Human capital theory, particularly as it is formulated by Mincer (1974) posits that individuals play a very active part in defining their own earnings. The differentiation of employees' earnings is due to their differential decisions to invest in themselves (the training they choose to acquire). According to this formulation, higher ability individuals often sacrifice the possibility of receiving higher initial earnings in order to acquire more training. These investments in training lead to even greater subsequent advantages in the human capital they possess and thus to greater subsequent earnings. In other words, individuals' decisions about self-investments are the primary cause of the increased variation in earnings over time.
This theory is elegant in its abstraction and simplicity, and it is powerful in the range of its explanatory power. However, observers of firms have argued that such a theory is too simple, for it ignores the complexities of organizations (Lester 1946; Doeringer and Piore 1971). Moreover, empirical evidence on career attainment processes over time, which would be useful for testing this or competing theories, have generally been lacking.
The social structural theory presented here suggests another view. It posits that the organizational selection system assigns new employees to different entry jobs and that early job assignments have a strong impact on subsequent attainments. Social structural theories vary in the way they interpret these forces. Some versions indicate that status attainments have effects because of their symbolic meanings and their effects on interpersonal interaction (Meyer and Rowan 1977; Collins 1979); other versions suggest that their effects are due to administrative rules or practices regulating career attainment (Doeringer and Piore 1971; Spilerman 1977). Thurow (1975) comes closest to using the human capital conception to explain structural effects in positing that jobs determine human capital because they confer training, but, as a result, they also determine employees' later careers. Regardless of interpretation, social structural theories concur in contending that early social attainments have strong effects on individuals' later career trajectories.
Thus, instead of explaining differentiation in terms of assumed individual self-investments, the social structural theory contends that there is never a possibility of differentiation not occurring, regardless of what actions individuals take. Even if no individuals chose to make additional self-investments, individuals will still be put in different jobs from the outset; and these jobs offer different career trajectories regardless of individuals' investments (within reasonable limits). While of course, not every individual in the better jobs may follow the highest subsequent career trajectory, and the better jobs might not lead to higher trajectories for anyone if no one in the better jobs satisfies minimum job requirements. However, social structural human capital theory posits that individuals' decisions about self-investments determine their attainment; the social structural theory posits that jobs have the primary influence on individuals' attainments.
Social structural theory also differs from human capital theory in suggesting that the most important selections occur at the outset. Human capital theory posits that employees' initial earnings show little variation and most of the variation in earnings only gradually appears over the first decade of employees' careers. Mincer attributes this "fanning-out" process to individuals' differential decisions to invest in themselves, and he stresses that the process is gradual. "Investments are spread out over time because the marginal cost curve of producing them is upward sloping within each period" (1974, p. 14). Consequently, the fanning out of earnings is a process which evolves over time rather than being predetermined from the outset.
In contrast, according to social structural theory, the important selections among individuals are made at the outset of their careers, these rankings remain stable over the rest of their careers, and any subsequent fanning out is only a magnification of the initial selections. In effect, the two theories suggest different timetables for the selections of individuals' ultimate career attainments. While social structural theory suggests that individuals' ultimate attainments are largely determined by initial attainments, human capital theory attributes more openness and less predictability to the process.
Finally, the two theories also differ in the kind of attainment they are trying to explain. Human capital theory focuses only on earnings attainment, and it has virtually nothing to say about jobs or job statuses. Social structural theory focuses on social status or authority levels, although it also contends that these social attainments are primary determinants of earnings. It is commonly assumed that these different kinds of attainment are so highly correlated that the two theories essentially explain the same outcomes, regardless of their emphasis. As shall become clear, although earnings and authority level are highly correlated in the data gathered in this study (r = .82), there are some very important differences in their determinants.
As might be expected, the two theories do not differ in their predictions about the effects of the most salient individual attributes. Education and work experience are the key variables in all analyses of both theories, and the theories show consensus in their prediction about these; nor is this surprising, for these are salient in U.S. culture, and theory assumes that, on average, early jobs will have an effect on later careers regardless of individuals' actions.
In contrast, factors which are not explicit components of compensation policy are more interesting for empirical tests of the theories. Since the nature of their influence is not explicitly stated, their impact is more problematic, and the theories often have different interpretations of the reason for their impact and the kind of impact they will have. The present article analyzes two such factors which are not ordinarily matters of explicit compensation policy: age and college selectivity.
