AGING AND THE LIFE COURSE

Social gerontology, or the sociology of aging, has two primary foci: (a) social factors during late life, and (b) social antecedents and consequences of aging. Thus, social gerontology includes examination of both the status of being old and the process of becoming old. Increasingly, theories and methods of the life course are replacing the earlier emphasis on late life as a separate topic of inquiry. This is a vast arena, and the sociology of aging is appropriately informed by the theories and methods of many sociological subspecialities ranging from macrohistorical and demographic perspectives to the microorientations of social psychology and interpretive sociology.

HISTORY OF THE FIELD

Historically, social gerontology emerged from a social-problems orientation and focused on the deprivations and losses that were expected to characterize late life (e.g., Burgess 1960; Cain 1959). Early research in the field focused on issues such as poverty during late life; old age as a marginal status, reflecting problems of social integration; the negative effects of institutionalization and poor quality of long-term care; and ageism and age discrimination. Early on, however, some investigators saw the dangers of allowing a crisis orientation to dominate the study of aging and focused attention on patterns of "normal" and "successful" aging (Havighurst 1963; Palmore 1970). A significant proportion of research also focuses on the problems of late life. Investigators remain concerned about social integration and adaptation to loss. The majority of funding for aging research is provided by the National Institute on Aging, which is mandated to support health-related research. Consequently, much aging research focuses on illness and the health care delivery system. The dramatic aging of the population (U.S. Bureau of the Census 1987)—a trend that will peak with the aging of the Baby Boom cohorts—leads to questions about the capacities of social institutions and public policies to meet the needs of an unprecedented number and proportion of older adults. Scholars using political economy theories focus on the ways in which societies respond to the dependency needs of older adults and the social implications of those responses.

Although much research remains focused on the problems of late life, sociologists now recognize the broader importance and implications of old age and aging. Two primary factors appear to have been the driving forces that account for this broader and more complex view. First, despite the social-problems orientation of most early research, empirical data failed to confirm a uniformly bleak picture of old age. For example, in spite of higher rates of illness and disability, the vast majority of older adults are competent and able to live independent lives (Kunkel and Applebaum 1992). Similarly, rather than representing involuntary loss of a treasured role, retirement is actively sought by the majority of older workers and seldom poses adaptive problems (e.g., Hardy and Quadagno 1995). In addition, some of the problems observed in early studies of older adults have been remedied by the increased resources that recent cohorts have brought to late life, as well as to effective public policies. Thus, although health care costs remain a burden for many older adults, Medicare and Medicaid substantially reduced barriers to health care among older people. Similarly, as a result of improvements in Social Security benefits and increased participation in private pensions, older Americans now are no more likely to live in poverty than younger adults and, indeed, are less likely to live in poverty than children (U.S. Bureau of the Census 1997). Such findings pushed social gerontology toward more complex and empirically defensible perspectives on old age and aging.

Second, sociologists came to recognize that age plays a fundamental role in social structure and social organization. As a parameter of social organization, age affects the allocation of social resources and social roles. Along with sex and race, age is an ascribed status. But age is unique among ascribed statuses in that it changes over time, and movement across age categories results in changing expectations for behavior, changing access to social resources, and changing personal and social responsibilities. The structural quality of age is best articulated in age stratification theory (Riley 1987). Age stratification refers to the division of society into meaningful age groups that differ in social value and the allocation of social resources. The concept of age stratification has proven to be useful in a variety of ways. At the broadest level, it reminds us that age is a fundamental parameter of social organization. Age stratification has been particularly useful in highlighting age-related roles and norms. It has a social-psychological facet as well: age consciousness or awareness and identification with members of one's age group (Day 1990). Moreover, the structural and individual facets of age stratification operate in both directions. On the one hand, the structural component of age stratification allocates roles and resources and assigns differential social value to age strata. Thus, ageism is largely an effect of age stratification. On the other hand, by promoting age consciousness, age stratification sets the stage for age-based public policies and collective efforts by older adults to protect or increase their share of societal resources (e.g., voting and lobbying efforts based on the self-interests of the elderly).

In the 1990s, the sociology of aging focused on change and stability across the life course. Life-course perspectives have enriched aging research in several ways (e.g., Elder 1995; George 1993). First, a life-course approach is attractive because it recognizes that the past is prologue to the future. That is, status and personal well-being in late life depend in large part on events and achievements experienced earlier in the life course. Second, life-course perspectives emphasize relationships across life domains, recognizing that, for example, family events affect and are affected by work and health. Traditionally, sociological research has focused on specific life domains (e.g., the sociology of work, the sociology of the family); life-course perspectives, in contrast, are person-centered rather than domain-centered. Third, life-course perspectives focus on the intersection of history and personal biography. Although the macro-micro schism remains difficult to bridge, life-course research has documented some of the complex ways that historical conditions affect personal lives both contemporaneously and over subsequent decades.

THE AGE-PERIOD-COHORT PROBLEM

Isolating the effects of age and characterizing the aging process are difficult tasks. Because many factors affect social structure and individual behavior, it is always difficult to isolate the effects of a specific factor. But this task is especially difficult with regard to age, because it is inherently confounded with the effects of two other factors: cohort and period. Age, of course, refers to time since birth, and age effects refers to patterns resulting from the passage of time or sheer length of life. Cohort refers to the group of persons born at approximately the same time (e.g., the 1920 cohort, the 1940–1944 cohort). There are two primary kinds of cohort effects. One type results from historical factors. For example, cohorts who lived through the Great Depression or World War II had different life experiences than cohorts who were not exposed to those historical events. And, as a further complication, the effects of historical events vary depending on the ages of those who experience them. The second type of cohort effect reflects compositional characteristics. For example, large cohorts (such as the Baby Boomers) may face greater competition for social resources than smaller cohorts (e.g., those born during the Great Depression, when fertility rates were low). Both types of cohort effects can have persistent effects on the life course, and, hence, late life. Period effects (also called time of measurement effects) result from events or situations that happen at a specific time, and tend to affect individuals regardless of age or cohort. For example, faith in government decreased in all Americans (regardless of age or cohort) at the time of the Watergate scandal.

Age, cohort, and period effects are intertwined. If one knows when an individual was born and also knows the time of measurement, simple subtraction provides accurate information about the individual's age. Similarly, if one knows an individual's age and time of measurement, one can easily calculate date of birth or birth cohort. Statistically, there are no easy methods for disentangling age, period, and cohort. In general, however, the most compelling research results are those that are based on examination of multiple cohorts at multiple times of measurement. If the same age patterns are observed across different cohorts measured at different times, those patterns are likely to reflect age effects. If patterns are not similar across cohorts and times of measurement, however, they are likely to reflect cohort or period effects.

The issue of age-period-cohort effects has lost some of its appeal; critics point out that simply knowing, for example, that there is a cohort effect leaves unanswered what it is about those cohorts that generated the observed differences. This is an appropriate criticism; nonetheless, it is immensely helpful in searching for causal explanations to know whether the underlying mechanism is consistent across time and place (an aging effect), had strong contemporaneous effects on persons of all ages (a period effect), or affected only specific cohorts (a cohort effect). Age, cohort, and period effects are all important in aging research. Age effects provide information about human development as it unfolds in social context. Cohort effects permit us to observe the social implications of shared history and cohort composition. Period effects provide information about the effects of contemporaneous events and situations on social structure and individual behavior. Distinguishing among age, period, and cohort effects also has important implications for the generalization of research results. Age effects are the same or highly similar across time and place; thus, they are generalizable. In contrast, cohort and period effects are, by definition, variable across time. Consequently, generalization is limited.

Examination of age-period-cohort effects requires large data bases in which multiple cohorts are observed on multiple occasions over long periods of time. Because of these stringent requirements, few aging studies focus specifically on disentangling these confounded factors. But recognition of these sources of confounding appropriately temper investigators' generalizations. In addition, this issue has sensitized researchers to the need to examine change over time, with the result that longitudinal studies have become the dominant research design in efforts to characterize the aging process.

No single theme nor easily summarized list of topics does justice to the scope and diversity of research on aging and the life course. However, three major research domains can provide a general sense of the current major avenues of aging and life-course research: aged heterogeneity, life-course dynamics, and life-course trajectories.

AGED HETEROGENEITY

The majority of research on older adults focuses on heterogeneity in that population. In many ways, this is the legacy of the problem orientation upon which initial aging research in the social sciences rested. Even now, most aging research focuses on differences among older adults and the complex configurations of social factors that can account for individual differences during late life. And, although there are many exceptions, a majority of research in this domain focuses on the social factors that explain individual behavior and personal well-being.

The specific topics that are examined in research on aged heterogeneity are extensive. Health and disability are major concerns in this research tradition, including physical illness; mental illness, especially depression; cognitive status and dementing illness; functional status and disability; and health behaviors. Health service utilization is a corollary emphasis; there are numerous studies on both utilization in general (e.g., doctors visits, hospitalizations) and specialized health care settings and providers (e.g., long-term care facilities, emergency room visits, screening programs and other preventative services, mental health professionals, dentists). There also has been extensive research on the use of social and community services and living arrangements—both of which are strongly driven by health during late life.

Along with health, socioeconomic status has been a high-volume, long-standing focus in research on aged heterogeneity. Economic status, in the form of income and, to a lesser extent, assets has been the primary emphasis of this research, with studies of the antecedents and consequences of retirement ranking as a close second. Pensions have been shown to have strong effects upon the adequacy of postretirement income, which has spurred substantial investigation of the determinants of pension acquisition and value.

From a more social-psychological perspective, there are strong research traditions examining multiple psychosocial states that can be subsumed under the umbrella of "quality of life." Life satisfaction and morale have received paramount attention in this regard. But several dimensions of the self (e.g., self-esteem, locus of control, sense of mastery) also have rich research traditions.

Two common elements of the wide range of research conducted on the topic of aged heterogeneity merit note. First, a common thread in this research is the desire to understand the processes that render some older adults advantaged and others disadvantaged. This is often described as the "applied" character of aging research. However, it also is a focus on stratification in the broadest sense, which is one of the most persistent and cherished issues in mainstream sociology. Research on aging and the life course has contributed much to our understanding of how broadly life chances and life quality are linked to social factors, as well as to the multiple powerful social bases of stratification.

Second, examination of heterogeneity among the elderly has proven to be a strategic site for testing and refining multiple middle-range theories and concepts that are central to sociological research. A few examples will illustrate this point. Research on health and disability during late life has been enriched by attention to issues of social stress and social support; conversely, the older population has proven to be ideal for testing theoretical propositions about the effects of stress and social support. Similarly, the issue of social integration, which has an honored tradition in sociology, has been reactivated in studies of older adults, who vary widely in number and quality of links to social structure. And issues such as the "feminization of poverty" and the "double/triple jeopardy hypothesis" (i.e., the potential interacting deprivations associated with being female, nonwhite, and old) have highlighted the extent to which socioeconomic stratification rests on ascribed rather than achieved statuses.

LIFE-COURSE DYNAMICS

While the issue of heterogeneity focuses attention on differences across individuals, life-course dynamics focus on the persisting effects of social factors over time and stages of life; that is, on intraindividual change. At least three types of life-course studies have received substantial attention in aging research.

The Intersection of History and Personal Biography. There now is substantial evidence that historical events can permanently alter personal lives. Most research to date has focused on two historical events: the Great Depression and World War II; both historical events have been linked to life circumstances during late life. Elder (1974, 1999) has compellingly documented, for example, that entering the labor market during the Depression had a permanent negative effect on occupation and income, which in turn affected socioeconomic status during late life. In contrast, younger men, who entered the labor force immediately after World War II experienced historically unparalled occupational opportunities (and, if they were veterans, government-subsidized college educations).

Other studies by Elder and colleagues focused on the life-course consequences of combat experience during World War II (Elder, Shanahan, and Clipp 1994, 1997). Exposure to combat was strongly related to subsequent health problems, not only immediately after the war, but also during late life. Social factors loom large in these dynamics. First, social factors were strongly predictive of which soldiers were exposed to combat. Second, social resources and deprivations were strong predictors of the onset or avoidance of health problems.

The effects of historical events on the life course appear on two levels. When an historical event is pervasive, one method of observing its effects is via cohort differences. That is, an event with wide-ranging effects will render those who experience it discernibly different from cohorts that come before and after it. At the same time, historical events do not have the same immediate or long-term consequences for all members of a cohort. For example, families were differentially affected by the economic dislocations of the Depression, with some suffering extensive economic deprivation and others experiencing little or no change in their economic fortunes. Thus, not everyone who "lives through" a major historical event will experience serious life-course consequences. One of the tasks of the life-course scholar is to identify the social factors that determine whether or not an historical event alters personal biography.

The Persisting Effects of Early Traumas and Deprivations. A second research focus that illustrates the power of life-course perspectives is investigation of the consequences of early traumas and deprivations on well-being in middle and late life. A growing body of research documents the farreaching effects of severe childhood experiences—including parental divorce and, to a lesser extent, parental death; childhood poverty; and childhood physical and sexual abuse—on the course of adulthood. Most research to date has focused on the implications of these traumas and deprivations on physical and mental health (e.g., Krause 1998; Landerman, George, and Blazer 1991). The robust relationships between events experienced during childhood and health fifty to sixty years later, controlling on contemporaneous risk factors for morbidity, is strong evidence of the power of fateful events to alter the life course.

Again, social factors are strongly implicated in the processes that account for the relationships between childhood traumas and health and well-being during late life. First, the availability of social resources at the time of the trauma can dampen its negative effects in both the short and the long term. For example, financial security and adequate supervision ameliorate most of the negative effects of parental divorce on subsequent socioeconomic achievements and physical and mental health (e.g., Kessler and Magee 1994). Second, two of the primary mechanisms by which childhood traumas generate poor health in later life are socioeconomic achievement and high-quality social relationships (e.g., McLeod 1991). That is, childhood traumas are often associated with lower socioeconomic status and poor-quality relationships during adulthood—both of which are risk factors for physical and mental illness. If, however, individuals who experienced childhood traumas manage to achieve adequate financial resources and supportive social ties during adulthood, their excess risk of illness in middle and late life is reduced substantially.

The Persisting Effects of Early Life Decisions. There now is substantial evidence that the decisions that individuals make during early adulthood have important consequences for their life circumstances in late life. Studies of retirement income provide perhaps the best illustration of this research domain. The strongest predictor of retirement income is occupational history. Throughout adulthood, individuals "sort themselves" into jobs that differ not only in income, but also in benefits (i.e., total compensation packages). Of these, the availability and quality of pensions is most important for retirement income. There is strong evidence that the provision of pensions differs not only by occupation, but also by industrial sector (Quadagno 1988). Thus, when individuals make occupational choices—including job changes throughout adulthood—they are inevitably determining, in part, their retirement incomes.

Research on women's retirement income has broadened our understanding of the life-course consequences of early decisions. Women and men tend to be concentrated in different occupations and different industrial sectors—and those in which women dominate have, on average, lower earnings and lower likelihood of pension coverage (O'Rand 1988). Moreover, family formation decisions strongly affect women's job histories. Compared to men, women are less likely to work full-time and work fewer total years, largely as a result of parental responsibilities. All of these factors combine to produce substantially lower retirement incomes for women than for men (O'Rand and Landerman 1984).

LIFE-COURSE TRAJECTORIES AND PERSON-CENTERED RESEARCH

The two research domains described above focus on interindividual differences in late life and intraindividual change over the life course. A third domain, less developed than the others but exciting in its scope, attempts to examine interindividual differences and intraindividual change simultaneously. At this point, two emerging research traditions illustrate the nature and potential of this approach.

Life-Course Trajectories. Trajectories refer to long-term patterns of stability and change. They can be examined at both the aggregate (e.g., the "typical" career, the modal pathway to nursing home placement) and individual levels. Thus, trajectories capture patterns of intraindividual change. Examination of heterogeneity can be pursued in two ways. In the first, the trajectory that best describes the sample or population is constructed. Subsequently, using techniques such as hierarchical linear modeling or growth-curve analysis, investigators can examine the extent to which factors of interest alter the shape of the trajectory. For example, a trajectory of earnings across adulthood can be constructed for a given sample. Investigators can then examine the degree to which factors such as sex, education, and race affect the shape of the earnings trajectory. In the second approach, multiple common trajectories are identified and investigators then determine the characteristics associated with those trajectories. Using this approach, for example, several common trajectories of earnings during adulthood could be identified (e.g., consistently increasing earnings, earnings peaking during mid-life and then decreasing, consistently decreasing earnings, a relatively flat earnings history). Factors such as sex, education, and race could then be examined to determine their association with these distinctive earnings trajectories.

Trajectory-based research, with its focus on both interindividual differences and intraindividual change, is very attractive. The major limitation to the use of this approach is the availability of data, because longitudinal data covering long periods of time are required if one wishes to understand life-course patterns. It should be noted, however, that trajectory-based research can also be useful for studying shorter processes (e.g., patterns of illness outcome, with relatively short-term trajectories of death, chronicity, recovery, and relapse).

Trajectory-based research relevant to our understanding of middle and late life is gradually accumulating. This is especially true for long-term patterns of health and functioning, providing evidence about both the dynamics of disability during late life (e.g., Maddox and Clark 1992; Verbrugge, Reoma, and Gruber-Baldini 1994) and long-term patterns of stability, improvement, and decline in health over the course of adulthood (e.g., Clipp, Pavalko, and Elder 1992). Important trajectory-based research on pathways to retirement (Elder and Pavalko 1993) and place of death (Merrill and Mor 1993) is also available.

The concept of trajectories has been valuable in theoretical development, as well as empirical inquiry. The theory of cumulative advantage/disadvantage has intersected well with studies of life-course trajectories. This theory posits that heterogeneity is greater in late life than earlier in the life course as a result of the accumulation of assets (advantage) or liabilities (disadvantage) over time. The theory of cumulative advantage/disadvantage has been especially useful in understanding socioeconomic heterogeneity in late life, especially differences in total net worth (e.g., Crystal and Shea 1990; O'Rand 1996). However, it can be applied to other sources of heterogeneity in late life as well (e.g., health). Investigators who prefer examination of the multiple distinctive trajectories within a sample would offer a caution to cumulative advantage/disadvantage theory, however. They would note that although trajectories of increasing and decreasing advantage are undoubtedly common, there are likely to be other important trajectories as well—e.g., a trajectory of cumulating advantage that is reversed as a result of a personal (e.g., serious illness) or societal (e.g., severe economic downturn) catastrophic event.

Person-centered research. A more recent contribution to the sociological armamentarium for understanding aging and the life course is person-centered research. As the label implies, the focus of this emerging research strategy is to analyze "people" rather than "variables." In practice, this means that members of a sample are first grouped into subsets on the dependent variable of interest. Subsequently, those categories are further disaggregated into groups based on distinctive pathways or life histories associated with the outcome of interest.

Work by Singer, Ryff, and Magee (1998) provides the richest example of person-centered aging research to date. The dependent variable of interest was mental health. In the first stage of their research, groups of middle-aged women were divided into groups that differed on levels of mental health. Subsequently, the large archive on longitudinal data obtained from these women over the previous three decades was examined to identify distinctive pathways associated with mid-life mental health. For example, the group of women who exhibited high levels of mental health and well-being were further subdivided into two groups that Singer et al. (1999) labeled the healthy and the resilient. Healthy women were those who had life histories that were relatively free of major stressors or traumas, enjoyed adequate or higher levels of social and economic resources, and exhibited stable patterns of robust mental health. Resilient women were those who had achieved robust mental health at mid-life despite earlier evidence of poor mental health and/or histories of stress and/or inadequate social and economic resources. Clearly the life histories of women who were mentally healthy at mid-life varied in important ways. Moreover, the healthy women illustrate the benefits of leading relatively "charmed" lives, while the resilient women help us to understand the circumstances under which "risky" life histories can be turned around to produce health and personal growth.

There are clear similarities between person-centered research and trajectory-based research. Both are based on long-term patterns of change and stability and both are designed to understand both life-course dynamics and heterogeneity in those dynamics. But there also are important differences between the two approaches. In trajectory-based research, the trajectory or pathway itself is the "dependent variable" of interest and the "independent variables" are factors that have the potential to alter the shape(s) of those pathways. In person-centered research, the "dependent variable" is an outcome of interest (e.g., mental health), and trajectories of the "independent variables" are constructed to explain that outcome. Both are currently at the cutting edge of research that attempts to simultaneously examine life-course dynamics and life-course heterogeneity.

Aging and the life course is an important sociological specialty. It provides us with information about aging and being "old," and about the antecedents and consequences of stability and change during late life. The increasing life-course focus of aging research is especially important in that it concentrates sociologists' attention on the dynamics of intraindividual change and on the intersections of social structure, social change, and personal biography. At its best, aging and life-course research effectively link processes and dynamics at the macrohistorical and societal levels with individual attitudes and behaviors. In addition, the life course in general, and late life in particular, provide excellent contexts for testing and, indeed, challenging some commonly held assumptions and hypotheses about the dynamics of social influence. In this way, it offers valuable contributions to the larger sociological enterprise.

references

Burgess, Ernest W. 1960 "Aging in Western Culture." In E. W. Burgess, ed., Aging in Western Societies. Chicago: The University of Chicago Press.

Cain, Leonard D., Jr. 1959 "The Sociology of Aging: A Trend Report and Bibliography." Current Sociology 8:57–133.

Clipp, Elizabeth C., Eliza K. Pavalko, and Glen H. Elder, Jr. 1992 "Trajectories of Health: In Concept and Empirical Pattern." Behavior, Health, and Aging 2:159–177.

Crystal, Stephen, and Dennis Shea 1990 "Cumulative Advantage, Cumulative Disadvantage, and Inequality among Elderly People." The Gerontologist 30:437–443.

Day, Christine L. 1990 What Older People Think. Princeton, N.J.: Princeton University Press.

Elder, Glen H., Jr. 1999 Children of the Great Depression (twenty-fifth anniversary ed.). Boulder, Colo.: Westview Press.

——1995 "The Life Course Paradigm: Historical, Comparative, and Developmental Perspectives." In P. Moen, G. H. Elder, and K. Luscher, eds., Examining Lives in Context: Perspectives on the Ecology of Human Development. Washington, D.C.: American Psychological Association Press.

——1974 Children of the Great Depression. Chicago: University of Chicago Press.

——, and Eliza K. Pavalko 1993 "Work Careers in Men's Later Years: Transitions, Trajectories, and Historical Change." Journal of Gerontology: Social Sciences 48:S180–S191.

Elder, Glen H., Jr., Michael J. Shanahan, and Elizabeth C. Clipp 1997 "Linking Combat and Physical Health: The Legacy of World War II in Men's Lives." American Journal of Psychiatry 154:330–336.

——1994 "When War Comes to Men's Lives: Life-Course Patterns in Family, Work, and Health." Psychology and Aging 9:5–16.

George, Linda K. 1993 "Sociological Perspectives on Life Transitions." Annual Review of Sociology 19:353–373.

Hardy, Melissa A., and Jill Quadagno 1995 "Satisfaction with Early Retirement: Making Choices in the Auto Industry." Journal of Gerontology: Social Sciences 50:S217–S228.

Havighurst, Robert A. 1963 "Successful Aging." In R. Williams, C. Tibbets, and W. Donahue, eds., Processesof Aging. New York: Atherton.

Kessler, Ronald C., and William J. Magee 1994 "Childhood Family Violence and Adult Recurrent Depression." Journal of Health and Social Behavior 35:13–27.

Krause, Neal 1998 "Early Parental Loss, Recent Life Events, and Changes in Health." Journal of Aging andHealth 10:395–421.

Kunkel, Susan R., and Robert A. Applebaum 1992 "Estimating the Prevalence of Long-Term Disability for an Aging Society." Journal of Gerontology: SocialSciences 47:S253–S260.

Landerman, Richard, Linda K. George, and Dan G. Blazer 1991 "Adult Vulnerability for Psychiatric Disorders: Interactive Effects of Negative Childhood Experiences and Recent Stress." Journal of Nervousand Mental Disease 179:656–663.

Maddox, George L., and Daniel O. Clark 1992 "Trajectories of Functional Impairment in Later Life." Journal of Health and Social Behavior 33:114–125.

McLeod, Jane D. 1991 "Childhood Parental Loss and Adult Depression." Journal of Health and Social Behavior 32:205–220.

Merrill, Deborah M., and Vincent Mor 1993 "Pathways to Hospital Death Among the Oldest Old." Journal ofAging and Health 5:516–535.

O'Rand, Angela M. 1996 "The Precious and the Precocious: Understanding Cumulative Disadvantage and Cumulative Advantage Over the Life Course." TheGerontologist 36:230–238.

——1988 "Convergence, Institutionalization, and Bifurcation: Gender and the Pension Acquisition Process." Annual Review of Gerontology and Geriatrics 8:132–155.

——, and L. Richard Landerman 1984 "Women's and Men's Retirement Income: Early Family Role Effects." Research on Aging 6:25–44.

Palmore, Erdman B. 1970 Normal Aging: Reports from theDuke Longitudinal Studies, 1955–1969. Durham, N.C.: Duke University Press.

Quadagno, Jill 1988 "Women's Access to Pensions and the Structure of Eligibility Rules: Systems of Production and Reproduction." Sociological Quarterly 29:541–558.

Riley, Matilda White 1987 "On the Significance of Age in Sociology." American Sociological Review 52:1–14.

Singer, Burton, Carol D. Ryff, Deborah Carr, and William J. Magee 1998 "Life Histories and Mental Health: A Person-Centered Strategy." In A. Rafferty, ed., Sociological Methodology, 1988. Washington, D.C.: American Sociological Association.

U. S. Bureau of the Census 1997 Statistical Abstracts of theUnited States, 1997. Washington, D. C.: U. S. Government Printing Office.

U. S. Bureau of the Census 1987 An Aging World. Washington, D. C.: U. S. Government Printing Office.

Verbrugge, Lois M., Joseto M. Reoma, and Ann L. Gruber-Baldini 1994 "Short-Term Dynamics of Disability and Well-Being." Journal of Health and SocialBehavior 35:97–117.

Linda K. George