As for participation in voluntary associations, EPIC respondents belong to an average of two local associations or groups, as noted in Table 1. The most common affiliation was with church. About a third of respondents did not belong to any local organization or group, and about a quarter belonged to only one group (most commonly church). Almost half in round one participated in two or more organizations or groups; this dropped to 41% in round two, although there are fewer total respondents in round two due to five households dropping out replaced by five comparable types of households (according to the stratification) but with fewer total persons. People who belong to two or more organizations we called Bridges (weak social ties). People who belong to only one or to no organizations, we called Non-bridges (Table 3).
Table 3. Bridges versus Non-Bridges
|
Group
Status |
R1 |
R1 |
R2 |
R2 |
|
|
Bridge |
48% |
N=75 |
41% |
N=59 |
Two
+ groups |
|
Non-bridge |
52% |
N=83 |
59% |
N=84 |
None
or 1 group |
We tested hypotheses supported in other studies that voluntary association membership (a typical form of civic participation) is positively correlated with socioeconomic status, home ownership, length of residence, marital status, and gender. We calculated the socioeconomic status (SES) variable as the sum of education and income. Education was measured on seven point scale (1= elementary to 7= completed graduate work) and Income was measured on six point scale (1= under $25,000 and 6= over $100,000). Home ownership also reflects socioeconomic status. The correlations between the number of organizations the subject belonged to (Membership) and these variables are shown in Table 4 (round one N=145-149; round two N=134-143).
Table 4. Demographics and Association Membership
|
Variables |
Round 1 |
Round 2 |
|
SocioEconomic
Status (SES) |
.325** |
.213* |
|
Years
lived in local community |
NS |
.188* |
|
Rent
your house |
-.181* |
NS |
|
Gender |
NS |
.184* |
*p<.05 **p<.01 NS= Not Significant
The correlations are moderate, but their significance supports the relationship on these standard associations between community attachment, community involvement and SES. Some variables were significant in only one round. SES was positively correlated with membership in both rounds. Renters were less likely than homeowners to be affiliated with local organizations. Length of residence (a measure of community attachment) was significant in round two only. Women were more likely than men to belong to multiple organizations. Married couples were just as likely to belong to organizations as people who were single.
Correlations also show that people who were members of voluntary associations and who were well informed about affairs of their community were more active politically than individuals who were less informed and had fewer memberships or none at all. We created a composite variable derived from the construct Informed and the number of local organizations to which the respondent belonged (Membership) and called it ‘Informed Membership’. Table 5 shows the correlations between Informed Membership and each variable we used to measure political activity (Round 1 N=146-156; Round 2 N=140-143).
Table 5. Correlations between Informed Membership and Political Activity
|
Variables |
Round
1 |
Round 2 |
|
Registered
voter |
NS |
.211* |
|
Voted
in last presidential election |
.203* |
.307** |
|
Voted
in last municipal election |
.238** |
.361** |
|
Voted
in last congressional election |
.218** |
.355** |
|
Have
ideas to improve things in the community |
.385** |
.372** |
|
Work
to bring change to the community |
.457** |
.471** |
|
Active
and involved |
.603** |
.540** |
|
Spend
time with community work |
.488** |
.328** |
|
Discuss
politics |
.215** |
.195* |
|
Attend
public meetings |
.504** |
.293** |
|
Attend
political meetings |
.253** |
.288** |
|
Attend
local group meetings |
.402** |
.434** |
*p<.05 ** p<.01 NS= not significant
Informed members are more likely to vote, to discuss politics and attend local meetings. Several of the questions shown in Table 5 comprise our construct Activism (shown in Table 7), specifically: have ideas for improving things in the community, work to bring about change in the community, spend time working to solve problems in the community. Informed Members are more likely than others to use the Internet to obtain local political information and to discuss politics. Further, Informed Members are more likely than others to report that they are more involved in the local community and in local issues that interest them since going online, and that they have been attending more local meetings since going online (Table 6, round one N=125-126; round two N=117-118). Thus, a person who is a member of multiple local groups who keeps informed on local news not only represents an important form of civic participation, but also is associated with online civic engagement. This type of citizen also reports increased involvement in civic life as shown in Table 6 since going online.
Table 6. Informed Membership and Internet Use
|
Variables |
Round
1 |
Round
2 |
|
Involved
in local issues of interest since online |
NS |
.202* |
|
Involved
in local community since online |
.320** |
.367** |
|
Use
Net to obtain local political information |
.345** |
NS |
|
Discuss
Politics online |
.193* |
NS |
|
Increased
attendance at local meetings since online |
.239** |
.203* |
*p<.05 **p<.01 NS= not
significant
Among organizational members for both rounds, the majority (almost two-thirds) participated in the broad category of religious/charitable/support organization. About a third participated in the category of civic/political organization, and a quarter in educational/professional organization. These memberships remained fairly even across both rounds. Participation in social types of organizations, however, dropped from 41% in round one to 34% in round two, although it is not clear why. The percentage total sums over 100% because some people belonged to more than one organization. Members of civic/political types of organizations had the highest mean level of education among the four types. This is consistent with studies showing education positively correlated with civic engagement. Average age was highest in educational/professional types of organizations.
To compare changes between round one and round two among the four main categories of voluntary association on measures of membership, level of involvement and types of communication, we analyzed only those cases (individuals) who were members of the same type of organization in round one and round two. If someone dropped out or joined up between rounds, we did not include them in the analysis on these measures. This gave us a stable population within which to investigate changes over time with longitudinal data (N=115, discussed below).
We looked for differences over time in respondents’ level of organizational involvement (leader, attendee, etc.) and forms of communication with group members (traditional versus electronic) using paired t tests. Considering these measures for all groups combined, there are significant increases in overall group involvement, mostly due to active forms of involvement (leader, attendee). There are also significant increases in electronic communication (meanR1= 0.73 (SD=0.69); meanR2=1.03(SD=1.13); t(114) =-3.145; p<.01). When we compare these to civic type of groups, we see a similar pattern for overall civic group communication (meanR1= 2.19 (SD=0.88); meanR2=3.08 (SD=0.79); t(25)=-4.07; p<.01) and specifically for electronic communication (meanR1=0.72 (SD=0.6); meanR2= 1.37 (SD=0.79); t(25)= -3.553; p<.01). Traditional types of communication (face-to-face, telephone) are also up among civic group members over time, although it is only approaching significance (meanR1= 1.47 (SD=0.65); meanR2= 1.71 (SD=0.52); t(25)= -1.92; p=.066). Members of civic organizations also show significant increases in the use of the Internet over time for civic purposes (meanR1=2.72 (SD=1.0); meanR2= 2.96 (SD=0.92); t(23)= -2.18; ,p<.05). Civic organization members also show significant increases on the measure of Participation over time, (meanR1=3.13 (SD=0.53); meanR2= 3.26 (SD=0.48); t(25)= -2.213; p<.05). Respondents’ civic interests and communication behavior are further tested with constructs we developed shown in Table 7. Finally, we use these measures to develop the path model (Figure 1) showing how variables affect directly and indirectly changes in civic involvement since getting on the Internet.
Table 9
Results from the series of
multiple regression analyses for the path model
Civic Outcome: More
Involved in local Issues since Online
|
Predictors |
Dependent variables % variance explained (R2) for
round1 & round 2 |
F values and p |
|
|
Round 1 |
Round 2 |
||
|
1. Education 2. Extroversion 3. Membership |
Staying Informed 14.1% 10.9% |
F(3,124)= 6.799 P <.01 |
F(3,110)=4.471 P<.01 |
|
1. Extroversion 2. Collective Efficacy 3.Education 4. Age 5. Age Squared |
Membership 16.9% 9.9% |
F(5,122)=4.974 P<.01 |
F(5,108)=2.384 P < .05 |
|
1. Extroversion 2. Education 3. Age 4. Age Squared |
Collective Efficacy 16.4% 15.2% |
F(4,123)=6.019 P<.01 |
F4,109)=4.89 P<.01 |
|
1. Staying Informed 2.Extroversion 3. Collective Efficacy 4. Membership |
Activism 43.9% 41.4% |
F(4,123)=24.102 P<.01 |
F(4,109)=19.286 P<.01 |
|
1. Activism 2. Staying Informed 3. Membership |
On line Civic 18.9% 18.1% |
F(3,123) =9.57 P < .01 |
F(3,116)=8.522 P < .01 |
|
1. Membership 2. Activism 3. Online Civic 4. Staying Informed |
Involvement in community Issues 38.5% 19.7% |
F(4,121)=18.918 P < .01 |
F(4,113)=6.952 P < .01 |