workers should not be ignored simply because they are underrepresented in currently available data sources.
We find another ambiguity in the question of how to regard productive activities that share characteristics with paid employment but fall outside the traditional definition of work. For example, individuals who are self-employed in small businesses or on farms are not covered by a number of worker protections, such as unemployment insurance and workers’ compensation benefits. Yet, as these self-employed workers age they will have some of the same health and safety needs as older workers who are hired as paid employees. Similarly, the unpaid caregiving and labor provided at home within families (typically by women) have not historically been recognized as work, even though the same activities are regarded as work when performed by employees of commercial services (Robinson, 1999). As discussed in Chapter 4, the fact that many women workers have added an unrecognized second shift to their paid employment throughout their lives has important implications for their income security, health, and ability to engage in paid employment in their older years.
Conceptions of retirement are changing. Traditional portrayals of working life have emphasized a standardized, normative pattern of three distinct life stages, through which members of an age cohort progress sequentially at roughly the same time: (1) a childhood education period that can be seen as preparation for work; (2) adulthood, involving a working career that may include some job changes but centers around one primary long-term job; and (3) retirement, a period of leisure without paid employment (Cain, 1964; Kohli, 1986; Best, 1980). While more a product of conventional wisdom than careful research, this simplistic three-stage model of working life has shaped many aspects of public policy (Myles, 1989; Myles and Street, 1995). For example, it is the foundation for age-related eligibility criteria for receiving pension benefits. This model also influences public perceptions as well, reinforcing the sense that a particular chronological age such as 65 is the appropriate point at which all workers should expect to retire (Marshall, 1995).
Research on transitions in and out of the paid workforce suggests that this normative model of retirement fits workers’ actual experience rather poorly. Instead of crisp exits from the workforce at a particular typical age, many workers make blurred exits that can occur at any chronological age (Mutchler et al., 1997). Marshall and Clarke (1998) suggest that the familiar three-stage life course model needs to be expanded to include some additional transitional periods such as a precareer series of preliminary work experiences; a preretirement period in which the worker prepares to leave the main career job; and a period following exit from the main career job involving bridge jobs in the contingent labor force and possible additional education and training, prior to full retirement from paid employment.