Although labor-force participation among older men has been increasing (U.S. Bureau of Labor Statistics, 2005), retirement is still largely viewed as a highly desirable life transition. Even so, there seems to be a popular perception that retirement closely precedes ill health and death and leads to what once was referred to as a "roleless role" (Burgess, 1960). But retirement is not a role; it is, rather, a life stage-one that has multiple pathways of entry and during which the boundaries between work and nonwork are somewhat fluid. Accordingly, the relationship between retirement and health is quite complex (Ekerdt, 1987). In this article, we explore in greater depth the complexity of the tetirement-health connection among men. Is retirement the source of morbidity and mortality, is it the wellspring of activity and health, or is it something in between? Answering this question requires careful examination of the social context in which men's retirement occurs.
To understand the social context of retirement and its relevance to health among men, one must first consider whether retirement represents a normative process. For example, the more retirement is institutionalized and accepted as a regularized feature of the life course supported by major social institutions, the more social resources are likely to be devoted to it, and the more likely it is that the transition from worker to retiree will be conducted in a manner that promotes the health and well-being of the retiring worker. Because retirement is also age normative-that is, expected to take place at a certain time in life, early retirement and late retirement are considered "off-time" transitions that may have consequences for the retiree's health, particularly mental health.
Second, we must consider the extent to which retirement affords meaningful opportunities for self-development and social interaction. Indeed, if retirement becomes a "roleless role" in which few social supports and meaningful interactional opportunities exist, then we would expect health and overall functioning to diminish.
Based on the existing research, scholars have found it difficult to determine the effect of retirement on health. Expectations as to what this effect might be differ, mainly because retirement has evolved from two opposing theories. On one hand, retirement is often conceptualized as a stressful, if not disruptive, life event that disengages men from the social activity of the workplace and, often, from the steady income that it can provide (Atchley, 19823; Ekerdt, 1987; Hochschild, 1975). In that view, the circumstances surrounding retirement-namely, loss of (discretionary) income or mandatory retirement regulations-are expected to lead to decreased physical and mental health.
But some scholars, like Minkler (1981), have cautioned that seeing retirement solely as a stressful event is a theoretical approach that "remains insufficient to the extent that it fails to move beyond the individual and his or her perceptions and coping mechanisms to include consideration of the broader sociopolitical context within which these perceptions and resources are shaped" (p. 120).
And, indeed, retirement is also conceptualized as a largely normative and anticipated life transition-a healthy time of relaxation and enjoyment (Holden, Burkhauser, and Feaster, 1988; Kim and Moen, 2002; Mein et al., 1988). Adherents of this perspective note that retirees would seem to have more time for leisure and less stress from work expectations. As a result, retirees' health may be better than that of their counterparts who are not retired.
Perhaps the apparent discrepancy between these perspectives is the result of the way that contextual factors beyond men's immediate control affect the quality of their retirement. Atchley (19823) chronicled a changing public perception of retirement from an infrequent business arrangement by the few lucky enough to reach later life, to modern retirement structures in which a limited number of jobs means that someone has to leave the labor force. …