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.