Parenthood, Marital Status, and Well-Being in LaterLife: Evidence from SHARE

Karsten Hank • Michael Wagner

Accepted: 25 September 2012� Springer Science+Business Media Dordrecht 2012

Abstract Using pooled cross-sectional data from the first two waves of the Survey of

Health, Ageing and Retirement in Europe, we address the question of whether and how

parenthood and marital status are associated with various dimensions of elders’ well-being,

which we define by elements of the individual’s economic situation, psychological well-

being, and social connectedness. The results of our multivariate analysis suggest that

childless individuals do not generally fare worse than parents in terms of their economic,

psychological, or social well-being. Although there is some indication for a ‘protective

effect’ of marriage, having a partner does not per se contribute to greater psychological

well-being: only those reporting satisfaction with the extent of reciprocity in their rela-

tionship report lower numbers of depression symptoms than their unmarried counterparts.

We observe no systematic associations between parenthood (marriage, respectively) and

individuals’ propensity to participate in social activities. These findings are fairly stable,

that is, they hold for both men and women as well as across various cohorts, and they do

not vary systematically between countries.

Keywords Parenthood � Partnership � Well-being � SHARE

1 Introduction

Population ageing results from steadily increasing life-expectancy and sustained below-

replacement fertility. Two factors that have been suggested to contribute, among many

others, to the latter are increasing levels of childlessness (e.g. Dykstra 2009; Rowland

2007) and decreasing union stability (e.g. Coppola and Di Cesare 2008; Klein 2003). This

is of particular concern against the background of a growing population of frail elders who

depend on support by others, and particularly on help provided by their families. That is,

K. Hank (&) � M. WagnerInstitute of Sociology, University of Cologne, Greinstr. 2, 50939 Cologne, Germanye-mail: hank@wiso.uni-koeln.de

M. Wagnere-mail: mwagner@wiso.uni-koeln.de

123

Soc Indic ResDOI 10.1007/s11205-012-0166-x

the absence of children and/or partners might constitute a serious threat to well-being in

later life, especially in countries where welfare state services are poorly developed (e.g.

Brandt et al. 2009; Kohli et al. 2009).

Using pooled cross-sectional data from the first two waves of the Survey of Health,

Ageing and Retirement in Europe (SHARE), we address the question of whether (or to

what extent) this is actually the case from three complementary angles, thereby providing a

more comprehensive account than most previous research. First, many studies on the role

of parenthood and partnership in later-life well-being focus on psychological well-being

(or other health outcomes; see, for example, Hughes and Waite 2009; Umberson et al.

2010). Our study, though, takes a broader perspective on well-being, which also considers

elements of elders’ economic situation and social connectedness (e.g. Dykstra and Wagner

2007; Keizer et al. 2010).1 Second, we explicitly account for the interplay between par-

enthood and marital status (e.g. Keizer et al. 2010; Wagner 1997), as well as for the role of

relationship quality (e.g. Koropeckyj-Cox 2002; Ryan and Willits 2007) in determining

older people’s well-being. Third, and finally, we investigate whether and how the societal

context in which individuals age impacts the micro-level association between parenthood

(marital status, respectively) and various dimensions of well-being (e.g. Huijts et al. 2011;

Moor and Komter 2012).

The remainder of this article is structured as follows: to begin with we provide a brief

overview of previous research on the parenthood/partnership and well-being nexus. Following

a short description of data and methods, we present our empirical findings. The final section

concludes, discussing implications of our findings and perspectives for future research.

2 Previous Research

Proposed mechanisms Various mechanisms have been proposed to explain potential

associations between elders’ parental and partnership status on the one hand and their well-

being on the other hand. First, there is a general consensus that children and partners carry

both potential benefits (e.g. support), but may also bring about costs (e.g. stress); cf. Ross

et al. (1990). It has been argued that relations between family members are ‘special’ in the

sense that they are characterized by a particular form of solidarity between spouses and

especially between generations (e.g. Bengtson and Roberts 1991). One important dimen-

sion of family solidarity is the exchange of instrumental and emotional support (‘functional

solidarity’). Children, especially if they live in close proximity to their parents, might be

confronted with strong social expectations to support their parents in order to fulfil norms

of intergenerational reciprocity (e.g. Schutze and Wagner 1991; also see Hank 2007). Such

expectations, however, also bear in them the potential for intergenerational conflict (e.g.

Bengtson et al. 2002). Moreover, the quality of parent–child relations might be challenged by

spousal conflict (e.g. Erel and Burman 1995), which also has a direct negative impact on

individuals’ well-being (e.g. Choi and Marks 2008; Hawkins and Booth 2005), thereby

counteracting the frequently suggested protective effect of marriage. Whether the net effect

of parental and partnership status on individuals’ well-being is positive or negative is difficult

to establish, because this will depend on the specific combination of benefits, costs, and well-

1 This approach is consistent with recent work by Hallerod and Selden (2012), for example. In a sample ofolder Swedes, they identified five different well-being arenas: health, psychosocial well-being, functionality,social relations, and economic hardship. All ‘‘were part of a multidimensional wellbeing condition, i.e. olderpeople tend to simultaneously suffer from several types of specific wellbeing problems.’’

K. Hank, M. Wagner

123

being dimensions under consideration. Moreover, it is dependent on the extent to which kin

and non-kin social relationships are complements or substitutes (e.g. Kohli et al. 2009) and

likely to vary over the individual’s life-course (e.g. Umberson et al. 2010).

Second, childbearing behaviours ‘‘off the normative life course track’’ (i.e. childless-

ness, but also teenage or non-marital parenthood) tend to be associated with unstable

partnership careers and might have long-term negative consequences for individuals’ well-

being, resulting from economic disadvantage, social exclusion, or stress induced health

insults (e.g. Hank 2010; Koropeckyj-Cox et al. 2007). Although it has been suggested that

‘‘the ramifications of not entering the parental role are greater for women than for men’’

(Dykstra and Wagner 2007: 1491), young fatherhood has also been shown to be associated

with subsequent disadvantage across the life course (e.g. Sigle-Rushton 2005). It thus

seems important to consider gender issues in investigations of the relationship between

parenthood and later life well-being. Moreover, societal context—in terms of welfare,

family, or social capital regimes—is likely to mediate the contribution of family related

(dis-)advantages on individuals’ well-being (see below for a more detailed discussion of

potential pathways and empirical findings).—Third, and finally, observed associations

might result from selection effects, that is, individuals equipped with different initial levels

of (economic, social, psychological) well-being may exhibit different propensities to enter

parenthood or marriage (e.g. Kiernan 1989; Waldron et al. 1996).

Empirical findings The following review of previous empirical studies investigating the

role of parenthood and marital status in older people’s well-being is organized along the

lines of three major dimensions of well-being: economic, psychological, and social well-

being (e.g. Dykstra and Hagestad 2007; also see Hallerod and Selden 2012). We consid-

ered it important to complement the psychological component of well-being with eco-

nomic and social outcomes, because they influence older adults’ life chances today andreflect acquisitions, investments, and socially structured opportunities over the individual’s

life course (Dykstra and Wagner 2007). While we acknowledge that different conceptu-

alizations exist for each of the three dimensions, we focus on those indicators that will be

addressed later on in our empirical analysis, namely (a) income and wealth, (b) depression,

and (c) social participation.

(a) Income and wealth: Old-age security has been suggested as a motive for fertility even in

societies with a functioning social security system (e.g. Rendall and Bahchieva 1998).

However, empirical investigations of the effect of children on younger families’

material well-being clearly show that children have a negative impact on a variety of

economic indicators, albeit with considerable cross-national differences (e.g. Aassve

et al. 2005; Sigle-Rushton and Waldfogel 2007). The financial disadvantage of

parenthood seems to persist into later life (e.g. Hofferth 1984; Plotnick 2009),

particularly if women who raised children outside of marriage are considered (Johnson

and Favreault 2004). Independent of parenthood, marital status is also associated with

individuals’ economic circumstances. For example, the joint taxation of married

couples, which is common in many countries (e.g. Dingeldey 2001), tends to discourage

women’s participation in the labour market, which reduces their life-time income and

often results in financial distress, especially after a divorce or following the death of their

partner (e.g. Burkauser et al. 2005; Dewilde et al. 2011; Holden and Kuo 1996). Thus, it

does not come as a surprise that, contrary to men, many studies show ‘‘a strong

socioeconomic position for never-married women—who were childless because they

remained single’’ (Dykstra and Hagestad 2007: 1520).

Parenthood, Marital Status

123

(b) Depression: Different from other major social roles, previous research could not

identify a clear relationship between parenthood and psychological well-being

(depression, respectively; e.g. Evenson and Simon 2005; Hansen et al. 2009; Zhang

and Hayward 2001). If there is statistically significant evidence at all, it tends to

suggest that older childless men and women exhibit lower levels of depressive

symptoms than parents, particularly if they are compared to those who had their first

child early (e.g. Henretta et al. 2008) or whose relationship with their offspring is of

poor quality (e.g. Koropeckyj-Cox 2002). A recent exception is the European study

by Huijts et al. (2011), who find that being childless is associated with worse

psychological well-being for men (but not for women). One’s marital status, though,

appears to be a more salient factor influencing depression than parenthood as such

(e.g. Buber and Engelhardt 2008; Bures et al. 2009). Those who are currently (and

continuously) married are least likely to suffer from depressive symptoms, whereas

those who experienced marital disruptions even do worse than the never-married (see

Hughes and Waite 2009, who report similar findings for other health outcomes as

well). Widowhood in particular has been shown to be associated with elevated risks

of depression for men and women alike (e.g. Lee and DeMaris 2007; Schaan 2009).

(c) Social participation: Children have often been suggested to serve parents as ‘bridges’

both to the wider society as well as to local networks, independent of their age (e.g.

Choi 1994; Furstenberg 2005). However, findings from a variety of institutional and

cultural contexts show that childless older adults, regardless of marital status and

gender, are equally likely as parents to be active in the community and in voluntary

organizations (Wenger et al. 2007). Never-married childless women even appear to be

particularly active socially (Wenger et al. 2007), which one might attribute partially

to the fact that close intergenerational family relations tend to be associated with a

lower propensity to engage in informal non-kin social interaction (i.e. outside of

formal organizations; see Kohli et al. 2009). Having a partner, however, is positively

linked to formal social participation, such as volunteering (e.g. Erlinghagen and Hank

2006; Rotolo and Wilson 2006). Investigations of the effects of widowhood on older

adults’ social participation provide evidence for an increase in informal social

relations following the loss of a partner (due to greater support from friends and

relatives), but find no significant relationship between widowhood and volunteerism

(e.g. Donelly and Hinterlong 2010; Utz et al. 2002).

Although the studies reported above not always indicated strong gender differences in

associations between parenthood (marital status, respectively) and later-life outcomes,

recent evidence showed that men should not be neglected in research on childlessness (e.g.

Keizer et al. 2010) and we will thus take a gendered perspective in our analysis. Moreover,

societal context is likely to play a role in determining the well-being of older parents and

childless individuals with or without partners (e.g. Huijts et al. 2011; Moor and Komter

2012). One might, first, expect to find different associations along the lines of particular

welfare state regimes, depending on how their institutions impact the potential benefit and

burden of parenthood and marriage (or lack thereof). Second, differences in family cultures

(or, more broadly, ‘social capital regimes’; cf. Pichler and Wallace 2007) might matter

because of differences in the role of families as a source of support (relative to non-kin

social networks or welfare state institutions) and because of differences in the degree to

which the childless are perceived as ‘others’ or ‘deviants’ in a society (e.g. Letherby 2002).

For example, it is well-known that family ties are stronger in Southern Europe, whereas

community ties are stronger in the Nordic countries (e.g. Kohli et al. 2009; Sundstrom et al.

K. Hank, M. Wagner

123

2009). It is therefore important to account for potential cross-national variations in the

(long-term) implications of parenthood and marriage for elders’ well-being.

3 Data and Method

We use baseline interviews from the first two rounds of the Survey of Health, Ageing andRetirement in Europe (SHARE; cf. Borsch-Supan et al. 2005, 2010), which is represen-

tative of the non-institutionalized population aged 50 or older in all 15 participating

countries: eleven countries—Austria, Belgium, Denmark, France, Greece, Germany, Italy,

the Netherlands, Spain, Sweden, and Switzerland—contributed data to SHARE’s first wave

in 2004–2005. Further data were collected in Israel during the years 2005–2006. Three

more countries – the Czech Republic, Ireland, and Poland—joined SHARE for its second

wave in 2006–2007, which also included refresher samples in those countries that already

participated in Wave 1 (see http://www.share-project.org for a detailed breakdown of

sample sizes by country, sex, and age). Our main analytic sample pools baseline interviews

of more than 9,000 men and roughly 11,000 women aged 65 or over from both waves

(including refresher samples). Supplementary analyses using additional information on

respondents’ relationship quality (see below) are based on a subsample of more than 9,000

respondents who filled-out the standard self-completion questionnaire distributed as part of

SHARE’s Wave 1 (i.e. the Czech Republic, Ireland, and Poland are not represented here).

Outcome variables We define four outcome variables. The first two are binary ones

capturing different dimensions of individuals’ economic well-being, namely income ade-

quacy and homeownership. Using these indicators has the advantage that we avoid

potential problems related to missing or imputed financial information. Perceived income

adequacy, which equals 1 if the respondent reported to make ends meet (fairly) easily, has

been shown to be a robust indicator of financial capacity in older age (see Litwin and Sapir

2009). Our second economic indicator, which equals 1 if the respondent reported to own

his or her dwelling, is important because in many countries owner-occupied housing

constitutes an important component of older households’ total wealth (see Christelis et al.

2009). The third dependent variable refers to depression as an important indicator of

individuals’ psychological well-being. It is based on the EURO-D scale, ranging from 0 to

12 self-reported depressive symptoms. The psychometric properties of the EURO–D have

been extensively investigated and criterion validity demonstrated in the cross-cultural

context of SHARE (see Castro-Costa et al. 2008). Our fourth, and final, outcome variable

is a binary indicator of general social participation. The variable equals 1 if the respondent

participated in at least one2 of the following activities in the month preceding the inter-

view: attended an educational or training course, went to a sport, social, or other kind of

club, took part in a religious, political, or community related organisation.

Explanatory variables The main variables of interest here are those informing us about

respondents’ parental status and marital status. To begin with, we distinguish childless

individuals from those who reported having 1, 2, 3, or 4? (biological or social) children

that are still alive (see Martınez-Granado and Mira 2005). While we might slightly

overestimate the proportion of childless individuals, particularly in the oldest cohorts (if

2 Note that only 5 % of the respondents in our sample reported having been engaged in more than oneactivity. We therefore refrained from estimating a model using a count variable as indicator of individuals’social participation.

Parenthood, Marital Status

123

parents outlived their children), a comparison of levels of childlessness in the SHARE

sample with external sources does not suggest any significant bias (see below). Maritalstatus is operationalized by a set of four dummy variables: married (reference category,

including a small number of registered partnerships), never-married, divorced, and wid-

owed. Obviously, legal marital status does not necessarily reflect individuals’ partnership

status. Although unmarried cohabitation is likely to be of increasing importance in future

cohorts of elders, it is not yet a quantitatively relevant phenomenon in our sample of

individuals born 1942 or earlier (for related studies see Brown et al. 2006; de Jong Gierveld

2004). Less than 3 % of our sample could be identified as cohabitors using information

from SHARE’s household grid. The vast majority (74 %) of these cohabitors reported to be

widowed, 13 % were divorced, and the remaining 13 % were never-married.

For a subsample of respondents (see above) we also have rough indicators of parent–

child and spousal relationship quality, which we include in a supplementary analysis. We

use the frequency of contact between parents and children as a proxy for relationship

quality, where we distinguish childless respondents (reference category), those who

reported having had any kind of contact with a child at least several times a week, and

those with less contact to children (e.g. Hank 2007). Marital quality is proxied by

respondents’ satisfaction with the level of reciprocity in their partnership, where we dis-

tinguish those who are unmarried (i.e. widowed, divorced, or never-married; reference

category) from married individuals reporting to be satisfied and from those who reported

low levels of satisfaction, respectively.

Control variables are the individual’s age as well as indicators of respondents’ general

physical health (which equals 1, if respondents perceive their health as very good or

excellent), level of education (three binary variables derived from ISCED, indicating low,

medium, and high educational degrees), and previous employment (a binary variable that

equals 1, if the respondent reported that he or she ever did any paid work). Finally, we

control for the year in which the interview took place (with survey years ranging from 2004

to 2007) and account for the individual’s country of residence by including country

dummies in all regressions. See Table 1 for descriptive sample statistics.

4 Results

Descriptive findings In the pooled SHARE sample, 10 % of men and 12 % of women

have no living children (see Table 2a), which is very similar to levels of childlessness in

older Europeans reported elsewhere (e.g. Dykstra 2009: Table 30.1; Rowland 2007).

However, the proportions of childless individuals vary a lot between the 15 countries

examined here. Looking at current childlessness among men, Israel stands out with an

extremely low share of only 3 % of older men reporting to have no children. The lowest

proportion of men’s childlessness in Europe is found in the Czech Republic (6 %), whereas

the prevalence of childlessness is highest in Austria (13 %) and Ireland (15 %). Turning to

the prevalence of childlessness among older women, we, again, find the highest proportions

in Austria (16 %) and Ireland (20 %), and the lowest ones in the Czech Republic (7 %),

closely followed by Poland (7 %) and Greece (8 %).

Elders’ childlessness also varies by cohort, at least if women are considered. Consistent

with previous research suggesting higher levels of childlessness in older cohorts (e.g.

Rowland 2007), we find that 15 % of women born between 1900 and 1928 are childless,

10 % of the cohorts 1929–1935, and only 8 % of the cohorts born 1936–1942. Some

K. Hank, M. Wagner

123

Table 1 Pooled descriptivesample statistics (unweighted)

Variable Percentage (except‘depression’ & ‘age’)

Dependent variables

(Fairly) Easy to make ends meet 58.4

Homeowner 68.8

Depression (range: 0–12) 2.6 (2.4)

Social participation 28.7

Explanatory and control variables

Gender (female) 54.2

Parental status

Childless 8.8

1 Child 17.8

2 Children 34.5

3 Children 20.3

4? Children 18.4

Marital status

Married 65.4

Never married 4.4

Divorced 3.9

Widowed 26.3

Age (in years) 73.8 (6.7)

Self-perceived health: very good or excellent 20.1

Level of education

Low 60.7

Medium 25.7

High 13.4

Ever did any paid work 88.1

Survey year

2004 54.4

2005 14.0

2006 9.6

2007 26.0

Country

Austria 4.6

Germany 8.9

Sweden 8.7

Netherlands 7.0

Spain 8.1

Italy 8.6

France 8.7

Denmark 6.0

Greece 7.6

Switzerland 3.7

Belgium 9.0

Parenthood, Marital Status

123

cautions seems necessary, though, because there might be some bias in the results for the

oldest cohort, because our measure of ‘current’ childlessness does not account for children

who died before the time of the SHARE interview. In any case, we find indication for a

remarkable change of gender differences in childlessness: in the oldest cohort women are

more often childless than men (15 vs. 10 %), whereas in the youngest cohort the reverse

picture (8 vs. 10 %) appears to be true.

Eventually, we also contrasted men’s and women’s marital status (see Table 2b). Due to

women’s significantly higher life-expectancy, the proportion of older men being (still)

Table 2 Childlessness and marital status by gender, cohort, and country (95 % confidence intervals inparentheses)

(a) Childless (b) Married

Men Women Men Women

Pooled sample .098 (.092–.104) .112 (.107–.118) .769 (.760–.778) .418 (.409–.428)

Cohort

1900–1928 .094 (.083–.105) .151 (.139–.163) .682 (.663–.699) .224 (.211–.238)

1929–1935 .099 (.089–.110) .101 (.091–.111) .797 (.783–.810) .493 (.477–.509)

1936–1942 .105 (.094–.116) .082 (.073–.092) .819 (.806–.833) .624 (.608–.641)

Country

Austria .132 (.097–.166) .164 (.133–.194) .734 (.689–.779) .324 (.285–.362)

Germany .106 (-085–.126) . 110 (.090–.130) .750 (.721–.779) .402 (.371–.434)

Sweden .094 (.074–.113) .110 (.090–.131) .713 (.683–.743) .411 (.379–.443)

Netherlands .078 (.058–.098) .130 (.106–.155) .751 (.719–.784) .445 (.409–.481)

Spain .095 (.074–.116) .132 (.109–.153) .796 (.767–.825) .436 (.404–.469)

Italy .111 (.090–.132) .136 (.114–.159) .753 (.724–.782) .419 (.386–.452)

France .103 (.081–.125) .118 (.098–.138) .796 (.767–.825) .441 (.411–.472)

Denmark .082 (.059–.105) .101 (.078–.124) .727 (.688–.763) .418 (.381–.455)

Greece .080 (.060–.100) .081 (.063–.100) .836 (.808–.863) .420 (.387–.453)

Switzerland .112 (.079–.146) .123 (.091–.155) .771 (.727–.816) .460 (.411–.508)

Belgium .107 (.086–.128) .105 (.086–.124) .800 (.773–.828) .517 (.486–.548)

Israel .031 (.016–.045) .104 (.079–.129) .809 (.777–.841) .500 (.459–.541)

Czech Republic .060 (.039–.081) .066 (.047–.085) .761 (.724–.798) .364 (.378–.400)

Poland .075 (.052–.098) .069 (.048–.090) .772 (.735–.810) .360 (.321–.399)

Irelanda .153 (.105–.201) .198 (.149–.247) .698 (.637–.759) .465 (.404–.526)

Source: Baseline interviews from SHARE, 2004–2007; cross-sectional weights applieda No weights available

Table 1 continuedVariable Percentage (except

‘depression’ & ‘age’)

Israel 5.7

Czech Republic 5.8

Poland 5.3

Ireland 2.4

N 20,377

K. Hank, M. Wagner

123

married is much higher than among their female counterparts (77 vs. 42 %). Unsurpris-

ingly, this pattern is consistent across cohorts and countries. We do observe, however,

some marital status differences between countries. While in Belgium and Israel, for

example, roughly 80 % of men and 50 % of women reported being married, only 72 % of

Scandinavian men and 36 % of Czech and Polish women have been married at the time of

the SHARE interview.

Multivariate analysis Our main set of models suggests only relatively minor differences

in the well-being of parents and their childless counterparts (see Table 3). With regard to

economic-well-being, for example, only parents with four or more children are signifi-

cantly less likely to make ends meet easily than men and women without any children.

Turning to psychological well-being, we observe a non-linear association between the

number of children and the number of depressive symptoms: parents of two children report

the lowest level of psychological distress, whereas those with fewer or more children do

not differ significantly from the childless. Finally, mothers of three children—and only

those—are more likely than childless women to report having participated in any social

Table 3 Multivariate regression results for various well-being outcomes in later life, accounting forparental and marital status

Income adequacy Homeowner Depression Participation

Male Female Male Female Male Female Male Female

Parental status

Childless Ref. Ref. Ref. Ref. Ref. Ref. Ref. Ref.

1 Child 0.89 0.96 0.89 0.87 -0.17 -0.06 0.91 0.85

(0.100) (0.086) (0.099) (0.078) (0.097) (0.093) (0.101) (0.080)

2 Children 0.99 0.97 1.08 0.95 -0.25** -0.20* 1.10 1.01

(0.103) (0.081) (0.113) (0.081) (0.089) (0.087) (0.112) (0.088)

3 Children 0.92 0.95 1.05 0.99 -0.12 -0.14 1.13 1.22*

(0.101) (0.085) (0.116) (0.090) (0.095) (0.093) (0.122) (0.112)

4? Children 0.70** 0.74** 0.90 0.86 -0.09 -0.09 1.22 1.04

(0.079) (0.067) (0.102) (0.079) (0.097) (0.095) (0.134) (0.099)

Marital status

Married Ref. Ref. Ref. Ref. Ref. Ref. Ref. Ref.

Never married 0.90 0.52** 0.34** 0.32** -0.10 0.09 1.06 1.31*

(0.135) (0.062) (0.049) (0.038) (0.132) (0.125) (0.160) (0.160)

Divorced 0.79 0.32** 0.30** 0.22** 0.27* 0.42** 0.86 1.10

(0.113) (0.032) (0.040) (0.023) (0.125) (0.108) (0.119) (0.114)

Widowed 1.16 0.64** 0.51** 0.47** 0.50** 0.40** 1.12 1.21**

(0.093) (0.032) (0.039) (0.024) (0.070) (0.052) (0.089) (0.063)

N 9,193 10,838 9,193 10,840 8,989 10,563 9,236 10,897

Source: Baseline interviews from SHARE, 2004–2007

OLS estimates for the dependent variable ‘depression’, odds ratios for all others. Standard errors are inparentheses. All models control for age, health, education, employment, country, and interview year

**p \ 0.01, *p \ 0.05

Parenthood, Marital Status

123

activity; there are no statistically significant differences at all between fathers and childless

men.

Marital status appears to be related somewhat more systematically to elders’ well-being

(Table 3). The level of economic well-being of the unmarried (i.e. widowed, divorced, or

never married) generally tends to be significantly lower than among the married. The only

exception here is men’s perception of income adequacy, where we do not observe any

differences by marital status. Marital disruption (i.e. widowhood and divorce) is paralleled

by higher scores on the EURO-D scale in women and men. The never-married, however,

do not exhibit higher risks of depression than the currently married. Moreover, never

married and divorced women exhibit higher odds of participation in social activities.

Based on a subsample of SHARE, we eventually turned to the issue of whether par-

enthood and marital status as such bear associations with older people’s well-being, or

whether (to what extent, respectively) relationship quality matters here (cf. Ryan and

Willits 2007); see Table 4. There is no indication for a role of parent–child relationship

quality in elders’ well-being. Turning to spouses’ relationship quality, the most important

finding is that it does not seem to be the mere presence of a spouse that protects against

depression in later life, but that only men and women living in a marital union charac-

terized by reciprocity enjoy greater psychological well-being than the unmarried (see Choi

and Marks 2008, for a related discussion). We did not detect any further systematic well-

being differences along the lines of different levels of partnership quality.

Table 4 Multivariate regression results for various well-being outcomes in later life, accounting forparental and marital status as well as relationship quality

Income adequacy Homeowner Depression Participation

Male Female Male Female Male Female Male Female

Parental status and relationship quality

Childless Ref. Ref. Ref. Ref. Ref. Ref. Ref. Ref.

Parents w/out 0.65** 0.89 1.12 0.86 0.14 0.12 0.97 0.95

Frequent contacts (0.096) (0.107) (0.154) (0.102) (0.115) (0.120) (0.130) (0.113)

Parents w/ 0.66** 0.88 1.23 0.91 -0.07 -0.02 1.11 0.94

Frequent contacts (0.089) (0.091) (0.155) (0.094) (0.105) (0.103) (0.136) (0.097)

Marital status and relationship quality

Unmarrieda Ref. Ref. Ref. Ref. Ref. Ref. Ref. Ref.

Married and 1.08 1.16 2.15** 2.25** -0.05 0.29* 0.92 0.90

unsatisfied w/reciprocity (0.153) (0.135) (0.304) (0.280) (0.114) (0.118) (0.122) (0.105)

Married andsatisfiedw/reciprocity

1.16 1.69** 2.17** 2.64** -0.56** -0.58** 1.01 0.86*

(0.112) (0.126) (0.202) (0.204) (0.078) (0.074) (0.091) (0.063)

N 4,178 4,901 4,178 4,902 4,152 4,851 4,207 4,924

Source: Baseline interviews from SHARE, 2004–2007

OLS estimates for the dependent variable ‘depression’, odds ratios for all others. Standard errors are inparentheses. All models control for age, health, education, employment, country (excl. Czech Republic,Ireland, and Poland), and interview year

**p \ 0.01, *p \ 0.05a The reference group ‘unmarried’ comprises never married, divorced, and widowed respondents

K. Hank, M. Wagner

123

5 Discussion

While the majority of elders in contemporary Europe have at least one child, a non-

negligible minority of nearly 11 % is childless at age 65, with a somewhat higher prev-

alence of childlessness in women than in men. Around this mean value, our analysis of

SHARE data revealed considerable cross-country variation: while childlessness is an

almost unknown phenomenon among Israeli men, for example, almost 20 % of Irish

women do not report any living offspring. We also observe lower levels of childlessness in

‘younger’ cohorts (born around 1940) compared to the oldest cohorts in our study (born in

the first quarter of the twentieth century). However, the proportions of childless men and

women in more recent cohorts have increased again and are expected to rise even further in

future generations (e.g. Dykstra and Hagestad 2007; Rowland 2007). It has been the main

aim of this study to investigate whether this development, paralleled by decreasing union

stability, should give reason for concern regarding individuals’ well-being in later life. The

main value added by our study to the literature comes from the simultaneous consideration

of (a) various dimensions of well-being, (b) the interplay between parenthood and marital

status, and (c) the role of societal context in the empirical analysis.

Despite the undoubtedly central role of partners and children in providing instrumental,

emotional, and economic support in old age (e.g. Brandt et al. 2009), our main findings are,

by and large, reassuring: First, childless individuals do not generally fare worse than

parents in terms of their economic, psychological, or social well-being. Second, the

observed correlation of marital status with elders’ economic and psychological well-being

suggests a ‘protective effect’ of marriage. For example, divorce and widowhood in par-

ticular are paralleled by a higher number of depression symptoms. However, different from

findings reported by Hughes and Waite (2009), for example, never-married men and

women do not seem to be under a greater risk of psychological distress than their currently

married counterparts, which might reflect adaptation processes across the life-course. That

is, especially if individuals remained unmarried voluntarily throughout their life, they are

likely to have developed coping strategies independent of support from a spouse. Such an

interpretation is consistent with recent evidence suggesting that individuals embedded in

network types characterized by greater social capital tend to express a superior sense of

subjective well-being (Litwin and Shiovitz-Ezra 2011), that is, although family relations

are characterized by a particular form of solidarity they are by no means an exclusive

source of support for elders and can—to varying degrees—be substituted by non-kin social

relationships (also see Kohli et al. 2009). Moreover, once we take into consideration

individuals’ relationship quality, it becomes quite clear that marriage per se does not

contribute to psychological well-being: only those reporting satisfaction with the extent of

reciprocity in their relationship exhibit lower odds of depression than their unmarried

counterparts. That is, union disruption as such might not be a challenge to well-being,

whereas marital conflict has been shown to be a significant risk factor for health among

older adults (Choi and Marks 2008).

A third important finding is that the results reported above appear to hold for a variety

of (sub-)populations (but see Huijts et al. 2011). Estimating all our models separately for

men and women, for example, did not reveal any noteworthy gender differences. Further

analyses (whose details we do not report here) neither provided any indication for effects

of a changing cultural meaning of childlessness across cohorts, i.e. over time (e.g. Umb-

erson et al. 2010: 614), nor did they suggest any systematic geographic patterns along the

lines of familiar welfare or family regime typologies. Our findings thus support the notion

Parenthood, Marital Status

123

of a context-independent association between parental status (marital status, respectively)

and the well-being measures considered here.

It might still be worth investigating further the potential role of societal context in

shaping the role of family ties in individuals’ well-being (e.g. Moor and Komter 2012).

Data limitations, however, prohibited us from pursuing this line of more in-depth country-

specific analysis. Unfortunately, the number of observations in SHARE’s national samples

is relatively small. Researchers might thus want to explore opportunities to exploit larger

national ageing surveys (such as the ‘English Longitudinal Study of Ageing’ or the

‘German Ageing Survey’), although the extent to which they are suitable for cross-

nationally comparative analyses might be limited. Also due to SHARE’s restricted sample

size (even if all countries are pooled!), we were unable to account for the potential role of

interactions between legal marital status and unmarried cohabitation in elders’ well-being.

Non-marital unions (including ‘living apart together’) are likely to spread more quickly in

Northern Europe’s older population than in the South. Moreover, previous research sug-

gests regional differences in the association of kin and non-kin social networks with elders’

well-being (Litwin 2010; see Sundstrom et al. 2009, for a related discussion). Therefore it

seems worthwhile to investigate further issues of complementarity and substitution

between different partnership modes (social network types, respectively) from a cross-

nationally comparative perspective (cf. Kohli et al. 2009). Finally, future studies should

also aim at considering more details of individuals’ increasingly complex life-histories,

which would allow investigating the relationship between the sequencing and timing of

family events (such as early or late entries into parenthood; e.g. Hofferth 1984) on the one

hand, and later life outcomes on the other hand.

Acknowledgments This paper uses data from SHARE release 2.3.0, as of November 13th 2009. SHAREdata collection in 2004-2007 was primarily funded by the European Commission through its 5th and 6thframework programs (project numbers QLK6-CT-2001- 00360; RII-CT- 2006-062193; CIT5-CT-2005-028857). Additional funding by the US National Institute on Aging (grant numbers U01 AG09740-13S2;P01 AG005842; P01 AG08291; P30 AG12815; Y1-AG-4553-01; OGHA 04-064; R21 AG025169) as wellas by various national sources is gratefully acknowledged (see http://www.share-project.org for a full list offunding institutions).

References

Aassve, A., Mazzuco, S., & Mencarini, L. (2005). Childbearing and well-being: A comparative analysis ofEuropean welfare regimes. Journal of European Social Policy, 15, 283–299.

Bengtson, V. L., Giarusso, R., Marby, J. B., & Silverstein, M. (2002). Solidarity, conflict, and ambivalence:Complementary of competing perspective on intergenerational relations? Journal of Marriage andFamily, 64, 568–576.

Bengtson, V. L., & Roberts, R. E. L. (1991). Intergenerational solidarity in aging families: An example offormal theory construction. Journal of Marriage and the Family, 53, 856–870.

Borsch-Supan, A., Hank, K., & Jurges, H. (2005). A new comprehensive and international view on ageing:Introducing the survey of health, ageing and retirement in Europe. European Journal of Ageing, 2,245–253.

Borsch-Supan, A., Hank, K., Jurges, H., & Schroder, M. (2010). Longitudinal data collection in continentalEurope: Experiences from the survey of health, ageing and retirement in Europe. In J. A. Harkness,et al. (Eds.), Survey methods in multinational, multiregional, and multicultural contexts (pp. 507–514).Hoboken, NJ: Wiley.

Brandt, M., Haberkern, K., & Szydlik, M. (2009). Intergenerational help and care in Europe. EuropeanSociological Review, 25, 585–601.

Brown, S. L., Lee, G. R., & Bulanda, J. R. (2006). Cohabitation among older adults: A national portrait.Journal of Gerontology: Social Sciences, 61B, S71–S79.

K. Hank, M. Wagner

123

Buber, I., & Engelhardt, H. (2008). Children’s impact on the mental health of their older mothers andfathers: Findings from the Survey of Health, Ageing and Retirement in Europe. European Journal ofAgeing, 5, 31–45.

Bures, R. M., Koropeckyj-Cox, T., & Loree, M. (2009). Childlessness, parenthood, and depressive symp-toms among middle-aged and older adults. Journal of Family Issues, 30, 670–687.

Burkauser, R. V., Giles, P., Lillard, D. R., & Schwarze, J. (2005). Until death us do part: An analysis of theeconomic well-being of widows in four countries. Journal of Gerontology: Social Sciences, 60B,S238–S246.

Castro-Costa, E., Dewey, M., Stewart, R., Banerjee, S., Huppert, F., Medonca-Lima, C., et al. (2008).Ascertaining late-life depressive symptoms in Europe: an evaluation of the survey version of theEURO-D scale in 10 nations. The SHARE project. International Journal of Methods in PsychiatricResearch, 17, 12–29.

Choi, N. G. (1994). Patterns and determinants of service utilization: comparisons of the childless elderly andelderly parents living with or apart from their children. The Gerontologist, 34, 353–362.

Choi, H., & Marks, N. F. (2008). Marital conflict, depressive symptoms, and functional impairment. Journalof Marriage and Family, 70, 377–390.

Christelis, D., Jappelli, T., Paccagnella, O., & Weber, G. (2009). Income, wealth and financial fragility inEurope. Journal of European Social Policy, 19, 359–376.

Coppola, L., & Di Cesare, M. (2008). How fertility and union stability interact in shaping new familypatterns in Italy and Spain. Demographic Research, 18, 117–144.

de Jong Gierveld, J. (2004). Remarriage, unmarried cohabitation, living apart together: Partner relationshipsfollowing bereavement or divorce. Journal of Marriage and Family, 66, 236–243.

Dewilde, C., Van den Bosch, K., & Van den Heede, A. (2011). Separation: Consequences for wealth in laterlife. In M. Brandt, K. Hank, M. Schroder, & A. Borsch-Supan (Eds.), The individual and the WelfareState: Life histories in Europe (pp. 103–114). New York: Springer.

Dingeldey, I. (2001). European tax systems and their impact on family employment patterns. Journal ofSocial Policy, 30, 653–672.

Donelly, E. A., & Hinterlong, J. E. (2010). Changes in social participation and volunteer activity amongrecently widowed older adults. The Gerontologist, 50, 158–169.

Dykstra, P. A. (2009). Childless old age. In: P. Uhlenberg (Ed.), International handbook of population aging(pp. 671–690). New York: Springer.

Dykstra, P. A., & Hagestad, G. O. (2007). Childlessness and parenthood in two centuries. Different roads—Different maps? Journal of Family Issues, 28, 1518–1532.

Dykstra, P. A., & Wagner, M. (2007). Pathways to childlessness and late-life outcomes. Journal of FamilyIssues, 28, 1487–1517.

Erel, O., & Burman, B. (1995). Interrelatedness of marital relations and parent-child relations: A meta-analytic review. Psychological Bulletin, 118, 108–132.

Erlinghagen, M., & Hank, K. (2006). The participation of older Europeans in volunteer work. Ageing &Society, 26, 567–584.

Evenson, R. J., & Simon, R. W. (2005). Clarifying the relationship between parenthood and depression.Journal of Health and Social Behavior, 46, 341–358.

Furstenberg, F. F. (2005). Banking on families: How families generate and distribute social capital. Journalof Marriage and Family, 67, 809–821.

Hallerod, B., & Selden, D. (2012). The multidimensional characteristics of wellbeing: how different aspectof wellbeing interact and do not interact with each other. Social Indicators Research [online advanceaccess].

Hank, K. (2007). Proximity and contacts between older parents and their children: A European comparison.Journal of Marriage and Family, 69, 157–173.

Hank, K. (2010). Childbearing history, later-life health, and mortality in Germany. Population Studies, 64,275–291.

Hansen, T., Slagsvold, B., & Moum, T. (2009). Childlessness and psychological well-being in midlife andold age: An examination of parental status effects across a range of outcomes. Social IndicatorsResearch, 94, 343–362.

Hawkins, D. N., & Booth, A. (2005). Unhappily ever after: Effects of long-term, low-quality marriages onwell-being. Social Forces, 84, 445–465.

Henretta, J. C., Grundy, E. M. D., Okell, L. C., & Wadsworth, M. E. J. (2008). Early motherhood and mentalhealth in midlife: A study of British and American cohorts. Aging & Mental Health, 12, 605–614.

Hofferth, S. L. (1984). Long-term economic consequences for women of delayed childbearing and reducedfamily size. Demography, 21, 141–155.

Parenthood, Marital Status

123

Holden, K. C., & Kuo, H.-H. D. (1996). Complex marital histories and economic well-being: The continuinglegacy of divorce and widowhood as the HRS cohort approaches retirement. The Gerontologist, 36,383–390.

Hughes, M. E., & Waite, L. J. (2009). Marital biography and health at mid-life. Journal of Health and SocialBehavior, 50, 344–358.

Huijts, T., Kraaykamp, G., & Subramanian, S. V. (2011). Childlessness and psychological well-being incontext: A multilevel study on 24 European countries. European Sociological Review [online advanceaccess].

Johnson, R. W., & Favreault, M. M. (2004). Economic status in later life among women who raised childrenoutside of marriage. Journal of Gerontology: Social Sciences, 59B, S315–S323.

Keizer, R., Dykstra, P. A., & Poortman, A.-R. (2010). Life outcomes of childless men and fathers. EuropeanSociological Review, 26, 1–15.

Kiernan, K. (1989). Who remains childless? Journal of Biosocial Science, 21, 387–398.Klein, T. (2003). Die Geburt von Kindern in paarbezogener perspektive (Fertility in male-female partner-

ships). Zeitschrift fur Soziologie, 32, 506–527.Kohli, M., Hank, K., & Kunemund, H. (2009). The social connectedness of older Europeans: Patterns,

dynamics and contexts. Journal of European Social Policy, 19, 327–340.Koropeckyj-Cox, T. (2002). Beyond parental status: Psychological well-being in middle and old age.

Journal of Marriage and Family, 64, 957–971.Koropeckyj-Cox, T., Pienta, A. M., & Brown, T. H. (2007). Women of the 1950s and the ‘normative’ life

course: the implications of childlessness, fertility timing, and marital status for psychological well-being in late midlife. International Journal of Aging and Human Development, 64, 299–330.

Lee, G. R., & DeMaris, A. (2007). Widowhood, gender, and depression. A longitudinal analysis. Researchon Aging, 29, 56–72.

Letherby, G. (2002). Childless and bereft? Stereotypes and realities in relation to ‘‘voluntary’’ and‘‘involuntary’’ childlessness and womanhood. Sociological Inquiry, 72, 7–20.

Litwin, H. (2010). Social networks and well-being: a comparison of older people in mediterranean and non-mediterranean countries. Journal of Gerontology: Social Sciences, 65B, 599–608.

Litwin, H., & Sapir, E. V. (2009). Perceived income adequacy among older adults in 12 countries: Findingsfrom the survey of health, ageing and retirement in Europe. The Gerontologist, 49, 397–407.

Litwin, H., & Shiovitz-Ezra, S. (2011). Social network type and subjective well-being in a national sampleof older Americans. The Gerontologist, 51, 379–388.

Martınez-Granado, M., & Mira, P. (2005). The number of living children. In A. Borsch-Supan, et al. (Eds.),Health, ageing and retirement in Europe—First results from SHARE (pp. 48–52). Mannheim: MEA.

Moor, N., & Komter, A. (2012). Family ties and depressive symptoms in Eastern and Western Europe.Demographic Research, 27, 201–232.

Pichler, F., & Wallace, C. (2007). Patterns of formal and informal social capital in Europe. EuropeanSociological Review, 23, 423–435.

Plotnick, R. D. (2009). Childlessness and the economic well-being of older Americans. Journal of Ger-ontology: Social Sciences, 64B, 767–776.

Rendall, M. S., & Bahchieva, R. A. (1998). An old-age security motive for fertility in the United States?Population and Development Review, 24, 293–307.

Ross, C. E., Mirowsky, J., & Goldsteen, K. (1990). The impact of family on health: The decade in review.Journal of Marriage and the Family, 52, 1059–1078.

Rotolo, T., & Wilson, J. (2006). Substitute or complement? Spousal influence on volunteering. Journal ofMarriage and Family, 68, 305–319.

Rowland, D. T. (2007). Historical trends in childlessness. Journal of Family Issues, 28, 1311–1337.Ryan, A. K., & Willits, F. K. (2007). Family ties, physical health, and psychological well-being. Journal of

Aging and Health, 19, 907–920.Schaan, B. (2009). Verwitwung, Geschlecht und Depression im hoheren Lebensalter (Widowhood, gender,

and depression at older ages). In A. Borsch-Supan, K. Hank, H. Jurges, & M. Schroder (Eds.), 50plusin Deutschland und Europa. Ergebnisse des survey of health, ageing and retirement in Europe (pp.115–131). Wiesbaden: VS Verlag.

Schutze, Y., & Wagner, M. (1991). Sozialstrukturelle, normative und emotionale Determinanten der Be-ziehungen zwischen erwachsenen Kindern und ihren alten Eltern (Social-structural, normative, andemotional determinants of the relations between adult children and their elder parents). Zeitschrift furSozialisationsforschung und Erziehungssoziologie, 11, 295–313.

Sigle-Rushton, W. (2005). Young fatherhood and subsequent disadvantage in the United Kingdom. Journalof Marriage and Family, 67, 735–753.

K. Hank, M. Wagner

123

Sigle-Rushton, W., & Waldfogel, J. (2007). The incomes of families with children: A cross-nationalcomparison. Journal of European Social Policy, 17, 299–318.

Sundstrom, G., Fransson, E., Malmberg, B., & Davey, A. (2009). Loneliness among older Europeans.European Journal of Ageing, 6, 267–275.

Umberson, D., Pudrovska, T., & Reczek, C. (2010). Parenthood, childlessness, and well-being: A life courseperspective. Journal of Marriage and Family, 72, 612–629.

Utz, R. L., Carr, D., Nesse, R., & Wortman, C. B. (2002). The effect of widowhood on older adults‘socialparticipation: an evaluation of activity, disengagement, and continuity theories. The Gerontologist, 42,522–533.

Wagner, M. (1997). Uber die bedeutung von partnerschaft und elternschaft im alter. In J. Mansel, G.Rosenthal, & A. Tolke (Eds.), Generationen-beziehungen, austausch und tradierung (pp. 121–136).Opladen: Westdeutscher Verlag.

Waldron, I., Hughes, M., & Brooks, T. (1996). Marriage protection and marriage selection—prospectiveevidence for reciprocal effects of marital status and health. Social Science and Medicine, 43, 113–123.

Wenger, G. C., Dykstra, P. A., Melkas, T., & Knipsheer, K. C. P. M. (2007). Social embeddedness and late-life parenthood. Community activity, close ties, and support networks. Journal of Family Issues, 28,1419–1456.

Zhang, Z., & Hayward, M. D. (2001). Childlessness and the psychological well-being of older persons.Journal of Gerontology: Social Sciences, 56B, S311–S320.

Parenthood, Marital Status

123