ArticlePDF Available

Life Satisfaction and Grandparenthood: Evidence from a Nationwide Survey

Authors:

Abstract and Figures

This paper tests whether there is a potential payoff to grandparenthood in terms of life satisfaction. Using the new nationwide survey for the UK, which consists of over 5,000 grandparents and 6,000 non-grandparents aged 40 and above, and a flexible multiple-index ordered probit model with varying thresholds, we find that being a grandparent to at least one grandchild is associated positively and statistically significantly with individuals reporting to be very satisfied with life overall. Parents with no grandchildren are no more satisfied with life compared to non-parents of the same age. The findings suggest that even though children may not contribute significantly to parents' satisfaction with life overall, there may well be long-term benefits to having children, provided that our children go on and have children of their own.
Content may be subject to copyright.
DISCUSSION PAPER SERIES
Forschungsinstitut
zur Zukunft der Arbeit
Institute for the Study
of Labor
Life Satisfaction and Grandparenthood:
Evidence from a Nationwide Survey
IZA DP No. 5869
July 2011
Nattavudh Powdthavee
Life Satisfaction and Grandparenthood:
Evidence from a Nationwide Survey
Nattavudh Powdthavee
Nanyang Technological University
and IZA
Discussion Paper No. 5869
July 2011
IZA
P.O. Box 7240
53072 Bonn
Germany
Phone: +49-228-3894-0
Fax: +49-228-3894-180
E-mail: iza@iza.org
Any opinions expressed here are those of the author(s) and not those of IZA. Research published in
this series may include views on policy, but the institute itself takes no institutional policy positions.
The Institute for the Study of Labor (IZA) in Bonn is a local and virtual international research center
and a place of communication between science, politics and business. IZA is an independent nonprofit
organization supported by Deutsche Post Foundation. The center is associated with the University of
Bonn and offers a stimulating research environment through its international network, workshops and
conferences, data service, project support, research visits and doctoral program. IZA engages in (i)
original and internationally competitive research in all fields of labor economics, (ii) development of
policy concepts, and (iii) dissemination of research results and concepts to the interested public.
IZA Discussion Papers often represent preliminary work and are circulated to encourage discussion.
Citation of such a paper should account for its provisional character. A revised version may be
available directly from the author.
IZA Discussion Paper No. 5869
July 2011
ABSTRACT
Life Satisfaction and Grandparenthood:
Evidence from a Nationwide Survey
This paper tests whether there is a potential payoff to grandparenthood in terms of life
satisfaction. Using the new nationwide survey for the UK, which consists of over 5,000
grandparents and 6,000 non-grandparents aged 40 and above, and a flexible multiple-index
ordered probit model with varying thresholds, we find that being a grandparent to at least one
grandchild is associated positively and statistically significantly with individuals reporting to be
very satisfied with life overall. Parents with no grandchildren are no more satisfied with life
compared to non-parents of the same age. The findings suggest that even though children
may not contribute significantly to parents’ satisfaction with life overall, there may well be
long-term benefits to having children, provided that our children go on and have children of
their own.
JEL Classification: J13
Keywords: life satisfaction, grandparenthood, grandchildren, generalized ordered probit,
understanding society, happiness
Corresponding author:
Nattavudh Powdthavee
Division of Economics
School of Humanities & Social Sciences
Nanyang Technological University
14 Nanyang Drive
Singapore, 637332
E-mail: n.powdthavee@ntu.edu.sg
3
1. Introduction
There are perhaps no other studies in the psychological literature on well-being deemed as
controversial by the general public as those which had found that children do not generally
make us happy or becoming more satisfied with our life overall. Apart from a few notable
exceptions (Kohler et al, 2005; Margolis & Myrskylä, 2011), research in social science has
found consistent evidence of either a zero or a negative correlation between the presence of
children in a family and the respondent‟s self-rated well-being. For example, studies in the
US and Europe have found that parents often reported slightly though statistically
significantly lower levels of cognitive well-being such as marital satisfaction (White et al,
1986; Twenge et al, 2003) and life satisfaction (Powdthavee, 2008; Clark et al, 2008; Stanca,
2011), as well as affective well-being such as happiness (Alesina et al, 2004), mental health
states (Cleary & Mechanic, 1983; Shields & Wheatley-Price, 2001; Clark & Oswald, 2002)
and moment-to-moment feelings (Kahneman et al, 2004; White & Dolan, 2009) compared to
nonparents. The finding‟s unpopularity has further been bolstered by the evidence that the
strains associated with parenthood are not only limited to the period during which children
are physically and economically dependent. For example, Glenn and McLanahan (1981)
found those older parents whose children have left home report statistically the same level of
happiness compared to nonparents of similar age and status.
One of the main reasons why many people have found the insignificant and
sometimes negative correlation between the happiness and life satisfaction of parents and
their fertility controversial is because it is deeply counterintuitive: Most of us would argue
that although raising kids is hard work, we are nevertheless happy with our children, for our
children and because of our children. Such beliefs are strong and prevalent across cultures,
age groups, and genders (see, e.g., Blake, 1979; Baumeister, 1991). They are, however,
4
frequently at odds with the scientific data. For example, when Americans were asked how
their children have affected their marital lives, the majority concluded that the presence of
children had affected their marriage in the most positive light possible. Yet when the same
individuals were asked to report how satisfied they are with their marriage they ended up
reporting lower levels of marital satisfaction than those who are either childless or childfree
(Glenn & McLanahan, 1982).
According to psychologist Daniel Gilbert (2006), the discrepancy between what we
believe (Children should make us happier!”) and what the data actually tells us could be
explained using evolutionary as well as psychological theories. For example, given that
people who believe that there is no joy in parenthood and who thus stop having children
are less likely to pass on their beliefs any further beyond their generation, the belief that
„children bring happiness and satisfaction will transmit itself much more successfully from
generation to generation than the belief that „children bring misery‟. In other words, only the
beliefs that have the best chance of transmission, even if they are faulty ones, will be passed
on. In addition to this, when we believe that something makes us happy, we are willing to pay
a high price for it. Yet it is also often the case that when we pay a high price for something,
we rationalize that its possession makes us happier than when we did not possess it. And
given that evolution passes on this unconditional and invariable compulsion to care for our
children, it is therefore not surprising that we tend to rationalize those costs and conclude that
our children must be repaying us with a deep sense of satisfaction.
More recently, the theory on focusing illusion (FT) has been put forward by a number
of scholars as one of the key psychological explanations for why we tend to over-predict the
impacts of many events in our lives, including the positive effects of children on how parents
evaluate how their life has turned out (see, e.g., Schkade & Kahneman, 1999; Powdthavee,
2009). When prompted to think about parenthood either imagining future offspring or
5
thinking about their current ones the majority of people will tend to focus more of their
attention on the good and salient things about being a parent (e.g., seeing our kids smile for
the first time) and less so on the bad and seemingly trivial things about being a parent (e.g.,
the time spent changing dirty nappies and the frequent anxiety about the welfare of our
children), partly because of the transmitted belief that children bring happiness. However,
this does not negate the fact that the less salient experiences of parenthood e.g., having
more housework to do (Sanchez & Thomson, 1997), time spent worrying about the
household finance (Stanca, 2011), and less quality time with spouse (Crohan, 1996; Lavee et
al, 1996) do add up and are therefore likely to have daily emotional consequences. And
given that too little weight will often be placed on the less salient experiences about being a
parent, it should be no surprise why we tend to over-estimate the impacts children have, or
could potentially have, on our overall life satisfaction whenever we are prompted to think
about them.
What FI is implying is that the relationship between parenthood and life satisfaction is
likely to be mediated by how the presence of children in the family affects parents‟ time-use
and how their attentions are normally allocated on a daily basis. When individuals spend
most of their time tending to the very core process of child care and being directly
responsible for the child‟s well-being, then it may well be the case that they will always
report significantly lower levels of life satisfaction compared to non-parents irrespective to
what they might say when asked how much of their happiness can be credited to their
children. This raises an important social science question: If our prediction about the impact
of children on parents‟ life satisfaction is made incorrectly, could our prediction about the
impact of grandchildren on grandparents‟ life satisfaction be any better? The null hypothesis,
according to the FT theory, is that they should. Provided that most grandparents especially
in the western society are often free from the direct financial and non-financial
6
responsibilities to care for their grandchildren‟s welfare, e.g., they do not have to tend to the
very core process of childcare with their grandchildren as they would have with their own
children, it may well be that the net effect of children on grandparents‟ satisfaction with life
overall is in fact positive and statistically significantly different from zero.
Due to data limitation, empirical evidence on the relationship between life satisfaction
and grandparenthood is scarce and outdated. Two notable studies in this area are Kivnick
(1982) and Thomas (1989). Based on qualitative data gathered from 30 grandparents and
quantitative data gathered from 286 grandparents, Kivnick found activity level in the
grandparent role to be generally unrelated to life satisfaction. By contrast, Thomas (1989)
interviewed 301 grandparents about their relationship with their grandchildren and found that
there is a positive relationship between the components of being a grandparent e.g.,
indulging grandchildren and the feeling of immortality through their grandchild and life
satisfaction. Yet the above studies are based on very small samples of grandparents with no
appropriate control groups (i.e., non-grandparents) and primarily qualitative in nature.
The current study aims to fill this research void by using the new large-scale British
longitudinal data set to test whether there is any significant relationship between life
satisfaction and grandparenthood. It estimates, perhaps the first of its kind, generalized
ordered model of life satisfaction with a set of parenthood and grandparenthood statuses as
explanatory variables. Thus, this paper attempts to contribute to the existing literature by
exploring not only the possibility that grandchildren can impact our overall life satisfaction,
but also whether the effect is greater at reducing dissatisfaction or at increasing the
probability that the individual will report to be very satisfied with his or her life.
The research question of whether grandchildren contribute positively and significantly
to the life satisfaction of grandparents is also important in its own right. Evidence of a
7
positive relationship between grandparenthood and life satisfaction will imply that there is
indeed a long-run psychological payoff to the investment of children, one that skips a
generation.
2. Data
The data in this study comes from Wave 1 of the Understanding Society survey
(http://www.understandingsociety.org.uk/). The Understanding Society survey is a major
longitudinal study designed to provide new evidence about the people in the UK, their lives
experiences, behaviours and beliefs. Starting from December 2009, the study follows 100,000
individuals from 40,000 households in Great Britain. The dependent variable used in the
current study come from the responses to the life satisfaction (LS) question.
The LS question, which prompts survey respondents to rate themselves on a 7-point-
scale based on how satisfied or dissatisfied they are with their life overall, is considered by
psychologists as a standard gold measure of a person‟s cognitive well-being (Diener et al,
1985). It is formally defined as a global assessment of a person‟s quality of life according to a
standard which each individual sets for him or herself (Shin & Johnson, 1978). Responses to
life satisfaction question are elicited using the following question: “All things considered,
how satisfied or dissatisfied are you with your life overall using a 1-7 scale? 1 = very
dissatisfied, …, 7 = very satisfied”.
Our main explanatory variable is derived from (i) the “grandchildren living
elsewhere” variable (a_lvrel5), the “children living elsewhere” variable (a_lvrel3), and (iii)
the “relationship within household” variable (a_relationship). Here, the grandchildren
variable takes a value of 1 if the respondent has at least one grandchild living in other
8
households and/or there is at least one identifiable grandchild living in the same household as
the respondent, and 0 otherwise. Similarly, the children variable takes a value of 1 if the
respondent has at least one child living in other households and/or there is at least one
identifiable child living in the same household as the respondent, and 0 otherwise. The two
dummies are then interacted with each other to generate the “children and grandchildren”
statuses, which consists of the following categories:
(a) Neither have children or grandchildren
(b) Have children but not grandchildren
(c) Have grandchildren but no alive children
(d) Have children and grandchildren
The (c) category is generated in order to allow for the confounding factor of losing one‟s
child where there is at least one grandchild present. We restrict the sample to contain
individuals aged 40 and above. This leaves us with a total sample of 11,942 observations. Of
those, 5,246 are males, and 4,930 are over the age of 60. Approximately 36% (N = 4,275)
reported not to have children and grandchildren who are still alive and living together with
the respondent or residing in other households; 21% (N = 2,538) reported to have children
but no grandchildren; 1% (N = 127) had grandchildren but no children who are alive; and
approximately 42% (N = 4,998) of the full sample reported to have both grandchildren and
living children. It is possible to further differentiate the children and grandchildren status into
those living with the respondent and those living in other households. However, given that
the number of individuals who are living with at least one of their grandchildren is very small
(around 6% of the total number of individuals who reported to have at least one grandchild,
i.e. 317 from 5,125), differentiating them further into sub-groups is unlikely to influence our
final estimates in a significant way and the decision is to keep the groupings as they are.
Summary of descriptive statistics can be found in Table A1 in the Appendix.
9
3. Empirical strategy
Most empirical work on the determinants of subjective well-being uses either linear
regression or single-index ordered probit or logit estimators. The current study, however,
follows the empirical strategy outlined in Boes and Winkelmann (2010) and Mentzakis
(2011) and uses the generalised ordered probit (GOPROBIT) to estimate the effects of
grandchildren on LS of grandparents for different parts of the LS distribution. The
GOPROBIT estimator is preferred to other single-index ordered probit or logit models simply
because it relaxes the assumption of implicit cardinalization such that the grandchildren effect
must be constant across the distribution of life satisfaction responses (Boes & Winkelmann,
2006), and therefore enables unrestricted grandchildren effects for low and high levels of LS
to be estimated. In other words, the current study allows for the possibility that grandchildren
may help reduce dissatisfaction i.e., the probability that the respondent will place himself
on the lower rungs of the life satisfaction scale more than they help increase satisfaction for
the respondents.
To formulate the GOPROBIT model, let
},...,{JLSi1
denote self-reported LS of
individual
ni ,...,1
, and
i
x
to represent a vector of covariates which includes the respondent‟s
grandchildren (or grandparenthood) dummies. The relationship between LSi and xi can then
be written in terms of cumulative conditional probabilities as followed:
)(Φ);|( jijii θxθxjLSP
(1)
where
(.)
denotes the cumulative density function of the standard normal distribution, and
j
represents a vector of category-specific parameters including a constant term. In order to
10
ensure positive cell probabilities, it is required that
j
θ
fulfill the strict inequalities
.... 11
Jii θxθx
Rewrite
ji θx
as
jiiji βxαθx~
(2)
Equation (2) becomes the standard ordered probit when all the slope parameters are restricted
to be equal across the well-being distribution, i.e.
11 ...
J
. However, in the generalized
ordered probit such a restriction is not imposed. Rather, a set of coefficients on the covariates,
i
x
, is estimated for each of the J-1 points of the LS scale (for a detailed description of the
GOPROBIT model, see Boes & Winkelmann, 2006). All regressions are estimated using
STATA version 11.1 with robust standard errors.
4. Results
Are grandparents more satisfied with their life compared to non-grandparents? To provide a
first pass to this question, we present in Figure 1 a summary of mean LS scores by
grandchildren statuses for those who are aged 40 and above. A naïve comparison of these
averages tells us that (i) there is no statistical difference in the average LS scores between
non-parents and parents who are non-grandparents, (ii) respondents who have grandchildren
but no living children are significantly dissatisfied with their life, and (iii) grandparents with
living children are clearly the most satisfied people out of the four groups. The grandparents
with living children have an average LS score of 5.45 (S.E. = 0.021), while the average LS
scores for nonparents, parents with no grandchildren, and parents with grandchildren but no
living children are 5.27 (S.E. = 0.013), 5.23 (S.E. = 0.027), and 4.82 (S.E. = 0.142),
respectively. The relationships are, however, likely to be confounded by the age effects, i.e.,
11
older cohorts are likely to become grandparents and are also likely to be more satisfied with
their lives compared to younger cohorts (Blanchflower & Oswald, 2008).
The standard ordered probit (OPROBIT) estimates and the generalized ordered probit
(GPROBIT) estimates on the relationship between LS and grandparenthood are reported in
Table 1. In the generalized model, six different parameters vectors
61 θθ ,...,
are estimated
(where each vector contains coefficients for all the explanatory variables). With no additional
controls (Panel A), the estimated OPROBIT coefficients on “Have children but no
grandchildren” and “Have both children and grandchildren” are positive and statistically
significant at the 1% level, while the coefficient on “Have grandchildren but no living
children” is negative and statistically well-determined at the 5% level. By contrast, the
GOPROBIT coefficients vary in terms of signs and statistical significance across the
parameters. For example, the estimated coefficients on Have both children and
grandchildren” are positive and statistically well-determined at conventional levels for mid to
high LS, i.e.,
6543 θθθθ and ,,
, and negative and statistically significant for low LS, i.e.,
21 θθ and
.
Panel B of Table 1 controls for exogenous variables, adding age, age-squared, race
and gender of the respondent. Apart from the estimated coefficient on “Have grandchildren
but no living children” which continues to be negative and statistically significant at the 1%
level, the OPROBIT coefficients on both i) parents but no grandchildren and ii) parents and
grandparents have now lost their statistical significances once age, age-squared, gender, and
race are held constant in the estimation. By relaxing the parallel regression assumption,
however, the coefficients obtained from the GOPROBIT model suggest that having children,
regardless of whether or not the respondent also has at least one grandchild, continues to be
negative and statistically significant at the 5% level for
21 θθ and
. Yet the point estimates for
12
“Have children but no grandchildren” for
64 θθθ and , 5
turn statistically insignificant when
exogenous variables are controlled for in the estimation. However, the estimated coefficient
on “Have both children and grandchildren” continues to be positive and statistically
significant at the 1% level for
6
θ
(0.118 with a robust standard error of 0.041), while the
estimated coefficients for those who have grandchildren but no living children continue to be
negative and statistically well-determined at conventional levels for
1
θ
through to
5
θ
.
Panel C of Table 1 turns to OPROBIT and GOPROBIT regressions with full
specification, adding dummy variables for marital statuses, employment statuses, subjective
health statuses, and regions, as well as log of household income and age left full-time
education. Whilst many of the coefficients have lost significance, it is interesting to see that,
even with full controls, the point estimate for “Having both children and grandchildren” in
the GOPROBIT model hardly changes; it remains positive and statistically significant at the
5% level for
6
θ
(0.112 with a robust standard error of 0.044). See Table A2 in the appendix
for the estimates of the control variables.
With regards to the model selection, a Wald test on the GOPROBIT model with full-
specification suggests that we can reject the null hypothesis of equal slope parameters (
)., 9572214
2
219 χ
, thereby rejecting the implicit assumption of cardinalization under the
standard model in favour for the generalized model. The result thus suggests that parameters
are heterogeneous with respect to the LS responses.
Since the OPROBIT and GOPROBIT coefficients are not straightforward to interpret,
we present in Table 2 the estimated marginal probability effects (MPE), or
,
obtained from the regressions in Panel C of Table 1 (Boes & Winkelmann, 2006). Comparing
the MPEs among the standard and the generalized models over all outcomes produces the
13
following conclusions. First, whilst the standard model predicts statistically insignificant
effects of “Have children but no grand children” and “Have both children and grandchildren”
for each category of LS, the generalized model predicts that being a grandparent to at least
one person is associated, on average, with an increase in the probability of individual
reporting “7” or “very satisfied” on the LS scale by approximately 2.4-percentage-points.
Second, having only grandchildren but no living children is predicted in the standard model
to increase the probability of responses “5” and lower, whilst reduces the probability of
responses “6” and higher. The equivalent effects are, however, statistically significant only
for high LS responses (6-7) in the generalized model.
Following the work by Margolis and Myrskylä (2011) and Stanca (2011), who
explored the interaction effects between children and the parent‟s socio-economic status on
LS, Table 4 turns to the estimated MPEs of grandchildren on LS by gender (Panels A-B),
education (Panels C-D), income group (Panels E-F), and age group of the respondents (Panel
G-H). Looking across panels, the effects of “Have both children and grandchildren” on high
LS, i.e. LS = “7”, are positive and statistically well-determined only for women, people who
completed more than 15 years of full-time education, individuals with income above the
sample average, and those aged between 40 and 59. For instance, having both children and
grandchildren increases the probability of reporting “7” on the LS scale by 3.1-percentage-
points for women, 2.4-percentage-points for the (relatively) highly educated, 4.1-percentage-
points for the (relatively) rich, and 4.2-percentage-points for the younger cohorts. These
results are consistent with what had been found in previous studies with respect to the
impacts of own children on LS (Margolis & Myrskylä, 2011; Stanca, 2011). Nevertheless,
apart from the “40<age
60” age group, we find the effects of own children without
grandchildren on LS to be mostly insignificant here in our sub-sample regressions. In
addition to this, it is interesting to see how the estimated grandchildren effect on high LS is
14
more positive when the respondents (and implicitly, his or her grandchildren) are still
relatively young.
In summary, these results suggest that there is a positive and statistically important
relationship between life satisfaction and grandparenthood, conditioning on own children
being alive. However, there seems to be considerable heterogeneity on the grandchildren
effects with respect to the LS distribution, i.e., the estimated impacts are significant only for
high LS.
5. Concluding remarks
This paper investigates the relatively unexplored relationship between grandparenthood and
self-rated life satisfaction. Using the new Understanding Society survey for the UK, which
surveyed more than 5,000 grandparents and over 6,000 of non-grandparents (aged 40 years
and above) nation-wide, we find that there is indeed a positive and statistically important
correlation between having grandchildren and how the respondent rated his or her satisfaction
with life overall. However, it appears that the positive relationship is only confined to the
very top of the life satisfaction distribution. What this implies is that the presence of at least
one grandchild conditioning on parents being alive does not significantly reduce one‟s
dissatisfaction. Rather, it increases the probability of the individual reporting to be “very
satisfied” with his or her life overall. Such a heterogeneous effect of grandchildren on life
satisfaction is deemed consistent with the recently established notion that the determinants of
positive and negative well-being are not necessarily the same (see, e.g., Huppert &
Whittington, 2003; Boes & Winkelmann, 2010). By contrast, the relationship between
children and life satisfaction, conditioning on having no grandchildren, is generally
15
statistically insignificant at least at the 5% level in all but one sub-group regression, i.e., the
“40
age<60” group.
Overall, our results indicate that being a grandparent to at least one grandchild is
associated positively with becoming more satisfied with life even when parenthood itself may
not. The suspected impact is, however, marginal rather than applied averagely across the
distribution of life satisfaction. Implicitly, our findings imply that an investment in children
may have a long-term psychological payoff, providing that our children also go on to have
children of their own.
This paper is, however, not without limitations. The first is that it is hard to establish
causality between life satisfaction and grandparenthood. The exogeneity of grandchildren is,
by and large, debatable. While the choice of whether or not our children will go on to have
their own children is arguably weakly linked to our life satisfaction, it remains possible that
there are some omitted third variables that correlate with our desire (and our children‟s
desire) to pass on our genes and the way we evaluate how satisfied we are with our lives
overall. In addition to this, due to data limitation, we are unable to control for the number of
grandchildren in our regression equations. It may well be the case that quantity matters as
much as quality. Future research should return to examine whether the positive correlation
between grandparenthood and life satisfaction at the higher end of the life satisfaction score
varies according to the number of grandchildren the respondent has.
16
Reference
Alesina, Alberto; Di Tella, Rafael; and MacCulloch, Robert. 2004. Inequality and happiness:
are Europeans and American different? Journal of Public Economics, 88, 2009-2042.
Baumeister, Roy F. 1991. Meanings of life. New York: Guilford Press.
Blake, Judith. 1979. Is zero preferred? American attitudes toward childlessness in the 1970s.
Journal of Marriage and the Family, 41, 245-257.
Blanchflower, David G., Oswald, Andrew J. 2008. Is well-being U-shaped over the life
cycle? Social Science & Medicine, 66, 1733-1749.
Boes, S., Winkelmann, R. 2006. Ordered response models. Allgemeines Statistisches Archiv,
90, 165179.
Boes, Stefan, Winkelmann, Rainer. 2010. The effect of income on general life satisfaction
and dissatisfaction. Social Indicators Research, 95, 111-128.
Clark, Andrew E.; Diener, Ed; Georgellis, Yannis; Lucas, Richard E. 2008. Lags and leads in
life satisfaction: a test of the baseline hypothesis. Economic Journal, 118, F222-243.
Clark, Andrew E., Oswald, Andrew J. 2002. Well-being in panels. Unpublished manuscript,
Department of Economics, University of Warwick.
Cleary, Paul D., Mechanic, David. 1983. Sex differences in psychological distress among
married people. Journal of Health and Social Behaviour, 24, 111-121.
Crohan, Susan E. 1996. Marital quality and conflict across the transition to parenthood in
African American and white couples. Journal of Marriage and the Family, 58, 933-944.
Diener, Ed; Emmons, Robert A.; Larsen, Randy J.; and Griffin, Sharon. 1985. The
satisfaction with life scale. Journal of Personality Assessment, 49, 71-75.
Gilbert, Daniel T. 2006. Stumbling on Happiness. HarperCollins.
17
Glenn, Norval D.; McLanahan, Sara. 1981. The effect of children on the psychological well-
being of older adults. Journal of Marriage and Family, 43, 409-421.
Glenn, Norval D.; McLanahan, Sara. 1982. Children and marital happiness: a further
specification of the relationship. Journal of Marriage and Family, 44, 63-72.
Huppert, Felicia A., Whittington, Joyce E. 2003. Evidence for the independence of positive
and negative well-being: Implications for quality of life assessment. British Journal of Health
Psychology, 8, 107-122.
Kivnick, Helen Q. 1982. Grandparenthood: an overview of meaning and mental health.
Gerontologist, 22, 59-66.
Kohler, Hans-Peter; Behrman, Jere R.; and Skytthe, Axel. 2005. Partner plus children =
happiness? The effects of partnerships and fertility on well-being. Population and
Development Review 31, 407-445.
Lavee, Yoav; Sharlin, Shlomo; and Katz, Ruth. 1996. The effect of parenting stress on
marital quality: an integrated mother-father model. Journal of Family Issues, 17, 114-135.
Margolis, R., Myrskylä, M. 2011. A global perspective on happiness and fertility. Population
and Development Review, forthcoming.
Mentzakis, Emmanouli. 2011. Allowing for heterogeneity in monetary subjective well-being
valuations. Health Economics, 20, 331-347.
Powdthavee, Nattavudh. 2008. Putting a price tag on friends, relatives, and neighbours: using
surveys of life satisfaction to value social relationships. Journal of Socio-Economics, 37(4),
1459-1480.
Powdthavee, Nattavudh. 2009. Think having children will make you happy? The
Psychologist, 22, 308-311.
Sanchez, Laura; Thomson, Elizabeth. 1997. Becoming mothers and fathers: parenthood,
gender, and the division of labour. Gender & Society, 11, 747-772.
18
Shields, Michael A., Wheatley-Price, Stephen. 2001. Exploring the economic and social
determinants of psychological and psychosocial health. IZA Discussion Paper #396.
Shin, D.C., Johnson, D.M. 1978. Avowed happiness as an overall assessment of the quality of
life. Social Indicators Research, 5, 475-492.
Stanca, Luca. 2011. Suffer the little children: measuring the effects of parenthood on well-
being worldwide. Journal of Economic Behavior & Organization, forthcoming.
Thomas, Jeane L. 1989. Gender and perceptions of grandparenthood. The International
Journal of Aging and Human Development, 29, 269-282.
Twenge, Jean M.; Campbell, W.Keith; and Foster, Craig A. 2003. Parenthood and marital
satisfaction: a meta-analytic review. Journal of Marriage and Family, 65(3), 574-583.
White, Lynn K.; Booth, Alan; and Edwards, John N. 1986. Children and marital happiness:
why the negative correlation? Journal of Family Issues, 7, 131-147.
White, Mathew P.; Dolan, Paul. 2009. Account for the richness of daily activities.
Psychological Science, 20, 1000-1008.
19
Figure 1: Life Satisfaction, Children, and Grandchildren
Note: 4-standard-error bands (95% C.I.) are reported: two s.e. above and two below.
4.3
4.5
4.7
4.9
5.1
5.3
5.5
5.7
No children
and
grandchildren
Have children
but no
grandchildren
Have
grandchildren
but no (alive)
children
Have both
children and
grandchildren
Life Satisfaction
20
Table 1: Standard and Generalized Ordered Probit Life Satisfaction Models with
Grandchildren as Explanatory Variables
Panel A:
No control
variables
OPROBIT
Overall
GOPROBIT
θ1
θ2
θ3
θ4
θ5
θ6
Have children but no
grandchildren
0.0850**
-0.133*
-0.0594
0.0344
0.0884*
0.105**
0.156**
[0.0256]
[0.0676]
[0.0471]
[0.0388]
[0.0346]
[0.0318]
[0.0400]
Have grandchildren but
no living children
-0.194*
-0.360+
-0.415**
-0.218+
-0.276*
-0.255*
0.0378
[0.0974]
[0.195]
[0.141]
[0.128]
[0.116]
[0.113]
[0.148]
Have both children and
grandchildren
0.245**
-0.167**
-0.0157
0.121**
0.182**
0.246**
0.399**
[0.0219]
[0.0568]
[0.0400]
[0.0330]
[0.0292]
[0.0266]
[0.0324]
Panel B:
Exogenous variables as
controls
OPROBIT
Overall
GOPROBIT
θ1
θ2
θ3
θ4
θ5
θ6
Have children but no
grandchildren
-0.00742
-0.163*
-0.102*
-0.0396
0.000924
-0.00186
0.0395
[0.0269]
[0.0709]
[0.0495]
[0.0404]
[0.0362]
[0.0334]
[0.0423]
Have grandchildren but
no living children
-0.390**
-0.397*
-0.486**
-0.366**
-0.451**
-0.465**
-0.228
[0.0962]
[0.198]
[0.146]
[0.131]
[0.118]
[0.112]
[0.150]
Have both children and
grandchildren
0.00418
-0.235**
-0.138**
-0.0756+
-0.0445
-0.0154
0.118**
[0.0279]
[0.0685]
[0.0493]
[0.0405]
[0.0367]
[0.0338]
[0.0409]
Panel C:
Full
controls
OPROBIT
Overall
GOPROBIT
θ1
θ2
θ3
θ4
θ5
θ6
Have children but no
grandchildren
0.00913
-0.154+
-0.122*
-0.0191
0.0223
0.0175
0.0463
[0.0279]
[0.0806]
[0.0530]
[0.0430]
[0.0387]
[0.0354]
[0.0446]
Have grandchildren but
no living children
-0.354**
-0.357
-0.415**
-0.270*
-0.390**
-0.373**
-0.290+
[0.102]
[0.220]
[0.161]
[0.129]
[0.128]
[0.123]
[0.157]
Have both children and
grandchildren
0.0469
-0.152*
-0.0701
-0.0225
0.0261
0.0565
0.112*
[0.0293]
[0.0770]
[0.0550]
[0.0443]
[0.0398]
[0.0362]
[0.0437]
Note: N=11,849. Life satisfaction is measured on a 7-point-scale, with 1=very dissatisfied, ..., and 7=very
satisfied. Control variables in the full specification include age, age-squared, gender, race dummies, subjective
health statuses, age left full-time education, log of household income, employment dummies, marital statuses,
and regional dummies. Robust standard errors are in parentheses. +<10%; *<5%; **<1%.
21
Table 2: Marginal Probability Effects by Satisfaction Level
Marginal probability effects
Satisfaction level
1
2
3
4
5
6
7
A. Ordered Probit
Have children but no grandchildren
-0.0003
-0.001
-0.001
-0.001
-0.001
0.001
0.002
[0.001]
[0.001]
[0.002]
[0.002]
[0.002]
[0.004]
[0.006]
Have grandchildren but no living children
0.02
0.028
0.032
0.031
0.026
-0.076
-0.062
[0.008]*
[0.009]**
[0.010]**
[0.008]**
[0.004]**
[0.026]**
[0.014]**
Have both children and grandchildren
-0.001
-0.003
-0.004
-0.004
-0.004
0.007
0.010
[0.001]
[0.001]
[0.002]
[0.002]
[0.003]
[0.004]
[0.006]
B. Generalized Ordered Probit
Have children but no grandchildren
0.007
0.008
-0.011
-0.009
0.000
-0.003
0.010
[0.004]+
[0.006]
[0.006]+
[0.007]
[0.010]
[0.014]
[0.010]
Have grandchildren but no living children
0.020
0.043
-0.003
0.063
0.021
-0.092
-0.053
[0.018]
[0.026]+
[0.022]
[0.034]+
[0.039]
[0.046]*
[0.024]*
Have both children and grandchildren
0.006
0.002
-0.004
-0.011
-0.014
-0.003
0.024
[0.003]+
[0.005]
[0.006]
[0.007]
[0.010]
[0.014]
[0.009]*
Note: The marginal probability effects are based on the estimated coefficients obtained from Panel C, Table 1. Robust standard errors are in parentheses. +<10%; *<5%;
**<1%.
22
Table 3: Marginal Probability Effects by Satisfaction Level and by Group
Marginal probability effects
Satisfaction level
GOPROBIT models
1
2
3
4
5
6
7
A. Men (N=5,197)
Have children but no grandchildren
0.00773
0.00529
-0.0161
-0.0156
-0.0073
0.0229
0.00307
[0.00621]
[0.00888]
[0.00994]
[0.0113]
[0.0157]
[0.0201]
[0.0132]
Have grandchildren but no living children
0.305*
0.370**
-0.725**
0.0493
0.0738
-0.0532
-0.0200
[0.131]
[0.136]
[0.169]
[0.0558]
[0.0675]
[0.0762]
[0.0410]
Have both children and grandchildren
0.00458
-0.00026
0.00374
-0.0287*
-0.0269
0.0268
0.0206
[0.00521]
[0.00882]
[0.0107]
[0.0122]
[0.0169]
[0.0210]
[0.0133]
B. Women (N=6,652)
Have children but no grandchildren
0.00457
0.0120
-0.00688
-0.00333
0.00606
-0.0263
0.0138
[0.00304]
[0.00894]
[0.00920]
[0.00980]
[0.0148]
[0.0196]
[0.0143]
Have grandchildren but no living children
-0.0076**
-0.0137
0.0388
0.0949+
0.00653
-0.0567
-0.0622+
[0.000931]
[0.0218]
[0.0325]
[0.0498]
[0.0542]
[0.0643]
[0.0333]
Have both children and grandchildren
0.00305
0.00492
-0.00593
0.000375
-0.0063
-0.0266
0.0305*
[0.00203]
[0.00758]
[0.00809]
[0.00920]
[0.0142]
[0.0189]
[0.0136]
C. Age left school <= 15 years (N=4,485)
Have children but no grandchildren
0.0141
-0.00130
-0.00669
-0.0226
-0.0148
0.0167
0.0146
[0.0113]
[0.0123]
[0.0118]
[0.0152]
[0.0204]
[0.0282]
[0.0235]
Have grandchildren but no living children
0.0320
0.0147
0.0128
-0.00825
-0.0154
0.0296
-0.0654
[0.0369]
[0.0380]
[0.0347]
[0.0409]
[0.0530]
[0.0663]
[0.0458]
Have both children and grandchildren
0.0128+
-0.00770
0.00774
-0.0223
-0.0219
0.00564
0.0257
[0.00673]
[0.00920]
[0.00971]
[0.0138]
[0.0187]
[0.0244]
[0.0192]
D. Age left school > 15 years (N=7,364)
Have children but no grandchildren
0.00322
0.0115+
-0.0163*
-0.00543
0.00480
-0.0108
0.0130
[0.00362]
[0.00656]
[0.00779]
[0.00881]
[0.0126]
[0.0162]
[0.0103]
Have grandchildren but no living children
0.0365
0.0699
-0.0384
0.123*
0.0404
-0.190**
-0.0414
[0.0365]
[0.0467]
[0.0409]
[0.0583]
[0.0598]
[0.0650]
[0.0315]
23
Have both children and grandchildren
0.00250
0.00695
-0.0135
-0.0122
-0.0084
0.00113
0.0236*
[0.00353]
[0.00681]
[0.00844]
[0.00991]
[0.0138]
[0.0181]
[0.0114]
E. Income<=Mean income (N=5,849)
Have children but no grandchildren
0.0159*
0.0163
-0.00127
0.00618
-0.0081
-0.0237
-0.00534
[0.00802]
[0.0111]
[0.0123]
[0.0132]
[0.0182]
[0.0224]
[0.0165]
Have grandchildren but no living children
0.0106
0.0585
-0.0311
0.0900+
-0.0446
-0.0397
-0.0437
[0.0241]
[0.0418]
[0.0332]
[0.0502]
[0.0509]
[0.0645]
[0.0389]
Have both children and grandchildren
0.0106*
0.00214
-1.98e-05
0.00113
-0.03+
0.000584
0.0154
[0.00532]
[0.00857]
[0.0101]
[0.0114]
[0.0165]
[0.0207]
[0.0150]
F. Income>Mean income (N=6,000)
Have children but no grandchildren
0.000435
0.00939
-0.027**
-0.0198*
0.00140
0.0133
0.0220+
[0.00399]
[0.00803]
[0.00769]
[0.00882]
[0.0130]
[0.0178]
[0.0122]
Have grandchildren but no living children
0.0290
0.0502
-0.00821
0.0490
0.0809
-0.132+
-0.069**
[0.0292]
[0.0421]
[0.0399]
[0.0453]
[0.0617]
[0.0683]
[0.0259]
Have both children and grandchildren
0.00164
0.00642
-0.00827
-0.0207*
-0.0074
-0.0124
0.0407**
[0.00363]
[0.00816]
[0.00827]
[0.0101]
[0.0149]
[0.0197]
[0.0135]
G. 40<=Age<60 (N=6,699)
Have children but no grandchildren
0.00709+
0.00919
-0.0159+
-0.0170+
-0.0005
-0.00678
0.0240*
[0.00400]
[0.00734]
[0.00883]
[0.00975]
[0.0131]
[0.0161]
[0.00977]
Have grandchildren but no living children
0.00859
0.0341
0.000515
0.0977+
0.0604
-0.136*
-0.066**
[0.0206]
[0.0372]
[0.0383]
[0.0572]
[0.0639]
[0.0661]
[0.0206]
Have both children and grandchildren
0.0101*
0.0130
-0.00876
-0.00427
-0.0248
-0.0268
0.0416**
[0.00454]
[0.00847]
[0.0102]
[0.0118]
[0.0152]
[0.0187]
[0.0121]
H. Age>60 (N=5,150)
Have children but no grandchildren
-0.00875
0.00496
-0.00770
-0.00458
0.00932
0.0464+
-0.0396+
[0.00616]
[0.0108]
[0.0103]
[0.0124]
[0.0184]
[0.0270]
[0.0213]
Have grandchildren but no living children
0.0120
0.0442
-0.0378
0.0497
0.0195
-0.00871
-0.0787+
[0.0229]
[0.0378]
[0.0341]
[0.0345]
[0.0451]
[0.0624]
[0.0413]
Have both children and grandchildren
-0.00914
-0.00535
-0.00273
-0.0149
0.0132
0.0346
-0.0157
[0.00727]
[0.00922]
[0.00904]
[0.0117]
[0.0159]
[0.0237]
[0.0195]
24
Table A1: Descriptive Statistics
Full
sample
Life satisfaction level
Mean
1
2
3
4
5
6
7
Life satisfaction
5.31
Have children but no grandchildren
0.21
0.22
0.23
0.21
0.20
0.22
0.22
0.19
Have grandchildren but no (alive) children
0.01
0.01
0.02
0.01
0.02
0.01
0.01
0.01
Have both children and grandchildren
0.41
0.47
0.37
0.34
0.36
0.36
0.42
0.55
Control variables
Age
58.11
58.30
55.33
54.08
55.68
56.40
58.42
62.67
Men
0.43
0.49
0.43
0.46
0.46
0.45
0.45
0.38
Years of education
15.96
15.58
15.85
16.08
15.90
15.95
16.10
15.66
Log of household income
7.76
7.47
7.64
7.63
7.67
7.77
7.86
7.65
Race: Mixed
0.004
0.006
0.015
0.005
0.007
0.003
0.085
0.004
Race: Indian/Pakistani/Bangladeshi
0.013
0.019
0.028
0.017
0.022
0.018
0.010
0.009
Race: Chinese & Other Asians
0.004
0.000
0.004
0.003
0.012
0.010
0.003
0.002
Race: Blacks
0.012
0.016
0.012
0.014
0.021
0.019
0.010
0.007
Race: Other Groups
0.006
0.010
0.011
0.008
0.006
0.010
0.004
0.005
Self-employed
0.08
0.02
0.07
0.08
0.09
0.07
0.09
0.06
Unemployed
0.04
0.05
0.09
0.09
0.06
0.04
0.02
0.03
Retired
0.34
0.33
0.24
0.20
0.26
0.29
0.36
0.52
Married
0.62
0.45
0.52
0.49
0.53
0.59
0.69
0.66
Separated
0.03
0.08
0.06
0.06
0.05
0.03
0.03
0.02
Divorced
0.13
0.19
0.17
0.22
0.17
0.16
0.11
0.10
Widowed
0.09
0.16
0.09
0.07
0.10
0.09
0.09
0.14
N
11,942
309
540
782
1,043
1,864
5,524
1,880
25
Table A2: Standard and Generalized Ordered Probit Estimates for the Control
Variables (Panel C, Table 1)
Dependent variable:
Life satisfaction
OPROBIT
Overall
GOPROBIT
θ1
θ2
θ3
θ4
θ5
θ6
Age
0.0325**
0.0170
0.00802
0.0252*
0.0333**
0.0421**
0.0453**
[0.00817]
[0.0225]
[0.0149]
[0.0121]
[0.0110]
[0.00994]
[0.0120]
Age-squared/100
-0.0163*
-0.00926
0.000412
-0.0107
-0.0147
-
0.0217**
-
0.0275**
[0.00674]
[0.0183]
[0.0120]
[0.00988]
[0.00899]
[0.00814]
[0.00941]
Male
-0.108**
-0.160**
-0.0226
-0.0478
-
0.0796**
-
0.0820**
-0.177**
[0.0205]
[0.0578]
[0.0369]
[0.0316]
[0.0282]
[0.0256]
[0.0308]
Health: Good
-0.241**
0.193*
0.0697
-0.0140
-0.151**
-0.226**
-0.393**
[0.0326]
[0.0935]
[0.0630]
[0.0534]
[0.0472]
[0.0405]
[0.0419]
Health: Fair
-0.458**
0.191*
0.0614
-0.164**
-0.434**
-0.537**
-0.599**
[0.0331]
[0.0944]
[0.0645]
[0.0526]
[0.0462]
[0.0405]
[0.0437]
Health: Poor
-0.762**
0.0452
-0.129+
-0.474**
-0.792**
-0.938**
-0.823**
[0.0379]
[0.0982]
[0.0675]
[0.0564]
[0.0502]
[0.0456]
[0.0526]
Health: Very poor
-1.217**
-0.492**
-0.675**
-1.023**
-1.322**
-1.357**
-1.018**
[0.0523]
[0.103]
[0.0764]
[0.0668]
[0.0613]
[0.0591]
[0.0739]
Log of household income
0.0501**
0.0266
0.00633
0.0959**
0.110**
0.100**
0.000914
[0.0152]
[0.0415]
[0.0288]
[0.0220]
[0.0206]
[0.0184]
[0.0208]
Age left full-time education
-0.0227**
0.0603*
0.0414*
-0.00745
0.00294
0.0139
-
0.1000**
[0.00855]
[0.0277]
[0.0173]
[0.0138]
[0.0119]
[0.0107]
[0.0123]
Self-employed
0.0236
0.285+
0.0361
-0.00323
-0.0383
0.0687
-0.00266
[0.0368]
[0.146]
[0.0776]
[0.0606]
[0.0528]
[0.0483]
[0.0615]
Unemployed
-0.202**
-0.0302
-0.351**
-0.359**
-0.327**
-0.257**
0.0527
[0.0555]
[0.147]
[0.0873]
[0.0729]
[0.0685]
[0.0675]
[0.0868]
Retired
0.242**
-0.0110
0.0439
0.186**
0.223**
0.299**
0.278**
[0.0360]
[0.108]
[0.0765]
[0.0580]
[0.0512]
[0.0441]
[0.0501]
Married
0.140**
0.198*
0.220**
0.183**
0.194**
0.218**
-0.00183
[0.0349]
[0.0956]
[0.0634]
[0.0516]
[0.0460]
[0.0434]
[0.0564]
Separated
-0.196**
-0.273+
-0.127
-0.250**
-0.233**
-0.0747
-0.264*
[0.0648]
[0.145]
[0.0998]
[0.0859]
[0.0789]
[0.0776]
[0.104]
Divorced
-0.0847*
0.0573
0.0835
-0.0563
-0.0651
-0.0916+
-0.132+
[0.0419]
[0.108]
[0.0722]
[0.0602]
[0.0545]
[0.0520]
[0.0678]
Widowed
-0.0685
-0.135
-0.0245
-0.0201
-0.0609
-0.0331
-0.0984
[0.0512]
[0.127]
[0.0856]
[0.0719]
[0.0653]
[0.0604]
[0.0715]
Race: Mixed
-0.181
-0.184
-0.526*
-0.458*
-0.494**
-0.258
0.216
[0.163]
[0.350]
[0.210]
[0.200]
[0.180]
[0.180]
[0.217]
Race:
Indian/Pakistani/Bangladeshi
-0.231**
-0.229
-0.329*
-0.246+
-0.245*
-0.364**
0.0544
[0.0884]
[0.208]
[0.148]
[0.129]
[0.107]
[0.109]
[0.145]
Race: Chinese & Other Asians
-0.109
3.627**
0.141
0.523*
-0.0508
-0.508**
-0.0323
[0.114]
[0.167]
[0.377]
[0.249]
[0.202]
[0.180]
[0.291]
26
Race: Blacks
-0.0161
-0.217
0.0966
0.180
0.0426
-0.119
0.0778
[0.0848]
[0.235]
[0.147]
[0.137]
[0.118]
[0.109]
[0.158]
Race: Other ethnic groups
-0.0867
-0.345
-0.302
-0.0551
-0.0263
-0.184
0.0463
[0.137]
[0.289]
[0.201]
[0.174]
[0.167]
[0.160]
[0.204]
Cut_1
-1.186**
[0.296]
Cut_2
-0.668*
[0.296]
Cut_3
-0.256
[0.295]
Cut_4
0.123
[0.296]
Cut_5
0.637*
[0.296]
Cut_6
2.073**
[0.297]
Constant
0.460
0.490
-0.342
-1.142**
-2.019**
-0.680
Observations
11,849
11,849
11,849
11,849
11,849
11,849
11,849
Note: All regressions include regional dummies. Robust standard errors are in parentheses. +<10%; *<5%;
**<1%.
... There is some evidence showing a positive association between becoming a grandparent and wellbeing for women, but not for men (Arpino et al., 2018;Di Gessa et al., 2019;Powdthavee, 2011). While Di Gessa et al. (2019 found no additional improvement in life satisfaction for subsequent grandchildren, Arpino et al. (2018) show that life satisfaction is positively correlated with someone's number of grandchildren. ...
... Using data on ten European countries from Waves 1, 2, 4, 5 and 6 of the Survey of Health, Ageing and Retirement in Europe, we use an individual level fixed effects approach to investigate the effect of transitioning into grandparenthood on the health and well-being of older individuals. Unlike previous correlational evidence (Arpino et al., 2018;Di Gessa et al., 2019;Powdthavee, 2011) Our results thus show that grandparenthood on average is considered burdensome, fitting to the "role strain" theory (Goode, 1960). This average effect is not large, though. ...
Article
Full-text available
Older individuals commonly go through a few major life transitions which can impact their health and well-being. While transitions like that into retirement have been extensively investigated, little research focused on the transition into grandparenthood. Understanding effects of this highly common event is not only important from a descriptive viewpoint, but is also informative for the active aging policies that are increasingly pursued to deal with aging populations. Using data from ten Western European countries, we show that grandparenthood on average leads to a reduction in well-being while hardly impacting physical, cognitive and mental health. The effect is heterogeneous by family closeness, though. Grandparenthood reduces well-being for those having relatively little family contact and not providing child care. But it leaves well-being unaffected while improving health along some dimensions among those with the opposite profile. The only exception to the latter are grandmothers providing daily child care, for whom grandparenthood appears to be burdensome. This pattern of results suggests that involving grandparents non-intensively in child care may lead to beneficial side-effects. Becoming a grandparent induces people to retire, but retirement seems no relevant channel for well-being and health effects.
... Recently, some research has shown that having grandchildren comes at the cost of fewer social activities and lower labor supply of grandparents, especially grandmothers (Backhaus & Barslund, 2019;Rupert & Zanella, 2018;Frimmel et al., 2021). A few correlational studies found that having grandchildren is associated with higher life satisfaction (Arpino et al., 2018;Powdthavee, 2011), while other research found that having grandchildren comes with increases in mortality, especially among those who become grandparents at a younger age (Christiansen, 2014). Some studies specifically focused on how caregiving to grandchildren correlates with grandparental health, although most of these studies are not able to determine causality, and results are mixed (Brunello & Rocco, 2019;Arpino & Bordone, 2014;Ku et al., 2012). ...
... A few mostly correlational studies investigated the health and well-being effects of becoming a grandparent per se. There is some evidence showing a positive association between becoming a grandparent and well-being for women, but not for men (Arpino et al., 2018;Di Gessa et al., 2019;Powdthavee, 2011). While Di Gessa et al. (2019 found no additional improvement in life satisfaction for subsequent grandchildren, Arpino et al. (2018) show that life satisfaction is positively correlated with someone's number of grandchildren. ...
... In terms of reverse time support, some studies have shown that there is a strong positive relationship between the intergenerational care of grandchildren and the life happiness of the elderly, which means that intergenerational care is an important link to enhance the intergenerational connection between paternity and offspring. An important manifestation of active aging is to make it easier for older people to discover the meaning of life and realize their selfworth (Powdthavee, 2011;Zhang and Chen, 2014;Arpino et al., 2018;Cheng and Jiang, 2019). In terms of reverse spiritual support, the elderly provide rich and powerful emotional support to their children to show their wisdom and reflect on their value, to enhance their sense of self-affirmation and satisfaction (Kuause and Shaw, 2000). ...
Article
Full-text available
Intergenerational support is bidirectional, and reverse intergenerational support refers to parents providing financial support, time support, and spiritual support to their offspring. The emergence of reverse intergenerational support has created role conflicts among different groups of older adults. Based on survey data from 3,170 elderly people in eight sample provinces in China, this paper empirically investigates the relationship between reverse intergenerational support and the happiness of the elderly in contemporary China and the moderating effect of role conflict in it, using an ordered logit model. It was found that, first, reverse economic support reduces the happiness of the elderly, and reverse time support and reverse spiritual support can significantly enhance the happiness of the elderly. Second, in the presence of role conflict, the effect of reverse time support and reverse spiritual support on the enhancement of older adults’ happiness was suppressed; in the presence of role enhancement, the effect of reverse economic support on the reduction of older adults’ happiness was mitigated. The above findings provide new empirical evidence for understanding the relationship between reverse intergenerational support and the happiness of the elderly, which is prevalent in contemporary China, and offer new insights for enhancing happiness.
... Being grandparents plays a main role in late life and is one of the factors that contributes to the quality of life of older adults. 39,40 Relationships with grandchildren may be even more significant for older adults recovering from AUD, since it may be a way to compensate for the problematic relationship they had with their own children because of their addiction. Being grandparents and having meaningful relationships with their grandchildren is another component of successful aging. ...
Article
Full-text available
Objective: The population of older adults suffering from alcohol use disorder (AUD) is increasing worldwide. Recovery from AUD among older adults is a challenging process which can lead to amelioration in these individuals' physical, mental, familial and social domains. However, little is known about the life experiences of older adults who have recovered from AUD. Method: A qualitative-naturalistic approach was implemented. Semi-structured in-depth interviews were conducted with 20 older adults, age 60 +, who had recovered from AUD for periods ranging from 1 to 9 years. Results: Three main categories emerged from the content analysis: a) Regrets, self-forgiveness and a desire to remedy past wrongs; b) successful aging and eagerness to live; c) enduring challenges. These categories reflect the complex and multidimensional experiences of older adults who have recovered from AUD. Conclusion: Older adults who recover from AUD report experiencing successful aging. They are willing to engage in new ventures in late life, live actively and age healthfully. However, despite their positive outlook, older adults recovering from AUD are a vulnerable population, especially when they experience marginalization as post-AUD older adults. This underscores the need to reach out to this population and the host of challenges they face to provide supportive treatments and interventions from interdisciplinary professionals who can guide their recovery from AUD and help them flourish in late life.
... Though little literature is available on the positive relationship between being a grandparent and aging, few studies have found that having a grandchild is related to increased happiness and higher life satisfaction (Arpino et al., 2018;Danielsbacka & Tanskanen, 2016). That is why grandchildren served to be an emotional reservoir for grandparents (Silverstein et al., 2003), as having grandchildren is an overwhelming experience (Arpino et al., 2018) that positively affects grandparent's well-being (Powdthavee, 2011) which ultimately leads to successful aging (SA) (Fuller-Iglesias & Antonucci, 2016). ...
Article
The current study aimed to determine the reciprocal associations between the familial values and successful aging of grandparents and grandchildren and the extent to which quality of life accounts for these reciprocal associations. Data was collected from 270 grandparent–grandchild pairs living together. Actor-Partner Independence Model indicated that familial values of both grandparents and grandchildren showed significant associations with their own successful aging (actor effects), whereas familial values of grandchildren showed significant associations with the successful aging of grandparents (partner effects). Grandchildren’s quality of life mediated the actor and partner effects of familial values on the successful aging of grandchildren; whereas grandparents’ quality of life only mediated the actor effect of familial values on the successful aging of grandparents. These findings showed the interdependence of grandparents and grandchildren while emphasizing the importance of grandparents-grandchildren familial values and quality of life in enhancing successful aging.
Article
Full-text available
Portugal has been identified as the European country with the most rapid evolution of Pediatric Palliative Care provision, where approximately 7800 children have life-limiting conditions. This is a highly complex experience not only for the children and their parental caregivers, but also for their healthy siblings and grandparents. The present descriptive-exploratory study seeks to contribute to the understanding of the psychological experience of life-limiting conditions in grandparents. A total of 19 families, consisting of 15 grandmothers and 4 grandfathers, completed a sociodemographic and clinical data sheet and a semi-structured interview was conducted in which they shared their testimony. The results of the thematic analysis highlighted an integrated view on 10 important dimensions in the grandparental experience and promoted creative responses by means of their own perspective. However, it has some limitations, such as the small sample size and the data collection procedure via telephone. The results contribute to the design of specific intervention methodologies in an ecosystemic approach and suggest further research to explore more protective factors and communication with health professionals. For psychological intervention, it is suggested considering the identification of individual and family resources that contribute to the activation of key processes in resilience and posttraumatic growth.
Article
Using a representative sample from Japan and a difference‐in‐differences strategy, we investigate whether the effect of having grandchildren on the happiness of grandparents varies with the gender of their (own) single child. In line with our expectations, we find that maternal grandmothers have more to lose or less to gain from having grandchildren than paternal grandmothers. In contrast, grandfathers’ changes in happiness do not depend on their own child's gender. This result is explained by the fact that grandmothers are more likely to be involved in childrearing when their daughter has a child.
Article
Background : Caring for grandchildren has been a normative, pragmatic and prevalent practice in China, while its association with the well-being of grandparents remains inconclusive. This study aims to illustrate whether and how the provision of grandchild care is related to grandparents’ life satisfaction, and how the relationship is contingent on the features of grandparents. Methods : Using data from the 2018 wave of China Health and Retirement Longitudinal Study, we apply ordered logistic model to examine the link between caregiving intensity and life satisfaction. Results : Provision of moderate grandchild care is associated with better life satisfaction, as caregiving grandparents participate in social activities more actively and feel more confident in the future. The association between intensive child care and life satisfaction remains insignificant, as its detrimental effect might wipe out the benefits. The positive association between moderate caregiving and life satisfaction is more pronounced among male, younger, rural grandparents and grandparents not co-residing with their children. Conclusion : Our findings suggest that caring for grandchildren is not necessarily beneficial for grandparents. Attenuating the load of child care as well as enhancing social and economic support to caregivers would help them better reap the benefits of grandchild care.
Article
The current study aimed to investigate the association between grandparenting and cognitive function over time in noncustodial grandparents in China and the United States. Lagged dependent variable (LDV) approach and linear regression models were applied to analyze a sample of 1,411 Chinese and 6,579 American adults aged 65 and above from the China Health and Retirement Longitudinal Study (CHARLS, 2011-2013) and the U.S. Health and Retirement Study (HRS, 2012-2014). Grandparenting involvement was associated with less decline in episodic memory for grandparents and greater level of grandparenting had no negative effect on mental status and global cognitive function in noncustodial grandparents in China and the United States. The impact of grandparenting on cognitive function was conditioned on caregiving intensity, gender, urban/rural residence, and nation. Findings of the study suggest that greater attention on grandparenting facilitation might yield improved research, social support, policy, and interventions on cognitive health for the general older population.
Article
Full-text available
This paper uses panel data to study human wellbeing. It finds that fixed-effects equations have a similar structure to cross-section equations. This is potentially important, because nearly all work in the field has been forced to rely on cross-section information, and critics have argued that the omission of controls for person-effects makes the literature's conclusions open to doubt. Our paper follows a random sample of 7000 British individuals through each year of the 1990s. The paper calculates the relative importance of economic and non-economic events to psychological health. It puts dollar values --positive or negative --on the 'happiness' value of health, marriage, unemployment, children, and education. Widowhood is the worst life event. The paper also makes a first stab at identifying what lies behind the large fixed-effects in people's subjective well-being.
Article
Full-text available
There is substantial evidence in the psychology and sociology literature that social relationships promote happiness for the individual. Yet the size of their impacts remains largely unknown. This paper explores the use of shadow pricing method to estimate the monetary values of the satisfaction with life gained by an increase in the frequency of interaction with friends, relatives, and neighbours. Using the British Household Panel Survey, I find that an increase in the level of social involvements is worth up to an extra £85,000 a year in terms of life satisfaction. Actual changes in income, on the other hand, buy very little happiness.
Article
The effects of the presence of children on their parents' marital happiness were estimated, with data from six U. S. national surveys conducted from 1973 through 1978, for subpopulations delineated on the basis of sex, race, level of education, religious preference, employment status, and stated ideal number of children for a family. No evidence for distinctly positive mean effects was found for any subpopulation, the estimated mean effects ranging from distinctly negative (for most of the subpopulations) to near zero for whites who said that the ideal number of children for a family is four or more. The findings contribute to a rather large body of accumulating evidence which indicates that on the average children adversely affect marital quality, both in the total population of married persons in the U. S. and in several large and important subpopulations. Some implications of the findings are discussed.
Article
Changes in marital quality and marital conflict are explored in a longitudinal study of White and African American couples making the transition to parenthood within the first 2 years of marriage (38 White, 27 African American) and a group of couples who remained childless (75 White, 31 African American). Results show that White and African American spouses who become parents report lower marital happiness and more frequent conflicts after the transition than before. White parents also report higher marital tension and a greater likelihood to become quiet and withdrawn during conflict after the birth of their child. For new mothers and fathers, conflict behaviors predict marital happiness in similar ways—attacking the spouse and leaving the scene of the conflict predicted lower marital happiness, and becoming quiet and withdrawn predicted higher marital happiness. These patterns are similar for both African American and White spouses.
Article
Is a large increase in voluntary childlessness likely to contribute to extremely low fertility rates in the United States? To address this issue, this paper uses special questions on attitudes toward childlessness that were commissioned on a Gallup survey of voting-age adults in February, 1977. A Likert-type scale of attitudes toward childlessness was constructed and major sociodemographic, plus some attitudinal, predictors of scale values were studied by means of Multiple Classification Analysis. There is a high level of consensus that nonparenthood is not an advantaged status and, although offspring are not regarded as economic investments, they are viewed as being socially instrumental—they are not solely consumption goods. A desire for some children is thus not as vulnerable to cost factors as one might think on the basis of a consumer model of reproductive motivation. Less advantaged groups in the population are more likely than others to regard reproduction as a social investment, thus helping to explain the frequently-discovered inverse relationship between socio-economic status (including educational status) and either preferred or actual family size. A major finding of our work is that men are significantly more likely to regard childlessness as disadvantageous than are women.
Article
Data from six U.S. national surveys, conducted from 1973 through 1978, are used to estimate the effects of having had offspring on the global happiness and five dimensions of satisfaction of persons age 50 and older who have no children under age 18 living at home. Separate estimates are provided for each sex-race subpopulation and, for whites, each sex-education and sex-marital status subpopulation. All estimated effects on global happiness are either negligible or negative, the strongest evidence for negative effects being for high-education white males. The estimated effects on the dimensions of satisfaction are mixed and generally moderate in magnitude, being most distinctly negative for black males and high-education white males and most distinctly positive for unmarried white females. Overall, there is little evidence that important psychological rewards are derived from the later stages of parenthood.
Article
Increasing evidence from the empirical economic and psychological literature suggests that positive and negative well-being are more than opposite ends of the same phenomenon. Two separate measures of the dependent variable may therefore be needed when analyzing the determinants of subjective well-being. We investigate asymmetries in the effect of income on subjective well-being with a single-item measure of general life satisfaction. Using data from the German Socio-Economic Panel 1984–2004, and a flexible multiple-index ordered probit panel data model with varying thresholds, we find that income has only a minor effect on high satisfaction but significantly reduces dissatisfaction.
Article
The literature on fertility and happiness has neglected comparative analysis. We investigate the fertility/happiness association using data from the world values surveys for 86 countries. We find that, globally, happiness decreases with the number of children. This association, however, is strongly modified by individual and contextual factors. Most importantly, we find that the association between happiness and fertility evolves from negative to neutral to positive above age 40, and is strongest among those who are likely to benefit most from upward intergenerational transfers. In addition, analyses by welfare regime show that the negative fertility/ happiness association for younger adults is weakest in countries with high public support for families, and the positive association above age 40 is strongest in countries where old-age support depends mostly on the family. Overall these results suggest that children are a long-term investment in well-being, and highlight the importance of the life-cycle stage and contextual factors in explaining the happiness/fertility association.
Article
Recent research on 'happiness' regression equations has shown how monetary values can be put on the well-being effects of many life events (like health problems, marriage or the death of a spouse). Potentially, such work has practical implications for policy-makers and the courts. However, this article argues that we need to be careful in such work to use the appropriate statistical method. It goes beyond previous research and allows for heterogeneity in the subjective well-being scales. Using less restrictive models than the current literature, the article argues that standard linear or ordered response models seem consistently to overstate valuations. With data from the UK, it provides new monetary estimates of the well-being consequences of a number of health problems, social capital indicators, marital status changes and social relationships.