Cohort Change

views updated


The concept of cohort change, which is an attempt to link the "biological rhythm of human existence" (Mannheim, 1952) with the "evolution of the social order" (Parsons, 1951), is a prime example of a tool designed to analyze linkages between the micro (i.e., individual) and macro (i.e., societal) levels of human reality. The phenomena of cohort replacement creates opportunities for societies to rethink, and perhaps redefine, individuals' roles, rights, responsibilities, and rewards. This connection between changes in culture and the gradual and continuous processes of individual aging, birth, and death has led sociologists to investigate how these demographic processes may intersect with personality development.

Although cohorts can be defined in a variety of ways (e.g., a new cohort of graduate students, a new cohort of employees), questions of cohort change, cohort replacement, and cohort succession are primarily about birth cohortspeople born within a given time period. Ryder (1965) described the process of cohort change as an illustration of "demographic metabolism." Social change and population processes are interdependent because the composition of society (aggregate characteristics of the set of individual members) is always in flux. Fertility and in-migration infuse a society with new members, while mortality and out-migration deplete a society of its members. The continuing change in a society's membership creates a dynamic at the macro level that is different from the micro-level dynamic of individual aging. Aging as an individual, biological process also has social implications, because as people age they move from one set of roles or positions to another. In addition, aging is connected to population dynamics through fertility, mortality, and in- and out-migration. Therefore individual aging creates connections between the pace of demographic change and societal transformation. Cohort continuity through the life span provides the element of stability, whereas the perpetual entry, aging, and exit of successive cohorts provide flexibility.

Although initially framed as "the problem of generations" (Mannheim, 1952), the issue under consideration was essentially that of cohort successionthe entry, aging, and exit of successive cohortsand how it is linked to social change. In 1965, Norman Ryder formulated one answer to this question by describing how cohort characteristics (e.g., size, race/ethnic composition, average level of education) could provide the impetus for social change. The impetus for social change in his demographic approach concentrated on an easily observable feature, the equilibration of supply and demand. For example, an unusually large cohort (e.g., the baby-boom cohort) would require building more schools, hiring more teachers, and perhaps the development of community activities for children. In contrast, an unusually small cohort might motivate an emphasis on small classroom size or changes in immigration policy. Although these stimulus-response pairings seem logical, they represent a subset of a much larger number of possible pairings. Why was the response to the baby boom a school construction program rather than a massive shift to home schooling? And if a small cohort cannot satisfy the demand for labor, why not change expectations and policies governing the average workweek, the frequency of vacations, and/or the child labor laws? These questions illuminate the heart of the puzzle. For it is not only the objective features of a cohortby itself or relative to those of adjacent and/or coexisting cohortsthat determine the societal response to the demographic stimulus. Subjective factors are also important, for it is here that the link between the perceived importance of the stimulus and the appropriate response (e.g., relative to constraints, expectations) is forged.

That the biological process of aging is embedded in a social context creates variability in the experience of aging. This variability can be expressed as the difference between individuals, between cohorts, and over time. But it is not simply time (or historical location) that accounts for this variability, for shared experiences require a common framework of perception, interpretation, and reaction. The methodology of cohort analysis requires that people be classified into groups initially based on quantitative markers such as age, date of birth, date of marriage, or date of hire. Group boundaries are created relative to start dates and similar to any grouping exercise, boundary problems are encountered. If it is a birth-year cohort that is of interest, should that year run from January 1 through December 31, from September 1 through August 31, or some other start and end date? How is it sensible to assign two people born on consecutive dates to different cohorts?

While birth cohorts are often used to organize individuals, for analytic purposes the subjective experiences of cohort members must also be considered. The fact that each cohort is defined by a unique intersection of biography and history provides the possibility for commonality of subjective experience, but this potential is actualized only if the subjective experience of history is the same or similar across cohort members. Events must arouse the same kind of outrage or celebration; ideas must elicit similar levels of inspiration or aversion; public figures must be viewed with similar levels of admiration or revulsion.

Early development of the concept of cohort

Cohort approaches to the study of aging and social change have historically focused on the relationship between theoretical and empirical uses of the concept. During the 1970s, researchers continued to struggle with an operationalization of cohort study that did not invite ambiguities in the interpretation of research findings. For example, when comparing people of different ages at one point in time, cohort (based on year of birth) and age (the difference between current year and year of birth) cannot be distinguished. When extending the information across time such as comparing people of the same or different ages at different points in timecohort, age, and year (as the measure of historical time) are intertwined. Age equals current year (history) minus birth year (cohort). Irving Rosow argued in 1978 that focusing attention on methodological solutions to the confounding of age, period, and cohort resulted in the neglect of conceptual and empirical development of the concept of cohort. His question, "What actually is a cohort?" pointed to uncertainties in the definition of cohort and the difficulties inherent in measuring and modeling the complex processes and historical circumstances that cohort members collectively experience. Nevertheless, Rosow conceded that until the mechanisms that translated cohort characteristics into different attitudes and behaviors could be specified in models, cohort, measured simply as shared birth year, remained a useful proxy measure.

Norval Glenn, an early proponent of cohort analysis, conducted much research on social change in attitudes and behaviors. He also warned of the dangers of complete reliance on technical expertise at the cost of careful theoretical development in cohort analyses (see Glenn and Grimes, 1968; Glenn and Zody, 1970; Glenn, 1974)sophisticated statistical techniques alone would not solve the linear restrictions of age, period, and cohort effects in quantitative analysis. Glenn maintained that while Ryder's classic essay developing the connection between cohort succession and social change was extremely influential, it contained few specifics on the techniques of research.

In addition to this emphasis on cohort analyses, a second branch of cohort literature attempted to ground cohort research more firmly in a consideration of social structural constraint and facilitation. Age stratification theory, an attempt to link a structural perspective to the study of individual behavior, emphasized the manner in which the age structure of societal roles organized members into hierarchies (see Riley, 1973; Riley, Johnson, and Foner, 1972; Riley, 1987). Matilda White Riley combined a cross-sectional approach to age group differences with a longitudinal perspective that considers what happens to particular cohorts as they move through time.

From age stratification theory grew the life-course perspective, which is now the dominant approach in social gerontology. The life-course perspective draws on diverse intellectual sources to study differences in aging across cohorts by capturing individual biography within the context of social structure and historical circumstance. Glen Elder was one of the first to conduct micro-level research using longitudinal data on children's lives to systematically study change in families and children over time. Since then, the use of the life-course perspective to connect trajectories of individual lives to larger societal changes has been fueled by the development of new analytic techniques. These techniques allow researchers to represent underlying dynamics of aging, to assess the multilevel contextual impact of environments on individual level outcomes, and to model the individual-level processes that correlate individual outcomes with individual characteristics over time (see Tuma and Hannan, 1987; Kreft and de Leeuw, 1998).

Examples of cohort diversity

Compared to cohorts who lived in the 1800s, cohorts who came of age during the 1900s had a different experience of aging. In 1900, the median age for men was 23 years old; for women 22. By 1999, the median age had risen to 34 for men and 37 for women, in part a reflection of changes in fertility rates, but also linked to changes in life expectancy. In 1900, life expectancy for men was 46 years, for women, 48 yearscompared to 74 and 79, respectively, in 1999. The proportion of men and women aged 65 and older more than tripled (from 4 to 13%) during the twentieth century. Compared to earlier cohorts of 40 year olds, the cohorts of today's 40 year olds do not view their lives as almost over: the expectation of longevity has allowed people to contemplate second careers, and retirement is often viewed as an enjoyable time of life.

Increased life expectancy not only adds years to life; the anticipation of living to older ages transforms the subjective experiences of younger people as well. So, with regard to fertility, for example, not only did the fertility rate drop from an early 1900s high of 3.6 births to 2.1 births in 2000, but the timing of births moved to somewhat older ages. Comparisons of relatively recent fertility behavior demonstrates this point: in the 1960s, most childbearing occurred among women in their early twenties; while in 2000, birth rates for women in their early twenties and late twenties are almost equal, declining only among women in their thirties. The lives of contemporary women are organized much differently than the lives of their mothers and grandmothers. Women are bearing more children out-of-wedlock, marrying later, staying in school longer, becoming increasingly active in the labor force, and working in a wider variety of positions.

Net change and gross change

Studies of cohort change emphasize a dynamic different from that employed by studies of individual change and development. As an example, consider polls that measure approval ratings for elected officials. To examine change at the individual level, it is necessary to interview and reinterview the same respondents. Each time a respondent registers an approval level, the question of change requires a comparison to that individual's previous responses. Changes can be linked to respondents' experiences during the interval. Respondents could alternate between positive and negative reactions to an elected official as the official's actions in turn pleased and then displeased them. Although this type of volatility in approval ratings may carry certain disadvantages, so long as an equal number of people are switching directions, they cancel each other out. The official could therefore enjoy a majority approval rating on a continuous basis, even though the vast majority of people voiced disapproval sometime during the measurement period. In other words, the balanced magnitude and direction of gross change (change at the individual level), results in no change in the approval rating (zero net change). While the composition of those citing favorable versus unfavorable opinions may change, at any given time the official may continue to do a "good" job according to the majority of constituents. Gross change, therefore, refers to the volume of change at the individual level (how often individual evaluations shift from positive to negative or vice versa), whereas net change refers to a change at the aggregate level (i.e., whether the overall approval rating is higher or lower).

Under conditions of relative stability, these two measures of change may not be very different. For example, if one assumes that retirement is an absorbing state (i.e., a state that, once entered, cannot be vacated), then once a member of a particular cohort shifts from employment to retirement, no subsequent shifts occur. As a consequence, the labor-force participation rate characterizing that cohort progressively declines. However, if retirement (at least within certain age ranges) is a temporary status, then the labor-force participation rate for a given cohort can remain stable, even though individual members of the cohort are regularly moving in and out of employment. In contrast, comparing different cohorts who occupy the same age range at different points in time may reveal the impact of cohort replacement. If the average age of retirement is gradually declining, then comparing different cohorts at ages sixty to sixty-four, for example, will also show a decline in labor-force participation rates, as progressively more members of successive cohorts move into retirement at younger ages, demonstrating that the timing of the retirement transition is changing.

Challenges in studying cohort change

In trying to disentangle the various clocks that govern aging, careful conceptualization, rich data, and advanced analytic techniques are all essential. Because of the tendency to use chronological time as the measure for all the various clocks, ambiguous results are often produced: Cohort and period differences become confounded in trend data because different cohorts are compared in different historical periods; age and cohort differences are confounded in cross-sectional data because people who share cohort membership are also the same age; and age and period effects are confounded in intra-cohort trend data because people age into new historical periods.

The key to understanding the linkage between cohort change and social change will be an explanation of how a shared common understandinga shared social consciousnessis created by cohort members, and how this understanding resembles or contradicts that of earlier cohorts or same-aged people of a different mind.

Melissa Hardy Andrea E. Willson

See also Age; Age-Period-Cohort Model; Baby Boomers; Life Course; Population Aging; Theories, Social.


Elder, G. H., Jr. Children of the Great Depression: Social Change in Life Experience. Chicago: University of Chicago Press, 1974.

Elder, G. H., Jr. "Age Differentiation and the Life Course." Annual Review of Sociology, 1 (1975): 165190.

Elder, G. H., Jr. "Perspectives on the Life Course." In Life Course Dynamics: Trajectories and Transitions 19681980. Edited by G. H. Elder, Jr. Ithaca, N.Y.: Cornell University Press, 1985. Pages 2345.

Firebaugh, G., and Haynie, D. L. "Using Repeated Surveys to Study Aging and Social Change." In Studying Aging and Social Change: Conceptual and Methodological Issues. Edited by M. A. Hardy. Thousand Oaks, Calif.: Sage, 1997. Pages 148163.

Glenn, N. D. "Aging and Conservatism." Annals of the American Academy of Political and Social Science 415 (1974): 176186.

Glenn, N. D. "Cohort Analysts' Futile Quest: Statistical Attempts to Separate Age, Period, and Cohort Effects." American Sociological Review 41 (1976): 900904.

Glenn, N. D., and Grimes, Michael. "Aging, Voting, and Political Interest." American Sociological Review 33 (1968): 563575.

Glenn, N. D., and Zody, R. E. "Cohort Analysis with National Survey Data." The Gerontologist, 10 (1970): 233240.

Kreft, I., and de Leeuw, J. Introducing Multilevel Modeling. Thousand Oaks, Calif.: Sage, 1998.

Manheim, K. "The Problem of Generations." In Essays in the Sociology of Knowledge. Edited by P. Kecskemeti. Boston: Routledge and Kegan Paul, 1952. (Original work published in 1927).

Marshall, V. W. "The State of Theory in the Social Sciences." In Handbook of Aging and the Social Sciences, 4th ed. Edited by R. H. Binstock and L. K. George. New York: Academic Press, 1996.

Parsons, T. The Social System. New York: Free Press, 1951.

Riley, M. W. "Aging and Cohort Succession: Interpretations and Misinterpretations." Public Opinion Quarterly 37 (1973): 3549.

Riley, M. W. "On the Significance of Age in Sociology." American Sociological Review 52 (1987): 114.

Riley, M. W.; Johnson, M.; and Foner, A. Aging and Society, Vol. 3: A Sociology of Age Stratification. New York: Russell Sage, 1972.

Rosow, I. "What is a Cohort and Why?" Human Development 21 (1978): 6575.

Ryder, N. B. "The Cohort as a Concept in the Study of Social Change." American Sociological Review 30 (1965): 843361.

Tuma, N. B., and Hannan, M. T. Social Dynamics: Models and Methods. San Diego: Academic Press, 1984.

U.S. Bureau of the Census. Statistical Abstract of the United States: 1999. Washington, D.C.: U.S. Bureau of the Census, 1999.


See Geriatric medicine