Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2014 Oct 1.
Published in final edited form as: J Marriage Fam. 2013 Sep 3;75(5):1194–1202. doi: 10.1111/jomf.12065

Dating Relationships in Older Adulthood: A National Portrait

Susan L Brown *, Sayaka K Shinohara *
PMCID: PMC3848887  NIHMSID: NIHMS499794  PMID: 24319296

Abstract

Dating in later life is likely common, especially as the proportion of older adults who are single continues to rise. Yet there are no recent national estimates of either the prevalence or factors associated with dating during older adulthood. Using data from the 2005-2006 National Social Life, Health, and Aging Project, a nationally representative sample of 3,005 individuals ages 57-85, the authors constructed a national portrait of older adult daters. Roughly 14% of singles were in a dating relationship. Dating was more common among men than women and declined with age. Compared to non-daters, daters were more socially advantaged. Daters were more likely to be college educated and had more assets, were in better health, and reported more social connectedness. This study underscores the importance of new research on partnering in later life, particularly with the aging of the U.S. population and the swelling ranks of older singles.

Keywords: aging, couples, dating, gender, gerontology, social support

A growing share of the U.S. population is unmarried, and this trend is especially pronounced among middle-aged and older adults (Kreider & Ellis, 2011). The proportion of middle-aged adults who are single has risen 50% since 1980. Today, roughly one third of Baby Boomers are currently unmarried (Lin & Brown, 2012). Among adults 65 and older, a majority is unmarried (Manning & Brown, 2011).

Unmarried older adults often form partnerships, although they increasingly do so outside of marriage (de Jong Gierveld, 2004). Cohabitation has accelerated rapidly among older single adults, who are now equally likely to form cohabiting or marital unions (Brown, Bulanda, & Lee, 2012). The rise in unmarried individuals suggests the availability of a large pool of potential partners for non-coresidential relationships, too. Dating in later life may offer many of the benefits of a close intimate relationship without the constraints entailed by coresidence. For others, dating may be a stepping stone to cohabitation or marriage.

Regardless, researchers lack a basic understanding of who dates in later life; the most recent national, empirical study on the older adult dating population was conducted more than 20 years ago (Bulcroft & Bulcroft, 1991). With the growth in single older adults, an examination of today’s later life daters is overdue. The purpose of this study was to provide a national portrait of older adult daters using data from the 2005-2006 National Social Life, Health, and Aging Project (NSHAP).

Nearly all of the research on older daters comes from qualitative, in-depth interview studies that explore the reasons why older adults (often only women and only widows) pursue or avoid intimate relationships (Bulcroft & O’Connor, 1986; Dickson, Hughes, & Walker, 2005; McWilliams & Barrett, 2012; Stevens, 2002; Watson & Stelle, 2011). These studies illustrate the range of meanings of dating in later life. One study from the 1980s revealed that many daters were involved in serious, long-term relationships (Bulcroft & O’Connor, 1986). But more recent research has suggested many women view dating as a social activity that provides a unique form of companionship not achieved through friendships with other women (Davidson, 2001; Watson & Stelle, 2011). These women desired a close companion but at the same time wanted to be autonomous and ultimately were not interested in a long-term, conventional commitment (e.g., marriage or cohabitation; Dickson et al., 2005; McWilliams & Barrett, 2012).

There appear to be notable gender differences in the meanings of later life dating relationships. Specifically, men tend to be more interested than women in formalizing these relationships through marriage (McWilliams & Barrett, 2012; Stevens, 2002), although the prospect of remarriage can be stressful for older men and women alike (de Jong Gierveld, 2002). Women are often reluctant to marry, preferring instead to enjoy the companionship afforded through dating without incurring the potentially heavy caregiving burdens than marriage can entail in old age (Dickson et al., 2005). Many women also cite maintaining their autonomy as the reason why they do not wish to enter into a coresidential relationship (Dickson et al., 2005; Stevens, 2002). In fact, older widowed women’s interest in remarriage declines with age, but the likelihood of having a male confidante does not (Moorman, Booth, & Fingerman, 2006).

To our knowledge, only one empirical study has provided a portrait of older daters, and it relied on data from the 1987-1988 National Survey of Families and Households. Bulcroft and Bulcroft (1991) compared older daters and non-daters to determine the characteristics associated with dating in later life. They found that daters were younger than non-daters and more likely to be men. Daters also had better health and driving ability and greater involvement in organizational activities than non-daters, on average. Such patterns indicate that daters tend to be a socially advantaged group. These individuals are certainly the most attractive potential partners. At the same time, they are arguably the most active and engaged older adults, which may lead them to seek social interaction through dating relationships.

Older adults often pursue dating relationships to mitigate feelings of loneliness, and this is particularly true for men (Bulcroft & O’Connor, 1986; Carr, 2004; de Jong Gierveld, 2002; Stevens, 2002). Women enjoy wider circles of social support and therefore tend to be less eager to date (Watson & Stelle, 2011). In her study of older widow(er)s’ desire to date and remarry, Carr (2004) found that men with lower levels of social support were particularly likely to express interest in forming a new partnership, suggesting that dating was perceived as a mechanism through which men could minimize feelings of loneliness and isolation. Men with higher levels of support were more similar to women in their weaker desires to date following widowhood. For some older adults, forming a new partnership that provides close ties is preferable to relying on their adult children for social support and interaction (Stevens, 2002).

The present study extends prior research by drawing on a large, recent national sample to provide estimates of the prevalence and composition of the older dating population in the U.S. Specifically, we assessed whether and how today’s daters and non-daters differ in terms of demographic characteristics, economic resources, health, and social ties. Drawing on Bulcroft and Bulcroft (1991), we anticipated that daters are younger and more likely to be men. Age and gender were expected to interact such that the gender gap in dating widens with age. Bulcroft and Bulcroft found no racial variation in dating; neither did dating vary by ever having divorced. This latter nonsignificant association may reflect the rarity of divorce among older adults 25 years ago. Since 1990, the divorce rate has doubled among those ages 50 and older (Brown & Lin, 2012). We assessed whether divorced, widowed, and never-married singles differ in their dating behavior and expected the divorced to be most likely to date, because prior work indicates they are most likely to cohabit (Brown, Lee, & Bulanda, 2006).

Economic resources, including education, employment, and assets, presumably make one more attractive in the dating market and are indicators of success. Although Bulcroft and Bulcroft (1991) did not find any linkages among these three measures and dating, family behaviors for today’s cohorts are increasingly polarized by socioeconomic status (Cherlin, 2010). Economic factors may be more salient for men than women.

Similarly, health should increase one’s attractiveness as a dating partner and possibly raise one’s interest in dating. Bulcroft and Bulcroft (1991) found that comparative health (i.e., relative to one’s peers) and driving ability were positively associated with dating, whereas disability was unrelated to dating. We examined overall comparative health and driving ability.

Finally, social ties encompass both behavioral and subjective indicators of social connectedness and support. There are competing hypotheses about the relationship between social ties and dating (Talbott, 1998). The complementarity hypothesis suggests that individuals with the most social connections are most likely to date because they are more interested in and adept at forming social ties. In contrast, the compensatory hypothesis indicates that lower levels of social connectedness lead individuals to seek ties through intimate relationships, and thus social connectedness is negatively associated with dating. Qualitative research suggests the complementarity hypothesis is likely to hold for women, who date because they seek a unique form of companionship that friends and family cannot provide (Davidson, 2001; Stevens, 2002; Watson & Stelle, 2011). In contrast, the compensatory hypothesis is expected to characterize men, who want to date because they lack other forms of social ties (Carr, 2004).

METHOD

Data came from the 2005-2006 NSHAP, a nationally representative sample of 3,005 community-dwelling persons ages 57 to 85 (i.e., persons born between 1920 and 1947). Fielded by the National Opinion Research Center and the University of Chicago, the sample design was developed by the Health and Retirement Study, using their household screening process. Of the 4,400 persons selected from screened households, 92% were eligible for inclusion in the NSHAP and, of those, 76% completed the NSHAP interview (O’Muircheartaigh, Eckman, & Smith, 2009). The NSHAP included an in-person interview, a self-administered questionnaire, and a biomeasures collection. Topics covered by the NSHAP included demographic characteristics, sexual and union histories, social networks, physical and mental health, well-being and illness, and social and cultural activities.

A key advantage of the NSHAP is that it included a question about non-coresidential partners, allowing the measurement of dating among older adults. The analytic sample for this study was composed of unmarried, non-cohabiting adults (N = 1,144), of which 152 reported they were in a dating relationship. Note that the analytic sample excluded all 60 cohabiting respondents because they were in a coresidential union, which prior research has indicated is similar to marriage among older adults (Brown & Kawamura, 2010; King & Scott, 2005). Missing data were minimal. Mean substitution was used to handle missing values.

Measures

Dating was measured by the question “Do you currently have a romantic, intimate, or sexual partner?” This question was asked only of respondents who did not report that they were married or living with a partner. Responses were coded 1 = yes and 0 = no.

Several factors associated with older adult dating, including indicators of demographic characteristics, economic resources, health, and social ties, were included as covariates.

Demographic characteristics included age, race, and marital status. Age was coded in years. Race was dummy coded as (a) Black, (b) Other, and (c) White (reference group). Marital status was captured by a series of dummies: (a) never married, (b) widowed, and (c) divorced or separated (reference group).

Economic resources encompassed education, employment, and assets. College education differentiated those with a college degree (coded 1) from others (coded 0). Employment distinguished between respondents who report being currently employed (1 = yes) versus those who were not currently working (0 = no). Assets was a measure of the respondent’s financial resources, capturing household assets minus any debts. Respondents were to confirm that the numeric value referred to their net worth. The measure was logged to adjust for skewness.

Health was gauged by two measures. Comparative health tapped the respondent’s assessment of his or her health relative to peers of about the same age, ranging from 1 (“much worse”) through 3 (“about the same”) to 5 (“much better”). Drive was an indicator of whether the respondent had no trouble driving a car during the day (1 = yes, 0 = no).

Social ties were captured by two scales, one of which tapped into a behavioral dimension (connectedness) and the other a subjective dimension (support; Cornwell & Waite, 2009). Social connectedness was a three-item scale (α = .66) that summed the respondent’s reports of frequency of involvement during the past year in organized activities, volunteer activities, and time spent with family and friends. For each item of the scale, values ranged from 0 (“never”) through 3 (“several times a year”) to 6 (“several times a week”). Values on the social connectedness scale ranged from 0 to 18. Social support was a four-item scale (α = .63) that summed how often the respondent can do the following: rely on family for help if they had a problem, rely on friends for help if they had a problem, open up to family members to talk about worries, and open up to friends to talk about worries. For each item, values ranged from 1 (“hardly ever [or never]”) through 2 (“some of the time”) to 3 (“often”). Thus, the scale ranged from 4 to 21, with higher values indicating greater perceived social support.

Analytic Strategy

Our primary aim was to generate a national portrait of today’s older daters, distinguishing them from other unmarried, non-dating individuals. First, we documented the prevalence of dating and how it differed among men and women as well as by age group. Second, we examined the characteristics of daters versus non-daters, both for all unmarried individuals and separately by gender. Third, we estimated logistic regression models predicting dating among unmarried individuals to evaluate the covariates of dating in a multivariate framework. Because our objective was to describe the population of older adult daters, we were less concerned about causal order and acknowledge that some of the factors examined may be antecedents of dating, whereas others may be consequences of dating. The NSHAP complex sampling design means that the sample is not self-weighting. Thus, all analyses were conducted in Stata using svy procedures to generate corrected standard errors that adjust for the complex sampling design. Postestimation goodness-of-fit F tests were conducted using the svylogitgof command (Archer & Lemeshow, 2006).

RESULTS

Roughly 5% of older Americans are currently in a dating relationship (result not shown). Among older unmarried individuals, the population eligible to date, 14% were in a dating relationship. The proportions dating varied considerably by gender, with more than one quarter (27%) of older men reporting dating partners versus just 7% of older women (results not shown). For both men and women, the prevalence of dating declined with age, as depicted in Figure 1. Nearly one-fifth (18%) of singles ages 57 through 64 reported a dating partner, whereas just 14% and 9% of singles ages 65 through 74, and 75 through 85, respectively, reported that they were dating someone. For men, the proportions ranged from 32% among singles ages 57 through 64 to 27% for those ages 65 through 74 and 24% for singles ages 75 through 85. For women, 11% of singles ages 57 through 64 reported a dating partner versus 7% of single 65- to 74-year-olds and a mere 3% of 75- to 85-year-olds.

Bivariate Results

A comparison of daters and non-daters, both for the total sample and separately by gender, is provided in Table 1. Daters and non-daters differ in terms of demographic characteristics, economic resources, health, and social ties. Daters were about 3 years younger (68), on average, than non-daters (71). Not surprisingly, daters were disproportionately men (62%). Because women tend to date men the same age or older than themselves, whereas men typically date same-age or younger women, men have more options, especially at this stage in the life course given men’s shorter life expectancy. Most daters were divorced or separated (57%), whereas most non-daters were widowed (56%). The education distribution of daters differed from that of non-daters, with 37% of daters holding college degrees versus just 16% of non-daters. A higher proportion of daters (40%) than non-daters (25%) were working. Daters also tended to be wealthier; the logged value of daters’ assets was 9.02 (about $521,000) versus 6.75 (roughly $162,000) for non-daters. The health indicators for daters also were more favorable. Relative to their similar-age peers, daters (4.10, representing more than “somewhat better” but less than “much better”) rated their own health more favorably than did non-daters (3.8, representing more than “about the same” but less than “somewhat better”). Whereas 90% of daters reported they could drive a car safely, just 76% of non-daters were confident about their driving skills. Social connectedness among daters was greater, on average, than non-daters. The mean value for daters of 9.71 corresponded with interactions more than several times per year, versus the value of 8.71 for non-daters, indicating interaction more than once or twice a year but less than several times a year. Daters and non-daters reported comparable levels of perceived social support.

Another relevant comparison is that of daters versus non-daters within gender. Among men, a larger share of daters was divorced and a smaller share had never been married compared to non-daters. Dating men were also economically advantaged relative to non-dating men in that they were more likely to have a college degree and to be employed and held more assets. They were also healthier in that they rated their comparative health more favorably and a larger proportion still drove. Dating men reported greater social connectedness than non-dating men.

Among women, daters were younger and disproportionately likely to be divorced and unlikely to be widowed. The economic advantage was less pronounced among women, although daters were more likely to have a college degree. Dating and non-dating women did not differ in terms of health. Dating women reported greater social connectedness than non-dating women.

Table 1 also includes boldface coefficients that indicate significant gender differences among either daters or non-daters. Among daters, the characteristics of men and women were overwhelmingly similar. The only significant gender difference was that a larger share of dating men than dating women reported being able to drive safely during the day (95% vs. 83%).

In contrast, there were several notable gender differences among non-daters, perhaps because the larger sample size yielded greater statistical power. Non-dating women were about 1 year older than non-dating men. Women were predominantly widowed (62%), whereas men were similarly likely to be widowed (41%) or divorced (40%). Non-dating men were much more likely to have graduated from college than were non-dating women (23% vs. 13%). Non-dating women were less likely to be comfortable driving during the day than non-dating men (73% vs. 85%). Among non-daters, women reported more social connectedness and social support than men.

Multivariate Results

The odds ratios from logistic regression models predicting dating among the entire sample, as well as men and women separately, are shown in Table 2. Beginning with the model estimated for the whole sample, the odds of dating declined marginally with age (p = .06) and were 77% lower for women than men. The rate of decrease in the likelihood of dating by age was marginally greater among women than men (interaction term for gender × age, odds ratio = 0.94, p = .07) which aligned with our expectations. Blacks were more likely to be dating than Whites. Relative to divorced individuals, never-married and widowed people were less likely to be dating. Economic resources were associated with dating: The odds of dating were 80% greater for persons with a college degree compared with those with less education. Also, the higher one’s assets, the more likely it is that one was dating. We expected that the role of economic resources might be more pronounced among men, but gender interactions with college degree and assets did not achieve significance (results not shown). Comparative health was positively related to dating, indicating that those who are most robust compared with their peers are most likely to be dating. Social ties were linked to dating, with greater social connectedness positively associated with dating. This pattern aligned with the complementarity hypothesis, according to which those with the most ties would be most likely to date. The inclusion of an interaction term for gender × social connectedness was not significant, indicating that the complementarity hypothesis holds for women and men alike.

We estimated separate models for men and women to explore possible gender differences in how factors are related to dating. Although the gender interactions in the full model failed to achieve statistical significance at conventional levels, this may reflect a lack of statistical power associated with the modest number of daters. Among men, the odds of dating were 2.5 times higher for Blacks than Whites. Never-married men were less likely to be dating than divorced men. The odds that college-educated men were dating were more than twice as high as those for men with less education. Wealthier men were more likely to be dating than men with fewer assets. Men’s comparative health was positively related to dating. Also, men who reported being comfortable driving during the day had odds of dating roughly 2.5 times higher than men who were not comfortable driving. Social ties were not related to dating among men (the inclusion of economic resources reduced the positive association between social connectedness and dating to nonsignificance, result not shown).

Turning now to women, younger women were more likely to be dating than older women. The odds that widowed women were dating were more than 50% lower than those of divorced women. In contrast to the pattern found among men, economic resources and health were unrelated to dating among women. Social connectedness was positively associated with women’s dating, supporting the complementarity hypothesis.

DISCUSSION

This study provides a national portrait of dating in later life, a topic of growing importance with the rise in single older adults. Using a recent, national sample of adults age 57 through 85, we documented the prevalence and correlates of later life dating. Approximately 14% of older unmarried individuals were in dating relationships. Nearly two thirds of older adult daters were men. Stated differently, more than one quarter of older single men were in a dating relationship, compared with less than 10% of single older women. The gender asymmetry in older adult dating is not surprising considering that older men enjoy a larger pool of potential dating partners than women, reflecting gender differences in life expectancy and norms prescribing that men should date younger women.

The findings from our study confirmed our expectations that older daters are an advantaged group. Relative to non-daters, daters tended to be younger, had more economic resources, were in better health, and enjoyed greater social ties. Our analyses also provided preliminary evidence that the social advantages related to dating may differ by gender. The age gradient in dating was modestly larger for women than men. Also, economic resources and health appeared more salient for men, whereas social ties were more central for women. Although gender interactions in the full model did not achieve statistical significance, the disparate findings in the gender specific models nonetheless are suggestive of unique patterns for men and women that would likely be supported in the full model with a larger sample size of daters.

Our national portrait of older daters is largely consistent with that constructed by Bulcroft and Bulcroft (1991) more than two decades ago. Age, gender, health, and social ties are related to dating among today’s older adults just as they were in the late 1980s. There were also a couple of notable differences. First, Bulcroft and Bulcroft anticipated that a less traditional marital history would be positively associated with dating, but they found no evidence to support this assertion. Among today’s older adults, we found that those who were either widowed or never married were less likely to be in a dating relationship than those who were divorced. Second, Bulcroft and Bulcroft found that economic resources were unrelated to dating. For contemporary older adults, education and assets were positively associated with dating, which is consistent with research showing family patterns are increasingly stratified by socioeconomic status (Cherlin, 2009). Bulcroft and Bulcroft did not formally test for significant gender differences in the correlates of dating.

Much of the literature on dating in later life emphasizes the role of social ties. According to a recent study on the desire to date, men with low social support are more likely to want to date, whereas men who enjoy high social support are more comparable to women in their relatively weaker interest in dating (Carr, 2004). But our study revealed that social connectedness was associated with an increased likelihood of dating. This finding arguably holds for women and men alike, because the interaction between gender and social connectedness was not significant. Gender-specific models showed that connectedness was not related to dating among men and was positively associated with dating among women. The positive association between social connectedness and dating among men was reduced to nonsignificance with the inclusion of economic resources. Our results support the complementarity hypothesis that those with the most ties were the most adept at forming and maintaining intimate relationships (Talbott, 1998). This logic is also consistent with the conclusion from qualitative research that women form dating relationships to achieve a type of companionship that is not available through friends and family (Davidson, 2001; Watson & Stelle, 2011).

Our study has some limitations. First, our measure of dating was conservative, because the respondent had to identify a sexual, intimate, or romantic partner, signaling a steady relationship. Some singles may be in the market searching for a partner (i.e., dating), but they are not defined as being in a dating relationship. Second, our sample size of daters was modest (n = 152) and may have contributed to Type II errors (i.e., failing to reject the null hypothesis when it is actually false) due to a lack of statistical power. Third, we were not able to distinguish among unmarried respondents on the basis of a desire to have a dating relationship. Many of the singles who were not dating did so by choice, not circumstance. Finally, the findings should not be construed to suggest causal pathways from demographic characteristics, economic resources, health, or social ties, to dating. Indeed, dating actually may be influencing some of correlates. The cross-sectional data did not permit us to disentangle the causal linkages between these factors and dating. Still, this study provides substantial insight into the characteristics and composition of the dating population in older adulthood. Because a larger share of older U.S. adults is unmarried and therefore eligible to form a dating relationship, an important first step is to establish a national portrait of daters in later life.

This study lays the groundwork for future work on later life dating. For example, what are the relationship dynamics characterizing these relationships? Do older adult daters follow a traditional path to marriage, or do they prefer to remain single? For some older daters, the next step may be either a living-apart-together (LAT) relationship (Karlsson & Borell, 2003) or a cohabiting union (Brown et al., 2012), both of which allow individuals to enjoy many of the benefits of marriage without the legal entanglements. These relationship types also offer more flexibility in terms of gender roles and expectations for caregiving, which is of particular concern for women (Karlsson & Borell, 2005). As the range of partnership options widens, researchers must expand their lens when examining how intimate relationships are linked to individual health and well-being. Cohabiting partners are less likely to provide care to frail partners than are married spouses (Noel-Miller, 2011). Do the various types of unmarried relationships offer benefits comparable to marriage?

Dating is a common experience among older singles, especially for men. In general, it appears that daters tend to be socially advantaged in that they are better educated and enjoy a larger net worth, are healthier than their peers, are more socially connected, and are younger, on average, than singles who are not dating. This national portrait of today’s older daters underscores the need for additional research to address the dynamics and consequences of dating in later life, particularly with the aging of the U.S. population and the swelling ranks of older singles.

FIGURE 1. Age Variation in the Weighted Percentages Dating Among Older Unmarried Respondents, by Gender.

FIGURE 1

Open in a new tab

Table 1. Descriptive Statistics (WeightedMeans or Percentages) for All Variables, by Dating Status.

Daters Non-daters
Variable Total Men Women Total Men Women
Demographic characteristics
Age (in years) 67 74*** 68.53 66.45*** 70.92 69.85 71.30
Gender (1 = woman) 37.81%*** 73.84%
Race
 White 77.27% 76.67% 78.26% 79.39% 80.16% 79.12%
 Black 18.80% 17.87% 20.34% 14.33% 12.82% 14.87%
 Other race 3.92% 5.46% 1.39% 6.11% 7.02% 5.79%
Marital status
 Divorced 56.59%*** 54.81%* 59.51%*** 32.80% 40.15% 30.20%
 Widowed 35.78%*** 37.44% 33.06%*** 56.29% 41.27% 61.61%
 Never married 7.63% 7.75%* 7.43% 10.91% 18.58% 8.19%
Economic resources
 College degree 36.50%*** 42.83%*** 26.09%* 15.59% 22.57% 13.12%
 Employment (1 = working) 39.91%** 40.29%** 39.29% 25.40% 24.74% 25.63%
 Assets logged 9 02*** 9.56*** 8.14 6.75 7.25 6.57
Health
 Comparative health 4 10*** 4 14*** 4.03 3.80 3.77 3.82
 Drive 90.44%*** 95.26%** 82.50% 75.90% 85.43% 72.59%
Social ties
 Social connectedness 9 71** 9.21** 10.54* 8.71 7.83 9.03
 Social support 9.13 9.00 9.33 9.24 8.60 9.46
N 152 99 53 992 257 735
Open in a new tab

Note: Asterisks denote significant differences between daters and non-daters Significant (p < .05) gender differences among daters or non-daters are indicated by boldface coefficients.

*

p < .05.

**

p < .01.

***

Table 2. Odds Ratios From Logistic Regression Models Predicting Dating.

Variable Total Men Women
Demographic characteristics
Age (in years) 0.97a 0.99 0.93**
Gender (1 = woman) 0 23***
Race
 White (ref.)
 Black 1.87* 2.51* 1.26
 Other race 0.72 1.48 0.15
Marital status
 Divorced (ref.)
 Widowed 0.58* 0.64 0.44*
 Never married 0.34** 0.27** 0.36
Economic resources
 College degree 1.80* 2.14* 1.55
 Employment (1 = working) 1.10 1.22 0.94
 Assets logged 1.06* 1.09** 1.02
Health
 Comparative health 1.40* 1.61** 1.21
 Drive 1.32 2.49* 0.86
Social ties
 Social connectedness 1.06* 1.03 1.10*
 Social support 0.97 1.03 0.87
Goodness-of-fit F test (9, 42) 568.90*** 724 47*** 113.82***
N 1,144 356 788
Open in a new tab

Note: ref. = reference category.

*

p < .05.

**

p < .01.

***

p < .001.

a

p = .06.

Acknowledgments

An earlier version of this article was presented at the 2010 annual meeting of the National Council on Family Relations, Minneapolis, MN. This research was supported in part by the Center for Family and Demographic Research, Bowling Green State University, which has core funding from the Eunice Kennedy Shriver National Institute of Child Health and Human Development (R24 HD050959-09).

REFERENCES

  1. Archer KJ, Lemeshow S. Goodness-of-fit test for a logistic regression model fitted using survey sample data. Stata Journal. 2006;6:97–105. doi:10.1016/j.csda.2006.07.006. [Google Scholar]
  2. Brown SL, Bulanda JR, Lee GR. Transitions into and out of cohabitation in later life. Journal of Marriage and Family. 2012;74:774–793. doi: 10.1111/j.1741-3737.2012.00994.x. doi:10.1111/j.1741-3737.2012.00994.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Brown SL, Kawamura S. Relationship quality among cohabitors and marrieds in older adulthood. Social Science Research. 2010;39:777–786. doi: 10.1016/j.ssresearch.2010.04.010. doi:10.1016/j.ssresearch.2010.04.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Brown SL, Lee GR, Bulanda JR. Cohabitation among older adults: A national portrait. Journals of Gerontology: Series B: Psychological Sciences and Social Sciences. 2006;61:71–79. doi: 10.1093/geronb/61.2.s71. [DOI] [PubMed] [Google Scholar]
  5. Brown SL, n I-F. The gray divorce revolution: Rising divorce among middle-aged and older adults, 1990-2009. National Center for Family & Marriage Research Working Paper Series No. WP-12-04; Bowling Green State University, Bowling Green, OH: 2012. Retrieved from http://ncfmr.bgsu.edu/pdf/working_papers/file108701.pdf. [Google Scholar]
  6. Bulcroft RA, Bulcroft KA. The nature and functions of dating in later life. Research on Aging. 1991;13:244–260. doi:10.1177/0164027591132007. [Google Scholar]
  7. Bulcroft K, O’Connor M. The importance of dating relationships on quality of life for older persons. Family Relations. 1986;35:397–401. [Google Scholar]
  8. Carr D. The desire to date and remarry among older widows and widowers. Journal of Marriage and Family. 2004;66:1051–1068. doi:10.1111/j.0022-2445.2004.00078.x. [Google Scholar]
  9. Cherlin AJ. The marriage-go-round. New York: Knopf: 2009. [Google Scholar]
  10. Cherlin AJ. Demographic trends in the United States: A review of research in the 2000s. Journal of Marriage and Family. 2010;72:403–419. doi: 10.1111/j.1741-3737.2010.00710.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Cornwell E, Waite LJ. Measuring social isolation among older adults using multiple indicators from the NSHAP study. Journals of Gerontology: Series B: Psychological Sciences and Social Sciences. 2009;64B(Suppl. 1):i38–i46. doi: 10.1093/geronb/gbp037. doi:10.1093/geronb/gbp037. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Davidson K. Late life widowhood, selfishness and new partnership choices: A gendered perspective. Ageing and Society. 2001;21:297–317. doi:10.1017/S0144686X01008169. [Google Scholar]
  13. de Jong Gierveld J. The dilemma of repartnering: Considerations of older men and women entering new intimate relationships in later life. Ageing International. 2002;27:61–78. doi:10.1007/s12126-002-1015-z. [Google Scholar]
  14. de Jong Gierveld J. Remarriage, unmarried cohabitation, living apart together: Partner relationships following bereavement or divorce. Journal of Marriage and Family. 2004;66:236–243. doi:10.1111/j.0022-2445.2004.00015.x. [Google Scholar]
  15. Dickson FC, Hughes PC, Walker KL. An exploratory investigation into dating among later-life women. Western Journal of Communication. 2005;69:67–82. doi:10.1080/10570310500034196. [Google Scholar]
  16. Karlsson S, Borell K. Reconceptualizing intimacy and ageing: Living apart together. In: Arber S, Davidson K, Ginn J, editors. Gender and ageing: Changing roles and relationships. Open University Press; Maidenhead, UK: 2003. pp. 47–62. [Google Scholar]
  17. Karlsson S, Borell K. A home of their own: Women’s boundary work in LAT-relationships. Journal of Aging Studies. 2005;19:73–84. doi:10.1016/j.jaging.2004.03.008. [Google Scholar]
  18. King V, Scott ME. A comparison of cohabiting relationships among older and younger adults. Journal of Marriage and Family. 2005;67:271–285. doi:10.1111/j.0022-2445.2005.00115.x. [Google Scholar]
  19. Kreider RM, Ellis R. Current Population Reports, P70-125. U.S. Census Bureau; Washington, DC: 2011. Number, timing, and duration of marriages and divorces: 2009. [Google Scholar]
  20. Lin I-F, Brown SL. Unmarried boomers confront old age: A national portrait. The Gerontologist. 2012;52:153–165. doi: 10.1093/geront/gnr141. doi:10.1093/geront/gnr141. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Manning WD, Brown SL. The demography of unions among older Americans, 1980-present: A family change approach. In: Settersten RA, Angel JL, editors. Handbook of sociology of aging. Springer; New York: 2011. pp. 193–212. doi:10.1007/978-1-4419-7374-0_13. [Google Scholar]
  22. McWilliams S, Barrett AE. Online dating in middle and later life: Gendered expectations and experiences. Journal of Family Issues. 2012 Advance online publication. doi:10.1177/0192513X12468437. [Google Scholar]
  23. Moorman SM, Booth A, Fingerman KL. Women’s romantic relationships after widowhood. Journal of Family Issues. 2006;27:1281–1304. doi:10.1177/0192513X06289096. [Google Scholar]
  24. Noel-Miller C. Partner caregiving in older cohabiting couples. Journals of Gerontology: Series B: Psychological Sciences and Social Sciences. 2011;66:341–353. doi: 10.1093/geronb/gbr027. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. O’Muircheartaigh C, Eckman S, Smith S. Statistical design and estimation for the National Social Life, Health, and Aging Project. Journals of Gerontology: Series B: Psychological Sciences and Social Sciences. 2009;64B:i12–i19. doi: 10.1093/geronb/gbp045. doi:10.1093/geronb/gbp045. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Stevens N. Re-engaging: New partnerships in late-life widowhood. Ageing International. 2002;27:27–42. doi:10.1007/s12126-002-1013-1. [Google Scholar]
  27. Talbott MM. Older widows’ attitudes towards men and remarriage. Journal of Aging Studies. 1998;12:429–449. doi:10.1016/S0890-4065(98)90028-7. [Google Scholar]
  28. Watson WK, Stelle C. Dating for older women: Experiences and meanings of dating in later life. Journal of Women & Aging. 2011;23:263–275. doi: 10.1080/08952841.2011.587732. doi:10.1080/08952841.2011.587732. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES