1
Participation in Voluntary Organizations and Volunteer Work as a Compensation for the
Absence of Work or Partnership? Evidence From Two German Samples of Younger and
Older Adults
Maria K. Pavlovaa and Rainer K. Silbereisen
b
a Jena Graduate School “Human Behaviour in Social and Economic Change” (GSBC),
Friedrich Schiller University Jena, Germany
b Department of Developmental Psychology and Center for Applied Developmental Science
(CADS), Friedrich Schiller University Jena, Germany
Corresponding author: Maria K. Pavlova, GSBC, Friedrich Schiller University Jena,
Bachstraße 18k, 07743 Jena, Germany.
E-mail: maria.pavlova(at)uni-jena.de
This is a pre-copy-editing, author-produced PDF of an article accepted for publication in the
Journals of Gerontology, Series B: Psychological Sciences and Social Sciences following
peer review. The definitive publisher-authenticated version
Pavlova, M. K., & Silbereisen, R. K. (2012). Participation in voluntary organizations and
volunteer work as a compensation for the absence of work and partnership? Evidence from
two German samples of younger and older adults. The Journals of Gerontology, Series B:
Psychological Sciences and Social Sciences, 67, 514-524.
is available online at: http://psychsocgerontology.oxfordjournals.org/cgi/reprint/gbs051?
ijkey=9Va8OqUiKuYBPvz&keytype=ref
2
Abstract
Objectives. We tested whether formal volunteering, in terms of its associations with mental
health, compensates for the absence of major work and family roles among older adults or
rather complements such roles among both younger and older adults.
Methods. Two cross-sectional samples of younger (age 18–42, N = 2346) and older (age 56–
75, N = 1422) German adults were used. We regressed mental health indicators on control
variables, two indicators of formal volunteering (participation in voluntary organizations and
volunteer work), and their interactions with employment/partnership status.
Results. Participation in voluntary organizations was associated with higher positive affect,
higher life satisfaction, and fewer depressive symptoms in younger adults. In older adults, it
was related to higher life satisfaction only among working individuals, although the
difference from non-working individuals was not significant. Volunteer work was associated
with higher positive affect in both age groups. In younger adults, it had no relation to life
satisfaction and depressive symptoms. In older adults, it was related to higher life satisfaction
among non-working individuals and to fewer depressive symptoms among those without a
steady partner.
Discussion. Volunteer work but not participation in voluntary organizations yielded
compensatory effects on mental health among older adults.
3
Participation in Voluntary Organizations and Volunteer Work as a Compensation for the
Absence of Work or Partnership? Evidence From Two German Samples of Younger and
Older Adults
There is a general consensus that formal volunteering (i.e., uncoerced and unpaid
individual activity that is structured by an organization and directed towards a community
concern; Morrow-Howell, 2010) is beneficial not only for the society but also for the
volunteers themselves. In particular, several longitudinal studies have shown that formal
volunteering has salutary effects on physical and mental health (Li & Ferraro, 2005; Oman,
Thoresen, & McMahon, 1999; Thoits & Hewitt, 2001). Researchers have also started to
document moderating factors in these relationships, such as age, availability of other social
roles and activities, and economic and psychosocial resources (Greenfield & Marks, 2004;
Hao, 2008; Morrow-Howell, Hinterlong, Rozario, & Tang, 2003; Musick & Wilson, 2003;
Piliavin, 2010; Sugihara, Sugisawa, Shibata, & Harada, 2008; Van Willigen, 2000). Indeed, a
better understanding of differential effects of volunteering is needed to maximize positive
outcomes via targeted recruitment (Morrow-Howell, 2010).
Formal volunteering is generally believed to be more beneficial for older than for
younger or middle-aged adults (Dovidio, Piliavin, Schroeder, & Penner, 2006; Wilson, 2000).
One possible reason for this may be the need for older adults to compensate for major age-
related role losses, such as retirement from work or death of a spouse (Baltes, 1997;
Heckhausen, Wrosch, & Schulz, 2010), whereby formal volunteering provides opportunities
for social contact and activities that may function as a compensation (Dovidio et al., 2006).
Non-working younger adults (students, homemakers, and unemployed) are usually expected
(and want) to enter or return to employment, just as younger individuals without a steady
partner, be they single or divorced, are supposed (and want) to find a mate (Havighurst, 1972;
Heckhausen et al., 2010). Thus, formal volunteering can hardly be regarded as a substitution
4
for major work and family roles in young and middle adulthood, neither by the society nor by
individuals themselves. In contrast, older adults have fewer opportunities to reenter the labor
market if they do not work or to find a new partner if they have none than younger adults
experiencing similar role losses. Consequently, mental health benefits of formal volunteering
may be greater in older than in younger adults, and especially in those older adults who do
not work or do not have a steady partner or a spouse (vs. those who do).
On the other hand, combining volunteering with work and family roles may have its
own benefits. Given that volunteers are most often recruited through the network of relatives,
friends, and acquaintances (Verba, Schlozman, & Brady, 1995; Wilson, 2000), employed or
married individuals are more likely to volunteer, and in particular, to volunteer together with
someone they know and like, which may bring additional gratification. Moreover, employed
individuals, especially those in high-status, complex jobs, may be more effective as
volunteers because they possess the organizational and communication skills important for
success in both paid and unpaid work (Verba et al., 1995; Wilson, 2000). These examples
illustrate how different social roles may overlap and how the enactment of one of them may
facilitate the enactment of others. According to role accumulation theory (Sieber, 1974;
Thoits, 1983), multiple social roles enable the accumulation of privileges and resources,
increase overall social status security, and serve personality enrichment. As multiple roles are
usually not independent, they do not necessarily require the separate investment of time and
other resources, so that the benefits of performing them outweigh possible drawbacks of role
strain and role conflict. As a corollary, the more roles an individual undertakes, the greater
the gain in well-being. From this perspective, formal volunteering complements major work
and family roles rather than compensates for their absence, that is, its positive effects on
mental health may be more pronounced in those working (vs. those not working) or in those
with a steady partner (vs. those without one), irrespective of age.
5
Pertinent empirical evidence is mixed. Some studies did show compensatory effects
of formal volunteering on mental health with respect to major role-identity absences in older
adults (Greenfield & Marks, 2004; Piliavin, 2010), whereas other studies provided support
for role accumulation theory at all ages (Adelmann, 1994; Hao, 2008; Moen, Dempster-
McClain, & Williams, 1992). Moreover, Sugihara et al. (2008) reported both compensatory
and complementary effects of formal volunteering at an older age, whereas Morrow-Howell
et al. (2003) found neither. Finally, few researchers who actually compared younger and
older adults reported age differences in the expected direction but found no evidence that the
stronger effects of formal volunteering on mental health in older adults were due to age-
related role losses (Musick & Wilson, 2003; Van Willigen, 2000). These authors offered
other explanations of the age differences, such as higher altruistic motivation, more
discretionary nature of volunteering, and involvement in more rewarding (e.g., church-
related) types of volunteering among older adults in comparison to younger individuals.
Previous research has typically considered either age or social roles, such as work and
partnership, but not both, as potential moderators of the effects of formal volunteering on
mental health. Moreover, relevant data come primarily from U.S. samples; evidence from
other countries with fairly high volunteering rates, such as Germany, is lacking. The present
study aimed to fill this gap: We investigated how the associations between formal
volunteering and mental health depended on employment and partnership status in two cross-
sectional samples of German adults aged 18–42 and 56–75.
According to the German Microcensus, in 2009, labor force participation (i.e., the
percentage of employed and job-seeking unemployed individuals in the population) among
German adults aged 55–64 amounted to 60.7% (52.7% among females) in comparison to
3.9% (2.6% among females) in those aged 65+ (i.e., beyond the statutory retirement age;
Statistisches Bundesamt Deutschland, 2011). These figures show that the productive potential
6
of the young-old is underused in the German labor market even among those younger than
65, which is at odds with the widely acknowledged need to foster active aging (Pavlova &
Silbereisen, 2012; World Health Organization, 2002). Furthermore, in 2009, 28.7% of
German adults aged 55–74 were single, divorced, or widowed (34.3% among females;
Statistisches Bundesamt Deutschland, 2011), thereby forming a substantial part of the
population. Concerning rates of self-reported formal volunteering (operationalised in the
German Survey on Volunteering [Freiwilligensurvey] as being a member of a voluntary
organization and doing volunteer work for this organization; Gensicke & Geiss, 2010), within
the past decade, they stabilized at 34–36% of the entire population older than 14, and among
those aged 55–74 there was a visible increase from 30.5% in 1999 to 34.5% in 2009.
Given these numbers, it is tempting to think of formal volunteering in this age group
(i.e. the young-old) as a possible compensation for a rather early exit from the labor force and
as a means of maintaining social integration among unpartnered individuals, whereas for
young and middle-aged adults, similar role losses are not irreversible and therefore cannot be
truly compensated by formal volunteering (Baltes, 1997; Greenfield & Marks, 2004;
Heckhausen et al., 2010; Piliavin, 2010). However, according to role accumulation theory
(Adelmann, 1994; Hao, 2008; Moen et al., 1992; Sieber, 1974; Thoits, 1983), those who are
better integrated in social life (i.e., enact multiple roles) may gain more from formal
volunteering, irrespective of age. Hence, we tested the following alternative hypotheses.
Hypothesis 1. Formal volunteering has a compensatory function in older adults.
Hypothesis 1a. At age 18–42, positive associations between formal volunteering and
mental health will not depend on employment and partnership status.
Hypothesis 1b. At age 56–75, positive associations between formal volunteering and
mental health will be most pronounced in those not working or without a partner.
Hypothesis 2. Formal volunteering has a complementary function at all ages.
7
Hypothesis 2a. At age 18–42, positive associations between formal volunteering and
mental health will be most pronounced in those working and in those with a partner.
Hypothesis 2b. At age 56–75, positive associations between formal volunteering and
mental health will be most pronounced in those working and in those with a partner.
As formal volunteering, by definition, implies doing unpaid work on behalf of an
organization, we considered two indicators of formal volunteering that represent different
sources of benefits for the volunteer: active participation in voluntary organizations, which
may provide status, social contact, and support (“social gratifications”; Verba et al., 1995),
and volunteer work itself, which may satisfy one’s altruistic motives (“civic gratifications”;
Verba et al., 1995) and give a sense of “mattering” (Owens, 2006; Thoits & Hewitt, 2001).
All these benefits are important for mental health and may therefore account for the salutary
effects of formal volunteering (Dovidio et al., 2006; Morrow-Howell, 2010). Separate
assessment of participation in voluntary organizations and volunteer work is possible as some
voluntary organizations, such as sport clubs, exist only to benefit their own members, who
attend meetings and interact with others regularly, thereby enjoying social gratifications of
organizational involvement without doing actual work for the wider community. At the same
time, some types of volunteer work, such as at a hospice, do not involve rewarding social
interactions and may only yield civic gratifications (Thoits & Hewitt, 2001; Verba et al.,
1995). Concerning outcome variables, we used three indicators of mental health and well-
being: life satisfaction, positive affect, and depressive symptoms.
Methods
Participants and Procedure
The present study was based on the secondary analysis of two cross-sectional surveys
covering 16–42 and 56–75 years of age from the Jena Study on Social Change and Human
Development (Silbereisen et al., 2006). These surveys, which investigated individual
8
responses to social change, focused originally on the growing uncertainty in the domains of
work, family, and leisure, and its implications for psychosocial adjustment at different ages.
The method of probability sampling was identical in both cases: Data were collected in four
German federal states, two representing the East (Mecklenburg-Western Pomerania and
Thuringia) and two representing the West (Baden-Württemberg and Schleswig-Holstein).
Sampling points were selected at random from the stratified area sample provided by the
Association of German Market and Social Research Institutes (ADM). Within each sampling
point, target households were identified using a random route technique. The interviewers
(trained personnel of a field research agency) made up to four attempts to approach each
household to find out whether an eligible individual was available for interview. Standardized
face-to-face interviews lasting about 1–1.5 hours were conducted in German. The interviews
slightly differed between the two surveys but contained many identical measures.
The younger sample was surveyed in October–December 2005 (response rate 77%).
In comparison to the German Microcensus 2004, unemployed individuals were slightly
overrepresented whereas single individuals and foreigners were underrepresented; otherwise
this sample was fairly representative of the same age population of respective federal states
(Reitzle, 2008). The older sample was surveyed in July–August 2009 (response rate 53%). In
comparison to the German Microcensus 2008, relatively older females and younger males, as
well as better educated and non-married individuals, were somewhat overrepresented in this
sample. From both samples, we excluded the participants with missing data on
sociodemographic variables and the indicators of formal volunteering (110 cases in the
younger and 86 cases in the older sample, respectively). Minors were also excluded from the
younger sample. Resulting sample sizes were: Nyounger = 2346, Nolder = 1422.
Measures
9
All measures described below were identical in both samples unless specified
otherwise. Participation in voluntary organizations (referred to as simply “participation”
throughout Method and Results) was assessed with one item: “I belong to a club, church
fellowship, or other initiative where I can make a contribution or where I am needed” (1 =
does not apply, 7 = fully applies). As a Likert scale, this item captured both membership of
an organization and the degree of (emotional) commitment to it, but it did not explicitly ask
whether the respondent had actually undertaken unpaid work for this organization. In the
younger sample, an additional response option (0 = is not applicable to me) was used by
many respondents (n = 376). To ensure comparability between the samples, we z-
standardized scores on this item within each sample.
Volunteer work was assessed with two items (adopted from Andolina, Keeter, Zukin,
& Jenkins, 2003): “Have you ever spent time participating in any community service or
volunteer activity? By volunteer activity, I mean actually working in some way to help others
for no pay” (yes/no). If the answer was affirmative, participants were asked: “Was this within
the last 12 months?” (yes/no). From this, we created a binary indicator of volunteer work
within the past 12 months (0 = no, 1 = yes). As follows from the item wording, it was not
explicitly specified whether the activity in question was a personal endeavor or a group
initiative. However, the terms “community service” (gemeinnützige Arbeit) and “volunteer
activity” (Ehrenamt) in their most common usage do not cover giving informal help to friends
and neighbors, therefore our measure of volunteer work refers to formal volunteering,
although not necessarily in conjunction with a permanent membership of an organization.
Participation in voluntary organizations and volunteer work were moderately intercorrelated
(.37 in the younger sample and .48 in the older sample), which testified to their distinctness.
For the purposes of regression analyses, employment status was coded as a binary
variable: working (employed full- or part-time; coded 0) and not working (coded 1).
10
Partnership status also comprised two categories: with a steady partner (irrespective of legal
status; coded 0) and without a steady partner (coded 1). More details on the sample
composition in terms of employment and partnership status will follow in the Results section.
Concerning mental health indicators, general life satisfaction was measured with a
single item (“How satisfied are you at present with your life altogether?”; 1 = very
dissatisfied, 7 = very satisfied). The 10-item positive affect subscale of the PANAS (Watson,
Clark, & Tellegen, 1988) rated how often participants had experienced certain positive
emotions within the last month (e.g., “enthusiastic”; 1 = never, 7 = very often; = .90 in both
samples). Depressive symptoms were measured with five items from the Brief Symptom
Inventory (Derogatis, 1993) assessing prevalence of the symptoms in the last month (e.g.,
“feeling hopeless about the future”; 1 = not at all, 7 = very strongly; = .89 in both samples).
Among the control variables, we included region (0 = West, 1 = East) as the
development of civil society still lags behind in the former East Germany with its communist
past (Howard, 2003) and continuing economic difficulties (Gensicke, Olk, Reim, Schmithals,
& Dienel, 2009), at least in terms of participation in voluntary organizations (64% in the East
vs. 73% in the West as of year 2009; Gensicke & Geiss, 2010). In addition, we controlled for
community size (1 = under 2000 inhabitants, 7 = more than 500 000 inhabitants), gender (0 =
male, 1 = female), school attainment (1 = 8 years of schooling, 2 = 10 years, 3 = 12-13
years), net income per person in the household in euro, age in years, and parenthood (0 = no
own children, 1 = has own children). Finally, we used two health indicators. Physical
handicap was measured with one item (“Are you permanently physically handicapped either
since birth or due to an accident?”; 0 = no, 1 = yes). General health was assessed with four
items from the German version of the SF-36 Health Survey (Bullinger & Kirchberger, 1998;
e.g., “I am as healthy as anybody I know”; 1 = completely disagree; 7 = completely agree; α
= .83 and .69 in the younger and in the older sample, respectively; mean score was used).
11
Analytical Approach
We regressed each mental health indicator on the predictor variables using a two-
group design, whereby we estimated two models for each regression equation: one with all
regression coefficients free to vary across the two samples and one with the coefficients of
interest constrained to be equal in both samples. A chi-square difference test indicated
whether the respective regression coefficients differed significantly between the younger and
the older samples. The predictors were all the control variables, employment and partnership
status, participation (z-standardized), volunteer work, and selected interactions among these
variables. Namely, we tested the hypothesized compensatory versus complementary effects
with respect to employment and partnership status by estimating the two-way interactions
between participation/volunteer work and employment/partnership status.
To avoid small cell sizes where interactions among multiple categorical variables
were involved, we tested interactions with employment and partnership status separately.
Given the large number of analyses (in particular, we tested 24 interaction effects), we set the
alpha level for regression coefficients to p < .01. We used a hierarchical approach to obtain
R2 change for the models with main effects only and with interaction effects added. Where
significant interactions emerged, we did not consider the main effects of participation and
volunteer work from the first step but calculated their simple effects at different levels of
moderator variables instead (see Aiken & West, 1991; Edwards, 2009). Analyses were
conducted using Mplus v.6 (Muthén & Muthén, 2010), wherein missing values on dependent
variables and covariates (income and general health) were handled with the full information
maximum likelihood algorithm. To compensate for a lack of multivariate normality, we used
logged scores on income and obtained bootstrapped standard errors for all regression
coefficients and simple effects.
Results
12
Descriptive Statistics
Descriptive statistics for the study variables are given in Table 1. In comparison to the
younger sample, participants from the older sample were more likely to live in larger
communities, reported lower school attainment and higher income, were more likely to have
children and less likely to be employed, reported more often to be physically handicapped,
scored lower on self-reported general health, and reported slightly higher life satisfaction and
slightly lower positive affect (all differences significant at p < .001). Concerning the
prevalence of participation and volunteer work, it was higher in the older than in the younger
sample (p < .001). In both samples, our global measures of formal volunteering yielded
somewhat lower prevalence rates than those reported by the German Survey on Volunteering
(Gensicke & Geiss, 2010), which employed more inclusive and differentiated indicators.
[Table 1 about here]
The composition of the two samples in terms of age and employment/partnership
status is presented in Table 2. It shows that the non-working category was largely comprised
of students in the youngest age group (18–25), replaced by the unemployed and homemakers
at age 26–42 and by retired individuals in the older sample (age 56–75). The high number of
non-working individuals in the younger sample is characteristic of the year 2005, reflecting
the then high unemployment rates (11.6% at ages 20–40) and the relatively low labor force
participation in this age group (81.1%; Statistisches Bundesamt Deutschland, 2011).
Regarding those without a steady partner, never married individuals prevailed in this category
at age 18–34; from age 35, those who were divorced became more visible; and at age 66–75,
widowed individuals formed the majority. Because of these age differences, we could not use
the subcategories of employment and partnership status in the regression analyses without
losing the comparability of the samples and compromising power of the analyses.
[Table 2 about here]
13
Bivariate correlations between the study variables (see Supplementary Material 1)
showed that sociodemographic indicators were weakly related to participation and volunteer
work, but the associations were in the expected directions. For instance, in both samples,
participation and volunteer work were related to higher school attainment, and West Germans
reported higher levels of participation than East Germans. Not working and not having a
steady partner were unrelated or weakly negatively related to participation and volunteer
work. As expected, correlations of participation and volunteer work with the mental health
indicators were positive (negative for depressive symptoms), although small in size.
Findings for Life Satisfaction
Regression analyses for life satisfaction are presented in Table 3, which shows
unstandardized regression coefficients. In the model with main effects (see Table 3, Model
1), after sociodemographic variables and health indicators had been controlled for, only
participation, and only in the younger sample, was significantly positively associated with life
satisfaction. However, before interpreting this effect, we had to consider possible interactions
with employment and partnership status (Aiken & West, 1991; Edwards, 2009). Here, our
analyses yielded different patterns of findings for the two indicators of formal volunteering.
With regard to participation in voluntary organizations, its interaction with
employment status was only marginally significant in the older sample (see Table 3, Model
2). More important, the first-order effect of participation (showing its effect at zero level of
the moderator, that is, in working individuals, coded 0 on employment status; Aiken & West,
1991) became significant in this model. This effect implied that, if anything, participation
was positively associated with life satisfaction among older working individuals, which
pointed at a possible complementary function of participation with respect to work. However,
no such trend was observed in the younger sample, where participation had clearly only a
main effect (see Table 3, Models 1 and 2). Thus, no evidence for compensatory effects of
14
participation was obtained, disconfirming Hypothesis 1, whereas Hypothesis 2, which posited
complementary effects, was partially supported in older adults.
[Table 3 about here]
A different picture emerged for volunteer work. In the older but not in the younger
sample, its interaction with employment status was significant (see Table 3, Model 2), which
is illustrated by Figure 1. At age 18–42, volunteer work had no significant relation to life
satisfaction, irrespective of employment status. In contrast, at age 56–75, volunteer work was
significantly associated with higher life satisfaction among non-working, but not among
working, individuals. Moreover, we found that this interaction effect differed significantly
between the younger and the older samples, 2(1) = 10.11, p < .01. These results were very
much in line with Hypothesis 1, suggesting that volunteer work had a compensatory function
with respect to employment status in older, but not in younger, adults. The alternative
Hypothesis 2, positing complementary effects in both age groups, was not supported.
[Figure 1 about here]
The two-way interactions with partnership status were not significant, nor were they
associated with significant R2 change in any of the samples (see Table 3, Model 3). Thus, we
found neither compensatory (Hypothesis 1) nor complementary (Hypothesis 2) effects of
participation and volunteer work on life satisfaction with regard to partnership status.
Findings for Positive Affect
Regression analyses for positive affect are shown in Table 4. In the model with main
effects (see Table 4, Model 1), both participation in voluntary organizations and volunteer
work were significantly related to higher positive affect in the younger sample, whereas in
the older sample, only volunteer work had a significant positive effect. The two-way
interactions with employment status were not associated with significant R2 change (see
Table 4, Model 2), whereas the two-way interactions with partnership status yielded
15
significant R2 change in the older sample (see Table 4, Model 3). However, none of the single
interaction effects was significant. Thus, we obtained only main effects of participation in
voluntary organizations and volunteer work (see Table 4, Model 1), neither of which differed
significantly between the samples: 2(1) = 1.12, ns, for participation and
2(1) = 0.06, ns,
for volunteer work. From this, we concluded that, in younger and older adults alike,
participation and volunteer work were associated with higher positive affect, but they had
neither compensatory (Hypothesis 1) nor complementary (Hypothesis 2) effects with respect
to employment and partnership status.
[Table 4 about here]
Findings for Depressive Symptoms
Regression analyses for depressive symptoms are shown in Table 5. In the model with
main effects (see Table 5, Model 1), participation in voluntary organizations had a significant
negative effect on depressive symptoms in the younger sample, whereas volunteer work had a
significant negative effect in the older sample. None of the interaction effects involving
participation was significant (see Table 5, Models 2 and 3), which suggested that
participation had neither compensatory (Hypothesis 1) nor complementary (Hypothesis 2)
effects on depressive symptoms with respect to employment and partnership status. However,
it had a negative main effect on depressive symptoms in the younger sample (see Table 5,
Model 1), which was significantly stronger than in the older sample: 2(1) = 4.86, p < .05.
Thus, participation was only related to fewer depressive symptoms in younger adults.
[Table 5 about here]
With regard to volunteer work, no significant interaction was found with employment
status (see Table 5, Model 2), but a significant interaction with partnership status emerged in
the older sample (see Table 5, Model 3). This interaction is illustrated by Figure 2. At age
18–42, volunteer work was not significantly related to depressive symptoms, irrespective of
16
partnership status, whereas at age 56–75, it was significantly related to fewer depressive
symptoms, but only among individuals without a steady partner. It should be noted, though,
that the interaction between volunteer work and partnership status (see Table 5, Model 3) did
not differ significantly between the younger and the older samples, 2(1) = 2.10, ns. Thus,
for volunteer work, Hypothesis 1, which posited compensatory effects of formal volunteering
in older adults, was partially supported with respect to partnership status. No evidence for
complementary effects of volunteer work (Hypothesis 2) was obtained. Concerning
employment status, neither Hypothesis 1 nor Hypothesis 2 was supported.
[Figure 2 about here]
Discussion
The present study comparing two independent samples of German adults (age 18–42
and 56–75) contributed to the mounting evidence on the benefits of formal volunteering for
the volunteers themselves. We distinguished between participation in voluntary organizations
and volunteer work and investigated how their associations with mental health (i.e., life
satisfaction, positive affect, and depressive symptoms) varied depending on employment and
partnership status. Moreover, our study was one of the few that directly tested commonly
held assumptions about the especially salutary effects of formal volunteering in older adults
by comparing age groups (cf. Musick & Wilson, 2003; Van Willigen, 2000).
We juxtaposed two views on the interplay between formal volunteering and other
social roles at different ages. According to the lifespan psychological perspective (Baltes,
1997; Heckhausen et al., 2010), formal volunteering may be most important for mental health
at an older age, when there is a need to compensate for irrevocable losses of work and family
roles (e.g., retirement and widowhood). Formal volunteering may thereby have compensatory
effects on mental health among non-working individuals and in those without a steady
partner in older, but not in younger, adults. In contrast, proponents of role accumulation
17
theory (Sieber, 1974; Thoits, 1983) would rather argue that formal volunteering has
complementary effects (i.e., it is more beneficial for those who do work or have a steady
partner) in all age groups, as enacting multiple social roles enables the accumulation of
privileges and resources, provides status security, and serves personality enrichment.
Our findings yielded limited support for role accumulation theory. Only in one case
did we find some indication that participation in voluntary organizations might have
complementary effects as it was particularly related to higher life satisfaction in older
working Germans. In comparison to their peers exiting the labor market, older Germans who
are still in work are typically better qualified and have higher occupational status (Radl,
2007). For many of this group, participation in voluntary organizations may involve holding a
leading position rather than ordinary membership; the very organizations in which they
participate may be of high prestige. This would certainly be an example of role accumulation.
Otherwise, participation in voluntary organizations had positive associations with mental
health irrespective of employment and partnership status, that is, participation in voluntary
organizations did not seem to have a compensatory role either. Its effects being sometimes
stronger among younger adults may be explained by the higher responsiveness of younger
individuals to the social gratifications of volunteering (cf. Omoto, Snyder, & Martino, 2000).
For volunteer work, a very different picture emerged, providing convincing evidence
for its compensatory, rather than complementary, function in older German adults.
Specifically, in the older sample, volunteer work was significantly associated with higher life
satisfaction only among non-working individuals and with fewer depressive symptoms only
among those without a steady partner (cf. Greenfield & Marks, 2004; Piliavin, 2010;
Sugihara et al., 2008). In contrast, in the younger sample, volunteer work had no significant
associations with life satisfaction and depressive symptoms, although the differences between
the age groups were significant only for life satisfaction. In addition, volunteer work was
18
related to higher positive affect in younger and older adults alike, irrespective of employment
and partnership status.
Thus, the two interrelated indicators of formal volunteering were associated with
remarkably different patterns of effects. Although there was some overlap in the wording of
the two items, we reasoned that including them simultaneously as predictors in a regression
equation should yield the unique effects of organizational involvement, with its presumed
social gratifications, and volunteer work, which had more to do with civic gratifications
(doing one’s duty, benefiting others; Verba et al., 1995). If participation in voluntary
organizations had stronger effects on mental health, this could indicate that social
gratifications matter more, and vice versa. Seen in this way, our results suggest that civic
gratifications are more important to older volunteers, who may be driven by community
concerns and altruistic motives more than their younger counterparts (cf. Musick & Wilson,
2003; Omoto et al., 2000). Apart from that, non-working older Germans may profit from
volunteer work insofar as it satisfies their need to remain productive, especially given that
they are bound to exit the labor force at age 65–67 (the statutory retirement age), but are
nevertheless under increasing pressure to maintain an active lifestyle and to contribute to the
common good (Pavlova & Silbereisen, 2012; van Dyk & Lessenich, 2009).
In turn, older adults without a steady partner, be they single, divorced, or widowed,
are faced with the prospect of remaining unpartnered for the rest of their lives (Heckhausen et
al., 2010). This group may benefit from volunteer work particularly because it gives them a
sense of “mattering,” that is, a feeling that one is needed and appreciated by other people
(Owens, 2006; Thoits & Hewitt, 2001). Organizational involvement as such, despite its
importance for the social integration of older adults, may be less effective in meeting such
needs than volunteer work as a purposeful activity with a tangible output (cf. Thoits &
Hewitt, 2001).
19
The differences that emerged across mental health indicators may also highlight
different functions of formal volunteering (Dovidio et al., 2006; Morrow-Howell, 2010). For
instance, satisfying the need to stay productive may enhance self-perceptions of well-being in
the first place, which explains the link between volunteer work and life satisfaction, a
predominantly cognitive evaluation of one’s life as a whole (Diener, 1994), in older non-
working individuals. In turn, a sense of “mattering” may buffer against depression (Owens,
2006), which accounts for the connection between volunteer work and reduced depressive
symptoms in older individuals without a steady partner.
Limitations
Our study had its limitations. The datasets that we used were cross-sectional and
contained only global measures of formal volunteering. Types of activities, time investment,
and duration of commitment were not specified, nor were particular features of the volunteer
experience. To a large extent, these topics are covered in the German Survey on Volunteering
(Freiwilligensurvey; Gensicke & Geiss, 2010), which, however, does not include mental
health indicators. We could have used the German Socio-Economic Panel, which has
advantages of a panel survey, but its focus is neither on volunteering nor on psychological
variables. Furthermore, as our study utilized cross-sectional data, causal inferences cannot be
drawn. It is possible that depressed individuals abstain from volunteering, which would lead
to volunteers reporting better mental health on average. However, concerning the moderator
effects we found, reversing their direction would imply that only those depressed individuals
who are older and have no steady partner (e.g., widowed) are less likely to volunteer, which
would be difficult to interpret. It does seem, therefore, that the pattern of our findings can be
more parsimoniously explained via compensation than via self-selection. In previous studies,
the longitudinal effects of formal volunteering on well-being have been established (Li &
Ferraro, 2005; Thoits & Hewitt, 2001), which also supports the suggested direction of effects.
20
The fact that we compared two independent samples of different ages was both a
strength and a potential limitation because the time of measurement also differed: The
younger sample was surveyed in 2005 whereas the older sample was surveyed in 2009. The
compensatory effect of volunteer work on life satisfaction among non-working individuals
that we found in the older sample could thus be attributed to the success of activation policies
between 2005 and 2009, which especially encouraged formal volunteering in the older
population (e.g., Bundesministerium für Familie, Senioren, Frauen und Jugend, 2009).
However, as activation policies targeting young and middle-aged non-working individuals
primarily aim at their reemployment (Jacobi & Kluve, 2007), volunteering is not likely to
gain prominence as an alternative to work in this group. Thus, the age differences we found
should still hold even if both samples had been surveyed in 2009.
Conclusions
We presented evidence for the compensatory role of volunteer work (but not of
participation in voluntary organizations) in older German adults. That is, we found the
positive association between volunteer work and mental health to be stronger among older
individuals who were not working or had no steady partner than among those working or
with a steady partner. In young and middle-aged adults, such compensatory effects were not
pronounced. Thus, volunteer recruitment programs in Germany should perhaps target older
non-working and unpartnered individuals in particular. Nevertheless, it should be borne in
mind that younger individuals also derive emotional benefits from formal volunteering.
Our study highlights individual differences in the benefits of volunteering for the
volunteers themselves. In the future, researchers should direct more attention to the role of
macrolevel variables, such as cultural and regional contexts, in these associations, as well as
to microlevel variations in the volunteer experience (e.g., program characteristics), which
may be decisive for the outcomes of volunteering (Morrow-Howell, 2010).
21
Funding. This work was supported by the German Research Foundation as a subproject of
the Collaborative Research Centre 580 “Social Developments in Post-Socialistic Societies:
Discontinuity, Tradition, Structural Formation” (SFB580-04-C6 to R. K. S.) and by the
Federal Program “ProExzellenz” of the Free State of Thuringia (to M. K. P.).
Acknowledgements. The authors are grateful to Verona Christmas-Best for her helpful
comments on the manuscript.
Author contributions. M. K. Pavlova planned the study, conducted statistical analyses, and
wrote the paper. R. K. Silbereisen helped to plan the study and supervised the preparation of
the paper.
22
References
Adelmann, P. K. (1994). Multiple roles and psychological well-being in a national sample of
older adults. Journal of Gerontology: Social Sciences, 49, S277-S285.
Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting
interactions. Thousand Oaks, CA: Sage.
Andolina, M., Keeter, S., Zukin, C., & Jenkins, K. (2003). A guide to the index of civic and
political engagement. College Park, MD: The Center for Information and Research on
Civic Learning and Engagement.
Baltes, P. B. (1997). On the incomplete architecture of human ontogeny: Selection,
optimization, and compensation as foundation of developmental theory. American
Psychologist, 52, 366-380. doi:10.1037/0003-066X.52.4.366
Bundesministerium für Familie, Senioren, Frauen und Jugend. (2009). Mitgestalten und
mitentscheiden: Das Programm “Aktiv im Alter” [Shape and decide together: The
program “Active in old age”]. Retrieved on November 24, 2011, from
http://www.bmfsfj.de/RedaktionBMFSFJ/Broschuerenstelle/Pdf-
Anlagen/mitgestalten-
mitentscheiden,property=pdf,bereich=bmfsfj,sprache=de,rwb=true.pdf
Bullinger, M., & Kirchberger, I. (1998). Der SF-36 Fragebogen zum Gesundheitszustand.
Handbuch für die deutschsprachige Fragebogenversion [The SF-36 questionnaire on
health status. Manual for the German version]. Göttingen, Germany: Hogrefe.
Derogatis, L. R. (1993). Brief Symptom Inventory (BSI): Administration, scoring, and
procedures manual. (3rd ed.) Minneapolis, MN: National Computer Systems.
Diener, E. (1994). Assessing subjective well-being: Progress and opportunities. Social
Indicators Research, 31, 103-157. doi:10.1007/BF01207052
23
Dovidio, J. F., Piliavin, J. A., Schroeder, D. A., & Penner, L. A. (2006). The social
psychology of prosocial behavior. Mahwah, NJ: Lawrence Erlbaum Associates,
Publishers.
Edwards, J. R. (2009). Seven deadly myths of testing moderation in organizational research.
In C. E. Lance & R. J. Vandenberg (Eds.), Statistical and methodological myths and
urban legends: Doctrine, verity and fable in the organizational and social sciences
(pp. 143-164). New York, NY: Routledge.
Gensicke, T., & Geiss, S. (2010). Hauptbericht des Freiwilligensurveys 2009:
Zivilgesellschaft, soziales Kapital und freiwilliges Engagement in Deutschland 1999 –
2004 – 2009 [Main report from the Survey on Volunteering 2009: Civil society, social
capital, and volunteering in Germany 1999 – 2004 – 2009]. Retrieved on November
24, 2011, from http://www.bmfsfj.de/RedaktionBMFSFJ/Broschuerenstelle/Pdf-
Anlagen/3._20Freiwilligensurvey-
Hauptbericht,property=pdf,bereich=bmfsfj,sprache=de,rwb=true.pdf
Gensicke, T., Olk, T., Reim, D., Schmithals, J., & Dienel, H.-L. (2009). Entwicklung der
Zivilgesellschaft in Ostdeutschland: Quantitative and qualitative Befunde
[Development of civil society in East Germany: Quantitative and qualitative
findings]. Wiesbaden, Germany: VS Verlag für Sozialwissenschaften.
Greenfield, E. A., & Marks, N. F. (2004). Formal volunteering as a protective factor for older
adults’ psychological well-being. Journal of Gerontology: Social Sciences, 59B,
S258-S264. doi:10.1093/geronb/59.5.S258
Hao, Y. (2008). Productive activities and psychological well-being among older adults.
Journal of Gerontology: Social Sciences, 63B, S64-S72.
Havighurst, R. J. (1972). Developmental tasks and education. (3rd ed.) New York, NY:
David McKay.
24
Heckhausen, J., Wrosch, C., & Schulz, R. (2010). A motivational theory of life-span
development. Psychological Review, 117, 32-60. doi:10.1037/a0017668
Howard, M. M. (2003). The weakness of civil society in post-communist Europe. Cambridge,
UK: Cambridge University Press.
Jacobi, L., & Kluve, J. (2007). Before and after the Hartz reforms: The performance
of active labour market policy in Germany. Zeitschrift für ArbeitsmarktForschung, 1,
45-64.
Li, Y., & Ferraro, K. F. (2005). Volunteering and depression in later life: Social benefit or
selection processes? Journal of Health and Social Behavior, 46, 68-84.
doi:10.1177/002214650504600106
Moen, P., Dempster-McClain, D., & Williams, R. M., Jr. (1992). Successful aging: A life-
course perspective on women’s multiple roles and health. American Journal of
Sociology, 97, 1612-1638.
Morrow-Howell, N. (2010). Volunteering in later life: Research frontiers. Journal of
Gerontology: Social Sciences, 65B, 461-469. doi:10.1093/geronb/gbq024
Morrow-Howell, N., Hinterlong, J., Rozario, P. A., & Tang, F. (2003). Effects of
volunteering on the well-being of older adults. Journal of Gerontology: Social
Sciences, 58B, S137-S145. doi:10.1093/geronb/58.3.S137
Musick, M. A., & Wilson, J. (2003). Volunteering and depression: The role of psychological
and social resources in different age groups. Social Science & Medicine, 56, 259-269.
doi:10.1016/S0277-9536(02)00025-4
Muthén, L. K., & Muthén, B. O. (2010). Mplus User’s Guide. Sixth Edition. Retrieved on
November 24, 2011, from
http://www.statmodel.com/download/usersguide/Mplus%20Users%20Guide%20v6.p
df
25
Oman, D., Thoresen, C. E., & McMahon, K. (1999). Volunteerism and mortality among the
community-dwelling elderly. Journal of Health Psychology, 4, 301-316.
doi:10.1177/135910539900400301
Omoto, A. M., Snyder, M., & Martino, S. C. (2000). Volunteerism and the life course:
Investigating age-related agendas for action. Basic and Applied Social Psychology,
22, 181-197. doi:10.1207/S15324834BASP2203_6
Owens, T. J. (2006). Self and identity. In J. Delamater (Ed.), Handbook of social psychology
(pp. 205-232). New York, NY: Springer.
Pavlova, M. K., & Silbereisen, R. K. (2012). Perceived level and appraisal of the growing
expectations for active ageing among the young-old in Germany. Research on Aging,
34, 80-99. doi:10.1177/0164027511416371
Piliavin, J. A. (2010). Volunteering across the life span: Doing well by doing good. In S.
Stürmer & M. Snyder (Eds.), The psychology of prosocial behavior: Group processes,
intergroup relations, and helping (pp. 157-172). Chichester, UK: Wiley-Blackwell.
Radl, J. (2007). Individuelle Determinanten des Renteneintrittsalters: Eine empirische
Analyse von Übergängen in den Ruhestand [Individual determinants of the age of
retirement: An empirical analysis of transitions to old age pensions]. Zeitschrift für
Soziologie, 36, 43-64.
Reitzle, M. (2008). Stichprobe [Sample]. In R. K. Silbereisen & M. Pinquart (Eds.),
Individuum und sozialer Wandel (pp. 37-54). Weinheim, Germany: Juventa Verlag.
Sieber, S. D. (1974). Toward a theory of role accumulation. American Sociological Review,
39, 567-578.
Silbereisen, R. K., Pinquart, M., Reitzle, M., Tomasik, M. J., Fabel, K., & Grümer, S. (2006).
Psychosocial resources and coping with social change (SFB-580 Reports, Volume
26
19). Retrieved on November 24, 2011, from http://www.sfb580.uni-
jena.de/typo3/uploads/tx_publicationlist/sfb_580_silbereisen_5.pdf
Statistisches Bundesamt Deutschland. (2011). GENESIS-Online Datenbank [GENESIS-
online database]. Retrieved on November 24, 2011, from https://www-
genesis.destatis.de/
Sugihara, Y., Sugisawa, H., Shibata, H., & Harada, K. (2008). Productive roles, gender, and
depressive symptoms: Evidence from a national longitudinal study of late-middle-
aged Japanese. Journal of Gerontology: Psychological Sciences, 63B, P227-P234.
Thoits, P. A. (1983). Multiple identities and psychological well-being: A reformulation and
test of the social isolation hypothesis. American Sociological Review, 48, 174-187.
Thoits, P. A., & Hewitt, L. N. (2001). Volunteer work and well-being. Journal of Health and
Social Behavior, 42, 115-131. doi:10.2307/3090173
van Dyk, S., & Lessenich, S. (2009). “Junge Alte”: Vom Aufstieg und Wandel einer
Sozialfigur [“Young-old”: On the emergence and change of a social figure]. In S. van
Dyk & S. Lessenich (Eds.), Die jungen Alten: Analysen einer neuen Sozialfigur (pp.
11-48). Frankfurt am Main, Germany: Campus Verlag.
Van Willigen, M. (2000). Differential benefits of volunteering across the life course. Journal
of Gerontology: Social Sciences, 55B, S308-S318. doi:10.1093/geronb/55.5.S308
Verba, S., Schlozman, K. L., & Brady, H. E. (1995). Voice and equality: Civic voluntarism in
American politics. Cambridge, MA: Harvard University Press.
Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief
measures of positive and negative affect: The PANAS scales. Journal of Personality
and Social Psychology, 54, 1063-1070. doi:10.1037/0022-3514.54.6.1063
Wilson, J. (2000). Volunteering. Annual Review of Sociology, 26, 215-240.
doi:10.1146/annurev.soc.26.1.215
27
World Health Organization. (2002). Active ageing: A policy framework. Retrieved on
November 24, 2011, from
http://whqlibdoc.who.int/hq/2002/WHO_NMH_NPH_02.8.pdf
28
Table 1
Descriptive Statistics for the Study Variables
Indicator Age 18–42 Age 56–75
N 2346 1422
Age, M (SD) 32.4(7.5) 65.8(5.9)
East Germany, % 50.2 50.2
Community size, M (SD) 3.5(1.7) 4.9(1.6)
Female, % 54.6 52.3
School attainment, %
8 years 24.0 46.6
10 years 53.6 31.6
12-13 years 22.4 21.9
Income in euro, M (SD) 868.0(580.2) 1245.0(689.8)
Has children, % 51.8 86.8
Physical handicap, % 3.5 6.1
General health, M (SD) 5.8(1.3) 4.7(1.3)
Not working, % 42.2 76.4
No steady partner, % 33.5 32.4
Participation in voluntary organizations, M (SD) 2.7(2.5) 3.7(2.6)
High participation (> M + 1SD), % 21.4 28.7
Volunteer work in the past 12 months, % 19.4 32.8
Life satisfaction, M (SD) 5.0(1.4) 5.2(1.3)
Positive affect, M (SD) 5.4(1.0) 5.1(1.0)
Depressive symptoms, M (SD) 1.8(1.2) 1.8(1.1)
29
Table 2
Self-Reported Employment and Partnership Status in the Two Study Samples
Categories
Younger sample, n (% within age group) Older sample, n (% within age group)
Age 18–25 Age 26–34 Age 35–42 Age 56–65 Age 66–75
Employed 219 (34.5%) 409 (60.8%) 727 (70.0%) 306 (44.2%) 29 (4.0%)
Not employed 415 (65.5%) 264 (39.2%) 312 (30.0%) 386 (55.8%) 701 (96.0%)
In education 252 (39.7%) 37 (5.5%) 4 (0.4%) 2 (0.3%) 0 (0.0%)
Homemakers, parental leave 42 (6.6%) 96 (14.3%) 119 (11.5%) 21 (3.0%) 2 (0.3%)
Unemployed 109 (17.2%) 126 (18.7%) 165 (15.9%) 68 (9.8%) 0 (0.0%)
Retired 0 (0.0%) 0 (0.0%) 0 (0.0%) 259 (37.4%) 694 (95.1%)
Other 12 (1.9%) 5 (0.7%) 24 (2.3%) 36 (5.2%) 5 (0.7%)
With a steady partner 289 (45.6%) 466 (69.2%) 804 (77.4%) 502 (72.5%) 459 (62.9%)
Without a steady partner 345 (54.4%) 207 (30.8%) 235 (22.6%) 190 (27.5%) 271 (37.1%)
Single 343 (54.1%) 187 (27.8%) 109 (10.5%) 52 (7.5%) 25 (3.4%)
Divorced/separated 2 (0.3%) 19 (2.8%) 118 (11.4%) 87 (12.6%) 64 (8.8%)
Widowed 0 (0.0%) 1 (0.1%) 8 (0.8%) 51 (7.4%) 182 (24.9%)
30
Table 3
Regression Results for Life Satisfaction
Model 1 Model 2 Model 3
Predictors Age 18–42 Age 56–75 Age 18–42 Age 56–75 Age 18–42 Age 56–75
East Germany -.310***
(.055)
-.182
(.073)
-.312***
(.055)
-.176
(.073)
-.311***
(.055)
-.187
(.073)
Community size -.015
(.016)
-.009
(.019)
-.014
(.016)
-.009
(.019)
-.014
(.016)
-.008
(.019)
Female .129
(.053)
.313***
(.062)
.126
(.053)
.318***
(.062)
.129
(.053)
.312***
(.062)
School attainment .293***
(.040)
-.013
(.040)
.291***
(.040)
-.009
(.039)
.294***
(.040)
-.009
(.040)
Income (logged) .338***
(.066)
.446***
(.085)
.338***
(.066)
.445***
(.085)
.337***
(.066)
.446***
(.085)
Age -.034***
(.005)
.032***
(.007)
-.034***
(.005)
.034***
(.007)
-.034***
(.005)
.033***
(.007)
Has children .172
(.077)
.141
(.094)
.171
(.077)
.147
(.094)
.168
(.077)
.132
(.094)
Physical handicap -.138
(.171)
.002
(.128)
-.132
(.170)
-.009
(.129)
-.131
(.171)
.005
(.127)
General health .208***
(.024)
.270***
(.027)
.209***
(.024)
.270***
(.027)
.208***
(.024)
.272***
(.027)
Not working -.524***
(.072)
-.038
(.094)
-.484***
(.078)
-.231
(.111)
-.522***
(.072)
-.034
(.093)
No steady partner -.429***
(.063)
-.587***
(.069)
-.429***
(.063)
-.591***
(.068)
-.402***
(.068)
-.504***
(.084)
Participation in voluntary
organizations (P)
.081**
(.027)
.076
(.038)
.058
(.033)
.249**
(.078)
.087**
(.032)
.031
(.045)
31
Volunteer work (V)
.047
(.066)
.161
(.082)
.130
(.077)
-.253
(.163)
.098
(.081)
.249**
(.095)
P x Not working
.055
(.058)
-.219
(.088)
V x Not working
-.193
(.151)
.530**
(.177)
P x No steady partner -.026
(.062)
.128
(.078)
V x No steady partner -.176
(.171)
-.271
(.176)
Total R2 .166*** .201*** .167*** .207*** .167*** .204***
R2 change .001 .006** .001 .003
Note. Cells represent B (SE).
** p < .01. *** p < .001.
32
Table 4
Regression Results for Positive Affect
Model 1 Model 2 Model 3
Predictors Age 18–42 Age 56–75 Age 18–42 Age 56–75 Age 18–42 Age 56–75
East Germany -.017
(.037)
.090
(.052)
-.015
(.037)
.089
(.052)
-.018
(.037)
.090
(.052)
Community size -.018
(.011)
-.037
(.017)
-.018
(.011)
-.037
(.017)
-.018
(.011)
-.036
(.016)
Female .098**
(.037)
.066
(.047)
.100**
(.037)
.066
(.047)
.098**
(.037)
.057
(.047)
School attainment .122***
(.028)
.050
(.031)
.123***
(.027)
.049
(.032)
.123***
(.028)
.050
(.031)
Income (logged) .138***
(.038)
.204***
(.057)
.137***
(.038)
.204***
(.057)
.137***
(.038)
.200***
(.057)
Age -.012***
(.003)
-.004
(.005)
-.012***
(.003)
-.004
(.005)
-.012***
(.003)
-.004
(.005)
Has children .092
(.050)
.225**
(.079)
.092
(.050)
.225**
(.079)
.090
(.050)
.233**
(.078)
Physical handicap .231
(.101)
.066
(.111)
.226
(.101)
.068
(.111)
.234
(.102)
.069
(.111)
General health .295***
(.016)
.289***
(.020)
.294***
(.016)
.289***
(.020)
.295***
(.016)
.289***
(.020)
Not working -.180***
(.044)
.000
(.063)
-.210***
(.048)
.029
(.074)
-.180***
(.044)
-.001
(.063)
No steady partner -.150**
(.044)
-.179**
(.053)
-.149**
(.044)
-.178**
(.053)
-.134**
(.047)
-.217**
(.065)
Participation in voluntary
organizations (P)
.103***
(.019)
.069
(.027)
.121***
(.024)
.041
(.052)
.102***
(.023)
.029
(.034)
33
Volunteer work (V)
.199***
(.044)
.217***
(.057)
.135
(.057)
.280**
(.107)
.228***
(.054)
.179
(.069)
P x Not working
-.042
(.039)
.035
(.061)
V x Not working
.147
(.093)
-.081
(.126)
P x No steady partner -.001
(.041)
.120
(.058)
V x No steady partner -.096
(.100)
.149
(.120)
Total R2 .233*** .221*** .233*** .221*** .233*** .226***
R2 change .000 .000 .000 .005**
Note. Cells represent B (SE).
** p < .01. *** p < .001.
34
Table 5
Regression Results for Depressive Symptoms
Model 1 Model 2 Model 3
Predictors Age 18–42 Age 56–75 Age 18–42 Age 56–75 Age 18–42 Age 56–75
East Germany .206***
(.046)
.072
(.059)
.207***
(.046)
.073
(.059)
.206***
(.046)
.069
(.059)
Community size .020
(.013)
.015
(.017)
.020
(.013)
.015
(.017)
.021
(.013)
.016
(.017)
Female -.030
(.047)
.085
(.051)
-.029
(.047)
.085
(.051)
-.030
(.046)
.093
(.051)
School attainment -.155***
(.035)
-.119***
(.032)
-.153***
(.035)
-.119***
(.032)
-.153***
(.035)
-.117***
(.033)
Income (logged) -.133
(.054)
-.178**
(.067)
-.133
(.054)
-.179**
(.067)
-.134
(.054)
-.175**
(.067)
Age .014**
(.004)
-.010
(.006)
.014**
(.004)
-.010
(.006)
.014**
(.004)
-.010
(.006)
Has children -.091
(.066)
-.064
(.081)
-.094
(.066)
-.063
(.081)
-.093
(.066)
-.080
(.080)
Physical handicap .115
(.157)
.036
(.125)
.114
(.157)
.035
(.126)
.121
(.157)
.036
(.123)
General health -.310***
(.022)
-.348***
(.022)
-.310***
(.022)
-.348***
(.023)
-.310***
(.022)
-.346***
(.022)
Not working .376***
(.059)
-.067
(.071)
.368***
(.065)
-.082
(.093)
.377***
(.059)
-.063
(.071)
No steady partner .510***
(.057)
.508***
(.064)
.511***
(.057)
.508***
(.064)
.532***
(.063)
.623***
(.080)
Participation in voluntary
organizations (P)
-.090***
(.022)
-.007
(.029)
-.068**
(.022)
-.005
(.058)
-.089***
(.023)
-.009
(.035)
35
Volunteer work (V)
-.046
(.053)
-.207**
(.060)
-.060
(.055)
-.239
(.116)
-.005
(.061)
-.087
(.071)
P x Not working
-.052
(.047)
-.002
(.068)
V x Not working
.034
(.116)
.042
(.141)
P x No steady partner -.007
(.052)
-.002
(.062)
V x No steady partner -.136
(.132)
-.401**
(.131)
Total R2 .224*** .260*** .224*** .260*** .224*** .265***
R2 change .000 .000 .000 .005**
Note. Cells represent B (SE).
** p < .01. *** p < .001.
36
Figure 1. Simple effects of volunteer work on life satisfaction in working and non-working
individuals. Adjusted means with all other predictors held constant are shown. ** p < .01.
37
Figure 2. Simple effects of volunteer work on depressive symptoms in those with and those
without a steady partner. Adjusted means with all other predictors held constant are shown.
*** p < .001.
38
Supplementary Material 1
Bivariate Correlations Between the Study Variables
Variable 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
1.East Germany – -.23 -.01 .16 -.27 -.12 .04 .06 .10 .10 .05 -.08 .06 -.11 .03 .07
2. Community size -.31 – .00 .08 .13 .00 -.13 -.04 -.12 -.01 .06 -.05 -.12 .00 -.09 .05
3. Female .03 .03 – .09 -.11 .07 .21 -.01 -.07 .12 -.12 -.01 .02 .03 .02 .01
4. School attainment .01 .07 -.13 – .14 -.04 -.11 -.01 .08 -.01 .01 .09 .11 .17 .15 -.12
5. Income (logged) -.35 .21 -.13 .29 – .13 -.33 -.04 .01 -.54 .06 .07 -.02 .24 .12 -.16
6. Age .02 -.01 .16 -.22 -.04 – .51 .05 -.14 -.30 -.27 .06 .04 -.05 -.05 -.01
7. Has children .16 -.09 .10 -.06 -.09 .07 – .01 -.01 -.03 -.42 .06 .07 -.02 .02 -.06
8. Physical handicap .02 .01 -.05 .01 -.06 -.09 -.04 – -.20 .05 .01 -.01 .03 -.09 -.05 .10
9. General health -.09 -.03 -.01 .08 .16 -.04 .06 -.08 – -.07 -.04 .00 .05 .24 .42 -.37
10. Not working .08 -.04 .14 -.19 -.27 .55 .01 .04 -.12 – .13 -.06 .01 -.25 -.15 .23
11. No steady partner -.03 .09 .25 -.09 .00 .12 -.22 .02 -.10 .10 – -.08 -.05 -.16 -.11 .24
12. Participation in
voluntary organizations -.14 .01 .01 .13 .11 .02 .06 .04 .08 .01 -.03 – .37 .11 .16 -.12
13. Volunteer work -.02 -.02 -.05 .21 .08 -.05 .02 .10 .08 -.02 -.08 .48 – .06 .16 -.08
39
14. Life satisfaction -.14 .02 .07 .05 .22 .12 .10 -.05 .33 -.02 -.21 .15 .13 – .40 -.50
15. Positive affect .00 -.08 -.01 .13 .15 -.05 .13 -.01 .43 -.09 -.15 .17 .19 .38 – -.50
16. Depressive symptoms .08 .03 .11 -.16 -.18 .00 -.08 .04 -.45 .06 .27 -.11 -.16 -.42 -.44 –
Myounger 0.50 3.49 0.55 1.98 6.58 32.40 0.52 0.04 5.75 0.42 0.34 2.74 0.19 5.03 5.40 1.79
SDyounger 0.50 1.74 0.50 0.68 0.62 7.53 0.50 0.18 1.29 0.49 0.47 2.48 0.40 1.41 0.96 1.19
Molder 0.50 4.90 0.52 1.75 7.00 65.80 0.87 0.06 4.71 0.76 0.32 3.73 0.33 5.21 5.12 1.84
SDolder 0.50 1.56 0.50 0.79 0.50 5.86 0.34 0.24 1.32 0.42 0.47 2.63 0.47 1.29 0.99 1.12
Note. Coefficients above the diagonal refer to the younger sample (N = 2059–2346), those below the diagonal refer to the older sample (N =
1317–1422). Pairwise deletion of missing values was employed. Coefficients significant at p < .05 are italicized, at p < .01 are in bold and
italicized, at p < .001 are in bold.
Top Related