In addition, as in a small community, these statuses can easily be communicated throughout the organization. Formal personnel records make the information very readily available, and the organization's size makes it easy to communicate shared understandings about how to interpret the status nuances connoted by particular jobs and job moves.
Organizational statuses also provide a better test of the extent that status might be associated with (and determined by) individual at-tributes. The finding of a modest relationship between individual at-tributes (e.g., education) and occupational statuses across many employers (Featherman and Hauser 1978), may occur either because different employers have different practices or because most employers have vague and diffuse practices. In contrast, within a single organization, the relationship will clearly indicate whether organizational selection practices are determined by these individual attributes.
Finally, these organizational status classifications provide a better test of the extent to which careers have a structure of their own. The effects of early statuses on later careers are more likely to operate with-in a small demarcated community in which these statuses have common social meanings and in which they are likely to be widely known and understood.
The consensus about the statuses associated with jobs, their extensive explicit and implicit implications for rewards, deference, and authority, and the small-community characteristics of organizations which make an individual's history of job statuses known and readily interpreted throughout the organization suggest that job statuses are likely to have particularly strong effects on individuals' career opportunities.
This study analyzes careers in terms of employees' earnings, status, and authority attainments. Surely there is more to careers than these materialistic and status outcomes. Individuals seek variety, autonomy, challenge, and interpersonal influence from all aspects of their lives, including their work (Argyris 1964; Hall 1976). In limiting its empirical analyses only to status, authority, and earnings attainments, does this study assume that all employees are "status seekers?" Not at all. Even though individuals seek many kinds of rewards from work besides status, authority, and earnings, organizations allocate these non-status rewards to jobs in ways that are related to job status. As this study will show, jobs are structured in organizations so that regardless of whether one wishes more variety, autonomy, challenge, or interpersonal influence in one's work, one can only get more of it by advancing in the status hierarchy.
Organizations sometimes make this explicit. Job evaluation systems actually formalize this relationship by using these non-status features of jobs as factors in the determination of job worth and job status. Jobs which offer greater variety, autonomy, challenge, or interpersonal influence are evaluated as having more value and conferring more status. To attain jobs with these non-status qualities, an employee must advance up the status hierarchy.
Although this study only measures career attainment in terms of status, earnings, and authority attainments, a wide range of non-status rewards are also at stake. This makes it all the more important to understand how status attainments are structured and allocated, for they are the basis by which individuals obtain many psychological rewards, including opportunities for personal growth at work.
Generalizability
However, it is not clear what further information would be useful for inferring generalizability. Even researchers who identify many features of the firms they study offer few speculations about generalizability implications (Wise 1975, Halaby 1978). Moreover, some of the general dimensions which are commonly used to specify the generalizability of organizational labor markets might not be very useful for that purpose. For instance, one of the main dimensions discussed in the theoretical literature, the distinction between "core" and "peripheral" industrial sectors, suggests predictions about personnel practices which are not supported by empirical research (Cohen and Pfeffer 1983; Jacobs 1983). Reviewing this literature, Baron (1984) concludes that many of the dimensions which have been used to describe organizational environments have been too coarse to explain differences in personnel practices. He proposes that research must focus on more specific dimensions. Ongoing work by Baron and Bielby, Bridges and Villemez, and Althauser and Kalleberg are designed to identify some pertinent dimensions which affect career systems, and may in the future be helpful in assessing the generalizability of these findings. But, at present, the dimensions on which generalizability depends have not yet been identified.
Certainly some of the general findings of this study match findings of other studies: the age-graded career patterns (Martin and Strauss 1959; Kanter 1977; Lawrence 1983, 1984), the job status effects on earnings (Malkiel and Malkiel 1973; Halaby 1978), and the effects of historical context on the careers of different cohorts (Grandjean 1981). Of course, for many of the specific findings, comparisons cannot be made because appropriate data have not previously been available. From everything that I know about ABCO, I think that it is reasonable to consider it typical of many other large, well-established corporations. Although the present findings seem unlikely to generalize to small firms, high technology firms (Kanter 1983), or other firms undergoing radical changes in their products or markets, these findings seem likely to generalize to many other well-established corporations. Surely 2Readers sometimes try to uncover the true identities of anonymous objects of study. Kanter has commented that many people have tried to identify Indsco, and I have heard several different companies identified as the corporation I studied. These rumors serve no useful purpose. If false, they mislead. If true, they only prove that researchers cannot promise anonymity, thus shutting off research possibilities for other researchers. If there is some kind of information about this firm that is needed for particular comparisons, I will try to provide it. In a few years, these data will become old enough to make it possible for me to identify the corporation.
this corporation was not alone in responding to economic growth, contraction, and affirmative action over the period studied, and, judging from reports about career policies, the dominant concerns motivating the career system in this corporation were also salient in other major corporations: Citibank, Uniroyal, AT&T, Macy's (Schaeffer 1972), General Electric (Miller 1979), and Sears (Wellbank et a J. 1978).
How does mobility in organizations compare with mobility in larger society? The extensive research on societal mobility since 1960 raises the question of generalizability with special force. That research de-scribes mobility in a larger context, and it would be desirable to place organizational mobility within this context. But the difficulty of making this comparison is that different kinds of social mobility are studied in the two domains. The status mobility studied in most societal surveys (changes in occupations) is not the most important or most common kind of status mobility within organizations. On the other hand, the main kinds of status mobility within organizations (job status and authority changes) are difficult to study in societal surveys. Although some work has begun to analyze changes in authority attainments (Robinson and Kelley 1979), it is difficult to get valid and comparable indicators of detailed authority and status categories across employing organizations. While this makes the findings of organization studies difficult to compare with survey findings, it also indicates a kind of mobility which survey research has largely ignored and which can best be described by studies like this one. The two kinds of research are comparable in studying earnings, but earnings is a lagged outcome which only becomes differentiated well after job status distinctions have been made. If this finding is generalizable, then the detailed study of the timing of career differentiations requires job status distinctions not easily available to survey researchers. Of course, this means that tests of the generalizability of these findings requires more longitudinal studies within organizations.
Survey research does address one aspect of generalizability raised in this study. The longitudinal analyses introduce an additional concern about generalizability: that the sample in the longitudinal analyses is unrepresentative in being more stably employed than most members of the labor force. Of course, this limitation is unavoidable. Careers in an organization can only be studied for employees who remain in the organization.
However, recent work suggests that this is not such a serious constraint on generalizability as one might have thought. Analyzing data from the Current Population Survey, Hall (1982) concludes that over a quarter of all workers in the U.S. labor force hold continuous employment with the same employers, who will last 20 years or more. Specifically, since the median entry age for male employees in this corporation was 25 years, what portion of the male labor force at this age would remain with the same employer for 13 years? Hall calculates that 27.0% of men in the 25-29 age range can expect 20 years of continuous employment with the same employer. Moreover, this percentage would be larger for a 13-year period and for employees of large corporations like the one studied here. Hall's analyses suggest that these longitudinal analyses, though restricting generalizability, still represent a substantial segment of the labor force.
