| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||||||||||
RESEARCH ARTICLE |
Department of Sociology and Center for Family and Demographic Research, Bowling Green State University, Ohio.
Address correspondence to Dr. Susan L. Brown, Department of Sociology and Center for Family and Demographic Research, Bowling Green State University, Bowling Green, OH 43403. Email: brownsl{at}bgnet.bgsu.edu
 |
Abstract |
|---|
 TOP Abstract MethodsResultsDiscussionReferences |
|---|
Methods. We used data from both Census 2000 and the 1998 Health and Retirement Study (
HRS; Health and Retirement Study, 1998
) to estimate the size and composition of the cohabiting population aged 51 and older. Also, using HRS data, we estimated multinomial logistic regression models to identify the correlates associated with cohabitation and remarriage (vs being unpartnered) among women and men who were previously married.
Results. More than 1 million older adults, composing 4% of the unmarried population, currently cohabit. About 90% of these individuals were previously married. We identify significant differences among cohabitors, remarrieds, and unpartnereds across several dimensions, including sociodemographic characteristics, economic resources, physical health, and social relationships. Cohabitors appear to be more disadvantaged than remarrieds, and this is especially evident for women.
Discussion. Older cohabitors differ from individuals of other marital statuses, and therefore future work on marital status should explicitly incorporate cohabitation.
THE PREVALENCE of heterosexual cohabitation in the United States has increased enormously in recent decades. The number of cohabiting couples grew by a factor of 10 between 1970 and 2000, from about 500,000 to around 5 million (U.S. Census Bureau, 2001). Today, even though only about 45% of all cohabiting unions eventuate in marriage, more than half of all marriages are preceded by cohabitation (Bumpass & Lu, 2000; Smock, 2000).
Most of the public attention surrounding cohabitation, as well as most of the research, has focused on younger adults, who usually cohabit as either a prelude or an alternative to first marriage (Smock, 2000). However, cohabitation is actually more common among the formerly married than the never married (Bumpass & Lu, 2000). According to the 2000 Cencus, of the roughly 10 million individuals currently cohabiting, more than 1 million are older than age 50. It is clear that the rising incidence of cohabitation crosses the age spectrum. Yet with few exceptions (Brown, Bulanda, & Lee, 2005; Chevan, 1996; De Jong Gierveld, 2004; Hatch, 1995; King & Scott, 2005), the research literature has ignored the cohabitation experiences of older adults (Cooney & Dunne, 2001). Research on the living arrangements of older persons (e.g., Angel, Angel, McClellan, & Markides, 1996; Wilmoth, 1998, 2001) has not considered cohabitation.
The objective of this study was to use Census 2000 data and national data from the 1998 Health and Retirement Study to describe the characteristics of cohabitors aged 51 and older and to show how these individuals compare with their contemporaries of different union statuses. As will be shown in the Results section, nearly all older cohabitors have prior marital experience; that is, they are divorced or widowed. Consequently, we will emphasize comparisons across the previously married: cohabitors, remarrieds, and unpartnereds. In other words, to explore the significance of union type, we will investigate how those who are formally remarried differ from those who are cohabiting in informal unions. This research contributes to an understanding of the antecedents and consequences of cohabitation in the later stages of the life cycle, issues that will become increasingly important to gerontologists as the number of older cohabitors continues to rise.
Background
Several demographic processes are responsible for the increase in cohabitation among middle-aged and older persons. A major cause is simple cohort replacement. Substantial increases in cohabitation among younger adults first occurred in the baby boom cohorts. These people are now aging into their later middle-age years, meaning that an increasing share of older adults is likely to be favorably disposed towards cohabitation (De Jong Gierveld, 2004). Moreover, divorce rates have been relatively high among these cohorts, and remarriage rates have decreased (Allen, Blieszner, & Roberto, 2000; Cooney & Dunne, 2001), yielding more adults who are unmarried and available for partnering. Cohabitation is also more common among Blacks and Hispanics than among non-Hispanic Whites (Raley, 1996; Wherry & Finegold, 2004), and these groups are increasing as a proportion of the older population. In consequence, a smaller fraction of the older population will be married in coming decades, and a higher proportion is likely to be cohabiting. Because the older population is growing, this will produce an even greater increase in the number of older cohabitors. There were just more than 25 million Americans aged 55 and older in 1950; by 2000 this number had increased to nearly 60 million (U.S. Census Bureau, 2002). Although growth in this age group has been modest over the past decade, it is expected to accelerate between 2010 and 2020 as baby boomers age into this group. Indeed, Chevan's (1996) estimates of cohabitation among persons aged 60 and older indicate that whereas fewer than 10,000 were cohabiting in 1960, more than 400,000 were cohabiting by 1990. Brown and colleagues (2005) report that more than 1.2 million persons aged 50 and older were cohabiting in 2000; we calculate that more than 500,000 of these were aged 60 and older.
Like the demographic trends, Americans' attitudes and values regarding cohabitation are changing. Research shows a dramatically increased tolerance of unmarried cohabitation among both younger and older persons (Thornton & Young-DeMarco, 2001), a trend that appears to comprise both cohort and period effects. Well over a decade ago, Bulcroft and Bulcroft (1991) reported that unmarried adults older than age 60 were as interested in cohabiting as in remarrying. These trends point to an increasing diversity of living arrangements among the older population that includes, notably, cohabitation (Cooney & Dunne, 2001).
From a theoretical standpoint, cohabitation likely has a unique meaning and plays a different role in the life course of older adults than it does in that of younger adults (Chevan, 1996; Hatch, 1995; King & Scott, 2005). Research on dating and remarriage among older adults suggests that theories developed to explain these behaviors among young and middle-aged adults are inadequate as they ignore the significance of life-course stage in structuring opportunities and outcomes (e.g., Bulcroft & Bulcroft, 1991; Bulcroft, Bulcroft, Hatch, & Borgatta, 1989). Similarly, the motivations for cohabitation among the older population are likely to differ from those of young adults.
Several researchers assert that older adults, particularly women, are not especially interested in remarriage (Bulcroft & Bulcroft, 1991; Bulcroft et al., 1989; Chevan, 1996; Hatch, 1995; Talbott, 1998). Chevan (1996) maintains that the disincentives for remarriage are greater among older adults, who are especially likely to have economic resources whose value may be undermined by remarriage. For instance, provisions governing social security and pension receipt typically depend on one's marital status. Also, adult offspring may encourage their parents to cohabit rather than to remarry in order to protect their estates (i.e., the offspring's inheritances).
In summary, there are important demographic and theoretical reasons for investigating the composition of the older adult cohabiting population and for assessing how this group compares with other union statuses. Despite the growth of the older population, which is increasingly composed of unmarrieds who may enter into cohabitation, research on this topic is scarce and dated. For instance, Chevan's (1996) study of cohabitors older than age 60 relied on decennial census data from 1960–1990, and Hatch's (1995) research on older cohabitors used census data from 1980. Not only were these analyses conducted at a time when cohabitation was much less common, but both studies were also undermined by significant data constraints, including the availability of only indirect measures of cohabitation and a very narrow set of sociodemographic correlates to explore. Until 1990, cohabitation had to be inferred from census data; respondents were not directly asked whether they were currently involved in an intimate, heterosexual, coresidential relationship (Casper & Cohen, 2000). Moreover, Chevan's analyses compared cohabitors only to other unmarrieds; marrieds were excluded. A few recent analyses have incorporated cohabitation into studies of mental health (Brown et al., 2005) and repartnering following marital dissolution in The Netherlands (De Jong Gierveld, 2004) among older adults, but these studies do not describe the older adult cohabiting population and how it compares with other union statuses.
The Present Study
The present study updates and extends the limited prior work in this area by drawing on two data sources. We will construct a detailed portrait of older cohabitors, providing explicit comparisons to those in other union statuses (i.e., marriage, remarriage, separation or divorce, widowhood, and the never-married state). By examining older adults' characteristics across these union statuses, we will document the composition of the older adult cohabiting population and the extent to which these older adults are similar to (and different from) those in other union statuses. We will investigate multiple features of older adults' lives according to union status, emphasizing factors that were shown in prior research to be related to union status among older adults: (a) demographic characteristics, (b) economic resources, (c) physical health, and (d) social relationships.
Demographic characteristics
Union status is associated with individual demographic characteristics, including gender, age, and race. Union formation and dissolution processes differ for men and women, particularly among the older population. This is largely an artifact of the skewed sex ratio: Unmarried women outnumber unmarried men by a ratio of 2.5:1 among the 55 and older population (U.S. Census Bureau, 2002: Table 48). This imbalance is exacerbated by gender-specific mate selection strategies, whereby men tend to partner with younger women whereas women often partner with older men. These dynamics indicate that older men will be more likely to form and maintain unions than women. Indeed, older cohabitors are also disproportionately men. Men compose only about 25% of the elderly unmarried population but account for nearly 60% of cohabitors older than age 60 (Chevan, 1996). Among older unmarrieds, age is negatively associated with cohabitation (Chevan, 1996; De Jong Gierveld, 2004). In the general population, cohabitation is especially common among African Americans and Hispanics (Raley, 1996). As the older population becomes more racially and ethnically diverse with the changing composition of the American population and the lengthening life expectancies of non-Whites, there is no reason not to expect that African Americans and Hispanics will continue to be more likely than Whites to cohabit. Among older persons, Blacks are more likely to reside in cohabiting unions than Whites, but Hispanics do not differ from Whites (Chevan, 1996; Hatch, 1995).
Economic resources
Among younger persons (especially men), economic stability is associated with marriage versus cohabitation (Clarkberg, 1999; Oppenheimer, 2003; Smock, Manning, & Porter, 2005; Xie, Raymo, Goyette, & Thornton, 2003). Cohabitation is particularly common among those with lower levels of education, although its popularity has increased among individuals of all educational strata. Among older persons, the association between education and cohabitation is unclear. Hatch (1995) documents a negative relationship between education and cohabitation, but Chevan (1996) finds no significant association. In the present study, we tested whether education was negatively related to cohabitation among older adults, as it is for younger adults. Chevan's analysis of cohabitation among older unmarried adults shows that poverty is positively related to cohabitation. Nonetheless, he also finds that labor force participation is positively associated with cohabitation among men, perhaps because it serves as a proxy for health and social integration. Or, economically active men may be more attractive partners. In contrast, Hatch's analysis of 1980 Census data indicates that cohabiting older men are less likely to be working and earn smaller incomes than either married or single older men. Women exhibit a distinct pattern; cohabitors have higher levels of employment than marrieds or singles. Older cohabiting women report higher earnings than single women but lower earnings than married women. Additional economic factors associated with cohabitation among older adults include renting (vs owning) a home and receipt of entitlement income (Hatch, 1995).
Physical health
Despite being younger, on average, than marrieds, cohabitors report poorer physical health (Brown et al., 2005). Relative to marriage, cohabitation may be selective of individuals in worse health. Those in poor health may be less attractive as potential spouses, leaving these less healthy individuals at risk of a cohabiting union, which typically requires a weaker commitment from a partner. Among older men, poor health is positively related to cohabitation versus being single (Hatch, 1995), perhaps because older men in ill health are seeking support. This finding is consistent with Talbott's (1998) assertion that older women are reluctant to assume the burden of caregiving that may follow from remarriage at older ages, particularly when men's traditional familial obligation of economic provision ends in old age. Existing studies of caregiving, even those focused on ethnic minorities among whom cohabitation is more common, essentially ignore cohabitation (e.g., Pinquart & Sorenson, 2005).
Social relationships
There are competing hypotheses regarding the association between social relationships and cohabitation among older adults (Hatch, 1995; Talbott, 1998). The compensatory hypothesis suggests that those individuals who do not receive adequate support from friends and family members will be especially motivated to form a union to achieve support through an intimate partner. Alternatively, the complementary hypothesis states that those individuals who enjoy the highest levels of social interaction with friends and family will be most interested in—and effective at—forming an intimate partnership, which these individuals view as offering unique, nonoverlapping forms of support. Widows with close friends are more interested than those without close friends in forming a new partnership, supporting the complementary approach to social relationships (Talbott, 1998). Additional support for this hypothesis comes from a study of dating among older unmarrieds that documents a positive relationship between organizational participation and dating (Bulcroft & Bulcroft, 1991). To the extent that social relationships enhance individual well-being and promote social integration, we expected that they would be positively associated with cohabitation. Religious commitment and affliliation are negatively associated with cohabitation not only because some religions frown upon living together outside of formal marriage, but also because religious persons tend to espouse more traditional, conservative family attitudes that are disapproving of cohabitation (Clarkberg, Stolzenberg, & Waite, 1995; Thornton, Axinn, & Hill, 1992). One study of repartnering among older adults conceptualized the absence of religious affiliation as an indicator of individualism, which was positively associated with forming a cohabiting union (De Jong Gierveld, 2004). Older adults' relationships with their children may influence their own relationship trajectories. Close relationships with children may weaken an older adult's interest in forming a cohabiting (or marital) union. Or, children may pressure parents to avoid the legal complexities that would accompany remarriage and may suggest cohabitation as an alternative (Bulcroft & Bulcroft, 1991; Chevan, 1996; Hatch, 1995). Indeed, Hatch contends that children actively discourage their widowed mothers from remarrying, and De Jong Gierveld posits that the parent–child relationship weakens following repartnering. Widowed older women with children are less likely than their childless counterparts to remarry, but children are not associated with older men's remarriage patterns (Wu, 1995).
In summary, we expect cohabitors, remarrieds, and those who were previously married but are now unpartnered (i.e., the divorced and widowed) to differ in terms of demographic characteristics, economic resources, physical health, and social relationships. Additionally, these union status differences are likely to be distinct for men and women. We extend prior work on remarriage following divorce or widowhood by including cohabitation as a competing risk.
|
METHODS |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
). Designed to examine health and retirement decision making as well as how older adults and their families respond to declining health in later life, the HRS is a nationally representative sample comprising 21,384 noninstitutionalized persons born in 1947 and earlier. Thus, at the time of the 1998 interview, respondents were as young as 51 years old. Of the 21,384 HRS respondents, 33 were eliminated from our sample because we were unable to determine their current marital status. Another 21 respondents were eliminated because they reported being in same-sex partnerships (although substantively of interest, the small number of cases prohibits meaningful analysis of this group). Finally, an additional 681 respondents were excluded because they did not have a sample weight (typically because they were institutionalized). We also limited our sample to individuals aged 51 and older, which resulted in the loss of an additional 922 respondents (i.e., younger spouses or partners of primary respondents), for a final sample size of 19,727.
Measures
Union status
HRS respondents reported their union status at the time of interview. Approximately 2.2% of the sample was cohabiting, yielding 437 cohabitors for analysis. About 48.7% of the respondents (n = 9,617) were in first marriages, 19.8% (n = 3,899) were widowed, 16.5% (n = 3,255) were in remarriages, 9.9% (n = 1,948) were separated or divorced, and the remaining 2.9% (n = 571) had never married.
Demographic characteristics
Gender was a dichotomous variable (1 = women). Race/ethnicity was composed of four dummy variables: non-Hispanic White (reference), non-Hispanic Black, Hispanic, and non-Hispanic other race. Age was coded in years. Previously widowed (a dummy variable coded 1 for those whose marriage ended through death of the spouse and 0 for those whose marriage ended in divorce) was controlled for in the analyses that were restricted to the previously married.
Economic resources
Several indicators of respondents' economic resources were included. Education was coded in years, ranging from 0 (no formal education) to 17 (post-college; i.e., more than 17 years of education). Household income was a constructed measure in the HRS data set that incorporated bracketed income responses using sophisticated imputation techniques. It was logged to minimize the effects of skewness. A dichotomous measure indicated whether the respondent owned his/her home (1 = yes). Employment status was dummy coded full-time employment (reference), part-time employment, unemployed, and not working. Retirement status indicated whether the respondent was retired (1 = yes). This measure was distinct from the employment status variable because some respondents were technically retired but working in a different job. We also included measures to tap receipt of Social Security and a pension (1 = yes). Health insurance coverage is gauged by four mutually exclusive dummy variables: private insurance (reference), Medicare, other insurance (e.g., Medicaid), and no insurance.
Physical health
The activity of daily living (ADL) measure is a scale composed of the responses to 11 items asking the respondent whether, because of health problems, he or she has any difficulty with various ADLs (e.g., walking several blocks, sitting for about two hours, picking up a dime from a table). For each question, answers of yes are coded 1 and answers of no are coded 0. Responses are then added together to form the scale. Alcohol measured whether and how much the respondent drank. Those who indicated that they did not drink or have never drunk alcohol (more than 50% of the sample) were coded 1; those who responded that they drank alcohol but less than once a week were coded 2; those who responded that they drank alcohol at least once a week but not daily were coded 3; and those who reported drinking alcohol every day were coded 4.
Social relationships
Social relationships were gauged by several measures. Neighborhood friends was coded 1 if the respondent reported having good friends living in the neighborhood and 0 otherwise. Relatives in the neighborhood was coded 1 if the respondent reported at least one relative residing in their neighborhood and 0 otherwise. (The term neighborhood is not explicitly defined by the HRS.) Religiosity is a measure of the reported significance of religion in the respondent's life, ranging from 1 = not too important to 3 = very important. Children is coded 1 if the respondent reported having at least one living biological child and 0 otherwise.
Analytic Strategy
We began by establishing the size and marital status distribution of the older adult cohabiting population using the Census 2000 data. These data provide a benchmark against which to compare the figures we obtained from the 1998 HRS. Next, we employed the HRS to compare older adults across six union statuses in terms of several sets of characteristics that are related to union status, including demographic characteristics, economic resources, physical health, and social relationships. Because roughly 90% of HRS cohabitors were previously married (i.e., had been divorced or widowed), the most appropriate comparison group for ascertaining the significance of union type was not individuals in marriages but rather those in remarriages. Previously published literature (e.g., Bulcroft et al., 1989; Burch 1990; Talbott 1998) has examined middle-aged and older adults' decisions to remarry following divorce or the death of a spouse, and we extended this line of inquiry by treating cohabitation and remarriage as competing risks. Thus, our final set of analyses was restricted to those who were previously married. Detailed comparisons by union type (i.e., cohabitation vs remarriage) were examined separately for women and men. We estimated multinomial logistic regression models of the competing risks of being in either a cohabiting union or remarriage (vs being unpartnered; i.e., being divorced or widowed) separately for women and men. Our objective is to describe the population of older cohabitors, so we are less concerned about causal order here; the set of independent variables likely includes both antecedents and consequences of union type. This analysis shows most clearly how older, previously married persons who cohabit differ from those who are remarried or unpartnered. These are critical comparisons because it is unclear whether cohabitation is better conceptualized as an alternative to marriage or to remaining unpartnered among older persons who have experienced divorce or widowhood. The complex sample design of the HRS means that the sample is not self-weighting. Thus, standard errors need to be adjusted to correct for design effects. Both descriptive and multivariate analyses were conducted in Stata (StataCorp LP, College Station, TX) using the survey procedure to correct for the complex sample design.
|
RESULTS |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
|
Table 2 shows the weighted means of the variables by union status. As anticipated, older adult cohabitors are less likely than first marrieds and remarrieds to be White and are more likely to be Black. Cohabitors are also younger, on average, than all other groups except the divorced. Although cohabitors have a greater likelihood of being employed full time, the household incomes of older cohabitors, first marrieds, and remarrieds do not differ significantly. Also, older cohabitors are less likely than either group of marrieds to own a house or to have health insurance. About 13% of cohabitors have no health insurance, and another 25% rely solely on Medicare. Although cohabitors, first marrieds, and remarrieds experience similar levels of ADLs (levels that are lower than those reported by divorced, widowed, and never marrieds), cohabitors report significantly higher average levels of alcohol consumption than any other group. In terms of social relationships, cohabitors do not fare especially well. They are least likely of any group to report having friends or relatives in the neighborhood (although they do not differ from the divorced in their reports of having friends or from remarrieds in their reports of relatives nearby). This suggests that cohabitors are more isolated socially. Cohabitors also report lower levels of religiosity than any other union status group. Not surprisingly, although cohabitors are slightly less likely than both groups of married persons to have living children they are much more likely than the never married to have children. Taken together, cohabitors appear most similar to the divorced group, which is consistent with our decision to emphasize here comparisons among those who had been previously married.
|
|
|
|
DISCUSSION |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
; Chevan, 1996). Nonetheless, the cohabitation experiences of older adults have been largely overlooked (Cooney & Dunne, 2001). Combining data from Census 2000 and the 1998 HRS, we constructed a national portrait of older adult cohabitors that reveals both the characteristics of this group and how this group compares with older adults in other union statuses. The data reported here suggest that, in terms of social and demographic characteristics, older cohabitors and remarried persons differ in many of the same ways that younger cohabitors and married persons do. Blacks are less likely to be married. Cohabitors (particularly women) have lower incomes and are less likely to own their homes than are remarried persons. They are also more likely than remarried to use alcohol. Finally, they score lower than older remarrieds on most of the measures of social relationships used in this study, including religiosity and having friends and relatives who live nearby.
In some respects, cohabitors compare favorably to the unpartnered. Cohabitors' household incomes are higher, and they are more likely to be employed full time and to have private health insurance than are the unpartnered. On the other hand, cohabitors (especially women) are more likely than the unpartnered to have no health insurance at all. Cohabitors also score lower than the unpartnered on all of the measures of social relationships used in this study, except having children.
The picture that emerges from these comparisons is one of fairly systematic disadvantage among older cohabitors, especially in comparison to their remarried counterparts. As is the case among younger persons (Clarkberg, 1999; Oppenheimer, 2003; Smock, 2000; Smock et al., 2005), cohabitation may be chosen by individuals whose economic circumstances make marriage more risky. This raises the question of whether cohabitation has similar consequences to remarriage for older persons. Studies show, for example, that older married persons (especially men) are more likely than unmarried persons to engage in healthy behaviors (Schone & Weinick, 1998); researchers do not know yet whether the same is true of cohabitors.
The disadvantage associated with cohabitation appears especially pronounced among women, who may derive fewer benefits than men do from cohabitation. Among men, cohabitors are more likely than unpartnereds to own a home; among women, cohabitors and unpartnereds do not differ in this respect. Among women, the proportion of cohabitors with no health insurance is 3 times larger than that of remarrieds and unpartnereds; among men, there are no significant union status differences in this respect. Although cohabiting and remarried men enjoy statistically similar high household incomes, cohabiting women report an average household income of roughly two thirds that of remarried women. It is therefore not surprising that more cohabiting women than either remarried or unpartnered women are employed full time (among men, cohabitors and remarrieds are equally likely to work full time). Taken together, our findings indicate that union status may be more consequential for women than men. Simply put, there are fewer union status differences among men than among women. As researchers continue to explore the gendered consequences of remarriage, divorce, and widowhood, our results reveal the importance of incorporating cohabitation into this research, too.
There are still many unknowns about later life cohabitation. Cohabitation, of course, entails less commitment than remarriage. Does this lesser commitment allow cohabiting persons to "bail out" when their partners become infirm or disabled, and do they take advantage of this possibility? Do older cohabitors share their resources in the same ways that married couples do, or do they hoard their own because the union may be fragile? This is important because cohabitors tend to have fewer resources per capita than remarried persons. To the extent that cohabitation is an alternative to remaining unpartnered, it may be a key source of support for older persons. Similar to marrieds, cohabitors presumably enjoy access to a regular sex partner. To the extent that cohabitation is a substitute for remarriage among the divorced and widowed, it may make older adults' lives less secure and their support networks more vulnerable to disruption.
In order to fully ascertain the implications of cohabitation for policy and practice, researchers need answers to at least some of these questions. For some older persons, a cohabiting partner may be a critical resource, providing help in times of need, material support that alleviates the strains of poverty, and emotional support to assuage loneliness. Or the partner may disappear just when the need for support is greatest. Until researchers know more about what motivates older persons to cohabit, as well as the factors that predict entry into and exit from cohabitation, it is difficult to accurately assess the implications of cohabitation.
As the older adult population continues to grow, and especially as the baby boomers age into this group, it becomes increasingly important that researchers assess the ramifications of cohabitation among older adults. Baby boomers were the first generation to cohabit in significant numbers as young adults, and consequently society can anticipate an increase in older adult cohabitation as the boomers age into older adulthood. Unfortunately, baby boomers are not represented in the 1998 HRS, which is a sample of persons born in 1947 and earlier. Early boomers (those born in 1948–1953) will be folded into the 2004 panel of the HRS.
The present study indicates that cohabitation among older adults is more common than researchers realized and is likely to increase in the future. We are not alone in this prediction; others researchers have noted that a declining share of older adults will be married in the coming decades, and thus a larger share will be at risk for cohabitation (e.g., Allen et al., 2000; Cooney & Dunne, 2001; De Jong Gierveld, 2004). The determinants of the formation and dissolution of older adult cohabiting relationships await future research. In the meantime, the present analyses demonstrate that cohabitors are a distinct group, comparable neither to remarrieds nor to other unmarrieds. As described in this article, the ramifications of this emerging relationship form for the well-being of older adults are not entirely clear. Moreover, because we were not able to ascertain whether older adult cohabitors weighed the decision to cohabit against remaining unpartnered or getting remarried, we compared cohabitors to both of these groups. Whether cohabitation typically operates as an alternative to singlehood or as a substitute for marriage is not readily apparent, because cohabitors—especially women—are unlike those in other union statuses across several key dimensions, including several indicators of economic resources and social relationships. For this reason, future gerontological research would benefit from expanding the formulation of marital status to include cohabitation so that researchers can begin to uncover the implications of cohabitation for the health and well-being of older adults.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received for publication June 23, 2005. Accepted for publication November 9, 2005.
|
References |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  H. Moustgaard and P. Martikainen Nonmarital Cohabitation Among Older Finnish Men and Women: Socioeconomic Characteristics and Forms of Union Dissolution J Gerontol B Psychol Sci Soc Sci, July 1, 2009; 64B(4): 507 - 516. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
