Marital Biography,Social Security Receipt,and Poverty

Abstract

Increasingly, older adults are unmarried, which could mean a larger share is at risk of economic disadvantage. Using data from the 2010 Health and Retirement Study, we chart the diverse range of marital biographies, capturing marital sequences and timing, of adults who are age eligible for Social Security and examine three indicators of economic well-being: Social Security receipt, Social Security benefit levels, and poverty status. Partnereds are disproportionately likely to receive Social Security and they enjoy relatively high Social Security benefits and very low poverty levels. Among singles, economic well-being varies by marital biography and gender. Gray divorced and never-married women face considerable economic insecurity. Their Social Security benefits are relatively low, and their poverty rates are quite high (over 25%), indicating Social Security alone is not sufficient to prevent these women from falling into poverty. By comparison, gray widoweds are the most advantaged singles.

Keywords

gender gray divorce marital biography poverty social security

Dramatic changes in U.S. family life are evident across the life course. Perhaps the most fundamental shift is that a growing share of midlife and older adults is unmarried (Lin & Brown, 2012; Sassler, 2010). Although widowhood is less common nowadays, gray divorce is on the rise (Brown & Lin, 2012). And, those who form unions in later life are increasingly likely to cohabit rather than marry (Brown, Bulanda, & Lee, 2012; Vespa, 2012). This retreat from marriage among middle-aged and older adults has wide-ranging consequences for individual health and well-being (Dupre & Meadows, 2007; Hughes & Waite, 2009; Pienta, Hayward, & Jenkins, 2000). An extensive literature links marriage with numerous benefits, and foremost among these is financial resources (Wilmoth & Koso, 2002; Zissimopoulos, 2013). The declining share of older adults who are married raises critical questions about the role of changing marital biographies for later life economic well-being.

As more adults confront old age alone, how do they fare economically? In particular, to what extent do public programs such as Social Security help to mitigate economic disadvantage for unmarried older adults? The increasingly varied marital biographies of the older adult population are potentially altering usage patterns for various federal programs such as Social Security (Iams & Tamborini, 2012; U.S. Government Accountability Office, 2014). At the same time, Social Security was designed during an era when most elders were married, a scenario that is less common today and is likely to be even less typical in the future. In fact, the decline in marriage is linked to reduced spouse and widow benefit eligibility for Social Security among women (Tamborini, Iams, & Whitman, 2009). This is particularly concerning, given the outsize role that Social Security plays in economic well-being for many Americans, particularly those who are poor. Moreover, poor elders are disproportionately single women (Torres, 2014).

Using data from the 2010 Health and Retirement Study (HRS), we address how marital biographies, capturing marital sequences and timing, are related to Social Security benefits and to poverty. Our goal is to describe marital biographies for adults who are age eligible for Social Security benefits (i.e., over age 62), emphasizing the diversity among the unmarried population. We move beyond the traditional, narrow emphasis of much gerontological research that has focused on marriage and widowhood to encompass other unmarrieds: cohabitors, the separated and divorced, and the never married (de Jong Gierveld, 2004). In addition to capturing the sequencing of marital status by examining marital history pathways (e.g., remarried following divorce vs. widowhood), we consider whether the timing of marital dissolution matters by differentiating between dissolutions that occurred before versus after age 50. This approach allows us to evaluate the potential impact of a gray divorce, which is an increasingly common but still off-time event, versus divorce that occurs earlier in the life course. Similarly, we can evaluate whether early widowhood (prior to age 50), a nonnormative event, is more detrimental to economic security in later life than widowhood after age 50. Unmarrieds are examined separately by gender because unmarried women tend to have lower economic well-being than unmarried men (Holden & Smock, 1991).

Background

Older adults enjoy the lowest overall levels of poverty largely because of the social insurance provided by programs such as Social Security, which are critical to economic well-being, especially for those who are economically disadvantaged. Nevertheless, poverty levels vary by marital status among late middle-aged and older adults with singles 4 times as likely to be poor as marrieds (Holden & Kuo, 1996; Lin & Brown, 2012). Further, single women are more likely to be poor than single men. These differentials partially reflect the rules governing the distribution of Social Security benefits, which rest on the now outmoded assumption that nearly all adults marry and remain married until spousal death. Couples who divorce can access spousal benefits, provided they were married at least 10 years and they do not remarry before age 60. But, the share of divorced women who qualify for spousal benefits is projected to decline for successive cohorts (Tamborini & Whitman, 2010). This is especially troubling because among older unmarried women, the divorced are most likely to be poor, followed by never marrieds and lastly widows (Butricia & Smith, 2012). And, we can expect the shares of single women who are either divorced or never married will increase in the coming years (Lin & Brown, 2012). Lowering this 10-year marriage rule by just a few years would substantially reduce poverty levels among low-income divorced women, according to a recent study (Tamborini & Whitman, 2010). The salient role of Social Security for the economic well-being of the most vulnerable elders underscores the need to examine how older adult marital biographies are linked to not only Social Security receipt and benefit levels but also poverty.

Marital Biographies

Prior research demonstrates that older adult economic well-being varies by marital status (Gustman & Juster, 1995; Seigel, 1993), but the work to date has not adequately captured the full range of marital histories characterizing older adults (Holden & Kuo, 1996; Wilmoth & Koso, 2002; Zissimopoulos, Karney, & Rauer, 2015). Moreover, most research on the enduring legacy of marital transitions is now rather dated (drawing on the 1992 HRS) and thus is not necessarily reflective of the financial circumstances of today’s older adults who have more diverse marital biographies compared with earlier generations (although see Addo & Lichter, 2013). Prior work has focused on sequencing of marital events but ignored the timing of transitions. Finally, very few studies have examined either Social Security or poverty status according to marital biography. Our goal is to consider how various marital biographies that encompass both the sequencing and timing of marital transitions are related to economic security among older adults.

From a life-course perspective, we anticipate that one’s marital biography is more salient for well-being in later life than a simple measure of current marital status because marital biography captures the sequencing and timing of life events that contribute to the accumulation of (dis)advantage (O’Rand, 1996; Wilmoth & Koso, 2002). One’s marital biography is also associated with eligibility for various types (e.g., spousal, survivor) of Social Security benefits, shaping the magnitude of the overall level of benefits received. Marital biography is relevant for poverty status. For example, among the currently married, those ever widowed are more likely to be poor than those in a first marriage. There is no comparable penalty for ever divorced currently marrieds (Holden & Kuo, 1996).

Partnered Older Adults

A majority of older adults are married and marriage is associated with economic advantage. The continuously married enjoy higher levels of economic well-being than remarried couples in late midlife (Addo & Lichter, 2013; Holden & Kuo, 1996; Wilmoth & Koso, 2002). Unlike those who have experienced marital dissolution, the continuously married have had a lifetime together to accumulate resources. In contrast, those in remarriages have had to surmount the financial shock that often accompanies marital dissolution. The remarried are also disadvantaged compared with the continuously married because of their shorter average marital durations and competition for resources that support prior unions. Nevertheless, all marrieds presumably benefit from economies of scale. In fact, men and women can often achieve predivorce levels of financial well-being through repartnering (Jansen, Motelmans, & Snoeckx, 2009; Wilmoth & Koso, 2002).

The number of older adults who coreside with a partner outside of marriage is skyrocketing with close to 3 million individuals aged 50 and older cohabiting in 2010 versus less than 1 million in 2000 (Brown et al., 2012). Cohabitation allows older adults to enjoy many of the benefits of marriage without the legal constraints (Chevan, 1996; Hatch, 1995). In particular, individuals point to the economic autonomy afforded through cohabitation that enables them to protect their assets for the next generation. Cohabitors can continue to receive a former spouse’s benefits, such as Social Security, that would not be available were they to remarry. One study indicated that cohabitors may be largely comparable to their married counterparts in terms of economic well-being (Brown, Lee, & Bulanda, 2006), although another found that cohabitors were more economically disadvantaged, on average, than either continuously married or remarried couples (Wilmoth & Koso, 2002). The prevalence of poverty among older cohabitors appears similar to that for single older adults (Chevan, 1996).

Unpartnered Older Adults

Roughly one in three baby boomers is unmarried in contrast to just one in five midlife adults in 1980 (Lin & Brown, 2012). Even at midlife, unmarrieds tend to be worse off compared with their married counterparts. The share of baby boomers living in poverty is nearly 5 times higher among unmarrieds (19%) than marrieds (4%). Whereas marrieds enjoy a median household income of over US$105,000, unmarried boomers are in households with a median income of about US$57,000. There was some variation among unmarrieds, with widowed women and never married men most vulnerable, on average (Lin & Brown, 2012).

The marital status composition of older adults has shifted in recent decades such that a larger proportion is divorced and a smaller share is widowed (Manning & Brown, 2011). Early research indicated that divorce and widowhood have comparable, negative ramifications for economic well-being in late midlife (Holden & Kuo, 1996). Couch, Tamborini, Reznik, and Phillips (2013) considered the earnings trajectories and initial Social Security benefits experienced by women who divorced prior to age 41 during the 1970s. They found that women who divorced and did not remarry enjoyed a higher earnings trajectory than either continuously married women or divorced women who remarried. Divorced women who did not remarry claimed Social Security benefits at later ages, on average, and had the highest initial benefit level too. Nonetheless, the overall economic picture for divorced women remains grim since they do not have a spouse who is also drawing Social Security. Continuously married and remarried women typically have more economic resources because of the husband’s contributions.

Prior research on economic well-being in later life has overlooked gray divorce (i.e., divorce after age 50), an increasingly common event. Gray divorce has doubled since 1990 even though the overall U.S. divorce rate remains stable (Brown & Lin, 2012). Gray divorce is more common among remarrieds, the less educated, and those with fewer economic resources. The consequences of gray divorce are unknown, but it is likely that this experience could have devastating financial consequences for those who have been out of the labor force (e.g., stay-at-home wives) or have few economic resources.

Timing

The timing of marital dissolution in the adult life course may have implications for postdivorce adjustment, including late life economic well-being. Divorce tends to be more normative at younger ages, whereas widowhood becomes increasingly likely with age. From a life-course perspective, the timing of an event can magnify or reduce its influence on well-being (Elder, 1994; Uhlenberg, 1996). Off-time events are associated with poorer outcomes than on-time events. Thus, divorce prior to age 50 may be less detrimental to economic well-being than divorce after age 50. Those who divorce earlier in adulthood have more time to recoup the financial losses divorce usually entails. In contrast, those who divorce later have fewer years of working life remaining and may not be able to fully recover economically from a gray divorce. Indeed, gray divorce appears to diminish wealth more than an earlier divorce (Zissimopoulos, 2013). Similarly, widowhood prior to age 50 is an off-time event that is not a normative life-course experience. Young widows are more likely to become poor compared with older widows (Holden, Burkhauser, & Feaster, 1988). Couples tend to be overly optimistic about the likelihood they both will survive to an old age (Holden & Kuo, 1996). Thus, they may not have adequately planned for this unlikely possibility and ultimately may be less able to recover fully.

The Role of Gender

Women are economically disadvantaged compared with men and this gender disparity widens with age. Despite rising female labor force participation in recent decades, a majority of women receive Social Security through spousal or widow benefits rather than on the basis of their own contributions (Meyer, Wolf, & Himes, 2006). Given the broader retreat from marriage, fewer older women may be eligible for spousal or widow benefits whether because they divorced less than 10 years into their marriage, they did not remarry, or they never married in the first place. These shifts in women’s marital biographies are linked to reduced benefits levels (Tamborini et al., 2009). For these reasons, we anticipate that marital biographies will play a larger role in the economic well-being of older women than men, an expectation that aligns with results from prior studies (Holden & Kuo, 1996; Wilmoth & Koso, 2002).

The Present Study

The current study documents the complex marital biographies characterizing today’s older adults and addresses how these biographies are linked to economic well-being in later life. Specifically, we examine the proportions of age eligible older adults who receive Social Security benefits and the average level of annual benefits by marital biography. Additionally, we consider poverty status because Social Security benefits play a large role in the economic well-being of those who are poor. Social Security benefits keep many older adults from falling into poverty. Our analyses control for factors linked with both marital biography and economic well-being, including age, race and ethnicity, education, work history, current work status, health status, home ownership, and receipt of Social Security prior to full retirement age. Also, we include a measure of whether the individual was married for at least 10 years (i.e., eligible for spousal Social Security benefits) and the presence of any dependents. Additional analyses address whether gender modifies the association between marital biography and economic well-being because prior work demonstrates women are more disadvantaged by divorce and widowhood than are men (Wilmoth & Koso, 2002).

Method

Data used in the analysis come from the HRS, a longitudinal study of a nationally representative, continuous cohort of individuals born before 1960. The original HRS cohort consists of respondents who were aged 51–61 in 1992. The panel has been refreshed every 6 years—in 1998, 2004, and 2010—with a new sample of persons aged 51–56. The HRS respondents and their spouses/partners are interviewed every other year until their death. The response rates for the original baseline interviews hover around 70–80%. The reinterview response rates are around 90% in each wave. We drew most information from the RAND HRS file (Version M, 2013) and combined it with the original HRS data to construct respondents’ marital biographies. The HRS is arguably the best available data source to study marital biography and economic well-being in late life because respondents are asked at initial interview about up to three previous marriages and their subsequent union transitions are tracked over time. The study also includes detailed questions about Social Security receipt and income.

The analytic sample is restricted to the HRS respondents and their spouses/partners who were aged 63 and older in 2010 because Social Security receipt is measured for the past year and to be eligible individuals would have to be at least 62 years of age. In 2010, 20,508 HRS respondents and their spouses/partners were interviewed, of which 10,134 were younger than age 63 (49%). We further excluded 33 respondents who were in a same-sex relationship (<1%), 370 respondents for whom their marital biographies cannot be ascertained from the data (<2%), and 322 respondents who moved into nursing homes by 2010 (<2%). In total, the analysis consisted of 9,649 individuals.

Measures

Marital biography was assessed using respondents’ reports on their previous marriages and subsequent union transitions. When respondents first joined the study, they were asked about their marital history. Ever-married respondents reported the beginning and ending years/months of three previous marriages and the most recent marriage.¹ The HRS also collected information on how each previous marriage ended—through divorce or widowhood. At each subsequent interview, respondents were asked whether their spouse/partner at the previous interview was still their current spouse/partner. If not, the ending year/month of the union and how the union was ended were recorded. For respondents who had not been partnered in the previous wave but were currently partnered, the starting year/month of the current union was asked. Based on the retrospective reports on previous marriages and the prospective follow-up on subsequent union transitions, we created 11 groups of respondents with various marital biographies: continuously marrieds, remarrieds after divorce (before vs. after age 50), remarrieds after widowhood (before vs. after age 50), cohabitors, widoweds (before vs. after age 50), divorceds (before vs. after age 50), and never marrieds.²

Social Security receipt (yes = 1, no =0) and benefit level were measured in 2010 dollars, including income from Social Security retirement, spousal, or survival benefits. For respondents who were partnered, this measure combined both respondents’ and their spouse/partner’s Social Security receipt and benefits.

Poverty status (yes = 1, no = 0) captures whether income for the last calendar year was below the poverty threshold, using the U.S. Census definition and poverty thresholds. HRS income was measured at the household level, including both respondent’s and spouse/partner’s earnings, capital income, employer pension or annuity, Social Security disability or Supplemental Security income, Social Security, unemployment or workers compensation, and other sources of income. The poverty measure adjusted for household size.

Other covariates comprise age (in year), gender (women = 1, men = 0), race and ethnicity (White, Black, Hispanics, and others), education (attended college = 1, did not attend college = 0), the number of years worked, current work status (working = 1, not working = 0), poor health status (fair or poor health = 1; excellent, very good, and good health = 0), length of the longest marriage is at least equal to 10 years (yes = 1, no = 0), home ownership (yes = 1, no = 0), presence of any family members who depended on the respondent for more than half of their support (yes = 1, no = 0), and whether respondents received Social Security before full retirement age (yes = 1, no = 0).

Analytic Strategy

Four analyses were conducted. First, we unpacked the complexity of marital biography by illustrating the various marital status pathways in old age. Second, we examined the proportion of individuals who receive Social Security, the mean value of Social Security benefits, and percentage poor across the 11 marital biography groups, paying attention to gender differences among the unpartnered. Third, we provided descriptive statistics for all sociodemographic characteristics according to marital biography. Fourth, we estimated a logistic regression model predicting receipt of Social Security to assess whether marital biography variation is an artifact of sociodemographic characteristics (full sample, N = 9,649 households). Similarly, we conducted ordinary least squares regression to estimate Social Security benefit levels among individuals who received Social Security (n = 8,965). We also used logistic regression to predict the likelihood of being poor across marital biographies, net of sociodemographic factors (N = 9,649).³ Because less than 2% of the analytic sample is missing in at least one of the measures, we used mode and mean to impute missing value in categorical and continuous variables, respectively. We also weighted all analyses to adjust for HRS’s complex survey design; the clustering of respondents and their spouse/partner; and the oversample of African Americans, Hispanics, and Florida residents (Willis, 1999).

Results

Among individuals who were age eligible for Social Security benefits in 2010, their marital biographies are diverse. As shown in Figure 1, nearly two thirds of individuals are partnered, with 96.6% married and 3.4% in a cohabiting union. Of those who are married, 70.1% are in a continuous first marriage and 29.9% are in a remarriage. The pathway to remarriage is more likely to have been through divorce (85.3%) than widowhood (14.7%). Divorce is more common than widowhood for those who experienced marital dissolution prior to age 50 (86.1% vs. 44.9%). Of those who are in a cohabiting union, roughly 6 in 10 are divorced, 3 in 10 are widowed, and 1 in 10 are never married. Divorced individuals are more often remarried than cohabiting if their divorce occurred prior to age 50 (86.1% vs. 41.6%), but they are more often cohabiting than remarried if their divorce occurred after age 50 (58.4% vs. 13.9%). A similar pattern is found among widoweds. Turning now to the more than one third of individuals who remain unpartnered, 53.3% are widowed, 32.9% are divorced, and 13.8% are never married. Remaining unpartnered is far more common for individuals who were widowed than for individuals who divorced after age 50 (89.2% vs. 38.6%). Marital biographies are distinctive for men and women, as illustrated by the gender-specific figures provided as online supplemental files (available at http://journals.sagepub.com/doi/suppl/10.1177/0164027516656139).

Figure 1.

Partnership pathways among people aged 63 and older in 2010 (unweighted N = 9,649).

Table 1 shows how older adults classified across 11 distinct marital biography categories fare on several indicators of economic well-being. The table presents the percentages of individuals who receive Social Security. And, for those who receive Social Security, it depicts the mean value of Social Security benefits. It also shows the percentage poor.

Table 1.

Social Security and Poverty by Marital Biography Among HRS Respondents Aged 63 and Older in 2010 (Unweighted N = 9,649).

Marital Biography	% Receive Social Security	Mean Social Security^b	% Poor	Unweighted N
Partnered
(1) Continuously married^a	90.86	22,607	3.39	4,220
(2) Remarried after early divorce^a	87.85	20,807¹	3.10	1,297
(3) Remarried after gray divorce^a	83.39¹	22,062	3.33	264
(4) Remarried after early widowhood^a	93.76	22,072	1.52	146
(5) Remarried after gray widowhood^a	97.53^1,3	22,830	0.98	247
(6) Cohabiting^a	93.37³	20,721	3.76	234
Unpartnered
(7) Early divorce (before age 50)	78.19	12,092	16.23	547
Women	77.07	11,518	18.56	424
Men	80.88	13,403	10.66	123
(8) Gray divorce (after age 50)	79.58	12,332	18.83	420
Women	81.73	10,995	26.89	229
Men	77.59	13,633	11.39	191
(9) Early widowhood (before age 50)	88.75	12,805	17.19	196
Women	88.11	12,745	17.52	183
Men	97.90	13,575	12.44	13
(10) Gray widowhood (after age 50)	92.81⁸	13,800⁸	13.32⁸	1,830
Women	93.51	13,462	13.68	1,438
Men	90.33	15,026	12.06	392
(11) Never married	76.42^9,10	11,804¹⁰	19.84¹⁰	248
Women	75.62	11,651	25.32	145
Men	77.34	11,974	13.60	103

Note. Numbers are weighted using HRS respondent-level weights. Superscript numbers indicate that the group significantly differs from the numbered group at p < .05 level. Bolded numbers indicate that the group significantly differs from the gray divorce group within the same gender at p < .05 level. Underlined numbers indicate that women and men in the same marital groups are significantly different at p < .05 level. HRS = Health and Retirement Study.

^aFor partnered households, both respondents’ and their spouses’/partners’ Social Security benefits are considered. ^bThe estimates are restricted to respondents who reported that they and/or their spouses/partners received Social Security (unweighted n = 8,965).

About 91% of continuously married couples receive Social Security. Compared with continuously marrieds, a smaller percentage of remarrieds after gray divorce (83%) and a larger percentage of remarrieds after gray widowhood (98%) receive Social Security. Relative to those who remarried following a gray divorce (83%), larger shares of those who remarried following gray widowhood (98%) or who are cohabiting (93%) receive Social Security benefits. Overall, the unpartnered are less likely to receive Social Security than their respective remarried counterparts. Relative to gray widoweds (93%), smaller shares of gray divorceds (80%) and never marrieds (76%) receive Social Security.

For those receiving Social Security, the mean benefit values are comparable between continuously marrieds and remarrieds except for one group: Remarrieds after an early divorce receive a lower mean level of benefits than continuously marrieds (US$20,807 vs. $22,607), suggesting an enduring influence of divorce across the life course. Among unpartnereds, there is considerable variation in benefits levels, with gray widoweds (US$13,800) enjoying higher benefits than either gray divorceds (US$12,332) or never marrieds (US$11,804). Some notable gender differences emerge among unpartnered individuals. Women’s benefits (US$10,995) are lower, on average, than men’s (US$13,633) among those who experienced a gray divorce. And among gray widoweds, widowers tend to receive higher benefits than widows (US$15,026 vs. $13,462).

Turning now to poverty status, unmarrieds are clearly at a higher risk of poverty than are marrieds. Poverty levels for marrieds are quite low, hovering around just 1–3%, illustrating the role Social Security plays in lifting older adults out of poverty. Nearly 4% of cohabitors are poor. By comparison, much larger shares of various unmarried groups are poor. Among the unpartnered, notable differences in poverty emerge that mirror those obtained for Social Security: Relative to gray widoweds (13%), larger shares of gray divorceds (19%) and never marrieds (20%) are poor.

The feminization of poverty is evident with a whopping 27% of gray divorce women poor. Just 11% of gray divorce men are in poverty. Gray divorce women are much more disadvantaged than other unmarried women. The percentage of gray divorce women living in poverty is higher than that for early divorce women (19%) and gray widowed women (14%). Gray divorced women (27%) are comparable to never-married women (25%).

The means or percentages of the individual’s sociodemographic characteristics are shown in Table 2 for partnereds and Table 3 for unpartnereds. The average age of continuously marrieds is 71. Those who remarried following an early divorce tend to be younger, while those who remarried after gray widowhood are older. Among those remarried after a gray divorce, most are men; just 28% are women. Similarly, 36% of those remarried after gray widowhood are women. The modal racial and ethnic category across all couple households is White, although cohabitors (11%), early widoweds (22%), and never marrieds (18%) are disproportionately Black. About one half of continuously marrieds have at least some college. Cohabitors tend to be less educated than marrieds and remarrieds after divorce. Early and gray widoweds are less educated at about one third with some college. Years worked ranges from lows of about 33 years for early and gray widoweds and remarrieds following early widowhood to highs of 41 years for those remarried after a gray divorce. About 28% of continuously marrieds are currently working. Those who remarried after gray widowhood or who are gray widoweds are unlikely to be working at about 15%. There is some marital biography variation in poor health, ranging from just 22% of continuously marrieds to about one third of early and gray widoweds and early and gray divorceds reporting poor health. By comparison, only 27% of never marrieds report poor health. Nearly all individuals have (had) a marriage that lasted at least 10 years, even among cohabitors and divorceds. About 7% of partnered have a dependent versus 5% of unpartnereds. More than three quarters of partnereds received Social Security prior to full retirement age, although the shares among cohabitors and remarrieds following a gray divorce were comparatively modest at 70% and 66%, respectively. Among unpartnereds, smaller shares of never marrieds (59%) and early (61%) and gray (64%) divorceds received Social Security early compared with early (72%) and late (73%) widoweds.

Table 2.

Weighted Means or Percentages of Sociodemographic Characteristics of HRS Partnered Respondents Aged 63 and Older in 2010.

	Total	Continuously Married	Remarried After Early Divorce	Remarried After Gray Divorce	Remarried After Early Widowhood	Remarried After Gray Widowhood	Cohabiting
		(1)	(2)	(3)	(4)	(5)	(6)
Age	70.79	71.01	69.22¹	71.81²	72.16²	76.56^1,3,4	70.40^4,5
Women	44.16	45.34	43.81	27.67^1,2	57.87^1,2	35.87^1,4	37.53^1,3,4
Race and ethnicity
White	86.15	85.85	88.48	84.43	83.99	89.84⁴	77.85^1,2,5
Black	5.46	4.77	6.69	6.63	7.11	3.24⁴	10.80^1.2.5
Hispanic	6.61	7.43	3.72¹	7.02²	6.67	5.16	8.88²
Others	1.79	1.95	1.10	1.91	2.23	1.76	2.48
Some college or higher	50.26	49.45	54.86¹	60.83¹	38.85²	42.79³	38.52^1,2,3
Years worked	37.95	37.41	39.68¹	40.90¹	34.39^1,2	38.22⁴	36.83^2,3
Currently working	28.44	27.38	33.10¹	33.05	26.74	15.39^1,3,4	25.83⁵
Poor health	21.64	20.86	22.77	21.10	25.46	24.09	26.81
Length of the longest marriage ≥ 10 years	98.07	98.84	98.24	98.67	97.71	100.00	79.74^1,2,3,4,5
Owns home	82.34	83.02	83.82	81.18	86.93	84.15	56.67^1,2,3,4,5
Any dependent	7.43	7.48	8.07	7.34	6.59	4.80	5.03
Receives Social Security before FRA	77.94	79.85	75.66	66.36¹	78.60	71.85¹	69.60¹
Unweighted n	6,408	4,220	1,297	264	146	247	234

Note. Numbers are weighted using HRS respondent-level weights. Superscript numbers indicate that the group significantly differs from the numbered group at p < .05 level. FRA = full retirement age; HRS = Health and Retirement Study.

Table 3.

Weighted Means or Percentages of Sociodemographic Characteristics of HRS Unpartnered Respondents Aged 63 and Older in 2010.

	Total	Early Divorce (7)	Gray Divorce (8)	Early Widowhood (9)	Gray Widowhood (10)	Never Married (11)
Age	72.86	69.30	71.10⁷	72.00⁷	75.57^8,9	70.73^7,10
Women	70.07	70.57	48.00⁷	93.49⁷	77.78^8,9	53.27^7,9,10
Race and ethnicity
White	78.54	76.64	79.22	64.19⁷	82.08⁹	74.54^8,10
Black	13.30	15.90	11.59	21.59⁷	10.26⁹	18.10^8,10
Hispanic	6.27	5.17	8.62	7.54	6.16⁸	5.57
Others	1.88	2.29	0.58	6.68⁷	1.50⁹	1.79⁹
Some college or higher	40.37	46.77	47.19	30.03⁷	33.80⁸	51.69^9,10
Years worked	35.26	37.31	37.63	33.07⁷	33.14⁸	38.31^9,10
Currently working	22.39	30.16	27.45	24.55	14.76^8,9	31.75¹⁰
Poor health	31.25	32.27	31.93	35.07	31.47	26.80
Length of the longest marriage ≥ 10 years	76.83	68.77	93.55⁷	80.79⁷	97.65^8,9	0.00
Owns home	59.25	54.36	48.18	58.43	65.84⁸	54.21¹⁰
Any dependent	5.12	6.35	6.05	5.99	5.14	2.06^7,8,10
Receives Social Security before FRA	67.63	61.27	64.47	71.76⁷	73.10⁸	59.33^9,10
Unweigthed n	3,241	547	420	196	1,830	248

Table 4 shows the multivariate models testing whether the marital biography differentials in economic well-being are due to compositional factors. Social Security receipt varies by partnership status, net of sociodemographic factors. Relative to the continuously married, remarrieds and cohabitors are similarly likely to receive Social Security with one exception: those who are remarried following gray divorce. The odds of receiving Social Security are about two thirds lower for remarrieds after gray divorce compared with continuously marrieds. In fact, remarrieds after gray divorce appear more similar to all unpartnered groups, whose odds of Social Security receipt tend to hover around one third of that enjoyed by the continuously married. The other covariates operate largely as expected, with the likelihood of Social Security receipt positively associated with age, being a woman, years worked, and negatively related to education and current work status. Contrary to our expectations, Social Security receipt is negatively related to poor health.

Table 4.

Weighted Regressions of Social Security Receipt (Logistic), Social Security Level (OLS), and Poverty (Logistic) on HRS Respondent’s Sociodemographic Characteristics (Unweighted N = 9,649).

	Social Security Receipt	Social Security Level^a		Poverty
	Odds Ratio	Coefficient	SE	Odds Ratio
Marital biography
Partnered
(1) Continuously married (reference group)
(2) Remarried after early divorce	0.91	−1,761**	592	1.08
(3) Remarried after gray divorce	0.37***²	−906	941	1.14
(4) Remarried after early widowhood	1.29	−441	1,145	0.40
(5) Remarried after gray widowhood	1.10	−379	968	0.28*
(6) Cohabiting	1.73³	−917	1,094	0.70
Unpartnered
(7) Early divorce	0.36***	−9,662***	543	4.20***
(8) Gray divorce	0.28***	−9,576***	436	5.56***
(9) Early widowhood	0.45*	−8,899***	984	3.31***
(10) Gray widowhood	0.30***	−8,872***	271	3.16***⁸
(11) Never married	0.34**	−9,246***	618	5.23***¹⁰
Age	1.37***	94**	26	0.99
Women	2.01***	1,006***	141	1.10
Race and ethnicity
White (reference group)
Black	0.84	−2,065***	386	3.12***
Hispanic	0.88	−2,965***	530	4.08***
Others	1.00	−986	1,508	1.20
Some college or higher	0.70**	1,186***	287	0.49***
Years worked	1.02***	50***	7	0.99***
Currently working	0.35***	−872*	353	0.30***
Poor health	0.72*	−850**	285	1.53**
Length of the longest marriage ≥ 10 years	1.20	1,195**	437	0.88
Owns home	0.98	1,449***	354	0.66**
Any dependent	0.76	−1,310	686	1.02
Receives Social Security before FRA	—	−84	416	0.60**
Constant	0.00**	11,433***	2,214	0.21*
F statistics	31.00	118.88		27.25

Note. Superscript numbers indicate that the group significantly differs from the numbered group at p < .05 level. FRA = full retirement age; OLS = ordinary least square.

^aOLS regression includes only respondents who received Social Security (unweighted n = 8,965).

*p < .05. **p < .01. ***p < .001.

Taking into account sociodemographic characteristics, there is little variation in Social Security benefits levels for partnered individuals. The one difference is that those who remarried after an early divorce receive fewer benefits than the continuously married. Relative to the continuously married, all unpartnereds receive smaller benefits. The associations of other covariates with Social Security benefits are consistent with our expectations. Age, gender, education, the length of employment history, a marriage lasting at least 10 years, and home ownership are positively associated with Social Security benefits. Minority status, current employment, and poor health are negatively related to Social Security benefits.

The logistic regression predicting poverty status yields similar conclusions in that the unpartnered face considerably higher odds of poverty than do marrieds and cohabitors. And, gray widoweds have lower odds of being poor than do gray divorceds. Never marrieds have higher odds of poverty than gray widoweds but do not differ from either early or gray divorceds. Gray widoweds who remarried have lower odds of being poor than the continuously married. The covariates operate in the expected directions, with minorities and poor health positively associated with poverty and education, years worked, currently working, home ownership, and receiving benefits prior to full retirement age negatively related to poverty.

Finally, we consider interactions between gender and marital biography with results shown in Table 5. All possible interactions were included in the models, but only those for the unpartnereds are shown in the table. We use gray divorced women as the reference category because they are arguably the most disadvantaged group (based on results shown in Table 1). The odds of Social Security receipt are largely comparable among unpartnered groups and only one significant gender difference within marital biography type emerges: Gray widows are much more likely to receive Social Security than gray widowers. But Social Security benefit levels vary such that relative to gray divorced women, all other unpartnereds, excluding early divorced women, experience significantly higher average Social Security benefit levels. Gender differences within marital biography type are few. The only other marital biography type in which there is a gender gap is early divorceds, with women faring worse than men. The pattern of findings for poverty status is similar, although never marrieds are no less likely to be poor than gray divorced women. The odds that gray divorced men are poor are about 67% lower than the odds for gray divorced women. Otherwise, there are no significant gender differences within marital biography type. Together, these marital biography and gender differentials underscore the economic disadvantage characterizing gray divorced women and to a lesser extent never married women and men.

Table 5.

Weighted Regressions of Social Security Receipt (Logistic), Social Security Level (OLS), and Poverty (Logistic) on HRS Respondents’ Sociodemographic Characteristics With Gender-Marital Biography Variation (Unweighted N = 9,649).^a

	Social Security Receipt	Social Security Level^b		Poverty
	Odds Ratio	Coefficient	SE	Odds Ratio
Unpartnered × Gender
(7) Early divorce, women	1.21	687^c	592	0.57*
Early divorce, men	1.47	3,072**^c	1,048	0.26**
(8) Gray divorce, women (reference group)
Gray divorce, men	0.73	2,425***	638	0.33**
(9) Early widowhood, women	1.87	2,336*	1,099	0.38**
Early widowhood, men	4.80	3,002*	1,223	0.23*
(10) Gray widowhood, women	1.41^c	2,085***	408	0.34***
Gray widowhood, men	0.59^c	2,690***	496	0.41**
(11) Never married, women	1.09	1,785*	839	0.66
Never married, men	1.11	1,853*	761	0.40
F statistics	24.60	85.11		23.32

Note. HRS = Health and Retirement Study; OLS = ordinary least square.

^aRespondents’ age, gender, race and ethnicity, education, number of years working, current work status, health, marriage duration, home ownership, any dependent, whether receiving Social Security before the full retirement age, and the interaction terms between gender and partnered groups are included in the model but are not shown. ^bOLS regression includes only respondents who received Social Security (unweighted n = 8,965). ^cWomen and men in the same marital groups are significantly different at p < .05 level.

*p < .05. **p < .01. ***p < .001.

Discussion

Our primary goal was to establish how the complex marital biographies of older adults are linked to key markers of economic well-being. Social Security is critical to keeping many older adults from falling into poverty. However, our results indicate that Social Security receipt and benefit levels as well as poverty vary dramatically by marital biography. This means that the social safety net is more effective for some groups than others. Social security receipt is more common, and benefits levels are significantly higher for couples than for singles. This difference reflects in part the number of adults in the household but recall many singles (primarily women) receive spousal or survivor benefits rather than benefits based on their own work history (Tamborini et al., 2009). Similarly, poverty is more widespread among singles than partnereds. Poverty levels for singles range from 13% to 20% versus a mere 1–4% for partnereds.

Among partnereds, those who are married enjoy largely comparable benefits regardless of whether they are in a first or remarriage. Moreover, whether the remarriage followed a divorce or widowhood (and the timing of that marital dissolution) is largely immaterial (although remarrieds following gray divorce are less likely to receive Social Security). For marrieds then, marital status arguably trumps marital biography. Cohabitors appear similar to continuously marrieds and remarrieds.

There is considerable variation in economic well-being among the unpartnered, illustrating the utility of capturing the marital biographies of singles. Relative to other singles, gray widoweds are most likely to receive Social Security and they receive the highest average levels of benefits, which is not surprising given that this program is designed to assist long-term marrieds who experience spousal death. This advantage extends to poverty status. Gray widoweds are much less likely to be poor than either never marrieds or gray divorceds. In fact, gray divorceds appear to be the most vulnerable group of singles with relatively low Social Security benefits and high poverty levels. This is a significant policy concern because even if the gray divorce rate does not rise in the next 20 years (an arguably conservative assumption), the number of older adults experiencing gray divorce is projected to increase by one third due to the aging of the population (Brown & Lin, 2012), resulting in a larger share of singles at risk of economic insecurity.

Gender plays a role in the economic well-being of singles. Gray divorced women face considerable economic disadvantage. They receive smaller Social Security benefits, on average, than all other single women and men. They also confront exceptionally high poverty levels at roughly 27%. By comparison, gray divorced men’s poverty is 50% lower. Gray widowed women also enjoy a 50% lower poverty rate. Among men, economic well-being is comparable for gray divorceds and gray widowers. Thus, it is gray divorced women who are left behind when a marital dissolution occurs in late life. Divorce timing matters for women: Early divorced women enjoy a 30% lower poverty rate than gray divorced women. Either early divorce is less financially devastating than gray divorce or early divorced women are able to recoup divorce-related financial losses over time. Nearly all gray divorceds were married at least 10 years, meaning they are eligible for spousal benefits. Yet, the Social Security benefits received by gray divorced women are insufficient to keep them out of poverty.

Gray divorced women are the most economically disadvantaged group, but never-married women are also economically insecure. The two groups are comparable in terms of poverty and mean Social Security benefits level. These stark findings elucidate vulnerable groups of singles. Traditionally, widowed women garnered concern; but for contemporary cohorts, it appears that two growing demographic groups—gray divorceds and never marrieds—are at greatest risk of economic insecurity. The fact that gray widowed women fare better than gray divorced and never-married women illustrates the effectiveness of social programs in minimizing economic disadvantage.

At the same time, these differentials expose gaps in current policy, as social programs have not been adequately reformulated to accommodate shifting demographics. Without changes to the eligibility rules for Social Security, we anticipate a growing share of older adults, especially women, will be economically disadvantaged in the coming years. Social Security is predicated on the traditional marital biography of long-term marriage followed by spousal death, yet this pathway is experienced by a declining proportion of older adults. Alterations to the eligibility rules could substantially improve the economic well-being of older adults, providing a stronger safety net for vulnerable elders regardless of their marital biographies. Modest rule changes, such as reducing the 10-year marriage rule by just a couple of years, have the potential to significantly improve the economic well-being of older adults (Tamborini & Whitman, 2010).

This study is enhanced by careful attention to the marital biography, but it has some limitations. Perhaps a key shortcoming is that we cannot directly account for selection into various marital biographies. For example, married couples who are economically disadvantaged are more likely to get divorced. And, the propensity of divorced individuals to remarry and the timing of remarriage also vary by sociodemographic characteristics and other unmeasured factors. Although we accounted for several possible confounding variables, the differences documented here may be attributable to selection rather than marital biography per se. Another limitation is that we do not account for couple-level marital biographies for marrieds and cohabitors. Future work should aim to decipher both the respondent’s and the spouse’s or partner’s marital biographies. Also, our approach assumes that couples commingle their finances and share resources. This is an entirely reasonable assumption for married couples but questionable for cohabitors, given that many choose to cohabit so they can keep their assets separate (Brown et al., 2006; Chevan, 1996). Also, our study is not generalizable to the oldest old (i.e., ages 85+) because of the cohort design of HRS. Only about 10% of persons aged 63 and older are over age 85 (authors’ calculations using American Community Survey data), and thus their inclusion would be unlikely to substantially change the results. Still, to the extent that they are disproportionately gray widowed women, their inclusion may alter the pattern of economic well-being observed for this group.

Our article adds to the growing literature on how the increasingly diverse marital biographies of older adults, especially women, are weakening the safety net and financial security of a growing share of elders. The Social Security system was designed during an era marked by nearly universal marriage and low levels of divorce. Protections are lacking for a growing share of adults, including never marrieds, those divorced but married less than 10 years, and even cohabiting couples. These gaps in protection correspond with the marital biographies of those groups that are disproportionately likely to be poor and thus rely heavily on Social Security benefits to make ends meet. In particular, gray divorced and never-married women confront high poverty levels despite receiving Social Security benefits. Moreover, both groups are expected to increase as a proportion of the older adult population in the coming years. In short, more older Americans are economically insecure, without an adequate safety net to respond to major life events, such as the onset of disability or chronic disease, that can be financially devastating. Future research should address how marital biographies are linked to the health and well-being of older adults.

Footnotes

Authors’ Note

An earlier version of this article was presented at the Social Insurance and Lifecycle Events among Older Americans Conference on December 5, 2014.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by a grant from the National Institute on Aging (R15AG047588) and by the BGSU Center for Family and Demographic Research, which is funded by the Eunice Kennedy Shriver National Institute of Child Health and Human Development (R24HD050959). Any opinions expressed here are solely those of the authors and not of the funding agency or center.

Notes

References

Addo

F. R.

Lichter

D. T.

(2013). Marriage, marital history, and Black–White wealth differentials among older women. Journal of Marriage and Family, 75, 342–362.

Brown

S. L.

Bulanda

J. R.

Lee

G. R.

(2012). Transitions into and out of cohabitation in later life. Journal of Marriage and Family, 74, 774–793.

Brown

S. L.

Lee

G. R.

Bulanda

J. R.

(2006). Cohabitation among older adults: A national portrait. Journal of Gerontology: Social Sciences, 61, S71–S79.

Brown

S. L.

Lin

I.-F.

(2012). The gray divorce revolution: Rising divorce among middle-aged and older adults, 1990-2010. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences, 67, 731–741.

Butricia

B. A.

Smith

K. E.

(2012). The retirement prospects of divorced women. Social Security Bulletin, 72, 11–22.

Chevan

(1996). As cheaply as one: Cohabitation in the older population. Journal of Marriage and the Family, 58, 656–667.

Couch

K. A.

Tamborini

C. R.

Reznik

G. L.

Phillips

J. W. R.

(2013). Divorce, women’s earnings, and retirement over the life course. In Couch

K. A.

Daly

M. C.

Zissimopoulos

J. M.

(Eds.), Lifecycle events and their consequences: Job loss, family change, and declines in health (pp. 133–157). Stanford, CA: Stanford University Press.

de Jong Gierveld

(2004). Remarriage, unmarried cohabitation, living apart together: Partner relationships following bereavement or divorce. Journal of Marriage and Family, 66, 236–243.

Dupre

M. E.

Meadows

S. O.

(2007). Disaggregating the effects of marital trajectories on health. Journal of Family Issues, 28, 623–652.

10.

Elder

G. H.

Jr (1994). Time, human agency, and social change: Perspectives on the life course. Social Psychology Quarterly, 57, 4–15.

11.

Gustman

A. L.

Juster

F. T.

(1995). Income and wealth of older American households: Modeling issues for public policy analysis (No. w4996). Cambridge, MA: National Bureau of Economic Research.

12.

Hatch

(1995). Aging and cohabitation. New York, NY: Garland.

13.

Holden

K. C.

Burkhauser

R. V.

Feaster

D. J.

(1988). The timing of falls into poverty after retirement and widowhood. Demography, 25, 405–414.

14.

Holden

K. C.

Kuo

H. H. D.

(1996). Complex marital histories and economic well-being: The continuing legacy of divorce and widowhood as the HRS cohort approaches retirement. The Gerontologist, 36, 383–390.

15.

Holden

K. C.

Smock

P. J.

(1991). The economic costs of marital dissolution: Why do women bear a disproportionate cost? Annual Review of Sociology, 17, 51–78.

16.

Hughes

M. E.

Waite

L. J.

(2009). Marital biography and health at mid-life. Journal of Health and Social Behavior, 50, 344–358.

17.

Iams

H. M.

Tamborini

C. R.

(2012). Implications of marital history change on women’s eligibility for social security wife and widow benefits, 1990-2009. Social Security Bulletin, 72, 23–36.

18.

Jansen

Mortelmans

Snoeckx

(2009). Repartnering and (Re)employment: Strategies to cope with the economic consequences of partnershop dissolution. Journal of Marriage and Family, 71, 1271–1293.

19.

Lin

I.-F.

Brown

S. L.

(2012). Unmarried boomers confront old age: A national portrait. The Gerontologist, 52, 153–165.

20.

Manning

W. D.

Brown

S. L.

(2011). The demography of unions among older Americans, 1980-present: A family change approach. In Settersten

R. A.

Jr Angel

J. L.

(Eds.), Handbook of sociology of aging (pp. 193–210). New York, NY: Springer.

21.

Meyer

M. H.

Wolf

D. A.

Himes

C. L.

(2006). Declining eligibility for social security spouse and widow benefits in the United States? Research on Aging, 28, 240–260.

22.

O’Rand

A. M.

(1996). The precious and the precocious: Understanding cumulative disadvantage and cumulative advantage over the life course. The Gerontologist, 36, 230–238.

23.

Pienta

A. M.

Hayward

M. D.

Jenkins

K. R.

(2000). Health consequences of marriage for the retirement years. Journal of Family Issues, 21, 559–586.

24.

Sassler

(2010). Partnering across the life course: Sex, relationships, and mate selection. Journal of Marriage and Family, 72, 557–575.

25.

Seigel

J. S.

(1993). A generation of change: A profile of America’s older population. New York, NY: Russell Sage Foundation.

26.

Tamborini

C. R.

Iams

H. M.

Whitman

(2009). Marital history, race, and social security spouse and widow benefit eligibility in the United States. Research on Aging, 31, 577–605.

27.

Tamborini

C. R.

Whitman

(2010). Lowering Social Security’s duration-of-marriage requirement: Distributional effects for future female retirees. Journal of Women & Aging, 22, 184–203.

28.

Torres

(2014). Aging women, living poorer. Contexts, 13, 72–74.

29.

Uhlenberg

(1996). Mutual attraction: Demography and life-course analysis. The Gerontologist, 36, 226–229.

30.

U.S. Government Accountability Office. (2014). Retirement security: Trends in marriage and work patterns may increase economic vulnerability for some retirees. Report to Chairman, Special Committee on Aging, U.S. Senate. Retrieved from http://www.gao.gov/assets/670/660202.pdf

31.

Vespa

(2012). Union formation in later life: Economic determinants of cohabitation and remarriage among older adults. Demography, 49, 1103–1125.

32.

Willis

R. J.

(1999). Theory confronts data: How the HRS is shaped by the economics of aging and how the economics of aging will be shaped by the HRS. Labour Economics, 6, 119–145.

33.

Wilmoth

Koso

(2002). Does marital history matter? Marital status and wealth outcomes among preretirement adults. Journal of Marriage and Family, 64, 254–268.

34.

Zissimopoulos

J. M.

(2013). Marriage and wealth changes at older ages. In Couch

K. A.

Daly

M. C.

Zissimopoulos

J. M.

(Eds.), Lifecycle events and their consequences: Job loss, family change, and declines in health (pp. 158–177). Stanford, CA: Stanford University Press.

35.

Zissimopoulos

Karney

Rauer

(2008). Marital histories and economic well-being (University of Michigan Retirement Research Center Working Paper, WP 2008-180).

36.

Zissimopoulos

J. M.

Karney

B. R.

Rauer

A. J.

(2015). Marriage and economic well being at older ages. Review of Economics of the Household, 13, 1–35.