Although the pyramid shape of hierarchies imposes limits on careers, the present analysis contends that it is not the only factor imposing limits, and it is not necessarily the most important constraint. Since jobs are continually being created and disbanded, vacancies are neither necessary nor sufficient for career advancements. Moreover, the finding that certain employee groups are so strongly and consistently preferred in good times and bad suggests that the career system might guarantee their advancement even if the required vacancies did not exist. Managers in this firm report that positions are sometimes created to advance the career of a valued employee. Granovetter (1974, p. 14) also reports that new positions are often created for new recruits. The interdependence of jobs and individuals is further suggested by Baron's (1982) finding that 47% of the jobs in a sample of 415 organizations had only one incumbent.
The central contention of the present inquiry is that careers are two-dimensional: Individuals' careers must be measured by their status attainments and by their timing. While the pyramid imperative may impose unavoidable constraints on maximum available status attainments, it says very little about which employees will be considered for those attainments, when individuals will be considered for advancements, or what their rate of advancement will be. To explain these features of careers in organizations, it has been posited here that organizations have "career systems" which result from explicit policies and implicit norms about how careers should unfold. Many of the problems which are often attributed to the pyramid imperative are also affected by aspects of the organization career system.
For instance, the strong effects of early jobs in limiting career advancement are not dictated by the hierarchy. Indeed, these early job effects actually contradict hierarchical models which predict ahistorical effects. Moreover, these effects are so strong that they undermine the possibilities for Horatio Alger successes, in spite of strong norms to the contrary. Given the great latitude this career system allows for special treatment of elites, it is curious that Horatio Alger types are not also treated as elites. Whatever the reasons, this is not dictated by the pyramid shape of the hierarchy.
The "sudden-death" quality of the tournament and the sharp promotion declines with age lead to radical discontinuities in the opportunities for college-educated employees. The hierarchy does not impose this age distribution on promotions. Indeed, non-college graduates have a different age distribution of promotions than college graduates at the same levels, and college graduates have similar age distributions of promotions over different hierarchical levels. These patterns seem to be determined more by the way the career system treats employee groups than by hierarchical shape.
Young employees often say that they must sacrifice their family commitments to achieve career advancement. This is often based on the fears that the top positions in the hierarchy will become filled up. The analyses presented in this volume suggest that the hierarchy per se is less to blame than the rigid timetables imposed by the tournament system.
The strong preferences for certain elite educational groups are certainly not dictated by the shape of the hierarchy. When organization growth is curtailed and the hierarchy provides fewer vacancies, the elite are minimally affected.
While the pyramid imperative imposes some limits on ultimate attainments, it says very little about features of the career system. The career system imposes additional constraints on attainments beyond the constraints imposed by hierarchical shape. Even if the organizational hierarchy were broadened to increase advancement opportunity at certain levels, the career system would still define which employees received the increased opportunity. On the other hand, changing the career system, even without changing the hierarchy, could increase the availability of advancement opportunity. A reform of career systems which increased the number of groups considered for advancement and which delayed selections would prolong advancement opportunity for many employees, even if the pyramid shape remained unchanged.
The key to doing this is an understanding of the organization's career system. Although organization policies describe some aspects of career systems, the analyses presented in this study provide examples of how the operation of organization career systems could be better under-stood. Describing the career trajectories of various groups and the ways they change over time can reveal the underlying career system which governs employees' careers.
The tournament model provides a way of conceptualizing the issues raised by career systems, a way of inferring their preconditions and outcomes, and a way of inferring policy alternatives for better achieving the system's goals.
The Tournament As a Model of Entire Career Systems
Organization planners and other employees desperately need a model of career systems, for career systems are not easy to see or understand. Selections in organizations occur through the operation of many diverse career programs and practices: programs for recruitment, career planning, succession planning, out-placement, and so on. Each is somewhat vague about the ways it makes its selections, its implications for later careers, and its relationships to other programs. In addition, these programs tend to exaggerate the amount of opportunity they allow, and they give widespread attention to Horatio Alger stories far out of proportion to their frequency of occurrence. These stories not only serve the interests of supervisors seeking to motivate employees, they also "ring true" with what employees (and supervisors) have been told in school and in the media, and they provide strong distractors preventing individuals from seeing the career system in perspective. Consequently, it is not surprising that employees' perceptions of career systems are piecemeal and often incorrect.
However, even if career systems cannot be seen directly, they can be described in terms of their outcomes. In much the way that physicists describe magnetic fields by the patterns they make on iron filings, career systems can be described by the patterns they make on the job mobility of an entering cohort. Morever, just as physicists use models to conceptualize magnetic fields, the tournament model can be used to conceptualize career systems.
The tournament model describes a selection mechanism that pervades the many different career programs in organizations. It describes how they are interrelated, how they create a coherent system, and how this system accomplishes some of the organization's goals. As such, this model can help individual employees to understand career realities and to make plans and choices about their careers. It may be a particularly good way of portraying this information because the model not only shows present choices, but also shows the probable long-term consequences of these choices.
This model could also be useful to organizations. While personnel managers surely have more information about career systems than do other employees, some kinds of crucial information can only be revealed by systematic research on career patterns. More important than information is the question of how it is conceptualized; the conflicting social norms operating in this corporation and in many others make it difficult for corporate employees to have a coherent picture of what is happening.
It seems quite conceivable that organization members, even those at high levels, are ignorant or confused about these issues. Perception is difficult on these matters and social norms are likely to distort perception. Of course, there is no way of knowing what top executives actually believe since there is no way to force them to be candid. But Vroom's example of the high-level personnel officer who took steps to eliminate structural barriers after learning the results of Vroom and MacCrimmon's (1968) research tends to suggest that this was new information and it was considered cause for concern.
The findings of this study and the tournament model provide a way of conceiving of careers and career-related issues at a structural level. They also permit identification of the sources of career problems at a structural level, rather than attributing them only to individuals' deficiencies.
Although the tournament is a good reconciliation between the opportunity and efficiency norms, its success in reconciling this conflict is based on a particular set of assumptions about preconditions. If these preconditions do not exist, then the tournament will fail to satisfy one or the other aspect of this conflict. In the following sections, the preconditions for opportunity and efficiency in career systems are identified and are assessed in terms of the findings of this and other studies.
