ABSTRACT
Organizational structure variables were first delineated in Weber’s (1946) theory of bureaucracy. They include the organization’s size, levels of hierarchy, centralization or decentralization, span of control, and division of labor. In earlier times (Porter & Lawler, 1965) there was no clear association between an organization’s functionality and its levels of hierarchy or centralization. Rather, the optimal number of hierarchical levels and degree of centralization was relative to what the organization was trying to accomplish. Inmore recent times, computer-based technologies have facilitated a drop in the necessary levels of hierarchy. The new forms of technological support could condense middle management jobs that involved collating information and passing it onward and upward. The combination of new means for speeding communication and the demand for greater speed has resulted in initiatives for reducing communication bottlenecks. The hierarchical aspect of organizational life is expanded further in Chapter 10. Larger organizations tend to have larger subunits and greater dysfunction in
terms of lower job satisfaction and higher rates of absenteeism, turnover, and accidents (Guastello, 1988; Porter & Lawler, 1965). Group productivity rates tend to follow the general rule of optimal size, however. The dysfunction associated with larger organizations or subunits may be related to poor coordination (Comer, 1995; this topic is expanded in Chapter 8) or to the division of labor being cut too narrowly. Historically, there have been two groups of responses to the division of labor
problem. One is sociotechnical systems theory, which originated with Trist and Bamforth (1951), and the other is job characteristics theory, which originated with Hackman and Oldham (1976). The central idea behind sociotechnical systems theory is that work should be rationalized from the viewpoint of the people who do the work, rather than from the viewpoint of the work itself. Under this general theme, the best way to divide work among people is to put a group of people who have the necessary capabilities together with the work, and then let the natural process of self-organization take over (DeGreene, 1991). The optimal arrangement of people and tasks will ensue. DeGreene also noted that in a changing work environment, the self-organizing properties of the sociotechnical systems approach offer the flexibility that a group needs to make adaptive responses. He also found, however, that the capability to be flexible is not alwaysapparent in the upper levels of the organizational hierarchy. A later product of sociotechnical systems thinking is the autonomous work
group. In those situations, work that was once organized as a fixed sequence of specialized processes, as in an assembly line, is now given over to a group. The
group establishes its own sequence of people and tasks, including the supervision or managing functions that used to be allocated to a supervisor. Although some type of supervision may be involved, a level of the organizational hierarchy may be deleted in some cases. The focus of job characteristics theory is to improve the motivational content
of jobs by increasing levels of skill variety, scope of the individual assignments (broad instead of narrow), importance of an individual’s task to the work of others, autonomy, and feedback from the job itself (Hackman&Oldham, 1976). Although there was a precedent, dating back to Herzberg, Mausner, and Snyderman (1959) to expect positive outcomes for work performance and attitudes, the positive outcomes from job redesign did not alwaysmaterialize, according toHackman (1992). One poignant explanation for the lack of effective of job redesign was that organizations were not willing to wait through the transition period, during which time work output would drop while workers regrouped their tasks and acclimated to the new work flow. During the transition there was also an opportunity for workers to become discouraged, and those who naturally showed little tolerance for ambiguity were least likely to acclimate to the new work systems. Another source of frustration was located in the group level process by which the contributions of individuals were meant to connect to a whole outcome. Later research on autonomous work groups sought to connect the best ideas of
sociotechnical systems theory with job characteristics theory (Cohen, Ledford, & Spreitzer, 1996; Pearson, 1992). Sociotechnical systems theory addressed the group level process in a vague sort of way without clarifying the specific reasons for its success, and its success was not consistently observed by those who tried to do so (Cordery, Mueller, & Smith, 1991). Job characteristics theory needed some semblance of a group process. Research now shows that positive attitudes toward autonomous work groups’
quality of work life does depend on the attentiveness to job characteristics; the group task design is preferred to traditional divisions of labor (Cohen et al., 1996; Cordery et al., 1991). On the other hand, Pearson (1992) reported improvements in job satisfaction, but not for role ambiguity or role conflict (stress-related variables) over time for autonomous groups. Pearson did report increases in role ambiguity, conflict, and accidents, and a decline in motivation among the nonautonomous group counterparts in that study. Cohen et al. (1996) found that ratings of productivity improved as a function
of groups’ ratings of their quality of work life and the use of group task designs. Management’s ratings of group productivity were negatively correlated with the extent to which a group encouraged the use of supervisory behaviors such as self-evaluation, self-criticism, goal setting, and encouraging rehearsals of group activities before executing them. Pearson (1992) found, in a more straightforward result, that newly implemented autonomous work groups outperformed conventional work groups after four weeks of implementation.
