Content uploaded by Elizabeth R. Cole
Author content
All content in this area was uploaded by Elizabeth R. Cole on Apr 17, 2015
Content may be subject to copyright.
PERSONALITY PROCESSES AND INDIVIDUAL
DIFFERENCES
Meanings of Political Participation Among Black and White Women:
Political Identity and Social Responsibility
Elizabeth R. Cole
Northeastern University
Abigail J.Stewart
University of Michigan
This study examined the correlates of midlife political participation among 64 Black and 107 White
women of the college classes of 1967-1973. Compared with White women, Black women scored
higher on political participation, generativity, power discontent, and politicization. Factor analysis
of personality and political attitude variables yielded three factors labeled Political Identity, Power
Discontent, and Social Responsibility Adult political participation was regressed on level of student
activism and index scores of political identity, power discontent, and social responsibility. For both
racial groups, social responsibility was associated with midlife political participation. For White
women, political identity was also related; for Black women, student activism bore a significant
relationship. The findings suggest that Black and White women's historical and political contexts
imbued their political activities with different meanings.
During the late 1960s and the early 1970s, three social move-
ments simultaneously developed and gained momentum in the
United States: resistance against the war in Vietnam; the wom-
en's liberation movement; and the Black power movement,
Elizabeth R. Cole, Departments of Psychology and African American
Studies, Northeastern University; Abigail J. Stewart, Department of
Psychology, University of Michigan.
A version of this article
was
presented as a poster at the 102nd Annual
Convention of the American Psychological Association (Division 35),
Los Angeles, August 1994. This article is based on Elizabeth R. Cole's
doctoral dissertation, completed at the University of Michigan,
The longitudinal-sample data collection and analyses reported in this
article were supported by a research grant from the Henry A. Murray
Research Center at Radcliffe College (which has archived previous
waves of the study) and Grant RO1-MH47408-01AI from the National
Institute of Mental Health. Collection of data for the African American
sample was supported by grants from the Center for Afro-Ameriean
and African Studies and Women's Studies Program at the University
of Michigan, the Ford Foundation, the Grants-in-Aid Program of the
Society for the Psychological Study of Social Issues, and the University
of Michigan Julia Lockwood Prize.
We thank the staff of the Women's Life Paths Study as well as Lauren
Duncan for help in collecting, coding, and analyzing the data for this
study. We are also grateful to Sandra Tangri for inviting our collabora-
tion in the follow-up of the longitudinal study of the class of 1967, Carol
Hollenshead and Jean Manis of the Center for the Education of Women
at the University of Michigan for their help in conducting the follow-up,
and Jill Y. Allen for her collaboration in collecting the data for the Afri-
can American sample. We thank Faye Crosby, Lauren Duncan, and Bill
Peterson for comments on an earlier draft of this article.
Correspondence concerning this article should be addressed to Eliza-
beth R. Cole, Department of Psychology 125-NI, Northeastern Univer-
sity, 360 Huntington Avenue, Boston, Massachusetts02115. Electronic
mail may be sent via the Internet to cole@neu.edu.
which grew out of the civil rights movement of the 1950s and
1960s. The goals and tactics of these latter two movements shared
much in common: Each aimed to redefine the roles and broaden
the privileges of historically disadvantaged groups and worked to
accomplish these ends through the redefinition of the constituent
groups' identities and political consciousness (Carroll,
1989;
Co-
hen, 1985; Morris, Hatchett, & Brown, 1989). The emphasis on
the transformation of the identity and consciousness of individ-
ual group members is well captured by each of the movements'
best known slogans: The Black power exhortation that "Black is
beautiful" rejects a conferred stigmatized racial identity, whereas
the feminist assertion that "the personal is political" recasts
women's private experiences within the traditional family in
terms of public relations of power.
Given the focus of both the women's movement and the civil
rights movement on issues of group identity, it might be ex-
pected that these movements would have had different mean-
ings and hence different long-term impact for individuals of
different social groups. Much of the research on these two
movements has been framed theoretically by Mannheim's
(1928/1952) essay on generations (e.g., Braungart & Braung-
art, 1990; Jennings, 1987). which argued that periods of youth
movement result from the combined effects of the circum-
stances of the historical moment in which a cohort comes of
age,
the shared cultural experiences of the cohort, and the particular
demographic characteristics of the cohort. Although the mem-
bers of a generation may be influenced by the same historical
events, Mannheim (1928/1952) suggested that within any
birth cohort there are subgroups, or "generation units [who]
work up the material of their common experiences in different
specific ways" (p. 304). Thus, for example, Stewart and Healy
(1989) pointed to the difference in the meaning of the Vietnam
Journal of Personality and Social Psychology, 1996, Vol 71, No I, 130-140
Copyright 1996 by the American Psychological Association, Inc. 0O22-J514/96/$3 00
130
WOMEN'S POLITICAL PARTICIPATION
131
War for those who fought it and those who opposed it (despite
the fact that it was laden with meaning for both groups). Al-
though several researchers have explored the long-term impact
of the social movements of this era on those who participated as
compared with those who did not (e.g., Abramowitz & Nassi,
1981;Fendrich, 1974, 1976, 1977; Franz & McClelland, 1994;
Hoge & Ankney, 1982; Jennings, 1987; Nassi, 1981; Nassi &
Abramowitz, 1979), little of this research has addressed the
unique experiences of women or explored the differential effect
of participation in these movements on the later lives of Black
and White
protesters.
The present study addressed these gaps in
the literature, exploring the ways in which social movements,
coinciding with personal development, may be related to
differences in political belief and involvement for members of
divergent subcultures with distinct political statuses.
Student Activism Among Women and
African Americans
Sherkat and Blocker( 1994) found that women were relatively
underrepresented among movement activists of this era; how-
ever, this difference was largely accounted for by differences in
gender socialization (including political efficacy and religiosity)
and lower rates of college attendance. In contrast, studies based
on interviews with women who were active (Cable, 1992;
Thorne, 1975) suggest that many women were dissuaded from
participation by being relegated to the most trivial and menial
tasks by the mostly male leadership. Longitudinal research that
followed the activists into adulthood suggests that despite the
fact that women faced gender discrimination even within liber-
atory social movements, participation in the movements of this
era had a long-term impact on both the later political participa-
tion and personal development of the women who were activists.
Franz and McClelland (1994) found that at 41 years of
age,
women who were activists as students seemed better adapted
than did their male activist peers, both in terms of occupational
status and levels of strain reported; they speculated that the
women's movement empowered women and increased their op-
portunities for satisfying work
lives.
McAdam's( 1992) research
following up applicants to the Freedom Summer program of
1964 found that women were less likely to participate in the
program, in part because of gender discrimination in the appli-
cation process. However, among those who did participate, the
experience was deeply meaningful. Indeed, despite the fact that
participation in the program was predictive of later activism for
men but not for women, women described the experience as
more personally significant than did men. McAdam attributed
the source of this discrepancy to the different historical meaning
of Freedom Summer for men and women. Because the develop-
ment of the women's movement subsequent to the Freedom
Summer program heightened the women's identities as activ-
ists,
they accorded their Freedom Summer experience greater
personal significance in their retrospective biographical ac-
counts than did their male peers. Braungart and Braungart's
(1991) findings based on interviews with women who had been
active in both left- and right-wing organizations during the
1960s similarly suggest that the movements had a long-term im-
pact on women's commitment to women's political concerns;
they observed that, in addition to pursuing the ideological com-
mitments of their youth, many of the former activists from both
groups had also become active in women's issues.
Little research has explored racial differences in the impact
of the social movements of this
era.
Fendrich (1976,1977) com-
pared a sample of White male students from the early civil
rights era with a contemporaneous sample of male African
American alumni from the same school 10 years after gradua-
tion. He found that variance in the degree of student activism
predicted later leftist attitudes and political behavior among
both the Black and White alumni. However, Fendrich (1976)
also found that student activism was a stronger predictor of
adult participation in leftist politics for White men than for
Black men.
Taken together, these few studies suggest that civil rights ac-
tivism in the 1960s bears a weaker relationship to adult political
participation for women and for African Americans than it does
for White men in this cohort; however, this is not because
women were relatively unaffected by their experience. On the
contrary, we argue that because women and African Americans
bore a different, and more direct, relationship to the civil rights
and women's movements, variables other than student activism
might be equally important in predicting their continued par-
ticipation later in life.
Predicting Midlife Political Participation
We expected that White women who graduated from college
in the late 1960s would view the women's movement as having
had a strong effect on their lives and that the civil rights move-
ment would have been significant as well (see evidence from
another similar sample presented in Stewart &Healy, 1989). In
contrast, because the leadership of the women's movement was
dominated by White women and much of
the
movement's en-
ergy focused on issues of much greater interest to middle-class
White women than to any other group (e.g., the critique of the
middle-class domestic role of housewife), we expected that
Black women graduating from college during this same era
would view the women's movement as having been a less im-
portant influence. Instead, because of the importance of the
civil rights movement to African Americans in general (see
Schuman
&
Scott, 1989), we expected that Black women would
perceive this series of events as having been an important influ-
ence on their later lives. We anticipated that for both groups,
though, involvement in political activity in late adolescence
would be related to their identities and later lives.
Two separate lines of theory and research suggest that student
movements of the 1960s that focused on gaining rights for
women and African Americans would strongly affect, or politi-
cize,
the identities of members of those groups who were on
college campuses at that time. First, theory and research in so-
cial psychology indicate that a central mechanism through
which social movements mobilize is the creation of a collective
identity that not only enlarges individual identity but also con-
nects the participant to the social group, cementing his or her
commitment (Gamson, 1992). When individuals share a com-
mon identity, and hence a sense of common fate with a group,
they act to protect group interests (Gurin & Townsend, 1986).
132
COLE AND STEWART
Theory and research that specifically address the women's
movement and the civil rights-Black power movement similarly
emphasize the centrality of collective identity. Cross (1991) ob-
served that the Black power phase of the Black social movement
attempted to transform the stigmatized Black identity into a
new one based on pride in African American culture and expe-
rience; he argued that this pattern of transformation character-
izes the development of individual Black identity as well (this
premise has been extensively tested empirically; see Helms,
1993;
Ponterotto
&
Wise,
1987). More generally, Cohen (1985)
argued that recent political struggles, notably feminism, differ
from earlier forms of leftist movement in that they invoke an
activist identity that shapes and is shaped by participants' own
experiences of group identity, rather than being strictly based
on class membership (see also Mueller, 1987).
Second, research based on a model of life span development
suggests that the period of late adolescence and young adult-
hood is a critical time for the development of
identity,
including
the commitment to values, ideologies, and groups. Stewart and
Healy (1986, 1989) suggested that events of historical signifi-
cance experienced during this period are likely to affect cohort
members' developing identities. Speaking specifically of this co-
hort, Erikson (1968b) argued that the student movements rep-
resented an attempt by that generation to develop its own ideol-
ogy and rituals of passage, making the connection between iden-
tity development, social change, and political participation
explicit. This premise found empirical support in Duncan and
Agronick's (1995) longitudinal analyses based on women of this
cohort. They found that women who actively explored different
identities available at age 21 (i.e., who were classified in the
stages of identity achievement and identity moratorium) rated
the women's movement as being more personally meaningful
at midlife. Similarly, Jennings's (1987) study of the "protest
generation" found evidence both for the persistence of the hy-
pothesized "cohort" definition through the life course and for a
particularly strong shaping effect of cohort or generation for
those who participated in student activism. This influence is
thought to be quite general and diffuse, including and integrat-
ing particular attitudes and stances toward social life (e.g., so-
cial dominance orientation; see Prato, Sidanius, Stall worth, &
Malle, 1994; Sidanius, Prato, &Bobo, 1994).
We use the term political identity to describe a pattern of
be-
liefs related to the social and structural relationships that con-
nect the individual to social
groups:
specifically, that human ex-
istence is interconnected, that disadvantaged groups are limited
by systemic obstacles rather than individual shortcomings, that
the political realm is personally relevant and meaningful, and
that collective actions are the best responses to social problems.
Because college-educated Black women of
this
cohort were ex-
pected to be powerfully influenced by both the civil rights and
women's movements, whereas White women were expected to
be more strongly influenced by the women's movement, we hy-
pothesized that Black women would show a higher level of po-
litical identity but that political identity would be associated
with midlife political activism for both groups.
Power discontent, or dissatisfaction with the power one's
group holds relative to other groups, provides an affective com-
ponent to political participation. This construct taps the indi-
vidual's view of her own personal stake in political action and
was hypothesized to be related to level of midlife political in-
volvement. We anticipated that Black women would have
higher levels of power discontent than White women because
they suffer both race- and gender-based disadvantages (con-
sistent with the racial differences found by Gurin, Miller, &
Gurin, 1980) but that power discontent would play a similar
role for both groups of women in motivating political participa-
tion. Both Gurin and Townsend's (1986) group consciousness
model and Crosby's( 1976, 1982) and others' (e.g., Gurr, 1968,
1970;
Runciman, 1966) relative deprivation models assume
that a sense of grievance is an important precondition for polit-
ical mobilization.
The construct of power discontent, like much of the tradi-
tional research on social movement participation, is based on a
conceptualization of participants as being motivated primarily
by their own personal material interests (hence the classic "free
rider" problem; see Olson, 1965). However, more recently,
scholars have argued that individuals profit from movement
participation not only through the tangible achievements won
but also in terms of the development of a shared collective iden-
tity, and this felt connection to others may motivate individuals
to act on behalf of group interests (Gamson, 1992). This obser-
vation suggests that a final aspect of political participation con-
cerns the desire to act individually for the benefit of the larger
group and the belief in one's own ability to do so effectively; we
label this construct social
responsibility.
The notion of social responsibility was addressed empirically
in a few articles during the 1950s and 1960s; however, recent
research on this topic
has
been scant except in discussions of the
related notion of generativity (for exceptions, see Chebat, 1986;
Witt, 1990). Early research by Gogh, McClosky, and Meehl
(1952) and by Berkowitz and Lutterman (1968) painted a dis-
tinctive portrait of the socially responsible personality. Such
people are concerned with social and moral issues, are commit-
ted to working for the good of groups rather than just for per-
sonal gain, and have a sense of trust in society in general. In this
way, they may be considered generative, which is in part defined
by a "belief in the species, concern for
the
next generation, [and
the] cultural demand" that individuals contribute their re-
sources to the long-term societal good (McAdams & de St. Au-
bin, 1992, p. 1004). Socially responsible people demonstrate a
form of personal efficacy in the public realm that Gogh et al.
(1952) characterized as "greater poise, assurance and personal
security" (p. 77). Perhaps most importantly, they may be de-
scribed as having a strong sense of community; they are active
participants in their communities, representing
the
antithesis of
the alienated citizen. In addition, these researchers speculated
that socially responsible individuals are likely to be highly con-
ventional, by virtue of their respect for and attention to com-
munally held social values and norms. However, they recog-
nized that other forms of social responsibility might exist, in
particular one characterized by the willingness to participate in
principled dissent, seen in the examples of Abraham Lincoln
and Mahatma Gandhi (Berkowitz & Lutterman, 1968).
Social responsibility is related to the Eriksonian (Erikson,
WOMEN'S POLITICAL PARTICIPATION 133
1963) concept of generativity, particularly as it may be ex-
pressed in political activity. Generativity is the desire to make a
lasting contribution to ensure the well-being of future genera-
tions,
which at midlife would provide a developmental press to
act
politically,
particularly for
those
who had made political and
ideological commitments in youth. Research in the
field
of
po-
litical science suggests that political activity does increase
throughout the adult life span, peaking at midlife (Conway,
1985;
Milbrath & Goel, 1982). This pattern is largely attrib-
uted to the decreased demands of family life that may be expe-
rienced during
this
period. Similarly, Flacks (1988) argued that
people participate politically to the extent that such activities
may be negotiated with the responsibilities of daily life. How-
ever, these mundane concerns are not incongruent with the ar-
gument that generative concern may press individuals to act po-
litically: Such relaxation of the everyday demands of family life
may allow individuals to consider what their contributions to
future generations will be (McAdams, 1988) and what they can
create that will "outlive the
self"
(Kotre, 1984). Indeed, re-
search has demonstrated that midlife adults express higher lev-
els
of generative concern than do either older or younger respon-
dents (McAdams, de St. Aubin,
&
Logan, 1992).
In the present study, we operationalized social responsibility
to include the elements of political efficacy, sense of community,
and generativity; we hypothesized that this measure would be
related to higher levels of political activism for both White and
African American women. Because there is considerable theo-
retical and empirical literature linking these constructs to mid-
life personality in general (see, e.g., Erikson, 1968a; Haan,
1989;
Mitchell & Helson, 1990; Neugarten, 1968)—although
there is virtually no empirical evidence from Black samples—
we did not anticipate racial differences.
This study tested the following hypotheses:
Hypothesis I: At midlife, Black women who were young adults
during the social movements of the late 1960s and the early 1970s
will report that the civil rights movement had a larger impact on
their lives than did the women's movement. Conversely, we expect
that White women of this cohort will endorse the women's move-
ment as having had a greater impact on their subsequent lives than
did the civil rights movement.
Hypothesis 2: Level of political participation at midlife will be pre-
dicted by student activism for both Black and White women.
Hypothesis 3: High scores on measures tapping the construct of
political identity will be associated with midlife political participa-
tion in women of both racial groups.
Hypothesis
4:
Black women will endorse higher levels of power dis-
content than will White women; however, power discontent will be
associated with midlife political participation for both groups.
Hypothesis 5: Social responsibility will be related to midlife politi-
cal participation for both Black and White women.
Method
Participants
The study was based on responses to mailed questionnaires adminis-
tered to two samples of women who were alumnae of the University
of Michigan. The measures discussed here were embedded in a larger
questionnaire that included items pertaining to respondents' occupa-
tional and family history, health, and life satisfaction; thus, the ques-
tionnaire was not obviously focused on political attitudes and behavior.
First, the measures discussed here were administered to the White
women in the sample as part of the fourth wave of the Women's Life
Paths Study, a longitudinal study initiated in 1967 by Tango (1969).
Current addresses for the alumnae were obtained with assistance from
the University of Michigan's alumni association. Of the original 200
women in the sample, 107 participated in this wave. Three of the origi-
nal sample members had died, and 48 were no longer locatable; thus,
the response rate from those receiving the questionnaire was 72%. Fol-
low-up respondents in this wave did not differ from nonrespondents
from the initial sample on six demographic variables available from the
1967
wave:
levels of mothers' and fathers
7
education, number of siblings,
parents
1
income, mothers
1
employment, and family religion.
Because the original longitudinal sample did not include any African
Americans, 64 African American participants from the same era were
recruited through the assistance of the alumni association, including a
solicitation in the university's Black alumni newsletter. Because of the
method of recruitment, there was no way to compute a response rate
analogous to that for the longitudinal sample. Women who graduated
between 1967 and 1973 were included in the sample, because substan-
tial numbers of Black women were recruited to the university only be-
ginning in 1970. The data were collected through questionnaires mailed
to the respondents in the fall of 1992.
Preliminary analyses indicated that the Black and White samples
were comparable on most demographic variables. There were no sig-
nificant differences between the groups' mean levels of education, per-
sonal income, and number of children. However, the White women in
the sample were more likely to be married or living with a partner (84%
for the White women vs. 56% for the Black women), x
2
(U N= 166) =
15.90,/J
<
.01.
White women were significantly older (M = 47.91 years,
SD = 0.43 years), than Black women (M = 44.52 years, SD = 2.49
years) F(l, 166) =
185.51,
p < .001, but the difference was not large.
Because of the discrepancies between the two samples on the mean and
the variance of
age,
we performed analyses to determine whether there
were significant age differences within the African American sample on
the variables of interest; t tests comparing older and younger African
American women (based on a median split) revealed no significant
differences in their mean scores on student activism, adult political par-
ticipation, political identity, power discontent, and social responsibility.
Thus,
the African American sample was treated as a single cohort.
Measures
Impact of the movements. To assess racial differences in experiences
of the civil rights and women's movements, we asked respondents
whether they had participated in each movement and to rate the impact
of these movements on their lives on a scale ranging from
1
(very little)
to 5 (very much).
Student activism and adult political participation. Student activism
and adult political participation were measured as continuous variables
on the basis of self-reports of the frequency with which respondents
engaged in specific political behaviors. The scale was developed by Fen-
drich and Lovoy (1988), who found that former civil rights activists
and student government leaders reported higher levels of political in-
volvement on this measure than did their less active classmates 10 years
after graduation.
For the measure of student activism, we asked respondents how often
as college students they had engaged in each of nine different political
behaviors that Fendrich and Lovoy (1988) characterized as protest and
community activism. They responded on a 4-point scale ranging from
0 (never) to 3 (regularly). Responses were summed and divided by the
134
COLE AND STEWART
number of items; therefore, scores ranged from 0 to 3. The behaviors
were as follows: joined in a protest march, attended protest meetings,
participated in any form of political activity that could lead to arrest,
was a candidate for office, worked with others on local problems, formed
a group to work on local problems, contacted local officials on social
issues, contacted a local state or federal official about a particular per-
sonal problem, and went with a group to protest to a public official.
Because this measure was administered retrospectively, we calculated
its correlation with an item administered to the White sample in 1970
measuring exposure to the women's movement. The item asked how
they had "heard about the new women's rights movement." Responses
ranged in proximity to the movement: from a low score of
1
(through
the media) to high scores of 4 (attended activities) and 5 (had helped
organize for the movement). This item was treated as a continuous vari-
able assessing relative proximity of exposure to and involvement in the
women's movement; its correlation with student activism was .37
{TV
=
85.
p
<•
.001). This single item served only as a proxy for a measure of
involvement; however, it lent some support to the validity of the retro-
spective measure of student activism.
Adult political participation was denned as encompassing a wider ar-
ray of behaviors, including participation in the electoral process, and
was measured similarly by using an expanded list of 17 political behav-
iors.
All of the behaviors listed above were included in this measure as
well as items tapping party and campaign work (e.g., "took an active
part in a political campaign'
1
and "participated in a political party be-
tween elections") and political communication (e.g., "kept informed
about politics" and "sent messages to a political leader when they were
doing well or poorly"). Respondents were asked to indicate how often
they had engaged in each type of behavior during the past 2 years. As
before, responses were summed and divided by the number of items.
Internal consistency reliability (a) for the student activism measure was
.85;
for the adult political participation index, it was .87.
Political identity. Political identity was conceptualized as the belief
that the political realm is personally relevant and meaningful, human
existence is interconnected, collective actions are the best responses to
social problems, and disadvantaged groups are limited by systemic ob-
stacles rather than individual shortcomings. Five measures were used to
operationalize this construct.
Political salience. Political salience was assessed using a measure
adapted from Stewart and Healy's (1989) study, in which respondents
were asked to judge how personally meaningful (1 = not at all, 2 =
somewhat, 3 = very) they found each of nine historical events of the
20th century, such as the Vietnam War and the freeing of Nelson Man-
dela. The reliability of this nine-item scale was .76.
Collective orientation. Collective orientation was measured using a
scale adapted from studies by Lykes{ 1984) and by Gurin et al. (1980)
lo measure the extent to which the respondents believed that human
existence is essentially relational, rather than individualistic (Lykes,
1984).
Seven items, stated either as proverbs (e.g.,
ll
'paddle your own
canoe' is a good principle to live by") or as generalized ideas about
human nature and existence (e.g., "if you think about life, you realize
that each person is a separate individual leading his or her own individ-
ual life"), were rated on 7-poinl scales (1 = strongly disagree, 7 =
strongly agree). Two items were taken from the 1972 National Election
Study (Gurin et al., 1980); these assessed agreement with collective
action as a strategy to improve the social and political situation of
Blacks and women. The reliability of the resulting nine-item measure
was .70.
System blame. Two measures of system blame, one pertaining to
race and the other to sex, developed by Gurin et al. (1980), gauged
the extent to which respondents located the cause of social inequalities
within social and political systems, rather than within individuals (on
7-point Likert-type scales). Seven items measured system blame with
respect to inequities experienced by women (a = .72); eight items as-
sessed system blame with respect to inequities faced by Blacks (a =
.76).
Left-right ideology. Finally, left-right ideology, the position on the
left-right political continuum, was rated by respondents on a 6-point
scale ranging from wry conservative to radical. Because the movements
under study were generally associated with "left" political attitudes (see
Jennings, 1987), women's endorsement of these political attitudes is an
important component of their politicization by these events. High
scores corresponded to a liberal or leftist political orientation; low scores
indicated more conservative views.
Power
discontent. An affective sense of relative deprivation, or griev-
ance,
was expected to be related to political involvement. This compo-
nent was operational ized as power discontent and was assessed using
items developed by Gurin et al. (1980). Respondents were asked to
judge the relative amount of influence in American life and politics held
by various groups. Specifically, they were asked to judge the relative
power held by women in general (gender-based) and by women of their
own racial group (race- and gender-based) on a scale ranging from 1
(low discontent) to 5 (high discontent).
Social responsibility. Social responsibility was conceptualized as in-
cluding the generative desire to improve the world for future generations
as well as a sense of political efficacy (or empowerment) and a feeling of
connection to a community. A six-item version of the Loyola Generativ-
ily scale (McAdams & de St. Aubin, 1992; Nakagawa, 1992) was ad-
ministered to measure generativity. Respondents were asked to rate
their agreement with the six items on 7-point scales. Items tapped gen-
erative concern with making a contribution that extended beyond the
self through teaching ("1 have important skills that I try to teach
others"), caring for future generations ("if
I
were unable to have chil-
dren of my own, I would like to adopt children"), or through the prod-
ucts of one's creative endeavors ("I have made and created things that
have had an impact on other people"). Internal consistency reliability
of this scale was .70.
Internal political efficacy. Internal political efficacy, the sense that
the individual can successfully affect the political system relative to the
ability of other individuals to do so, was measured using a scale devel-
oped by Craig and Maggiotto( 1982). The five-item measure asked par-
ticipants to rate their agreement with statements like "1 feel like
1
could
do as good a job in public office as most of the politicians we elect."
They responded on scales ranging from 1 (strongly disagree) to 7
{strongly agree). The internal consistency of the measure was .76.
Sense of community. Sense of community, encompassing feelings of
community attachment and belongingness, was chosen to capture the
communal aspects of social responsibility and was measured using a
scale developed by Bachrach and Zautra (1985). Respondents were
asked to select the community that was most important to them and lo
rate eight attitudes toward this community on a 7-point scale ranging
from
1
(not at all) to 7 (very much). Items concerned the sense of com-
munity belongingness, connection, and contentment as well as the feel-
ing that one shares commonalities with others in the community. Reli-
ability for the eight items was .84.
Intercorrelations among the variables hypothesized to predict midlife
political participation are presented in the Appendix.
Results
First, we consider evidence that the proposed predictor vari-
ables did assess the underlying constructs: political identity,
social responsibility, and power discontent. Next, we examine
evidence for racial differences in political activism and in the
measures of personality, attitudes, and political identity. Finally,
WOMEN'S POLITICAL PARTICIPATION
135
Table 1
Factor Analysis of
Predictor Variables
of
Midlife
Political Participation
Factor and variable
Political Identity
Political salience
Collective orientation
System
blame:
Sex
System blame: Race
Left-right ideology
Power Discontent
Gender within race
Gender
Social Responsibility
Generativity
Sense of community
Political efficacy
Factor
1
.56
.70
.47
.71
.74
.11
.01
.00
.03
.23
Factor
2
.18
-.19
.41
.03
.35
.80
.81
.02
-.11
.07
Factor 3
.36
.16
.05
.04
-.06
-.10
.07
.85
.67
.56
Note. Factor loadings in boldface type indicate variables included in
scale
scores.
Factor
1
= Political Identity; Factor
2
= Power Discontent;
Factor
3
= Social Responsibility.
we present the pattern of relationships between the hypothe-
sized predictors and midlife political participation.
Structure of Predictor Variables
The
five
variables assessing political identity, the two assess-
ing power discontent, and the three assessing social responsibil-
ity were factor analyzed using varimax rotation. Because the
two samples were relatively small, they were combined for this
analysis. Results presented in Table 1 show that the variables
did load onto three distinct factors (accounting for
55%
of the
variance). Moreover, all but one variable (system blame for
sex) loaded onto one and only one factor at above .40; system
blame for sex loaded on both Political Identity and Power Dis-
content. Because all but one of the variables showed a clear pat-
tern of loading on a single factor, and the exception made con-
ceptual sense, indices of standard-scored scales were created,
combining the standardized individual measures to create over-
all measures of Political Identity, Power Discontent, and Social
Responsibility. System blame for sex was included in Political
Identity for conceptual clarity.
Racial Differences in Political Activism and Predictors
The results of analyses pertaining to student activism and
participation in the civil rights and women's movements are
presented in Table
2.
Beginning with participation in the move-
ments, a chi-square analysis showed that significantly more
White women than Black women participated in the women's
movement; however, a t test indicated that the groups rated the
effect of the movement on their lives as equally important. Not
surprisingly, significantly more Black women reported having
participated in the civil rights movement than did their White
counterparts; similarly, Black women rated the movement as
having had a greater effect on their
lives.
T tests showed that compared with their White classmates,
Black women participated in more student activism and re-
ported higher rates of midlife political participation; however,
at midlife, the difference in their rates of participation had
narrowed.
Multivariate analyses of variance assessed racial differences
in political identity, power discontent, and social responsibility.
As is shown in Table 3, the hypothesized racial differences in
political identity and power discontent were revealed; also as
hypothesized, there was no racial difference in social responsi-
bility. All of the indicators of political identity
were
significantly
different in the two groups, except for collective orientation.
On power discontent, the measure of within-race power dis-
content showed a racial difference (with Black women endors-
ing higher discontent with the power of their own group than
did White women); there was no racial difference for the mea-
sure of discontent with the power of women in
general.
It should
be noted that both groups reported high levels of both types of
discontent.
Additional analyses revealed that Black women's scores on
three of the
five
variables constituting political identity—politi-
cal salience, system blame for
race,
and collective orientation—
showed significantly less variance (i.e., were more homo-
geneous) than White women's scores, using Levene's test for
equality of variance
(Norusis,
1990):
for political
salience,
F{
1,
167) =
4.21,
p <
.05;
for system blame, F(
1,
161) =
4.S5
f
p<
.05;
and for collective orientation, F{\, 142) = 3.85, p = .05.
There were no differences in the two groups' variance on politi-
cal orientation and system blame with respect to sex.
Table 2
Mean
Levels
of
Participation in Political Movements
of the
1960s
and
1970s,
by Race
Variable
White women
M SD
Black women
M SD
Significance of differences
Participated in women's movement 61
Rate effect of women's movement 3.4 1.3 3.3 1.2
Participated in civil rights movement 48
Rate effect of civil rights movement 2.9 1.3 4.7 0.6
Level of student activism 0.3 0.3 1.2 0.1
Level of midlife activism 0.9 0.5 1.1 0.1
42 x
2
(UN= 167) = 5.67**
F(\,
164) =
0.40,
ns
92 x
2
(l,A' = 168) = 33.88**
F{1,
165)= 108.54***
F(l, 157) = 93.17***
F(l,
160)
=
4.11*
*p<.05.
**p<.0l.
***/?<.001.
136
COLE AND STEWART
Table 3
Multivariate
Analyses
of
Variance on Political
Identity,
Power
Discontent,
and Social Responsibility
Variable
Political identity
3
Political salience
Collective orientation
System blame: Sex
System blame: Race
Left-right ideology
Total
Power discontent
Women in general
Women of own race
Total
Social responsibility*
Generativity
Sense of community
Political efficacy
Total
Mean score
White
women
1.95
-0.77
19.81
20.57
3.25
4.28
4.07
-0.59
45.93
21.80
Black
women
2.14
-0.21
21.66
22.64
3.58
4.21
4.46
0.48
46.43
23,13
F
8.41
0.40
5.57
6.09
4.72
2.73
0.35
8.70
6.65
2.77
0.17
2.36
1.28
P
.004
ns
.02
.02
.03
.02
ns
.004
.002
.10
ns
ns
ns
/Vow. Scores for collective orientation and generativity are aggregates
of standard-scored items, with means of 0 and standard deviations of 1.
a
For F values, dfs = 1, 139.
b
For F values, dfi = 1, 154.
c
For F
values, dfs^ 1, 151.
Correlates
ofMidiife
Political Participation
Table 4 presents the simple correlations of student activism
and the index scores of political identity, power discontent, and
social responsibility with midlife political participation for the
two samples separately by race. For both groups, student activ-
ism and midlife social responsibility were significantly and pos-
itively correlated with midlife political participation; for the
White women, political identity was also a significant positive
correlate.
Midlife political participation was regressed on level of stu-
dent activism and the index scores of political identity, power
discontent, and social responsibility, separately by race. These
results are presented in Table 4
as
well.
For White women, the multivariate model explained
29%
of
the variance in midlife political participation. Political identity
and social responsibility
were
significantly related to midlife ac-
tivism. Note that although level of student activism was signifi-
cantly and positively correlated with midlife activism, when all
variables were included in the model, student activism was not<