The Gerontologist
Vol. 41, No. 2, 250–256
Copyright 2001 by The Gerontological Society of America
The Effects of Membership in Church-Related
Associations and Labor Unions on Age
Differences in Voluntary Association Affiliations
Jon Hendricks, PhD,1 and Stephen J. Cutler, PhD2
Purpose: When compositional differences between age
groups are controlled, the bivariate, curvilinear pattern of
age differences in voluntary association memberships is
replaced by one in which peak membership levels occur
in late middle age and remain relatively stable thereafter.
We sought to determine if this adjusted pattern persists
when widespread memberships in two types of associations are removed: (1) church-related organizations, in
which membership increases with age but may be confounded with church membership per se; and (2) labor
unions, in which membership is also widespread but may
not be “voluntary.” Design and Methods: Total memberships in 14 types of associations were examined using
aggregated data from 12 national surveys conducted between 1974 and 1994. Results: The familiar curvilinear
relationship at the bivariate level appears, but in the multivariate analysis, membership levels rise from ages 18–
24 through 55–59 and persist close to that level through
ages 85 and older. Implications: The results reaffirm
that age differences in voluntary association affiliations
are due to age-based compositional differences; they
also show, however, that the pattern observed with controls is neither driven by membership in church-related
“associations” nor altered when membership in labor
unions is excluded.
other segments of the population. This research is
predicated on a contention that such claims risk mistaking cohort-related characteristics for willingness to
contribute and actual contributions. As for older persons themselves, participation in various forms of
voluntary association has salutary effects. Activity
levels per se and diverse forms of engagement are
thought to be supportive of life satisfaction, morale,
and even physical functioning. The rewards are seemingly both intrinsic and extrinsic, leading to a preservation of sense of well-being, solidarity, integration,
and health (Gecas & Burke, 1995; Moen, 1996;
Moen, Dempster-McClain, & Williams, 1989, 1992;
Monk, 1995; Musick, Herzog, & House, 1999; Pillemer, Moen, Wethington, & Glasgow, 2000).
Because it is widely maintained that both rates
and frequency of voluntary association membership
build to and then decline after a midlife peak (Fischer
& Schaffer, 1993; Goss, 1999; Herzog & Morgan,
1993; Hodgkinson & Weitzman, 1996; Rotolo, 2000),
our focus is on life-course patterns of voluntary association. Put simply, we assert that no such fall-off occurs and that rates remain relatively stable until age
85 or so. In a recent analysis of a composite National
Opinion Research Center (NORC) General Social Surveys (GSS) data set, we mustered evidence to counter
the presumed decline in membership among older
persons. Our data suggested that the decline is an artifact of compositional differences between cohorts.
That is, any late-life declines may be attributable to
factors known to differentiate among cohorts in the
first place. So, for example, the population of older
persons has lower levels of education, poorer health,
fewer family ties, and a higher likelihood of being
widowed than their younger counterparts. It is exactly these factors that are predictive of lower levels
of membership in voluntary associations. The conjunction of cohort factors and rates of voluntary affiliation has led some to assume that older persons
are less involved, and perhaps even in the process of
disengaging from active life styles. Our findings suggest that when crucial cohort-related compositional
factors are taken into account, rates of affiliation increase through the middle years and appear to re-
Key Words: Formal activities, Group memberships,
Voluntarism
The frequently asserted belief that older persons do
not engage in productive activities is false. In fact,
through various forms of voluntary contributions, elders provide a wide range of goods and services to
We thank Ann Donahue and Megan Johnson for their contributions to
the preparation of this article.
Address correspondence to Jon Hendricks, PhD, University Honors
College, Oregon State University, Corvallis, OR 97331-2221. E-mail:
[email protected]
1
University Honors College, Oregon State University, Corvallis.
2
Department of Sociology, University of Vermont, Burlington.
250
The Gerontologist
main fairly stable into the mid-80s (Cutler & Hendricks, 2000). Our results are generally consistent
with those of the MacArthur Studies on Successful
Aging, which suggest there is continuity in later life,
with more than 98% of older people who live in private households being engaged in some form of productive activities (Rowe & Kahn, 1998). It may be
that older persons continue their previous patterns of
engagement as an adaptive strategy for as long as
they are able, certainly into their mid-80s (Atchley,
1999; Hays et al., 1998; Mindel & Vaughan, 1978).
For these and other reasons, groups or organizations
charged with delivering services in the face of scarce
resources perceive a vast untapped pool of older volunteers (Freedman, 1997).
Those who posit a curvilinear relationship between
age and rates of voluntary association with a marked
downturn beyond age 65 or so, as well as those who
see no such decline, both point out conflating factors
not necessarily explained by compositional characteristics. One difficulty is the relative importance of two
widespread forms of association: church-related affiliations and union memberships. Putnam (1995) is of
the opinion that church-related associations are so
pervasive in the United States that those affiliations
misrepresent overall voluntary association rates. A
similar claim has been made for union memberships;
they are important, retained even after retirement,
and pervasive enough to distort global patterns (Curtis, Grabb, & Baer, 1992). Our question is simply:
Do these two types of voluntary association exert an
influence so ubiquitous that the entire pattern of
memberships during the adult years is affected?
because churches and religion promote a “do-gooder”
mindset or a “culture of benevolence” is difficult to
say; it may be that one form of participation abets
other forms (Goss, 1999; Musick, Wilson, &
Bynum, 2000; Wilson & Musick, 1997).
In the United States, church affiliations are strong
and appear to provide meaningful webs of social relations. In their analysis of the link between church
membership and volunteering, Wilson and Janoski
(1995) highlighted the complexity of the relationship
and discrepancies between denominations. They noted
that among Catholics, patterns form early and persist. Among liberal Protestant denominations (those
associated with social activism), volunteering does
not become part of the pattern or complex of values
until midlife. Among conservative Protestants, generalizable patterns are least apparent, and voluntarism
is largely limited to church-related activities (Wilson
& Janoski, 1995). Among other subgroups of the
population—African American elders, for example—
some research suggests that church affiliations are of
such salience that they can be used effectively to provide needed services (Hatch, 1991; Taylor, 1986).
Among older persons, some investigators suggest
that religious organizations account for the preponderance of volunteer work and may provide “the
most psychologically beneficial type of volunteer
work” (Van Willigen, 2000).
Another reason for focusing on church-related
groups is that there is conceptual and empirical disagreement about the appropriate boundaries to be
employed in delineating membership in this type of
voluntary association. On the one hand, Smith (1983)
has argued that churches (in the generic sense) are a
type of voluntary organization and, therefore, a
broad definition encompassing all forms of church
membership should be used. On the other hand,
most surveys have taken a more limited approach
and have asked about membership in church-affiliated or church-related groups. Yet, there is evidence
that the ambiguity in the way the question is posed
leads to confusion between memberships in churchrelated groups and church membership per se, with
the result that reported membership levels in churchaffiliated groups are inflated (Smith, 1990). It is also
not altogether clear what modes of religious activity
qualify as church-affiliated groups. Bible study and
youth groups would appear to fit the description, but
do church choirs or short-term preparatory classes
offered prior to a church-sponsored trip to the Holy
Land? In short, the need for conceptual clarification
provides yet another argument for the exclusion of
church-affiliated groups.
Importance of Religious and Union Membership as
Forms of Voluntary Association
Religion-Based Memberships.—Rates of religious
affiliation are higher in the United States than in
most industrialized countries. In their estimate of
church memberships in 15 countries, Curtis and associates (1992) ranked the United States first. Overall estimates are that 55% of all Americans are affiliated with church or religious groups, that 50–60%
of all voluntary participation occurs within religious
organizations, and that among older persons, about
65% identify church or synagogue as a primary focus of volunteer activity (Goss, 1999; Hodgkinson &
Weitzman, 1996; Independent Sector, 1998; Van
Willigen, 2000). In fact, “church-affiliated groups
may be the backbone of civil society in America, involving almost half of the population” (Galston &
Levine, 1997, p. 25; Lipset, 1990). Greeley (1997)
noted that religious-based memberships have an independent effect in explaining overall levels of voluntary association in virtually all 16 countries included
in the European Value Studies. Greeley reported that
religious affiliation (church attendance and membership in religious groups) accounts for 38% of the
variance found across the 16 countries. As he put it,
“[t]he American ‘advantage’ in volunteering is patently the result of the higher levels of American religious practice” (Greeley, 1997, p. 72). Whether it is
Vol. 41, No. 2, 2001
Union Membership.—Union membership has also
been a conspicuous form of voluntary association,
one that persists into retirement. Curtis and colleagues (1992) found that 13% of respondents in the
World Values Survey reported union membership. In
the GSS data set used here, 14.7% of respondents
who described themselves as “retired” reported belonging to a union as well—the fourth most common
251
form of association membership for all retirees (after
church-affiliated, fraternal, and veterans groups) and
third highest among male retirees (after church-affiliated and veterans groups). Continuing membership
may provide social interaction and recreation, a
venue for dealing with pension and benefit issues, and
an opportunity to remain involved in programs and
activities. According to Galston and Levine (1997),
union members are more likely to participate in other
voluntary associations or make charitable contributions than their nonunion counterparts. The hypothesis is that there is a “stepping stone” process—that
membership in preexisting voluntary or secondary associations leads to volunteer participation (Lofland,
1996, p. 187; Smith, 1994; Wilson & Musick, 1998).
ics. Through 1987, rotating items were included in
two of three surveys. Beginning with the 1988 survey,
the rotational system switched to one in which selected items were asked of two thirds of the respondents within each survey. Information about both
voluntary association memberships and self-perceived
health was sought from all respondents in the first six
surveys listed in Table 1. From 1988 through 1994,
however, these items were asked only of partially
overlapping subsets of two thirds of the samples, with
the effect that information on both association memberships and health was obtained from about one
third of the respondents in these years.
Although the final number of cases was considerably greater than is usually available for analyses of
voluntary association membership, we tested for possible biases stemming from the elimination of cases
by comparing the race, gender, and education distributions for the full sample and the analysis sample
and found negligible differences (see Cutler & Hendricks, 2000). Bias stemming from loss of cases,
then, does not appear to be a significant issue.
Methods
Data
For this analysis we created a composite data set
(N ⫽ 18,295) from 12 nationally representative
NORC General Social Surveys conducted between
1974 and 1994 (Davis & Smith, 1994). All surveys
in the GSS series (from 1972 to 1998) with the variables to be used in the analysis were selected for inclusion (see Table 1 for the survey years and initial
Ns). Samplings for the 1974 survey and half of the
1975 survey were based on a block quota design; the
remaining half of the 1975 survey and all other years
used full probability sampling designs. Each sample
was representative of the population of Englishspeaking persons 18 years of age and older living in
noninstitutional arrangements in the United States.
In addition to the exclusion of respondents with
missing data on any of the analysis variables, changes
in content coverage over the history of the GSS resulted in a further reduction of the total N for the 12
surveys from 18,295 to 11,789 (Table 1). Several of
the variables used in the analysis (age, gender, race,
education, employment status, marital status, and
presence of children in the household) are part of the
series of permanent items asked of all respondents in
each survey. Other items have been asked on a rotational basis to allow for an expanded number of top-
Variables
To assess rates of membership in any of 16 possible types of formal voluntary associations, we relied
on the question, “Now we would like to know something about the groups or organizations to which individuals belong. Here is a list of various organizations. Could you tell me whether or not you are a
member of each type?” Respondents were then read
a list that comprised fraternal groups; service clubs;
veterans groups; political clubs; labor unions; sports
groups; youth groups; school service groups; hobby or
garden clubs; school fraternities or sororities; nationality groups; farm organizations; literary, art, discussion, or study groups; professional or academic societies; church-affiliated groups; and any other groups.
Adopting the methodology used by Rotolo (1999)
and others, we obtained total counts of memberships
(corrected for a coding error for the years 1989–
1994 [Helliwell & Putnam, 1996]), with potential
scores ranging from 0 to 16 (with church-affiliated
groups and labor unions included) and 0 to 14 (with
memberships in these two types of associations excluded). If anything, this approach underestimates
total rates of affiliation as it depends on a count of
types rather than actual organizational memberships
(Baumgartner & Walker, 1988).
We recoded age into 5-year intervals, beginning
with 18–24, 25–29, and ranging up to 80–84 and
85⫹. Previous work (see Cutler & Hendricks, 2000,
for a summary) has shown that higher membership
rates are found among men and Whites; among those
in better health and with higher levels of education;
and among persons who are employed, married, or
have children residing in the household. Because distributional differences by age in these variables will
have the effect of dampening membership rates of
the older age groups, we controlled for gender, selfperceived health (excellent, good, fair, poor), education (0–8 years, 9–11 years, high school graduate,
Table 1. General Social Survey Years and Full and Reduced
Sample Sizes
Year
Full sample n
Reduced Sample n
1974
1975
1977
1980
1984
1987
1988
1989
1990
1991
1993
1994
Total
1,484
1,490
1,530
1,468
1,473
1,819
1,481
1,537
1,372
1,517
1,606
1,518
18,295
1,466
1,479
1,498
1,432
1,443
1,790
485
521
439
477
516
243
11,789
252
The Gerontologist
some college, college graduate or more), work status
(full-time, part-time, unemployed, retired, other), marital status (married, widowed, divorced/separated, never
married), presence of children in the household (yes,
no), and race (White, non-White).
support for using the composite of the 12 surveys.
Second, we also briefly note the bivariate and multivariate age differences in association memberships
when all 16 types are considered. Third, we consider
whether these latter findings are altered when memberships in church-affiliated groups and in labor
unions are excluded. Finally, we summarize the effects of the control variables.
In an effort to determine if period effects were influencing the composite pattern, our first step was to
determine whether it was appropriate to aggregate
the 12 surveys over the 20-year period. Based on an
ordinary least squares (OLS) regression analysis examining the effects on number of memberships of
survey year, age, age2, and an interaction term for
age2 and survey year, we found that neither survey
year nor the age2/survey year interaction was significant (see Cutler & Hendricks, 2000). From these results, we concluded that the nature of the relationship
between age and number of memberships did not
vary during the period in question, thus justifying the
aggregation of the 12 surveys in order to have as
large a number of cases as possible at the oldest ages.
In looking to overall number of associations in
which respondents claimed affiliation for the full set
of 16 types, we found that memberships did appear
to follow a curvilinear pattern by age when viewed in
bivariate relationship (Table 2). The gross mean
number of memberships was low among the youngest respondents, rose steadily to a peak at ages 40–
44, and then declined through ages 85⫹, where the
lowest average number of memberships was found.
On the face of it, this might be taken to confirm the
belief that there is a decline in affiliations during the
retirement years. However, we then controlled for
compositional effects and found a somewhat different pattern. Examination of the net mean scores for
Analysis
The primary objective of this study was to determine whether the adjusted pattern of increasing voluntary association membership rates through age
50–54, and high and generally stable rates thereafter,
is due to affiliations with church-related groups and
labor unions. To address this question, we used Multiple Classification Analysis (MCA; SPSS for Windows, Release 7.5; Andrews, Morgan, & Sonquist,
1967), a multivariate procedure that enabled us to
examine the bivariate effect of age on number of
memberships, as well as its impact on our dependent
variables after adjusting for the effects of the controls. MCA is particularly well suited to the objectives of this analysis in that no assumption of linearity was required, and we were able to obtain both
bivariate (gross) and adjusted (net) mean membership scores for each of the 14 age categories in order
to examine age patterns of membership. Eta and beta
coefficients can be used to evaluate strength of the
relationships at the bivariate and multivariate levels,
and F tests are available to determine whether any
given predictor (i.e., age) explains a significant proportion of the variance in the dependent variable before and after other predictors are held constant.
Results
Our discussion of the results of the analysis is divided into four sections. First, we briefly summarize
Table 2. Gross and Neta Effects of Age on Number of Memberships in 16 Types b and 14 Typesc of Voluntary Associations:
1974–1994 Combined
16 Types
Age
18–24
25–29
30–34
35–39
40–44
45–49
50–54
55–59
60–64
65–69
70–74
75–79
80–84
85⫹
Total
p Value
Gross Mean
1.44
1.53
1.83
1.95
2.09
1.97
1.93
1.96
1.59
1.61
1.48
1.45
1.34
1.21
1.74
Eta
14 Types
Net Mean
Beta
1.44
1.26
1.54
1.64
1.85
1.90
2.00
2.17
1.93
2.01
1.92
2.03
1.95
1.88
.125
⬍.001
.164
⬍.001
Gross Mean
1.12
1.16
1.39
1.46
1.57
1.40
1.34
1.34
1.02
1.07
.98
.91
.85
.77
1.25
a
Eta
Net Mean
Beta
1.10
.91
1.14
1.20
1.37
1.36
1.42
1.53
1.31
1.41
1.34
1.40
1.36
1.34
.120
⬍.001
X16(net)–X14(net)
.34
.35
.40
.34
.48
.64
.58
.64
.62
.60
.58
.63
.59
.54
.106
⬍.001
Controlling for gender, self-perceived health, education, work status, marital status, presence of children in the household, and race.
Including church-affiliated groups and labor unions.
c
Excluding church-affiliated groups and labor unions.
b
Vol. 41, No. 2, 2001
253
the full set of 16 types of associations showed that
memberships rise from ages 25–29 and peak at ages
55–59, after which there is a long plateau, lasting until ages 85⫹. Because the number of respondents at
even more advanced ages became so small at that
point, we were unable to make any comments about
what then occurs. For a more complete discussion of
the results for all 16 types, see Cutler and Hendricks
(2000).
Because our primary interest in this analysis is in
the effects of church-based and union memberships,
we first excluded one, then the other, and then both.
The principal conclusion was the same regardless of
whether the analysis was based on 15 types of memberships (with church-affiliated groups or labor unions
excluded) or 14 types (with church-affiliated groups
and labor unions excluded), so only the latter results
are presented here.
First, examination of the unadjusted (gross) mean
number of memberships in the 14 types of associations showed the now familiar curvilinear pattern of
age differences (Table 2). The highest level of affiliation occurred among persons 40–44 years of age.
Membership levels trailed off steadily on both sides
of this age group, with the lowest average number of
memberships found among the 85⫹ age group.
Again, however, it must be recalled that characteristics of older persons are differentially distributed
along the control variables. As detailed in Cutler and
Hendricks (2000), older persons were significantly
more likely to report their health as poor, have lower
levels of education, and be out of the labor force.
The older age groups also have higher percentages of
women, widows, persons living in households without children, and at ages 85⫹ non-Whites. These
very same characteristics dampen levels of association affiliation. The net means for the 14 types of associations yielded the pattern of age differences after
the effects of the varying compositional characteristics were controlled. This pattern proved to be quite
similar to what we previously found for the 16 types.
Although there was an initial decline between the
ages of 18–24 and 25–29, number of affiliations rose
thereafter through ages 55–59, dropped slightly at
ages 60–64, and then remained relatively stable
through ages 85⫹. Even with the deletion of membership in church-related groups and labor unions,
the net mean scores clearly indicate that a major
source of the lower levels of association membership
observed in the bivariate data among the older age
groups stemmed from their compositional characteristics.
To provide a rough approximation of the significance of the changes from the unadjusted to the adjusted mean scores for each age group, 95% confidence intervals (CI) were obtained for each of the
gross mean scores. The 95% CIs were used under the
assumption that whether an age group’s net mean
falls within or outside the interval for its gross mean
can be taken as an approximate indication of the significance of the change when compositional differences are controlled. The results (not presented here)
provided further support for our conclusions. Specifically, for the analysis of both the 16 types and the
14 types of associations, the adjusted means fell
within the 95% CIs for three of the age groups: 18–
24, 45–49, and 50–54. In these instances, the adjustment process made little difference. For each of the
remaining age groups, their net mean scores lay outside of the gross mean’s confidence interval, suggesting that controlling for compositional characteristics
appears to have had an appreciable effect.
As a way of summing up this part of the findings,
the net means reported in Table 2 for both the 16type analysis and the 14-type analysis are presented
in Figure 1. Apart from average membership levels
for the analysis of 14 types of associations being consistently but unsurprisingly lower for each age group
than was the case with the analysis based on 16
types, the age-based membership curves are quite
alike. Both show a decline in average number of
memberships between the ages of 18–24 and 25–29,
increasing levels through ages 55–59, a decrease
from 55–59 to 60–64, and a relatively constant number of memberships through the 85⫹ age group.
That the curve for the 14 types of associations appears to be flatter is supported by the respective betas (.164 for the 16 types and .106 for the 14 types),
but both patterns are statistically significant by the F
test for net effects (see Table 2). Figure 1 also suggests that the exclusion of membership in churchrelated groups and in labor unions makes less difference at the younger ages, leads to a widening gap up
through ages 55–59, and then to a constant difference thereafter. This impression is supported by the
data in the far right column of Table 2, which presents the simple differences between the net mean
scores for the analysis based on 16 types and that
based on 14 types. Generally, the magnitude of the
differences increased up through age late 50s and
then remained fairly steady, at least through the 80–
84 age group. In other words, excluding church-related
organizations and labor unions lowered the member-
Figure 1. Adjusted age differences in number of voluntary association memberships.
254
The Gerontologist
ship count of all age groups but had the least impact
on membership levels of younger persons, a growing
impact on overall affiliation rates up to the late middle-age period of the life course, and a uniform impact thereafter.
Finally, of the control variables, education exerted
the strongest net effect (Beta ⫽ .368), that is, the higher
the level of educational attainment, the greater the
number of memberships. Respondents with children
in the household belonged to more associations than
those without children (Beta ⫽ .060), men had a
greater number of memberships than women (Beta ⫽
.059), membership levels were higher the better one’s
self-reported health (Beta ⫽ .049), and Whites had a
higher number of affiliations than non-Whites (Beta ⫽
.043). For work status, the unemployed had the lowest number of memberships, part-time workers the
highest, and full-time workers, retirees, and other
statuses were intermediate (Beta ⫽ .037). Widowed
respondents had fewer memberships than persons
who were married, divorced/separated, or never married (Beta ⫽ .023). All relationships between the control variables and number of association memberships were significant beyond at least the .01 level.
berships or union memberships accounted for the
stable pattern of membership (e.g., Lofland, 1996;
Wilson & Musick, 1998).
To reach that conclusion, we aggregated data on
various types of association membership among 16
types of associations over the 20-year period from
1974 to 1994 as collected in 12 national GSS surveys. Had we looked merely at the bivariate relationship between age and number of memberships, there
would indeed have appeared to be a marked curvilinear relationship with a long, slow tapering off in the
later years. However, from a multivariate perspective, the relationship was positive through ages 55–
59 and, after a slight drop, continued pretty much as
a plateau until the respondents were in the 85⫹ category. Further, membership in neither church-related
associations nor unions appeared to exert a significant influence on the overall pattern.
One of the implications of our findings is that an
extensive pool of volunteers is found among older
segments of the population, and that they continue
to be engaged to the extent that their stamina and
vigor permit throughout the life course. In several respects, they are more engaged than many of their
younger counterparts. Until their mid-80s, many older
persons appear to remain involved. As has frequently
been pointed out (Atchley, 1999; Freedman, 1997),
older adults often search for options that may help
them perceive themselves as both engaged in and
contributing to some greater good. Our data imply
that this is the case, as even when church affiliations
and union memberships are controlled, high rates of
voluntary association memberships persist well into
the later years.
These results, of course, are based on an analysis
of the number of types of voluntary associations to
which people belong. As noted earlier, this approach
likely underestimates the actual number of affiliations, given that persons may belong to more than
one association within each type (Baumgartner &
Walker, 1988). More definitive estimates would require an enumeration of memberships within each of
these general types. Memberships themselves also
represent just one form of participation in voluntary
associations. Further research on age differences should
examine various modes and levels of involvement
and activity. Finally, participation in the types of formal, voluntary associations considered in this study
does not exhaust the range of ways in which voluntaristic contributions can be made. Only by casting the
widest possible net of types of voluntaristic activities
will we be able to fully appreciate the scope and
magnitude of productive activities in the later years.
Discussion
About half of all Americans are currently engaged
in some form of volunteer activity, including voluntary associations (Hodgkinson & Weitzman, 1996;
Sundeen, 1992; Van Willigen, 2000). Yet, in recent
years, some (Putnam, 1995, 2000; Rotolo, 1999)
have asserted that rates are waning overall or for
particular forms of participation. As far as older volunteers are concerned, some investigators point to
overarching patterns of decline, whereas others report that with increases in life expectancy have come
intensified searches for meaningful, perhaps even altruistic endeavors. Paradoxically, at the same time,
volunteering is “marketed” as a valuable activity, a
way to make a difference in deserving causes. In fact,
a monitoring group, Independent Sector (1998), asserted that in 1995, $77.2 billion worth of voluntary
services were contributed by older volunteers. Moreover, the number of nonprofit associations listed in
the Encyclopedia of Associations was more than
twice as great as it had been less than 30 years previously (Allard & Gruber, 1998–99). As if any further
evidence were needed, a spate of new popular books
provides ample testimony that volunteering is alive
and well. As far as older persons are concerned,
rather than withdrawing from active participation,
older persons seek and provide “pay back” through
voluntary participation of one form or another (Cohen, 2000; Freedman, 2000; Roszak, 1998).
Our data lead us to assert that voluntary association membership rises to a peak in ages 55–59 and
remains reasonably stable until age 85, religious and/
or union affiliations notwithstanding. Although our
findings are contrary to what some have maintained
(e.g., Fischer, Rapkin, & Rappaport, 1991; Greeley
1997), we did not find that church-affiliated memVol. 41, No. 2, 2001
REFERENCES
Allard, D. M., & Gruber, K. (1998–99). Encyclopedia of associations. Detroit, MI: Gale Research.
Andrews, F., Morgan, J., & Sonquist, J. (1967). Multiple Classification
Analysis: A report on a computer program for multiple regression using categorical predictors. Ann Arbor, MI: Institute for Social
Research.
Atchley, R. (1999). Continuity and adaptation in aging: Creating positive
experiences. Baltimore, MD: The Johns Hopkins University Press.
Baumgartner, F. R., & Walker, J. L. (1988). Survey research and member-
255
Moen, P., Dempster-McClain, D., & Williams, R. M., Jr. (1989). Social
integration and longevity: An event history analysis of women’s roles
and resilience. American Sociological Review, 54, 635–647.
Moen, P., Dempster-McClain, D., & Williams, R. M., Jr. (1992). Successful aging: A life course perspective on women’s roles and health.
American Journal of Sociology, 97, 1612–1638.
Monk, A. (1995). Volunteerism. In G. L. Maddox (Ed.), The encyclopedia
of aging (pp. 958–960). New York: Springer.
Musick, M., Herzog, A. R., & House, J. (1999). Volunteering and mortality among older adults: Findings from a national sample. Journal of
Gerontology: Social Sciences, 54B, S173–S180.
Musick, M., Wilson, J., & Bynum, W., Jr. (2000). Race and formal volunteering: The differential effects of class and religion. Social Forces, 78,
1539–1571.
Pillemer, K., Moen, P., Wethington, E., & Glasgow, N. (2000). Social integration in the second half of life. Baltimore: Johns Hopkins
University Press.
Putnam, R. (1995). Bowling alone: America’s declining social capital.
Journal of Democracy, 6, 65–78.
Putnam, R. (2000). Bowling alone: The collapse and revival of American
community. New York: Simon & Schuster.
Roszak, T. (1998) America the wise: The longevity revolution and the true
wealth of nations. Boston: Houghton Mifflin.
Rotolo, T. (1999). Trends in voluntary association participation. Nonprofit and Voluntary Sector Quarterly, 28, 199–212.
Rotolo, T. (2000). A time to join, a time to quit: The influence of life cycle
transitions on voluntary association membership. Social Forces, 78,
1133–1161.
Rowe, J., & Kahn, R. (1998). Successful aging. New York: Pantheon.
Smith, D. (1983). Churches are generally ignored in contemporary voluntary action research: Causes and consequences. Review of Religious
Research, 24, 295–303.
Smith, D. (1994). Determinants of voluntary association participation and
volunteering: A literature review. Nonprofit and Voluntary Sector
Quarterly, 23, 243–263.
Smith, T. W. (1990). Trends in voluntary group membership: Comments
on Baumgartner and Walker. American Journal of Political Science,
34, 646–661.
Sundeen, R. (1992). Differences in the personal goals and attitudes among
volunteers. Nonprofit and Voluntary Sector Quarterly, 21, 271–291.
Taylor, R. (1986). Religious participation among elderly black Americans.
The Gerontologist, 26, 630–642.
Van Willigen, M. (2000). Differential benefits of volunteering across the
life course. Journal of Gerontology: Social Sciences, 55B, S308–S318.
Wilson, J., & Janoski, T. (1995). The contribution of religion to volunteer
work. Sociology of Religion, 56, 137–152.
Wilson, J., & Musick, M. (1997). Who cares: Toward an integrated theory of volunteer work. American Sociological Review, 62, 694–713.
Wilson, J., & Musick, M. (1998). The contribution of social resources to
volunteering. Social Science Quarterly, 79, 799–814.
ship in voluntary associations. American Journal of Political Science,
32, 908–928.
Cohen, G. (2000). The creative age: Awakening human potential in the
second half of life. New York: Avon.
Curtis, J., Grabb, E., & Baer, D. (1992). Voluntary association membership in fifteen countries: A comparative analysis. American Sociological
Review, 57, 139–152.
Cutler, S. J., & Hendricks, J. (2000). Age differences in voluntary association memberships: Fact or artifact. Journal of Gerontology: Social
Sciences, 55B, S98–S107.
Davis, J. A., & Smith, T. W. (1994). General Social Surveys, 1972–1994:
Cumulative codebook. Storrs, CT: Roper Center for Public Opinion
Research, University of Connecticut.
Fischer, L., Rapkin, D., & Rappaport, J. (1991). Gender and work history
in the placement and perceptions of elder community volunteers. Psychology of Women Quarterly, 15, 261–279.
Fischer, L., & Schaffer, K. (1993). Older volunteers. Newbury Park, CA:
Sage.
Freedman, M. (1997). Towards civic renewal: How senior citizens could
save civil society. Journal of Gerontological Social Work, 28, 243–263.
Freedman, M. (2000). Prime time: How baby-boomers will revolutionize
retirement and transform America. New York: Public Affairs.
Galston, W., & Levine, P. (1997). America’s civic condition: A glance at
the evidence. Brookings Review, 15, 23–27.
Gecas, V., & Burke, P. (1995). Self and identity. In K. Cook, G. A. Fine,
& J. S. House (Eds.), Sociological perspectives on social psychology
(pp. 41–67). Boston: Allyn & Bacon.
Goss, K. (1999). Volunteering and the long civic generation. Nonprofit
and Voluntary Sector Quarterly, 28, 378–415.
Greeley, A. (1997). The other civic America: Religion and social capital.
The American Prospect, 32, 68–73.
Hatch, L. R. (1991). Informal support patterns of older African-American
and white women. Research on Aging, 13, 144–170.
Hays, J. C., Landerman, L. R., Blazer, D. G., Koenig, H. G., Carroll, J.
W., & Musick, M. A. (1998). Aging, health, and the “electronic
church.” Journal of Aging and Health, 10, 458–482.
Helliwell, J. R., & Putnam, R. D. (1996). The strange disappearance of
civic America. The American Prospect, 24 (Winter [http://epn.org/
prospect/24/24putn.html]).
Herzog, A. R., & Morgan, J. (1993). Formal volunteer work among older
Americans. In S. A. Bass, F. G. Carol, & Y.-P. Chen (Eds.), Achieving
a productive aging society (pp. 119–142). Westport, CT: Auburn
House.
Hodgkinson, V. A., & Weitzman, M. (1996). Giving and volunteering in
the United States: Findings from a national survey, 1996 edition.
Washington, DC: Independent Sector.
Independent Sector. (1998). America’s senior volunteers. Washington,
DC: Author.
Lipset, S. M. (1990). Continental divide: The values and institutions of
Canada and the United States. New York: Routledge.
Lofland, J. (1996). Social movement organizations. New York: Aldine de
Gruyter.
Mindel, C. H., & Vaughan, C. E. (1978). A multidimensional approach to
religiosity and disengagement. Journal of Gerontology, 33, 103–108.
Moen, P. (1996). A life course perspective on retirement, gender, and wellbeing. Journal of Occupational Health Psychology, 1, 131–144.
Received May 15, 2000
Accepted October 27, 2000
Decision Editor: Laurence G. Branch, PhD
256
The Gerontologist