Cross-cultural perspectives on the concept of retirement: An analytic redefinition

Abstract

A cross-cultural valid analytic definition of retirement remains elusive in gerontology despite a long tradition of research on the topic. Inadequate attention has been paid to consistently defining the key concepts used to examine retirement and to specifying its occurrence in non-Western, non-industrial societies. This paper critically reviews basic cultural tenets in the notion of retirement, and proposes a more comparatively valid definition. It then proposes a three part comparative categorization by exploring retirement in contemporary Western nations and comparing it with retirement-like practices from a range of non-Western cultures including Thai, Chinese, Ladak, Fulani, Lusi and Aymara.

Introduction

Precisely how to define “retirement” in a cross-culturally valid manner for comparative research, what to study, and where to study it, remains problematic for comparative research on social organization and the life course in later life. A seeming familiarity with the concept of retirement in our home culture challenges our ability to recognize its specific features and the problems it presents for research. Certain basic questions recur in the discourse on retirement. Do people in all cultures “retire?” Is the notion of retirement universal, or is it alien to non-Western cultures? What is (and is not) relevant subject matter for studies of this topic?

Retirement is a familiar part of the social universe for the current cohort of workers in Western industrialized nations. But, historically and cross culturally, the practice of retirement may be anything but common. Historians such as William Graebner (1980) and Jill Quadagno (1982) show that, in its current institutional form marked by state supported pensions, it is a relatively recent practice. The institution of retirement appears to be a product of the confluence of developing industrialization, surplus capital, and Protestant ideology. The demographic forces of decreased mortality, better health care systems, and increased longevity have contributed to larger numbers of elderly than ever before in our history. However, demographers and historians have taught us that there is no direct causal relationship between simplistic “hydraulic pressure” models of population growth and the development of forms of social institutions or the emergence of moral-conceptual categories. Reports in history and anthropology amply illustrate the many alternative pathways from cultural involution (Geertz 1975) to abandonment or neglect of portions of a population (Glascock & Finemann 1981). Gerontologists are now pursuing similar questions about how socio-political forces, not only population growth, shape state institutions for the elderly (Calasanti & Zajicek, 1993).

What, then, is the consensus in the literature about the presence of “retirement?” There are multiple, competing views:

Retirement is rare (in the sense of transfer of control as opposed to the withdrawal from productive activity). (Jack Goody. 1976:117ff)

All societies retire the elderly. (Foner and Kertzer, 1978).

Retirement is a rarity in tribal and village society, within Euro-American society it is a recent phenomena. Retirement can be examined in any culture … as a social institution … set of personal meanings. (Margaret Clark 1972:112, 121).

The emerging pattern of social life which we designate as retirement represents the development in modern society of a new and distinct role … distinct from previous patterns of old age. (Donahue et al, 1960:331, 334).

There is no retirement role in previous and non-Western societies. (Leo Simmons 1954:104).

Retirement is the unique creation of industrial society; in early societies the old were only supported so long as they were able to be productive. (Robert Atchley 1976:11ff).

Retirement is not a phenomenon peculiar to industrial societies … all societies have solutions to problems of supporting the aged … retirement existed in many non-industrial societies. (Kurt Back 1977:78).

Retirement as a sociocultural category is distinct from old age. As social categories, each is defined by different criteria. Not all retirees are old, nor are all elderly retired. The category of retiree is constituted according to criteria specific to the occupational domain; the label old/elderly is made by reference to criteria from several domains, including the occupational. Relevant factors include (among others) chronological age, grandparenthood, and age-stereotypical physical features such as gray hair, wrinkled skin, and slow gait.

There are two conceptual starting points for research on retirement. From one point of view retirement is a universal pattern of behavior observable in all societies; from the other it is a cultural construct specific to modern Western society. These different approaches have vital theoretical, methodological, and pragmatic implications. Why is there disagreement at the most fundamental level about the existence of such a seemingly obvious practice?

This paper examines these contrasting conceptions of retirement and proposes a definition that is more cross culturally accurate and useful. As it exists in the West, the practice of retirement represents a category of social relationships at the state level. Ideally it is available to all members of the society; the legal statutes that determine individual entitlements in retirement are not exclusively tied to any particular kinship, residential group, class, or religious factors. However, the interplay of local situations shaping the economic resources and health disparities of differing segments of the population will clearly affect how many people are actually able to enter into retirement. These factors are well documented and not the focus here, since the goal is to clarify the cultural definition, not the processes, of retirement nor its evolving population profile and diversity (ethnicity, gender, class) within those industrialized economies that condition the behavioral and population distribution of ‘retirement’ patterns (for example, O’Rand & Henretta 2002; Quadagno & Fobes, 1995, Calasanti, 1996; Zsembik & Singer, 1990).

The purpose of this paper is to critique and refine the analytic definition of retirement in the literature and to describe the ways that the modern Western conception of retirement is distinctive, perhaps unique. A set of selected cross cultural paradigmatic examples is discussed. There are two sections: In the first, various approaches to defining retirement are examined, the basis of their differences are determined and a definition proposed which will provide the key concept for the next section. In the second, we will examine the concept in a comparative framework by setting the American concept of retirement in the context of several other world cultures.

Objective patterns of behavior vs. social action: contrasting approaches to examining retirement

The following section will examine the two main perspectives discernible in comparative studies of retirement, the implications of each for research and theory, and show how one perspective in particular neglects crucial contextual data about behavior patterns.

Contemporary Western retirement is both an individually earned right to a period of leisure after a career of employment, and an age grade social obligation (to younger workers entering the work force; to the demands of a production-efficiency minded economy). Retirement may be defined in the United States as the age-fixed and socially mandated final phase in a career of employment in which a person is excluded from full time career jobs, is entitled to financial support without the stigma of dependency, and is personally responsible for managing his or her own life. Analytically, retirement is significant because it is a rightful entitlement of all adult persons wherein the activities and roles that were always critical to consolidating an identity as a full adult (e.g. working) may be relinquished with no erosion of adult status (Luborsky, 1994).

Modern retirement is a relatively new phenomenon. One organization argues,

“Just a few generations ago, older people could continue to be productive by contributing their insight, wisdom and knowledge of crafts and farming. By the early 1900s, however, America experienced a profound shift in the nature of work. The industrial revolution transformed many craftsmen and farmers into factory workers, and the sense of shared effort that marked the labors of earlier eras disappeared. Workers became, in a sense, replaceable cogs and eventually came to see their jobs solely as a source of income rather than as a way of life (Dumont 1965; 1977). Meanwhile, advances in medical treatment and safety on the job greatly prolonged the average worker’s life span.”

As workers demanded more security from their jobs, private pension plans became commonplace. The first private pension, instituted in 1875, was created by American Express. In 1935, the U.S. Government officially launched the national retirement program by inventing Social Security, thus creating a safety net for all citizens (American Express, 1998).

From an anthropological viewpoint, one of the most striking aspects of retirement in the US is the notion that a person is vested with the right to cease performing roles that are crucial to achieving and maintaining identity as an adult member of society, and yet may retain full social identity and jural-moral capacities. We will see later that this is a rare cultural construct for later life. The manifest change in activities which occur with increasing age in all human societies, including retirement in the West, is the source of differing interpretations by researchers. That is, viewed from an external, behavioral perspective, all humans can be observed to change their pattern of daily activities as they get older—often in a fashion seemingly analogous to the changes associated with retirement. One stream of research regards it as occurring world-wide. In that view, retirement’s current form in the West (institutionalized with pensions and determined by chronological age) is different from its form in other cultures merely in its degree of formal institutional elaboration and in the simple fact that greater numbers of people live long enough to take advantage of it. For most research, the focus of attention in definitions is on the chronological duration of the biological lifespan. Seeing that people everywhere age, and that with aging comes change in physical and cognitive capabilities, it is often (mistakenly) assumed that elders universally “retire” from certain ways of life and assume others.

World wide phenomenon

Retirement has been considered an aging “problem” primarily in industrial societies. However, retirement is not peculiar to these societies. It is generally the case that with diminishing physical powers the aged person becomes less able to perform what the society regards as “normal,” economically gainful, work (Savishinsky 2000). Since he or she can no longer maintain himself, society may choose to find some way to maintain an elderly person if he is to survive at all. The status of retirement and need for protection of the retiree are shown clearly in non-industrial societies (Back, 1977:78). Back argues that the status and meaning of retirement is transformed from its character in non-industrial societies “by the exigencies of an industrial economy.” Notably important influences on the character of retirement in the West are the rapid rate of technological innovation and obsolescence, bureaucratization, large nation states, the increased proportion of aged in our population, and the segregation of work and leisure (p. 82).

Back thus appears to confound the overt patterns of behavior. He confuses the way we would observe and interpret them within our Western ideological frame with the way members of the society in question might interpret them. Clearly, Back’s view obfuscates several aspects of retirement by considering them to be largely determined by the bio-physiological processes of aging and less by cultural constructs (cf. Sussman 1972: 37, 35, 30). Popular literature on retirement more clearly reveals key aspects of the contemporary folk model. Discussions there take the form of argument presented by Back. That is, while there were retirees in more primitive hunting, pastoral, or agricultural societies, their problems were different and generally less acute than those faced by retirees in modern industrial society. This line of reasoning hearkens back to discredited evolutionary theories of culture. It mistakenly claims that various societies’ cultural differences or distinctions represent their varying progression along a single universal line of putative cultural evolution (Katz 1996 p. 61). Several of the popular literature guides on how to prepare for retirement follow this mistaken assumption (Giles 1949, Wallis 1975).

Lastly, articles in the aging literature itself tend to bandy the terms about interchangeably and with imprecision. For instance, Foner and Kertzer (1978) compared the process of life course transitions in twenty-one African age-set societies and American society. In their article the successive relinquishing of roles over the life time of persons in age status systems is vaguely glossed as “retiring.” Even in those societies where “death does not draw the line” and the deceased are reckoned to remain active members of social life in their role as ancestors, the authors speak of the senior elders as being retired by the succession of junior age-sets to their new positions in the social structure. But, we must ask, is that nation of retirement true to the concepts held by the natives under discussion? Is this notion of retirement equivalent to the modern Western cultural concept?

In brief, this framework for conceptualizing retirement has at least three key points. First, retirement is used as a gloss to cover the institutional provisions in all societies for coping with the biophysiological changes that individuals undergo with increasing chronological age. Primary attention is directed to aging per se and the life course changes in general. Second, retirement is used to refer to the relinquishment of any activity or social role—generally, but not necessarily, associated with the passage into later life. There the focus is on changes in patterns of activity as apparent to outside observers. Third, the emergence and elaboration of an identity and role as retiree in the West is attributed to technological, economic, and demographic “pressures,” nonsocial factors, and studied in these terms.

Culture specific

Another stream of researchers regard the Western notion of retirement as a concept to be restricted to contemporary Western culture. Retirement per se— as distinct from the culturally defined and chronologically indexed category of old age—is a distinct cultural category and a social role to which all members of the society are entitled. Entry is carefully regulated. From a sociocultural framework, legitimate entry into this role entitles one to permanently abstain from paid employment, receive financial support from the state without the stigma of dependency, and also have the personal freedom and responsibility for the conduct of one’s life.

Retirement instantiates a relationship between individual citizens and the nation state. It is unmediated by other social structural considerations such as affiliations by kinship, residence, religion, class, or ethnic group. Ideally, none of those factors are relevant to judging entitlement to enter retirement. That is, in the US, all people are theoretically independent and free with equal rights to the leisure and rewards pursuant to completion of their major productive and reproductive tasks. Significantly, retirement is viewed as an entitlement of all workers who have invested their years of labor into the economy and the society at large. It thus contrasts with the similar category of “unemployment”: while retirees may abstain from work, the unemployed are obliged to search for employment; while retirees’ state-supported income (e.g. Social Security pensions) are seen as a “right” that they’ve “earned,” the unemployed who receive financial support from the state are pejoratively labeled as lazy or dependent.

Of course, the actual availability of retirement options and the quality of post-retirement life is not equally distributed across all social strata; clearly, retirement-aged members of some social classes enjoy more opportunities than do their poorer peers. But, as we see later, in many non-Western societies it is only through one’s status and role in a kinship group that practices analogous to retirement is possible.

Preservation vs. erosion of full personhood

There is a second distinctive feature. The concept of retirement creates a social role and identity where persons who have completed roles critical to the consolidation of their social identity as full adults are entitled—even obligated—to cease performing these activities, but theoretically and legally suffer no diminution in their rights as members of that society. Traditionally, for men the key roles are those of earner and provider, and for women, homemaker and mother. In many other cultures the completion of those key social obligations marks the end of full social identity and participation in the society. But in the US, the period after the dispersal of one’s family and the termination of employment signals the beginning of a new life stage. In this stage there is less emphasis on individuation within social groups (e.g. on the role of “father” in a family, or “employee” in a company) and increasing emphasis on the Western notion of a unique, autonomous, highly personalized individual (Dumont 1965, 1977; Luborsky 1994). Atchley (1976:55) proposes that this is a new and unprecedented part of an individual’s life span, the opportunity to exercise simultaneously the most prized of American values: personal autonomy, freedom, and independence.

Clark (1972), Donahue, Orbach and Pollack (1960) epitomize this second approach. Clark describes several cultures and points out how, in most technologically less developed societies, “respected elders” are generally provided with a new set of statuses and less arduous activities in the community which are commensurate with their declining strength. Yet, she is careful to argue, in none of the societies do people retire in the American sense. Simmon’s (1945) more extensive survey of the role of the aged in primitive and ancient societies concluded similarly: “There is no retirement role in previous and non-Western industrialized societies” (p. 104). In contemporary Western civilizations there is far less socially sanctioned expectation that individual retirees should engage in auxiliary jobs related to sacred or mundane areas of life.

Although an obvious feature of retirement in the West is that it occurs among the healthy aged, rather than being correlated with actual physical or intellectual decrements in old age (as it is in most other cultures), Western-style retirement is not wholly limited to industrialized nations. Examples of practices analogous to Western retirement, found in pre-industrial Japan, India, and Burma, are described below.

Donahue, Orback, and Pollack argue that retirement has virtually no precedence in existing or previous forms of social organization, that it represents not just a change in an aspect of life but a new form of social life distinct from previous patterns of old age (1960:331, 334).Although past societies have dealt with aged persons, they have not dealt with retired persons as such. Retirement is a unique socially legitimized relationship between individual citizens, as distinct jural entities, and the nation with which they are affiliated “apart from larger familial, class, or residential units” (Donahue et al (1960:340). In summary, this perspective contrasts markedly with the previous one which found retirement in all societies. This second, more culturally comparative point of view, clearly specifies criteria for defining retirement. As such, it focuses on the public meanings and values embodied in and expressed by the institution of retirement and the identity and interactional role of the retiree. The second perspective on retirement uses a definition which specifically distinguishes it from a broad class of other behaviorally similar processes and identities. The practice of retirement is not reducible to biology. The biophysiological alterations accompanying aging can in no way account for this institution, nor the variety of contrasting customs provided for the aging in various human societies. For example, contrast the roles and attitudes towards the elderly among the Fulani and the Karimojong. These societies are very similar: both are small, technologically simple, pastoral African societies. Among the Fulani (Stenning 1958) adults are deemed “socially dead” upon the marriage of their last child and stripped of virtually all rights and prerogatives. Thus marginalized, they have no voice in planning or regulation of community affairs. But among the Karimojong (Dyson-Hudson 1966), where named groups of males constituted according to closeness in age form the backbone of the social organization, members of older age sets are accorded greater respect and authority (Sangree 1966). Thus the attitudes, beliefs and roles that change as persons become older are coordinated by the cultural trends of that society—in these two instances creating vastly different situations for elderly adults. Biophysiological changes parallel the course of social development, but evidently do not dictate the form and character of social institutions (Katz 1996, Riley 1988).

Why do researchers define retirement so differently? Are these differences due solely to semantic disagreements? For example, when Back (1977) and Foner and Kertzer (1978) speak of people retiring from certain activities and roles throughout life, are they using the term in the same sense as Donahue et al (1960)? The development of arguments in the articles by Back and by Foner and Kertzer clearly indicate that the differences are not due to a simple confusion over terms. Back’s key concept is the process of retiring from “active work.” Significantly, “active work” itself is a culturally ladened notion drawn from the author’s own Western sensibilities. In this view, any status passage may be defined as retirement. In contrast, Donahue et al define retirement in a precise, culture-specific manner, referring to it as a native concept and social practice patterned by cultural tenets. Whereas Back constructs the Western model of retirement as a universal, Donohue et al understand it to be culture-specific and culturally delineated.

The behavioral approach posits that a change in a person’s daily activities (such as from more to less arduous tasks) matched to declining physical abilities necessarily indicates a kind of retirement analogous to that practiced in the West. Yet how does this differentiate the change in activities associated with increasing age? The transitions from childhood to adulthood, bachelor to husband, and son to father—are each of these retirement? Can the concept of retirement used by behavioral scientists be an accurate and useful tool for cross-cultural generalizations, or are their findings restricted to that one-third of the whole human population that resides in modern Western societies? Ad hoc conceptual adjustments and revisions are required for such an approach to answer such crucial questions.

Due to the dearth of detailed empirical data on retirement in non-Western societies, a variety of second hand data is needed. Despite the wealth of large-scale survey and panel-based studies of retirement, little research focusing on the meaning and contexts of retirement is available. Research in this area has been characterized as an “ethnographic vacuum” (Clark 1968:433), sparse and incomplete (Atchley (1976) and afflicted by cultural biases (Ross 1977); unfortunately, despite several excellent recent studies (Graebner 1980, Quadagno 1982, Guillemard 1991), this characterization remains largely accurate.

Analogies and contrasts to the Western concept of retirement: examples from non-Western cultures

Can we properly speak of retirement in other societies? Is there an analog to the Western idea of retirement in other societies? To start, I examine several cases of societies where there seem to be a striking similarity, as well as those where there is a patent contrast with the Western model of retirement.

A careful review of the existing ethnographic literature, including sources in the Human Relations Area Files, was conducted. Information was sought regarding beliefs about the later phases of the life cycle, alterations in jural and moral capacities and duties, the criteria for determining the timing of life course transitions, and the conceptions and social roles for post-parenthood adults. (See Glascock & Finerman (1981) for a critical review of the use of the HRAF for aging research.) It should be noted that the materials described here are drawn second hand from ethnographies written about other topics and thus depend upon the original ethnographers for the accuracy and quality of the data. Further, due to space considerations and because this is a pedagogical review meant to sharpen an analytic definition of retirement, much contextual ethnographic information cannot be included.

“Retirement is inconceivable”: two categories of contrast

Among the peoples of the following cultures, the notion of retirement at any time in the life course is inconceivable. In each group, differing statuses and activities define the crucial components of full adult personhood, yet in none is there the notion that one may still be a full member of the society unless engaged in these key roles. We will illustrate this argument with examples from two contrasting types of cultures in which there is no concept akin to the Western model of retirement: (1) among those groups where adult social identity and effective social participation ceases following a culturally mandated life course transition, and (2) groups where individuals are expected to, and desire to, continue to be fully incorporated in the community and to actively participate in group life. That is, in the first group there is no culturally patterned opportunity for Western-style retirement; in the second, though there is a shift in the areas of activity and tasks for older people, it is not acceptable to retire, as such a cessation of activity runs counter to central cultural values.

Retirement at death: Fulani and Lusi

Perhaps the most dramatic contrast to our notion of retirement occurs in societies such as that of the Fulani. A small West African pastoral society, the Fulani offer their elders what most Westerners would consider to be a bleak lifestyle. Following the marriage of their last child, a couple will transfer away the last of their cattle as part of the marriage transaction. The parents are now regarded as useless, having completed their duties to society in replenishing the stock of human capital. They must live as dependents with their eldest son, they have no voice in planning or executing activities and they retain only marginal rights in the community. The couple is regarded as “socially dead” (Stenning 1970). They must sleep, segregated by sex, on opposite sides of the outer margins of the compound. These quarters communicate their statuses, for they are literally sleeping over their future graves.

A key point is demonstrated here. What is crucial is that upon the marriage of their last child, the capacity of the couple as a socially functioning entity is regarded as terminated. Yet as biological individuals they may still be able to function for years to come. This highlights several operating interests of the social group: to replenish its human capital and to equip its new members with the necessary skills and knowledge to enable the society to continue into the future without interruptions or losses caused by the deaths of elders (Fortes 1958). Thus, the functional viability of a single person is defined by cultural criteria rather than biology.

Among the Lusi of Papua New Guinea, the social place of an elderly person is contingent on his or her independence, productivity and strength of kin relations. Dorothy A. Counts (1991) writes that old age is an exceedingly rare condition among most New Guinean societies and that if a person does reach old age, his or her place within the community is determined by how well he or she can remain independent and productive or claim support from kin. Significantly, the Lusi—who blame nearly all illnesses, accidents and deaths on witchcraft—believe that the only “natural” death is the death of a decrepit, dependent elder (Counts 1991). Thus decrepitude and dependency are equated with social death. Counts presents two compelling cases to illustrate this concept: in one case an elderly women remains active and socially viable, while another woman becomes dependent and is rendered socially nonviable.

The first of these women, a widow named Sapanga Biskit, was both productive and closely tied to her kin group. For example, she cooked and gardened for herself and contributed to her family’s daily well-being by caring for her grandchildren. Her kin ties and continued productivity maintained her social role and secured her place in the social fabric of the family and community.

The second widow, Mary, was blind, and totally dependent on others for help with daily activities. In addition to this physical impairment, she suffered from a social handicap of isolation since her children all lived in distant towns. She was relied completely on the families of her two foster children for her daily care. They viewed Mary as a burden and virtually ignored her, attending to only her most basic needs. One day, Mary was left alone in the village; she apparently tried to stir up the coals of a fire left burning to keep her warm and fell across the live coals. She was severely burned, and later died of her injuries.

These two widows’ situations exemplify how the Lusi are highly attuned to evaluating and treating older members according to the individual’s continued ability to maintain competence at socially normative daily activities and to remain embedded within a web of social ties. The Lusi’s cultural responses to old age further illuminates how the Western model of retirement is far from universal. Among the Lusi there is no socially mandated system of support for all elderly individuals; there is no universal role for the elderly to fill. For aged members of the Lusi community, continued social status is contingent on continued self-sufficiency and/or kin support; the lack of both can spell the end of social—or even biological life.

Shift to less arduous tasks: Andean peasants, Iban, Aymara and Hopi

Another, more frequent situation occurs in societies where the Western notion of retirement is alien, but a similar native institution exists. Four examples, the Andean peasants, the lban the Aymara and the Hopi, are cited to illustrate Simmon’s often quoted summary: “[A]ble bodied persons generally participated at one time or another in subsistence tasks, with increasing age they step aside and shift to auxiliary functions and less arduous tasks” (1945: 104ff).

Retirement is a word unknown to the peasant societies of the remote Andes (Holmberg 1961:89). The people of these societies live by arduous subsistence farming and the raising of small herds. Healthy people continue to labor, although in reduced form, for as long as they are physically able. By tradition, aging couples transfer ownership of their fields on to sons and daughters, but maintain a separate home and reserve rights to a smaller share of the crop to help assure that their own needs are met. The tasks appropriate to older household members include herding, weeding the fields near the house, keeping birds away from maturing crops (Brush, as cited in Gross 1984) and minding the house while younger adults are away on trading expeditions.

Freeman reports that among the lban of Sarawak, elderly women serve a similarly (unarduous) but crucial function in padi (rice) cultivation by “… watching over rice grains laid out in the sun for drying, and in warding off with long bamboo poles the ever-voracious domestic fowls” (1970: 220). In such labor-intensive household economies these are significant tasks that contribute to the continued health of the household.

The Aymara of Bolivia also have no indigenous institution of retirement. An individual remains active in adult capacities until death (Cole 1969). In this model of the life course, only marriage qualifies an individual for adulthood— literally translated, the Aymara term for marriage means “to become a person.” But despite this new “personhood,” for the Aymara economic independence comes only after the death of one’s parents, not after marriage and accession to adulthood. (A similar tradition existed among traditional Irish farmers; see Arensberg and Kimball 1940 and Gordon et al 1981.) In the Aymara culture, increasing age merely brings about a change in activities, but no concurrent notion regarding the decline of the person’s jural rights, social roles or moral capacities or fitness.

Similarly, among the pueblo dwelling Hopi of the American Southwest, the old expressed the desire to “keep working until death” (Simmons 1945:86, 1942). Retirement is impossible at any age; one cannot accumulate credits in order to support one’s self later in life, nor are there surplus food supplies or capital to be regarded as a pension. Even with sons and daughters willing to provide support, the elders continue to work as long as possible. The elderly assume greater importance in ritual affairs, and take on different household and agricultural tasks. Death alone, not retirement, marks the end of active work and effective participation of adults in the community.

In summary, these examples illustrate the second kind of socio-cultural system where retirement is inconceivable. In these cultures elders keep working until death; there are no beliefs that one need only fulfill a stint of employment or that one can cease performing roles central to adult social identity and yet retain full jural status and participation in the family and community. Thus, similar to the first example of the Fulani, there is no socially institutionalized retirement role. There can be no social identity for people within these societies apart from the continuous performance and/or production of key roles and statuses which define adult persons in that society. Among the Fulani all effective social identity and participation ceases after the end of a couples’ child-rearing and cattle-owning period in life. Among the Hopi the end of social identity only occurs at biological death. In the first case, the life-cycle of the social person is too brief to permit a retirement-like phase. In the second set of cases where it is biologically possible, it is not culturally permissible, being regarded as repugnant to fundamental central cultural values regarding production tasks, activity, and social identity.

Shift from economic to spiritual realms: Thai, Ladak, Burmese and Chinese

In contrast to the two types discussed above, there is a third type in which there is a role for adults in later life that is strikingly similar to the Western notion of retirement. Societies in this third category have customary beliefs which prescribe a shift in the arenas of social life and tasks appropriate to older persons. The shift is suggestive of the Western practice of retirement, but is embedded in a different ideology and framework of values regarding the normative life course and the nature of the individual human being in society. After describing each of these as found among the Thai, Ladak, Burmese, and Chinese, we will highlight the two principles which sharply differentiate them from Western practice.

Rustom (1961:128; cf. Vatuk 1980, 1989) describes how Burmese women and men are expected to abstain from economic activity and to devote themselves to religious duties during the last stage of life. They assume important roles in ceremonial and religious observances, and are expected to perform good deeds in the community and become progressively more involved in meditation. Expected to be disinterested in worldly affairs and to “eat in tranquility” (1961) they are provided for by their families. Barring ill health, this is reckoned to be a very happy time of life in Burma. Individuals are relieved of the burden of laboring in economic production, yet ideally there are no decrements in their status, prestige, or influence within the family and community. The pattern in Thailand and traditional China closely parallel this practice.

Lao-nien is the stage of “old age” in the more traditional life-cycle for both old men and old women in China. The institutional expectation is for adults to relax their “serious application to the business of living, and superintending of the living of others” (Levy 1949:83). In traditional China the status of the elderly is preeminent with respect to all other age categories. Adults are regarded as lao-nien from the age of 55, approximately. Levy described this culturally patterned beliefs that the aged are not to be bothered with the details of daily life as de facto retirement (1949:128–9). Time is for the elders’ pleasure and reflection. For example, when important family issues arise, they might be submitted to the male lao-nien who bears the symbolic responsibility for the family, but in actuality the mantle of responsibility falls upon the shoulders of younger adults in the family (Levy 1949). That is, the elders retain jural headship of the family in old age, but delegate the responsibilities and tasks to younger males in their own name. For women this is a period of release from the lifelong arduous routines of work and household management. There is also a lessening of their strict subordination to males. The prescribed change in activities for both men and women seems to mirror that of retirement in America.

A similar portrait can be drawn of the Thai. At age 60 individuals have achieved a place in the eldest age group, the fifth in a series of twelve year cycles. By this time most men have stepped down from active farm life, turning it over to a son or son-in-law (De Young 1955:66ff). Kingshill (1960:71) explains that around the age of 40 a man generally feel it is time to “retire” from the normal chores of tilling the fields provided that he has sufficient land and children to support him. He sends his children to do the work while he stays around the farmyard; he still does a full day’s work, but performs less arduous labor such as repairs, basketry and making netting. At this time in life women are also released from their major task, that of household management. But they continue to be present in household affairs, assisting with the small children, doing household tasks and taking part in the daily routines. As adults in this society approach the end of their biological lives, both men and women are directed to turn an increasing amount of attention and time to the pursuit of sacred and religious activities.

A final example are the Ladak. When the eldest child of a Buddhist Ladak family marries, the parents must leave their home, give up the house and property associated with it, and pass ownership rights over to the son or daughter who has the right to immediately enter into possession of the inheritance (Peter 1948:352). For Ladak adults this change does not have the same negative effect on social status as it would for their Fulani counterparts. Custom forces parents to exit from the active life and responsibilities of their households when they are still in their middle age, both biologically and socially. However, a small part of the estate is made over by the offspring to support their mother and several fathers, and unmarried daughters. In this manner the parents, who retreat to a small house near their former household, are provided with the means for their subsistence (Drew 1875:252).

The similarity in the local patterns of life described for the societies above can be in part attributed to an underlying cultural tradition. The need to approach the study of retirement with adequate data on cultural and ideological contexts is well illustrated in these cases. The Western term “Buddhism” is a simplistic gloss that belies an immensely diverse system of beliefs and practices varying according to its manner of acculturation and to particular local values and institutions, and interactions with other indigenous ideologies. Nevertheless, the ethics and traditions of Buddhism links these societies with regard to basic conceptions of the individual and their relationship with the living and their ancestors. Here it is considered as an underlying cultural tradition and not in terms of members who are formally affiliated with a specific Buddhist sect. Across the wide variation, a few core tenets characteristic of Buddhist life contribute to the cultural patterns of later life. A point of unity is the symbol of Buddha and the central element of personal commitment in faith. It is a universalistic religion, offering salvation to all people regardless of ethnic, geographic or social origin. Strongest emphasis is placed on personal accountability and individual efforts directed at achieving salvation from the human situation. In this system, individuals are believed to have a perpetual soul that undergoes a potentially endless cycle of rebirths in which it assumes a different physical form and status in the next life according to the moral quality of their previous life.

The socially expected life task of a person in this cultural context is to strive to become a self-reliant individual through the search for wisdom, moral uprightness, and meditation. By means of pursing these goals for worldly life, one seeks to achieve a middle road between the two extremes of sensuality and self-mortification. If one is successful, it is possible to achieve a release from the endless cycle of rebirths. Buddhists believe that true happiness is not attainable in this world; thus, the religion’s general orientation is one of earthly transcendence.

Note that this form of mandate differs from the Western belief that we “have but one life to live,” and that retirement is therefore the last and best chance for an individual to attain personal fulfillment and/or leave a mark on the community. Buddhist cultural patterns concerning the life course are thus embedded in a philosophy and way of life with wholly different orientation to the meaning and goals of an individual’s life and place within a socio-temporal context.

Summary and discussion

Much gerontological research has focused on aging while little has focused on the institution of retirement per se. Due to a lack of empirical data (Holmes 1976:63) concerned with the “semantic field and the symbolic world of the subjects without prior commitment to specific concepts of the situation” (Beeson 1975:56,57), there are sizable and serious gaps in our knowledge of retirement and other late-life social roles and cultural phenomena. The Western notion of retirement is complex and ambiguous. It rests in part on the tenet that individuals have earned the right to a period of freedom in the use of their time after the completion of their major social duties. Retirement is the only way individuals may legitimately abstain from work, permanently and of their own volition, with no social opprobrium. In this sense it is a privilege earned through a lifetime of labor. At the same time, retirement is a social obligation rooted in the erroneous but still entrenched folk beliefs that a worker’s abilities and skills decline with increasing age and that it is necessary to remove older workers from the scene in the interests of production efficiency. Retirees are in many respects defined as functionally impaired; they must give up their adult responsibilities when society defines them as “classificatory persons about to die” (Clark 19767:11). Yet this relinquishing of duties upon completion is not accompanied by any diminution of retirees’ jural status as full adults. It is on the other hand, an “opportunity to exercise simultaneously two of the most prized qualities in the Western world, individualism and autonomy” (1976:55), as well as the freedom to decide how to use one’s own talents and time in pursuit of a self-chosen destiny.

We have shown that while there are practices similar to retirement in non-industrialized and non-Western nations, the particular features of the Western notion and institution are unique. Among all the societies described in the second section of this paper there is an observable change in the pattern of daily activities of chronologically defined late middle aged adults which, from a strict behavioralist perspective, might be regarded as identical to the changes typical of retirement in America. These changes mirror those found in the American notion of retirement, as defined in the first section of this essay. Yet these changes in activities must be framed within the specific cultural contexts of the communities in which they occur. The members of the respective social systems interpret them with differing meanings and the changes in life activities are cued to differing cultural criteria. The meanings of these changes in patterns of behavior are culturally dictated, and diverge sharply for the Fulani, Karimojong, the Ladak and Aymara, and the Americans.

The cultural criteria regulating the timing within the social life course of persons vary from society to society: for the Fulani it is the last child’s marriage, for the Hopi it is decrements in physical capability, and for the Americans and Burmese it is the attainment of a certain chronological age. The social implications of the change vary as well: among the Fulani persons are deemed socially dead at old age, in America a second life begins at 65 and among the Thai old age initiates final preparation for the transition to another life course.

With these facts in mind we therefore assert that without attention to the underlying cultural beliefs and social meanings of human behaviors, analyses of retirement will be severely hampered and distorted. The form of the generalizations by Back (1977:81; Atchley 1976) and others quoted in this paper regarding the cross cultural nature of retirement or its development in the West must be more carefully examined for over-generalizations and lack of cross-cultural validity. Such arguments are too often facile and misrepresentative of what we know about the range of human cultural diversity.

Further, as economic globalization becomes increasingly entrenched and as various culture-specific localities are drawn into an ever more connected web of social, cultural and economic links, a cross-cultural perspective of retirement will become more and more vital to those concerned with issues related to aging and the aged. Gerontologists interested in the effects of industrialization and globalization on developing countries are now obliged to examine the intersection of the populations’ unique, culture-specific beliefs and practices related to aging and work. International efforts to integrate the economies of developing nations into the global economic grid must confront entrenched local traditions which differ from Western norms; the concept of retirement is but one example of this. Facile assumptions based on knowledge of Western-style retirement are no longer sufficient; relying on such assumptions only compromises the validity of researchers’ work (see Appadurai 1996, 2000; Becker 2001). Indeed, prior assumptions about homogeneity within industrial nations have given way to a wave of new recognition of ethnic, gender and regional diversity in the timing and attitudes towards productive aging and retirement (Zsembik & Singer 1990, Calasanti 1996).

To summarize, this paper has reappraised the analytic definition of “retirement” in Western society bringing into focus some of its more culturally distinctive features by means of comparisons with other cultures. It first described the contrasts and similarities between the behavioral and the cultural approaches to retirement studies and then specifically outlined how the application of culturally comparative anthropological concepts refines our notions of retirement in an accurate and useful perspective. It accomplished this by clearly differentiating between issues of aging in general retirement per se, and by distinguishing between the cultural ideologies and environmental, economic and demographic conditions that determine value orientations towards elders