (2002) opined that empowerment was not a universal panacea—its effectiveness was contingent on the extent of operational uncertainty. In discussing new product development projects, Forrester (2002) agreed that management should yield power to the development team when uncertainty is highest, as when a radical innovation is being sought. When there is less uncertainty about the product—as when a change is to be just an incremental improvement—empowering the team is less important. Also, Shamir and Howell (1999) suggested that followers can empower leaders by cooperating and showing respect and admiration. They can also provide the leader with resources.
Empowerment is similar to but not the same as delegation. It is a more exacting kind of involvement. It implies “the freedom and ability to make decisions and commitments, not just to suggest them or [only] be a part of making them” (Forrester, 2000, p. 67). Organizational structures are changed. In many cases, middle managers are eliminated. Their decisions are now made by the empowered first-line supervisor and shared with the workers or work team. For Lawler (1986), this was an important aspect of high-involvement management. Team members were granted more autonomy, more self-direction, and more control over how they accomplished their tasks. As a consequence, they had more positive attitudes about their responsibilities and more positive feelings about their work. For instance, at Ericsson General Electric, suggestions from gainsharing programs were reviewed by employee teams, not managers. These teams could accept suggestions, put them into action, and authorize limited expenditures (Filipczak, 1993). Conger and Kanungo (1988) linked a sense of empowerment to a sense of self-efficacy. When leaders were supportive and empowered by higher authority, Parker and Price (1994) found that subordinates were more likely to perceive that they had control over decisions affecting themselves. Keller and Dansereau (1995) reported that for 92 superior-subordinate dyads, empowerment gave subordinates more control and a perception of more fairness in their superiors. Among 194 employees in a human service organization, self-efficacy and intrinsic motivation were enhanced by fostering their autonomy (Seltzer & Miller, 1990).
Empowerment can be structural or psychological. Structural empowerment accrues from occupying a position that is visible and central to the organization’s goals and allows the occupant flexibility (Kanter, 1993). An employee or manager is also empowered by structures that give access to information, support, access to resources, and opportunities to learn and develop. Psychological empowerment is intrinsically task-motivating by providing meaning to one’s role in the organization, a sense of competence, a sense of self-determination, and awareness of the impact of what one is doing (Thomas & Velthouse, 1990). Spreitzer (1995) extracted and validated as dimensions of psychological empowerment: meaning (for example, “The work I do is very important to me”); competence (“I am confident about my ability to do my job”); self-determination (“I have significant autonomy in determining how I do my job”); and impact (“I have a great deal of control over what happens in my department”). Shea and Howell (1992) combined structural and psychological empowerment in their model, understanding of the relationship between actions and desired outcomes. However, although structural empowerment enhanced job satisfaction for 185 staff nurses, psychological empowerment had no effect on job satisfaction after structural empowerment was taken into account (Laschinger, Finegan, Shamian, et al., 2002).
Herrenkohl, Judson, Thomas, et al. (1999) validated three additional dimensions that discriminated between those with more and less empowerment, fairness in the recognition system, fairness in decision processes, and clarity of company goals for 28 work groups. Felt empowerment was higher among middle managers if there was little role ambiguity in their work unit and if the unit had access to information, strong support, and a participative climate. Leaders learned empowerment roles from models provided by higher-ups in their organization. Compared with 561 male counterparts, 351 women leaders in 25 organizations were likely to be more empowering, according to their 912 direct reports (Howard & Wellins, 1994). Empowered managers worked for a superior with a wide span of control. Empowering management recognizes that reducing organizational controls can increase the individual’s shift toward personal control and outcomes. Management arranges conditions to facilitate the shift. The empowered individuals are induced to engage not only in “role behaviors” but also in discretionary behaviors that go beyond their roles and beyond minimum job requirements. Empowered persons expect to have favorable effects on outcomes if they take action (Spreitzer, 1996).
Empowerment is seen to be of benefit to the empowering leaders, their followers, and their organization. The leaders are relieved of routine work and are more committed to their organization, according to a survey by Howard and Wellins (1994) of 32 senior managers, 29 middle managers, and 323 supervisors from 25 organizations. The 948 nonleadership employees reporting to these leaders felt more positive about the work and the organization and experienced less work-related stress. The organization also benefited from an improved customer focus and improved quality. Pfeffer (1995) suggested that the organization also benefits by gaining a comparative advantage over traditionally organized rivals, and in the greater commitment, satisfaction, and involvement of its employees and managers.
Moreover, the organization benefits from making it possible for every member to be a source of ideas, initiatives, and influence. Benefits increase over time. Empowered people feel that they have a greater impact on their work and are more satisfied with it (Thomas & Tyman, 1994). They feel that they have a greater impact on the organization (Brossoit, 2001). This also contributes to greater satisfaction with their work (Ashforth, 1989). For instance, among 612 technically skilled professionals and managerial hospital employees those who felt more empowered reported greater job satisfaction, work productivity, and intent to stay (Koberg, Boss, Senjem, et al., 1999).
Democratic organizational flexibility replaces hierarchical rigidity. The controlling manager becomes the enabling leader (Cook, 1994). Nonetheless, the expected effects of empowerment on effectiveness, work satisfaction, and job-related strain vary with how much each of the components of psychological empowerment—meaning, competence, self-determination, and impact—is involved (Spreitzer, Kizilos, & Nason, 1997).
Contingent Effects. In some studies, empowerment was found to only indirectly affect beneficial outcomes. Niehoff, Moorman, Blakely, et al. (2001) observed that empowerment indirectly increased the loyalty of 203 employees as a consequence of enriching their jobs. In 128 leader-employee exchange dyads, managerial trust contributed to empowerment, but only when the exchange relationship was good for both the management and the employees (Gomez & Rosen, 2001).
Effects of Operator Control of Machinery. In manufacturing operations, empowerment meant giving operators the power to control loading, unloading, and monitoring their machinery. They also took responsibility for maintaining their equipment and for programming changes and corrections that formerly had been the province of staff specialists. Operators developed into their own staff specialists. A survey of 584 machining manufacturing plants in 1987 showed that plants where the machine operators routinely wrote and edited their own programs were 30% more efficient than plants where the operators were not so empowered (Ralls, 1994).
Operator empowerment in a British electronics manufacturer resulted in a saving of 10 hours of production time per week in the operations of seven numerically controlled insertion machines. The operators reported increased intrinsic satisfaction with their work and less job pressure. The former staff specialists were freed to work on other projects (Jackson & Wall, 1991). Over time, machine reliability was increased as the operators developed more skill in preventing problems. Downtime was reduced by 20% and eventually by over 70% (Wall, Corbett, Clegg, et al., 1990).
Empowering industrial employees by involving them in management decisions—according to a survey of human resources managers at 30 U.S. mini steel mills—reduced scrap rates, reduced employee turnover, and reduced the number of labor hours to make one ton of steel (Arthur, 1994).
A British laundry and work wear business, in dealing with small problems, generally found improvements from empowerment; yet some workers did not want to be empowered—nor did some managers want to change their practices to accommodate empowering their workers (Seath & Clark, 1993). Forrester (2000) enumerated several reasons why empowerment might cause problems and might fail to improve performance and satisfaction:
1. Empowerment may be introduced precipitately without adequate preparation of the people who are to be empowered. At the lowest levels, work has been enriched—i.e., increased—by administrative duties, decision making, and coordination requirements, but the workers have been given no additional time and no training to carry out functions that were previously done full-time by better-paid persons. The hasty changes take away power from supervisors and managers, who now suddenly must supervise differently, also without adequate time to train and prepare.
2. While increasing intrinsic task motivation matters, other motivations need to be considered, such as individual differences in preferences regarding autonomy, structure, money, competitiveness, desire to help others, and being accountable for results.
3. Accountability and responsibility for results may be confused and distorted.
4. Managers and supervisors may not be willing or ready to accept their loss of control, their reduced sense of achievement, and their reduced recognition.
Managers may have a problem with being accountable without control. They may feel that their job security is threatened—as indeed it may be, with the flattening of the organization (Stewart & Manz, 1995).
Additionally, managers of technology may have been negatively indoctrinated about empowerment by their engineering education. Fewer than half of 100 engineering textbooks and equipment design books made any mention of the operators’ or workers’ roles in production. Where people were mentioned, they were always subordinate to technology (Ralls, 1994).
A survey of 282 MBA students revealed another possible reason for resistance to empowering subordinates. These students felt that the quality of work would be better if workers were supervised than if they were un-supervised. They thought the supervisor would be more personally involved than subordinates in the quality of the product (Pfeffer, Cialdini, Hanna, et al., 1998).
For successful empowerment of others, Burdett (1991) suggested that leaders need to manage the context, not the individuals. A common vision is required. Organization values have to be enunciated, and leaders have to facilitate opportunities for action. According to Payne, Cangemi, Fuqua, et al. (1997), an environment of support and decentralized decision making must be created, and certain values must be communicated: high quality, good service, and excellence. For Randolph (1995), the most important factors in successful empowerment were sharing information about the organization, promoting understanding about it, building trust through the sharing of sensitive information, and replacing hierarchy with teams. Paradoxically, forceful leadership must be used, and power must be used effectively to give it to others, with limits—letting go of controls but remaining ready to help when necessary (PriceWaterhouse Change Integration Team, 1996). Forrester (2000) suggested that empowerment needs to proceed top-down, cascading from higher to lower levels in the organization as each level becomes knowledgeable, familiar, comfortable, and successful with it. Those with power must have enough control to share it. There is a need to develop members at each lower level in knowledge, skills, and information as well as to provide more access to important people. At each level, initially, low-risk decisions should be delegated. Censoring mistakes should be avoided. Rather, mistakes should be corrected and treated as learning experiences. Roberts and Thorsheim (1987) suggested that in interaction with employees, leaders need to express their own doubts, concerns, and uncertainties; ask and listen; reflect feelings; identify common experiences; paraphrase and summarize; and acknowledge and use others’ ideas. Howard and Wellins (1994) noted in their large-scale multiechelon survey that the empowerment strategies used most frequently were reengineering, training, pushing down decision making to lower levels, reducing layers of management, and using high-quality problem-solving teams. Support was provided by satisfying, high-involvement leadership but with no special financial rewards.
Feelings of empowerment among 531 individuals in 53 organizations were positively related to the rated transformational behavior of their leaders (Masi, 1994). Among a randomized sample of 327 managers in Fortune 100 firms, Brossoit (2001) found that transformational leaders provided their empowered employees with a sense of meaning in their jobs, a perception of choice, a sense of progress toward goals, and the belief that they were influencing the system. However, while transformational leadership resulted in a sense of empowerment at 76 Israeli banks, according to their 932 employees, as expected, it also increased the sense of dependence on the managers (Kark, 2000). A converse effect has also been reported. If managers feel they are empowered, they are more likely to be transformational but not transactional, according to a survey of 314 diverse middle managers of Fortune 50 firms rated by themselves and their subordinates (Spreitzer & Janasz, 1994).
Cook (1994) suggested that empowerment suffers if senior management endorses it but does not set a positive example; fails to communicate the reasons for empowerment and how it will help the organization; fails to explain the leader’s role as more than getting rid of work; and fails to provide training and a support network. According to Howard and Wellins’s (1994) survey of 948 employees, 323 supervisors, and 61 higher-ups in 39 business units of 25 organizations, some leaders might discourage empowering their subordinates by failing to get higher management to act on subordinates’ suggestions or by failing to praise success publicly. Leaders often failed to match empowering their direct reports with the readiness of the direct reports to handle more responsibility. The leaders couldn’t let go. Faulty implementation of empowerment could be also due to other reasons, such as failure to provide sufficient direction and training to accompany the empowerment. Chakravarthy and Gargiulo (1998) noted that in the restructuring of a hierarchy, trust and the legitimacy of leadership may be lost. Dover (1999) mentioned two other reasons for failure: being impatient, and expecting that giving power would provide self-reliance, competence, and commitment. Empowerment might be the objective, but its opposite might appear, as learned helplessness (Campbell & Martinko, 1998) or as bogus empowerment (Ciulla, 1998), in which subordinates feel they are free to make decisions as long as the decisions are ones their boss wants; or when subordinates are given responsibility they did not expect nor want nor request (Stayer, 1990).
Burns (1996) argued that empowerment comes about from collective leadership, which occurs in a “web of relationships” in which different people take on roles of influence as needed. Some initiate action; some join with others to take action. Different reactions occur: some follow the initiator; some oppose or ignore the initiative. Mutual empowerment is seen when team members influence one another both positively and negatively. An opponent may consider alternatives to an initiative.
Political scientists view empowerment of minority voters as evidenced by increased political participation and success in elections (Bobo & Gilliam, 1990). For educational administration, empowerment is pushing decision making from the district office to the local school, and from the board and superintendent to the principals, teachers, parents, and students (Glickman, 1990).
One of the best examples of empowerment—long before the term came into vogue—is what happened during the Normandy landings on D-Day, June 6, 1944. The U.S. Army suffered more casualties on that one day than in any 12 months in Vietnam. Omaha Beach turned into a killing zone.
But what happened in the face of this was truly a decentralized empowered event. … For the most part, the beach assault was an unorganized mass of intermixed units with little or no functioning chain of command. … Individuals moved forward, took initiative, took risks, and displayed awesome courage. Ad hoc teams formed on the beaches and … inland where paratroopers had been scattered over the landscape miles away from their planned drop zones. Ulmer, 1994, p. 8.
Since the settling of America, groups, movements, and organizations have arisen—and continue to arise—who are self-empowered or empowered by local, state, and federal governments in communities to organize action to meet needs and satisfy common purposes not adequately addressed by other formal institutions (Bachrach & Botwinnick, 1992). Critics viewed them as bureaucratic self-defense, co-optation of government agencies, and surrogates for public policy (Couto, 1992). But much citizen action now depends on the empowerment of such voluntary assemblages, ranging from neighborhood to environmental organizations.
Two of the more prominent questionnaires about empowerment serve different purposes. The Leader Empowering Behavior Questionnaire (LEBQ), developed by Konczak, Stelly, and Trusty (1996), contains 22 statements to be rated from 1 (strongly disagree) to 7 (strongly agree), which assess leader behaviors that empower followers. The Empowering Instrument (EI), developed by Chiles and Zorn (1995), assesses the extent to which followers feel empowered. Its six statements are rated using the same seven-point scale of agreement as the LEBQ.
In the Howard and Wellins (1994) survey, modeling trust and facilitation were, on average, the most important leadership empowering behaviors (6.2 on a 7-point scale of importance). Many other behaviors were cited as of considerable importance (5.8); these included supporting, coaching, partnering, envisioning, inspiring, team building, and championing. Ensuring compliance (5.4) was the only transactional behavior seen as important. Other transactional behaviors—such as controlling the way work is done (3.9) and pointing out failures (3.1)—were seen as not important to empowerment. Heslin (1999), along with many others, called attention to the need for empowering leaders to raise the self-efficacy of subordinates through mentoring, coaching, stimulating, demonstrating, and providing satisfying outcomes. Empowering leaders, willing to relinquish control and authority, did so on the basis of reasoned planning and actions, according to a study of 502 external team leaders from 64 Australian organizations (Wu, Cordery, & Morrison, 2001). Stewart and Manz (1995) ascertained that supervisors with positive attitudes about empowerment, according to their subordinates, helped subordinates’ empowerment. Arad, Arnold, Rhoades, et al. (1995) obtained the best-fitting model of the behaviors of the empowered team leaders of 95 customer service representatives and 110 building products supplier employees. The five correlated factors were: (1) leading by example, (2) coaching, (3) practicing participative leadership, (4) informing, and (5) interacting frequently. A similar analysis by Konczak, Stelly, and Trusty (1996) based on 424 managers and their 1,309 direct reports yielded a best-fitting model of empowering with six factors: (1) delegation of authority, (2) accountability, (3) encouragement of self-directed decisions, (4) information sharing, (5) skill development, and (6) coaching for innovative performance. Ford and Fottier (1995) agreed that empowerment involved shared responsibility and sharing of information. For Manz and Sims (1990), empowering members of self-managed groups involved encouraging self-set goals, positive thinking, personal initiative, and self-directed problem solving. Leaders lead the members to be self-leaders.
Discrepancy between Leader and Follower Perceptions. Larmore and Ayman (1998) administered the LEBQ, the EI, and the 20 transformational leadership items from the Multifactor Leadership Questionnaire (MLQ) to 52 CEOs and their 217 direct reports. The CEOs rated their own empowering and leadership behaviors. The direct reports rated the empowering and transformational leadership behaviors of their own CEO. The direct reports also rated their own feelings of empowerment. Correlational and variant analyses indicated that although the CEOs as a whole did not overrate the extent to which they empowered their direct reports, results suggested considerable disagreement between particular CEOs and their teams of direct reports. When some CEOs thought they were empowering their direct reports, the direct reports felt otherwise. Nonetheless, a correlation of .56 was obtained between the individual direct reports’ ratings of their own CEO’s empowering behaviors (LEBQ) and ratings of their own felt empowerment (EI). Direct reports’ sense of empowerment correlated .49 with their ratings of their CEOs’ transformational leadership. The transformational leadership was found to have mediated the relationship, among the teams of direct reports, between their feeling of empowerment and their perceptions of the empowering behaviors of their CEOs. These findings were more a matter of team differences than individual differences. Keiser and Shen (2000) found that leaders believed they were empowering subordinates more than the subordinates felt they had been empowered. When 9,098 school principals and their 47,105 teachers were queried, the principals reported that they gave teachers more power than the teachers felt they had to evaluate teachers, hire new teachers, decide on expenditures, establish curricula, determine the content of in-service programs, and set disciplinary policy. Howard and Wellins (1994) reported similar discrepancies between leaders at various organizational levels and their direct reports.
Differences in Empowering, Laissez-Faire, and Delegating Leadership. Empowering, laissez-faire, and delegating leadership behaviors are similar but not the same. When the 10 laissez-faire items of the Multifactor Leadership Questionnaire (Bass & Avolio, 1991) were factored by themselves for diverse samples of leaders rated by their followers, two distinct factors emerged. Six of the items were, as intended, measures of laissez-faire leadership (e.g., “Likely to be absent when needed,” “Avoids making decisions,” and “Fails to follow up requests for assistance”). Two items on a second factor were closer in meaning to empowering leadership: “Avoids telling me how to do my job,” and “Resists expressing views on important issues” (Bass, 1994). Delegation is assignment by the leader to a subordinate of activities and responsibilities. Leaders pass on to others tasks or specific aspects of their own jobs. They delegate to free their time to do other things, to develop the subordinates, or for other reasons. The leaders need to provide necessary direction and power for subordinates to effectively carry out the assignments (Shackleton, 1990). Adequate and inadequate delegation discriminated between innovative and uninnovative groups of Dutch teachers in 10 schools, according to keyword analyses of interview statements about the groups (Geijsel, Sleegers, & van den Berg, 1999). The innovative capacities of the primary schools (Geijsel, van den Berg, & Sleegers, 1999) and secondary schools (van den Berg & Sleegers, 1996) were also calculated. For leaders in their schools to effectively delegate and to sort out “the failures that are the price of delegation,” clear statements of purpose and secure, strong leaders and followers were needed. Babbitt (1987) suggested that these are also needed in bureaucracies.
Power may be formally redistributed by assigning worker-representatives to management boards and committees. The redistribution may be voluntary on the part of management, as in the United States, or mandated by legislation, as in Germany. Although some attempts had been made earlier in the United States, and elsewhere, representative industrial democracy appears to have been initiated in Sweden about 1938. It arose from the deliberations of national representations of employees and workers in response to labor unrest that was more common to Sweden in the 1920s and 1930s than in most other countries. The occupying military authorities in Germany after World War II, interested in ensuring that resurgent German industry could not again be exploited for military purposes or for the subversion of democratic government, passed a law prescribing Mitbestimmung—the representation of workers on a firm’s supervisory board. The Betriebsverfassungsgesetz (Works Constitution Act) was also formally prescribed by law somewhat later, to provide for the participation of employees in a wide range of matters. To signal Yugoslavia’s break with Soviet state centralism in 1950, Marshal Tito introduced self-management, another early method to use legislation to guarantee employees representative participation in the management of production and administration (Stymne, 1986). In the United States, along with a century of trade unionism came efforts, especially in the 1920s, to establish representative democracy in industry. However, it has been within the past 30 years that legislated tax advantages have contributed to the growth of employee stock ownership plans (ESOPs), through which employees begin to share in the company’s ownership. By 1986, 7,000 U.S. firms had ESOPs (Meek, 1987). Also, representatives of employees and consumers have been elected to serve on the boards of directors of corporations. Finally, groups of employees, such as those of United Airlines, began to purchase control of their own firms, and employees began to elect the directors who are responsible for the firm’s management and policies. Thus, quite a few businesses became partly or fully employee-owned. But attempts to share profits or gains in productivity such as Scanlon Plans, ESOPs, suggestion schemes, and productivity bonuses may founder on conflicts of interest and power differences between management and labor (Collins, 1995).
Efforts to create power-sharing institutions may be only a semblance of worker participation that may not yield the same degree of commitment, motivation, or productivity. Workers may have sat on boards of directors in the former Yugoslavia, but in fact party members, financial experts, and technocrats held most of the power, initiative, and influence (Obradovic, 1975). In some countries, unionism has declined. In the United States, for example, it declined to 11% of employees in 2005 since its peak in the 1930s and appears to remain strong only in the public sector. In Europe, it has remained stronger and better aligned with government, often controlling major political parties.
Representative democracy at the workplace is different from empowerment of employees. In workplace empowerment, workers participate directly with their immediate supervisor in decisions that are of consequence to their own work and working situation. Industrial cooperatives, such as Mondragon in Spain, became community-and-employee-owned holding companies for subsidiary plants. In the United States, ESOPs gave employees voting privileges and sometimes control over management. In reaction to threatened plant closings, on occasion, employees purchased the ownership of their plants. Trade unions, such as Histadrut in Israel, became holding companies for large conglomerates of business and industrial establishments. But none of these formal developments necessarily means that a participative climate in superior-subordinate relations will develop as a consequence; nor do such developments necessarily produce a real equalization of power. For example, who owns the company may not be as important to a new employee entering the firm at a low level as to an older employee with an important position in the company.
Nonetheless, empowerment of employees can result in increased democratization of union-management relations. In the 1960s, Ed Dulworth, an engineer working for General Foods, discovered how much he could enhance a Gaines Pet Food plant by designing into it all the elements that subsequently became known as employee empowerment. Output rose 50% in four years, but the successful redistribution of power became a threat to General Foods’ top management and to managers in other plants, and Dulworth had to move on. After arriving in Scranton, Pennsylvania, he organized a labor-management council to spread the word about what he had learned. By 1983, 15 in-plant labor-management committees had been established in the area (Osborne, 1988).
One purpose of equalizing power is to increase the commitment of those who initially have little power, despite their desire for more. However, Styskal (1980) failed to find the expected increased commitment in three federally sponsored temporary educational organizations in New York City when subunits were given more decision-making authority over the curriculum, program, employees, budgets, and work assignments. Many other problems can be cited. Discussing efforts by U.S. corporations to introduce voluntary and obligatory participation, Kanter (1982b) noted such problems as the paternalistic imposition of democracy; the need that results of participation be visible; concern by both management and the union about losing power; the time required; the lack of relevant knowledge by many employees; the limiting of participation to more abstract decisions that are remote from the employees’ concerns at the workplace; tension between innovation and democracy; the constraints of previous decisions; the need for leadership; the resistance to change of various interest groups; tension between social, emotional, and business needs; and excessive expectations about gains that may accrue from participation. Management may resist sharing power, not only from a fear of losing its prerogatives but also because of a disinclination to learn new leadership styles involved in consensus building. Unions may not only fear a loss of power but may remain suspicious of the management’s motives. And employees may be satisfied with their current inefficiencies in getting the work done (Alexander, 1984).
Dixon (1984) proposed that some interrelated elements are necessary for any participatory program to succeed. First, an information system is required to provide feedback on the organization’s performance. Ownership of the information must reside with the group, feedback must be timely, and results must be visible to all employees. Second, employees at all levels need to be represented. Representation calls for groups of competent employees to identify and solve problems, to have the power to act on their decisions, and to agree on a common vision. Finally, the organization’s leadership needs to believe in the creativity and responsibility of the employees.
Selekman (1924) reported the results of a management sharing plan in a chemical firm. For management to gain a better understanding of the workers’ point of view, two boards were created. The board of operatives was composed of elected workers, and was responsible for working conditions, grievances, recreation, training, company housing, and the like. The board of management, which was composed of a representation of executives and stockholders, was responsible for production, wages, farm equipment, hours, and financial policies. Although the board of operatives was slow to accept responsibilities, it gradually participated in planning and decision making. Effects attributed to the experiment included a reduction in the turnover of personnel and greater cooperation to improve production. Given (1949), a company president, experimented with a “bottom-up management” plan, in which he delegated authority down the line to employees and encouraged the employees to use their initiative and to assume responsibility. Results were good. Another company president worked directly with employees to stimulate their responsibility and initiative at the work-place. His activities placed foremen under great conflict and stress, but he worked with them until they accepted his philosophy of participation. He concluded that his organization became more informal and his workers more responsible and involved (Richard, 1959). W. M. Blumenthal (1956) described the management of steel plants in Germany. Each plant was managed by a board of three men, one a union representative. The general trend was for the two management representatives to assume responsibility for production while the union representative assumed responsibility for personnel, wages, and grievances.
Joint Consultation. Wyndham and White (1952) found that workers in an Australian refinery tended to regard the elected representative to worker-management committees as an alternative to their foremen. The fore-men then regarded their positions as threatened, even though they were members of the joint consultation committees. H. Campbell (1953) concluded that for joint consultation to be effective, the foreman had to be involved in the process as an active participant.
A study by W. H. Scott (1952) of joint consultation in three British industrial firms also indicated that the procedure was subject to strains in different segments and levels of the organization. Although workers’ representatives were generally well accepted in their discussions with management, both the workers and their representatives reported that little was accomplished because only trivial issues were discussed. The workers felt that they were neither consulted nor kept informed by the representatives. The foremen reported that they were bypassed in communications between management and workers’ representatives. The workers and their representatives complained that the foremen failed to take action on many matters that could have been handled more effectively on the spot rather than being referred to joint consultation. Although the majority of workers’ representatives were also union stewards, the union members complained that they were not adequately represented. Scott concluded that joint consultation resulted in few advantages that could not be produced as well as by responsible leadership.
Jaques (1952) completed an intensive case study of a single British factory that formed works councils and work committees to participate in management. The workers tended to feel dissatisfied with the performance of their elected representatives. The representatives were under heavy stress because of their different, conflicting demands; because they had little authority for making decisions; and because of their increasing separation from the workers. Other, similar problems were also revealed, but the study threw greater light on details of the consultative procedure. Jaques’s findings suggested that the difficulties encountered resulted not so much from the unwillingness of either management or employees to make the system work, as from their confusion about their respective responsibilities. Management tended to feel that it should not trespass on the responsibilities of workers’ representatives on the works council, particularly in matters involving social relationships; hence, the managers tended to take passive roles. The workers’ representatives tended to regard operative problems and policies as the responsibility of management and adopted passive attitudes toward those matters. When the general manager of the factory accepted the idea that he was responsible for all aspects of organizational life and activity, and that giving orders is not necessarily autocratic or un-ethical, he was able to “assume more fully his leadership mantle.” In turn, his subordinates accepted theirs. With their responsibilities more clearly focused, they were able to initiate building a works council more broadly representative of the factory. Jaques’s study suggested that members cannot act effectively in their own interests or in behalf of the group unless the leader exercises the functions of his or her role and takes the initiative in clarifying the definitions of roles throughout the system.
Power Sharing Cost-Reduction Plans. The Scan-lon Plan is a means for sharing reductions in unit costs that are under the control of labor. Such reductions are determined first by establishing a normal labor cost for a factory. This labor cost may be arrived at by a joint worker-management agreement. Whenever labor costs fall below the norm, the workers receive from 50% to 100% of the amount. A second element of the Scanlon Plan is a suggestion system in which benefits that accrue from the reduction in labor costs are shared with all workers in the plant. Reports indicate that as many as 80% of the suggestions submitted in a Scanlon Plan plant are accepted, whereas plants with individual suggestion plans commonly accept only 25%. Furthermore, with the Scanlon Plan there is no reason for the workers to withhold suggestions that will increase productivity; the result is lower labor costs and consequently a bonus for everyone (Lesieur, 1958).
To determine the effects of the Scanlon Plan, Puckett (1958) surveyed 9 firms with 11 plants that made many different products, and whose workers’ skills ranged from manual labor to highly technical. In the 11 plants that were followed over a two-year period, the gain in productivity varied from 17% to 49%, with a mean gain of 23%. Bonuses averaged 17% of gross pay and were based on splitting the savings in labor costs, with 75% going to the workers and 25% to the company. Besides the direct benefits from increased production and higher profits, there were also better service to the customers, higher-quality products, an improved competitive position, and no decline in employment.
Many other power sharing cost-reduction practices now in evidence can be mentioned, some of which, such as suggestion systems, were first introduced by the Yale Manufacturing Company in 1892. The upward communication provided by quality circles and survey feedback systems are other examples.
Rosen, Klein, and Young (1986) completed a survey of 2,804 employee stock owners at 37 firms with ESOPs. A majority of the respondents reported that stock ownership increased their interest in their company’s financial performance, their personal identification with the company, and their desire to stay with the firm. However, the ESOP arrangement was seen as having less of an effect on their job satisfaction and their influence at the work-place or over company policy. Nor did it result in managers treating them more like equals.
French and Rosenstein (1984) conducted a survey in a prosperous plumbing installation and services firm that began an employee stock ownership plan in 1958. Its managerial employees held 76% of the shares, which have substantially increased in value. For 461 respondents, the amount of their equity in the firm correlated with their desire to be of influence in it. However, the extent to which they identified with the firm depended more on their status and perceived influence in the firm than on the amount of stock they held. Also the amount of their equity was unrelated to their job satisfaction, which correlated only with their perceived influence. But their interest in financial information about the firm did depend on their equity position and their white-collar status.
Legislation in the United States has been directed toward establishing fairness in collective bargaining of employees and management and in employment practices. Elsewhere, power sharing through the representation of employees on supervisory and directors’ boards was legislatively mandated. Countries have tended to concentrate on one or two types of power-sharing legislation (Walker, 1975). Figures 12.1 and 12.2 show the degree of legal support for individual workers and representative bodies in 12 European countries. For each country, the figures show the average right to participate in 16 specified decisions, ranging from such daily routine decisions as the assignment of tasks to workers, to such long-range decisions as major capital investments. Each of the decisions was rated from “no right” to “final say,” based on assessment of laws and collective bargaining agreements and on interviews with labor and legal experts, and employers about company policies. Then they were averaged (Industrial Democracy in Europe, International Research Group, 1981).
Figure 12.1 Workers’ Legal Rights in 1976
SOURCE: From “Industrial Democracy in Europe (IDE),” International Research Group, Industrial Democracy in Europe. Oxford: Clarendon Press, 1981.
Figure 12.2 Legal Rights of Representative Bodies of Workers in 1976
SOURCE: From “Industrial Democracy in Europe (IDE),” International Research Group, Industrial Democracy in Europe. Oxford: Clarendon Press, 1981.
Effectiveness and Acceptance. Evaluations of the operations and effectiveness of workers’ councils have been reported for the former West Germany (Hartmann, 1970; Wilpert, 1975), Norway (Thorsrud & Emery, 1970), Great Britain (K. Walker, 1974), the former Yugoslavia (Obradovic, 1970, 1975; Rus, 1970), Sweden (Link, 1971), and socialist countries (Sturmthal, 1961, 1964). Derber (1970) suggested that although effective power sharing implied in industrial democracy was considerable, it tends to be moderated by a number of factors. First, as managers become more professional, their professionalism imposes a barrier to involving employees in decision making. Second, the majority of employees are not strongly motivated to assume managerial responsibilities as long as their financial and other personal needs are satisfied. Third, when union officials and employee representatives become involved in management decision making, they stir up political and factional conflicts that weaken their influence with workers. Last, joint decision making involves too much of management’s time for operations to run efficiently.
A number of other studies pointed to differences among managers in their acceptance of industrial democracy and power sharing. In agreement with Derber, Norrgren (1981a) found that Swedish managers who were young, had little seniority but much formal education, and were high in the hierarchy had positive beliefs and intentions about participation. Managers who were involved in white-collar union activities had positive beliefs and intentions and evaluated participative activities highly. Managers from several Canadian companies expressed more willingness to share power with employees if the managers felt they had more job security and influence and if they regarded their employees as competent (Long, 1982). Bass and Shackleton (1979) called attention to the exclusion of the lower-and middle-management levels from the participatory process when representatives of workers meet with top management. These lower-level managers, often ignored and eliminated from the mechanisms of industrial democracy, disassociate themselves from the participatory process and undermine its potential effectiveness.
After a carefully planned observational study of worker-management council meetings in the former Yugoslavia, Obradovic (1975) concluded that initiatives came mainly from the engineers, financial experts, and members of the Communist Party, not from the rank-and-file employees serving as council representatives. The few attempts of ordinary workers to lead were unlikely to be successful. Consistent with this finding, Rubenowitz and Norrgren (1980) used survey questionnaires and interviews to study the effects of different forms of participation in 10 Swedish plants that varied in size from 40 to 830 employees. The actual participation of employees was of much more perceived consequence when formal agreements for power sharing included acceptance of a policy to promote decision making by employees in their daily work. Much less participation was perceived if it involved only formal participation by representatives of the employees. Rosenstein (1977) observed that just because a powerful trade union with a strong social mission is the entrepreneur and owner of industries, as is the case with the Israeli Histadrut, this does not necessarily mean that participation by employees will be encouraged. Actually, the various power centers of the Histadrut—the managers, the trade union officials, the ideologists, and the politicians—have had to be persuaded and educated about the nature of the joint management program that is in effect. As of 1977, Histadrut had established such joint management in 30 of its plants. Success depends on the continuing support of both top management and employees. Nevertheless, a 12-country longitudinal survey revealed extensive behavioral changes in organizational relationships that were derived from mandatory programs of industrial democracy (Industrial Democracy in Europe, International Research Group, 1979).
Selective Power Sharing. Bass and Rosenstein (1977) and Bass and Shackleton (1979) proposed that industrial democracy is most likely to work well for some issues, such as dealing with pay and benefits, career development, working conditions, and job security, but not for others, such as financial planning or marketing strategies. The 12-country consortium study (Industrial Democracy in Europe, International Research Group, 1981) revealed in detail the extent to which these arguments were valid. In the same way, a total of 103 top Scottish managers and shop stewards who were interviewed by Dickson (1980) tended to favor the direct participation of employees on safety committees and briefing groups, and in dealing with the job enrichment of groups. The managers were less enthusiastic about the employees’ participation through representatives on company boards, works councils, and plantwide committees. However, shop stewards felt a bit more positive than the managers about the value of such representative participation. The managers saw the direct involvement of employees as a way to enhance communication and the acceptance of decisions. The shop stewards thought that both direct and representative participation were a moral right and that both contributed to the employees’ morale if they were acknowledged by management.
A random sample of 7,832 employees from a total of 134 European firms in metal engineering, banking, and insurance were asked how much they were involved in the 16 decisions for which they had legal rights (refer back to Figures 12.1 and 12.2). Their legally mandated involvement was more important in determining power sharing about long-term decisions. For workers in the 134 firms, the laws and regulations accounted for 24% of the variance in their involvement in short-term decisions; 50%, in medium-range decisions; and 58% in long-term decisions. The legal right of unions and other representative bodies to participate collectively short-term in such decisions had a negative effect on the actual involvement of individual workers. For the involvement of workers in medium-and long-term decisions, the right of both workers and their representatives to participate had a positive effect. The degree of power sharing that could be explained by legal norms was much higher than what could be explained by the firm’s technology, its organizational structure, or market conditions. It was concluded that the actual participation of employees depends on “an intricate interrelation of internal management practices and externally promoted support systems based on formal laws or collective bargaining agreements” (Industrial Democracy in Europe, International Research Group, 1981, p. 292).
The success of introducing industrial democracy in different countries was mixed. Straus & Rosenstein (1970) found that most plans in India had died. Israel’s plant council and Norway’s work council did not survive. In En gland, shop floor bargaining between union stewards and workers replaced joint consultative committees. In France, important welfare and shop floor bargaining functions survived, although joint consultative activities were hampered by interunion rivalries and union-management conflicts. Industrial democracy was most successful in the former Yugoslavia, in Sweden, in the German iron and steel industry, and in some American Scanlon Plan companies. Industrial democracy gave labor a sense of having won something. It broadened the scope of bargaining and strengthened management to the extent that the leaders of unions and workers were co-opted to advocate common goals. It provided channels of communication and opportunities for training prospective new managers. In 2006, a central issue in the United States was how to share rising employee health insurance costs, which, unlike in many other industrial nations where health insurance is covered by single-payer government programs, are becoming such a burden that 16% of the population are uninsured.
Roethlisberger (1941) observed that any technological change by management may affect not only the physical but also the social location of an individual in the organization. The possibility of social dislocation within the organization constitutes a severe threat to many members. Individuals in leadership positions tend to feel more tolerant of change and stress than do those in follower positions. In addition to being more receptive, members in leadership positions usually have access to information that enables them to predict some consequences of change. The member in a follower position has less access to such information unless it is provided by the management. Rumor, imagination, and speculation often lead to grossly inaccurate evaluations of the effects of an announced change. Participation in planning change provides at least a minimum base of information on which a member may evaluate some possible effects of a given change on his or her work, status, and relationship to the organization. Workers will feel more receptive to change if they are involved in planning it. Reducing the power differences between superior and subordinate at the shop floor level in the work group is a main theme of behavioral theorists and practitioners’ interventions. “Every employee a manager” sums up M. S. Myers’s (1968) changeover of supervisor relations. Formerly, the powerful supervisor fully planned for and controlled the performance of the powerless operator, whose only responsibility was to carry out orders. Now, as much as possible, the operator becomes self-planning, self-directing, and self-controlling.
Legally mandated democracy is not a substitute for participation in decisions. The correlation between legally mandated democracy and participation in decisions that affect daily work was only .10 in the 12-country survey mentioned above (Industrial Democracy in Europe, International Research Group, 1981). The workers’ satisfaction with the firm was also unrelated to legally mandated democracy, which suggests that mandated democracy is not a substitute for participative management and democratic leadership. The Swedish Codetermination Act gave legal rights for employees to be consulted by their supervisors at the workplace. White-and blue-collar workers reported an increase in their own influence five years after the Swedish Codetermination Act was passed, but managers did not report increased power sharing with their bosses as a consequence of the act (Stymne, 1986).
Sharing the Power to Plan Change. Shifting the distribution of power results in a sharing of planning and control that has positive effects on the commitment, understanding, satisfaction, and productivity of workers. Structure still remains, but it is based not on differences in power but on a structural design to which the workers have contributed. The utility of such self-planning was reaffirmed in a series of experiments by Bass (1970a, 1977) and Bass and Leavitt (1963).
In a first simulation, Bass and Leavitt (1963) demonstrated the importance of self-planning, both to productivity and to satisfaction. Several simple exercises were devised. In each, trios of managers in training developed a plan for themselves and then exchanged plans with another trio. A counterbalanced order was used so that half the teams executed their own plan first and half executed the other team’s plan first. Then they executed both plans. The result was that managers were more productive and more satisfied with operating their own plan.
In a similar simulation, 1,416 managers in training settings completed Exercise Organization (Bass, 1975h). As expected, the objective output and efficiency were significantly greater when they executed their own plans. Moreover, self-planning resulted in greater felt responsibility, job satisfaction, and satisfaction with the plan. However, the payoff was much greater when participants with practice, experience, and knowledge made their own plans than when those without the advantage of practice did so.
The importance of maintaining some special elements that the supervisor can bring to the planning process, such as support, experience, and knowledge, was illustrated by Schlacter (1969), who studied six highway maintenance crews over 25 weeks. In two crews, the fore-men assisted their supervisor in planning and reported results back to their crews. Two other crews assisted their superiors in planning. The third pair of crews assumed full responsibility for planning. The first two crews maintained stable job satisfaction scores and production records throughout the 25 weeks. The third pair of crews gained significantly in job satisfaction, but lost significantly in productivity. Similarly, Latham and Saari (1979) showed experimentally that supportive leadership resulted in higher goals being set by student participants than did nonsupportive leadership.
Rationale. Reasons given by participants for the utility of reducing differences in power for the purposes of planning included: (1) it increases the sense of accomplishment when carrying out the plan; (2) it generates more efforts to confirm the validity of the plan by executing it successfully and generates more confidence to see that it can be done; (3) it may increase commitment to see that the plan works well; (4) it may increase flexibility, room for modification, and initiative to make improvements; (5) it is likely to generate a greater understanding of the plan; (6) it may allow for better use of human resources; (7) it may reduce communication problems and the consequent errors and distortions in following instructions; and (8) it may avoid competitive feelings among planners and those who must execute the plans (Bass, 1970a).
Utility of Self-Managed Teams. Increasingly, project teams are being formed with informal leadership in which no one member is appointed as leader. Rather, power is fully shared in planning, operations, and control. Manz and Sims’s (1980, 1984, 1986) research investigated the utility of such self-managed teams (Cohen, Chang, & Ledford, 1997). More will be said about these teams in Chapter 26.
Copeland and McLaughlin (2000) called for redistributing power and authority in schools from the principals to staff and teachers, to meet shared responsibilities and shared purposes. “The capacity for leadership lies in the larger school community … originating from many peoples’ personal resources, and flowing through networks of roles” (p. 2). The authors reported on the success of the redistribution over a five-year period in 86 schools in the San Francisco Bay Area. The principals found increases in teacher leadership (90%), teacher voice in school decisions (69%), staff discussion of teaching and learning (95%), and parent voice in school decisions (70%). A survey of the teachers in 27 of the same schools disclosed some agreement that there had been increases in shared leadership and the development of leadership capacities.
In countries with large power differences, power is redistributed usually by a peaceful or violent coup d’état in which those in power are replaced. Where differences in power are smaller, redistribution is a matter of election (Hofstede, 1997). Redistribution of power in a country may also be due to changes in the rules and in political control. The power of women in Afghanistan increased with the fall of the Taliban and with women’s suffrage. In Kuwait, women were given the right to vote in 2005, and in the 2006 election for legislators, about 78% did so. Expectations were that corruption would be reduced and economic development and women’s rights would improve as a consequence (New York Times, June 30, 2006).
House (1988) suggested that increased power sharing may come about through major shifts in environmental or technological demands that require the lodging of greater power with lower-status but expert employees. Power sharing may also come about through the replacement of managers and employees; by a deliberate redesign of the organization’s structures and systems, so that critical information is broadly disseminated throughout the organization; by creating multiple centers of control; and by decoupling parts of the organization with different interests and unconnected requirements. However, the voluntary redistribution of power is likely to come about slowly and only after prolonged debate and conflict. Thus Dalton, Barnes, and Zaleznik (1968) intervened to shift power downward from managers to scientists in departments of a research and development organization. The scientists at the different levels gained in functional power and favored acceleration in the downward shift. But at the same time, the managers at various levels experienced a reduction of their hierarchical power and consequently favored slowing the downward distribution of power. The privileged, dominant coalitions in an organization will be reluctant to surrender power. They may shift technologies or move the organization into different environments to maintain their own competence-enhancing power.
By the 1990s, the downward shift in power had become commonplace to meet the organizational needs for flexibility and innovation. Structural changes in hierarchical relations had shifted to more collaborative relations across divisions and organizations. More channels had developed for influencing and taking actions. Peer networks were becoming more important than chains of command. Higher-ups increasingly had to face redistributed power that required them to use new ways of motivating their subordinates (Kanter, 1989).
How power is distributed in groups, organizations, and societies determines relations between leaders and followers. Power distribution depends on the structure of role relations. The human relations point of view about the distribution of power calls for its equalization among members. Structuralists favor strengthening it. Personal and situational contingencies need to be taken into account in determining whether differences in power and structure need to be reduced or increased to promote effective operations and relationships. Different distributions of power are likely to exist side by side, and deal with the legitimacy, politics, expertise, and identification processes. The total power of a group, which can be increased or decreased, strongly affects the conformity of members, their persistence, and their changes in attitudes and behavior. Research about the power of the group to induce conformity among its members peaked between 1950 and 1970. Structure is necessary under conditions of stress, threat, and the pressure of tasks. Group and organizational members are more satisfied under moderate degrees of structure than under overly structured or totally unstructured conditions. Sharing power under normal conditions has many benefits. Opportunities to consider a proposed change facilitate the acceptance of the change by members who will be affected by it. Participation in planning and decision making regarding the change induces a still higher degree of acceptance. As a consequence, power-sharing programs and restructuring, both voluntary and legally mandated, have become common. Their acceptance depends on the nature of the problems and decisions to be reached, as well as on whether the power sharing is voluntary or legally mandated, direct or through representatives. The practice of industrial democracy has been increasingly institutionalized by legislation in western Europe since World War II. Industrial democracy in the United States has been more likely to take the form of voluntary power sharing and collective bargaining. Increases in legislation to promote the fairness of collective bargaining and to equalize the power of employers and employees came in the 1930s. But how power is distributed can be a major source of conflict.