Like leisure, retirement is difficult to define. Retire­ment is more difficult to define than just guessing from someone’s age (Henretta, 1997; Szinovacz & DeViney, 1999). One way to look at retirement is to equate it with complete withdrawal from the work­force. But this definition is inadequate; many retired people continue to work part time (Mutchler et al., 1997; Ruhm, 1990). Another possibility would be to define retirement as a self-described state. However, this definition does not work either, because some African Americans define themselves with labels other than “retired”’ in order to qualify for certain social service programs (Gibson, 1991).

Part of the reason it is difficult to define retire­ment precisely is that the decision to retire involves the loss of occupational identity (discussed earlier). What people do for a living is a major part of their identity. Not doing those jobs anymore means we either put that aspect of our lives in the past tense—“I used to work as a manager at the Hilton”—or say nothing at all. Loss of this aspect of identity can be difficult to face, so some people look for a label other than “retired” to describe themselves.

A useful way to view retirement is as a complex process by which people withdraw from full-time participation in an occupation (Henretta et al., 1997; Mutchler et al., 1997; Sterns & Gray, 1999). This withdrawal process can be described as either “crisp” (making a clean break from employment by stopping work entirely) or “blurred” (repeatedly leaving and returning to work, with some unem­ployment periods) (Mutchler et al., 1997). Bob is a good example of a “crisp” retirement. He retired from TWA at age 65; now in his late 80s, he has done nothing work-related in the interim.

Whereas many people think of retirement as a crisp transition, the evidence shows that less than half of older men who retire fit this pattern (Mutchler et al., 1997). Most men adopt a more gradual or “blurred” process involving part-time work in an effort to maintain economic status. Jack is one of these men. When he retired from DuPont at age 62, he and a friend began a small consulting company. For about five years, Jack worked when he wanted, gradually cutting back over time.

The lack of crisp retirement creates another com­plicating factor—the idea of a “normal” retirement age such as age 65 may no longer be appropri­ate (Cornman & Kingston, 1996; Mutchler et al.,

1997) . Instead, the notion of a typical retirement age changes to a range of ages, further blurring the meaning of “early” or “late” retirement (Cornman & Kingston, 1996).

The complexity of the retirement process must be acknowledged for us to understand what retire­ment means to people in different ethnic groups. For example, whereas middle-class European Amer­icans often use a criterion of full-time employment to define themselves as retired or not, Mexican Americans use any of several different criteria

Work, Leisure, and Retirement 479