Changes in late adolescents’ voting intentions during the election campaign: Disentangling
the effects of political communication with parents, peers, and media
Jan Šerek & Tomo Umemura
Faculty of Social Studies, Masaryk University
serek@fss.muni.cz
Published in European Journal of Communication:
Šerek, J., & Umemura, T. (2015). Changes in late adolescents' voting intentions during the
election campaign: Disentangling the effects of political communication with parents, peers,
and media. European Journal of Communication, 30(3), 285-300.
http://dx.doi.org/10.1177/0267323115577306
This version is an original submission to the journal with revisions after peer review accepted
by the editor.
Abstract
This article investigates the effects of political discussions with parents, political discussions
with peers, and exposure to political news during an election campaign on the voting
intentions and behaviour of first-time voters. Longitudinal data collected in the Czech
Republic is employed in the main analysis (N = 223). Results show that young people who
frequently discuss politics with their peers are characterized by higher voting intention and
subsequent electoral participation. On the other hand, political discussions with parents and
exposure to political news have no such effects. Furthermore, although it does not have an
impact on voting intentions, more frequent political discussions with parents predict increased
frequency of political discussions with peers. Overall, our results underscore the importance
of peers in late adolescents’ political socialization.
Keywords: First-time voters, political discussions, political news, political socialization,
voting intention.
Introduction
Compared to previous generations, today’s young people are less engaged in
conventional political participation, such as voting (Norris, 2003; Putnam, 2000). A clear cut
example is that in the Czech Republic, only one half of 14-year-olds indicate their future
voting in national election (Kerr et al., 2010). Although the lower voter turnout among young
people might be partially compensated by their greater involvement in unconventional
activities, such as demonstrations (Norris, 2003), it still poses a problem if a large part of
young generation does not participate in the selection of political representatives and
governmental policies. This trend is troublesome particularly because fundamental political
habits and attitudes are formed during this young age and often remain stable over the course
of a lifetime (Krosnick and Alwin, 1989; Sears and Levy, 2003). Therefore, it is important to
understand how young individuals develop their voting intentions and what factors influence
their actual voting behaviour.
Previous studies have found that young people’s intentions to participate in politics are
strengthened by personal involvement in political discussions with parents (e.g., Zukin et al.,
2006), their political discussions with peers (e.g., Ekström and Östman, 2013), and their
exposure to political news in media (e.g., Pasek et al., 2006). Because the effects of political
communications are particularly strong during election campaigns (Valentino and Sears,
1998), it is expected that individuals who have been exposed to these three socialization
agents during a pre-election period are more likely to intend to vote, compared to people who
have not. In addition, it has been suggested that these forms of political communication with
different socialization agents are positively associated with each other (Amnå et al., 2009;
McLeod, 2000; McLeod and Shah, 2009). For instance, the effect of exposure to political
news on political participation is mediated by interpersonal political discussions, during
which participatory intentions are developed (e.g., Cho et al., 2009; Shah et al., 2005).
However, some processes still remain unclear. Specifically, more evidence is needed on
whether parents or peers are more important agents for young persons’ development of voting
intentions during an election campaign. In addition, it is unclear whether young individuals’
political discussions with their parents provoke their later political discussions with peers or
vice versa. Similarly, little is known about whether following politics in the media stimulates
political discussions with parents, peers, or both. To better understand the developmental
processes of voting intentions in young people during political campaigns, this study employs
panel data from first-time voters, questioned three months before, several days before and one
month after a national election in the Czech Republic.
The effects of political discussions and media on youth political participation
Involvement in political discussions plays an important role in youth political
socialization. Previous research has consistently shown that young people who discuss
politics with others are more likely to participate in politics and civic organizations, compared
to those not involved in political discussions (McIntosh et al., 2007; Richardson, 2003; Zukin
et al., 2006). However, since most of the above-mentioned evidence comes from crosssectional
studies, the causality of these associations has been occasionally questioned (e.g.,
Amnå et al., 2009; McLeod, 2000). To understand whether young persons’ involvement in
political discussion results in their greater political participation or whether young people who
are already active in politics tend to discuss politics, several longitudinal and quasi
experimental studies have been conducted more recently (Klofstad, 2007; 2010; Lee et al.,
2013; Shah et al., 2005). The results of these studies have supported the causal link from
young individuals’ political discussions to an increase in their subsequent political
participation. Thus, taking part in political discussions seems to have consequences for
political participation in youth.
There are several reasons why involvement in political discussions leads to greater
political participation. First, taking part in political discussions brings about a greater
likelihood that others will persuade a person to participate (Gerber and Green, 2000; Green et
al., 2003; Teorell, 2003). Second, through political discussions, people learn about political
attitudes and behavioural intentions that are present in their social environment. In turn, those
who perceive a supportive atmosphere for political participation are likely to engage in
political activity (Glasford, 2008; Glynn et al., 2009). Finally, by discussing politics with
others, individuals gain greater knowledge about politics (Eveland et al., 2005; Eveland and
Thomson, 2006), which has been regarded as an important resource for facilitating political
participation (Glasford, 2008). For instance, greater knowledge about the electoral system,
candidates and their political programs – all of which can be gained through political
discussions – may increase individuals’ motivation to vote.
Besides, many studies have found a positive relation between exposure to political
news in the media (e.g., Internet, television, newspaper) and political participation of young
people (Bakker and de Vreese, 2011; Pasek et al., 2006; Quintelier and Vissers, 2008).
However, in terms of enhancing political participation, exposure to political news may be a
less important factor than political discussions. Previous studies revealed that personal
conversation increased citizens’ voter turnout more strongly than did direct mails and
telephones (Gerber and Green, 2000; Green et al., 2003). The authors of these studies found
that face-to-face persuasion (e.g., personal conversation) is more effective for boosting
citizens’ political participation than non-face-to-face forms of persuasion. Since political
discussions occur during face-to-face communications, whereas exposure to political news
does not, political discussions may be more efficacious than exposure to political news as a
way to increase political participation of the youth.
Discussions with parents versus peers
Regarding the effect of political discussions on participation, a question arises about
with whom a person discusses politics. In research on political socialization, parents have been
considered to be the main socialization agents (e.g., Gallatin, 1980; Jennings, 2004; Sears,
1975). However, as young people move from adolescence to adulthood, parental influence on
adolescents generally diminishes, which is accompanied by spending less time with parents
and developing greater personal autonomy (Arnett, 2006; Smetana, 2011). Simultaneously,
adolescence is characterized as an important developmental period in which individuals learn
about egalitarian communication styles through their interactions with peers. In these peer
interactions, adolescents acquire dispositions that are relevant to their political participation,
such as understanding moral concepts of reciprocity and equality (Piaget, 1965), a sense of
collective political efficacy (Beaumont, 2010), and trust in their fellow citizens (Flanagan,
2003). Moreover, young people also often prefer to discuss politics with peers than with
adults. For instance, research that involved qualitative interviews with young political
activists found that young activists stress the importance of political socialization in their peer
groups while being critical of adults’ attempts to structure their political development
(Gordon and Taft, 2011). Thus, discussing politics with peers may be more important than
discussing politics with parents in terms of the effect on first-time voters’ political
participation.
Although previous studies have separately demonstrated that political discussions with
parents (e.g., Zukin et al., 2006) and peers (e.g., Klofstad, 2007) are linked to political
participation, few studies have examined these two types of discussions simultaneously.
Hence, a unique contribution of political discussions with each social agent (i.e., parents and
peers) to young persons’ political participation remains unclear. For instance, Richardson
(2003) studied parental and peer discussions; however, because she analysed them separately,
the unique contribution of political discussions with parents versus peers was not tested. Lee
et al. (2013) distinguished between discussions with peers and discussions with adults outside
the family, but did not measure the frequency of discussions with parents in their study.
McDevitt and Kiousis (2007), in their longitudinal analysis, showed the effects of political
discussions with parents on adolescents’ voting, and of political discussions with peers on
adolescents’ political protest; however, they did not assess both types of discussion in one
common model. The most evidential comparison of parental and peer discussions comes from
Eström and Östman (2013), who showed that several civic outcomes, including expected
formal political participation, were more strongly related to civic talk with peers than civic
talk with parents in 13- to 17-year-old students. However, their findings were limited by their
use of cross-sectional data and their focus on expected (not actual) political participation.
Therefore, our study aims to overcome these limitations by employing longitudinal design and
measuring actual political behaviour to analyze the effects of political discussions with
parents versus peers.
Interrelations among political communications with different socialization agents
Moreover, the causal order, in which interactions with one socialization agent
provokes interactions with another, still needs to be fully comprehended (McLeod, 2000;
McLeod and Shah, 2009). Recent studies have shown that the exposure to political
information outside the family context (e.g., schools) can motivate young people to initiate
political discussions with their parents (McDevitt and Chaffee, 2000; McDevitt, 2006; Saphir
and Chaffee, 2002). However, these studies focused on younger adolescents who were not
eligible to vote and who might be more inclined to approach parents, compared to older
adolescents. Moreover, it is not sure whether the effect works in the opposite direction as
well; young people’s political discussions with their parents can serve as “playgrounds,” in
which young people practice their abilities to hold political discussions before they start to
discuss politics in more public settings (Kiousis et al., 2005; McDevitt and Chaffe, 2002). The
positive effect of adolescent-parent political discussions on adolescents’ involvement in
political discussions with peers was shown in the study conducted by Kiousis et al. (2005).
However, the cross-sectional design of their study renders it unable to provide convincing
evidence on the causality between parental and peer discussions. Thus, one unique
contribution of our study is to examine the causal order in which political discussions with
parents and peers occur, using longitudinal data.
Furthermore, the causal order between exposure to political news in media and
political talk has been a subject of interest in the recent years. Several studies have shown that
the effect of exposure to political news on greater political participation is only indirect,
mediated by greater involvement in political discussions with others, as well as by interactive
online messaging with others (Cho et al., 2009; Lee et al., 2013; Shah et al., 2005; Shah et al.,
2007). This model, referred to as a citizen communication mediation model, suggests that
informational uses of media stimulate interpersonal political communication, through which
participatory intentions are formed; however, it is not clear whether informational media use
stimulates political discussions with parents or peers. Most studies guided by the citizen
communication mediation model focus on adult populations (Cho et al., 2009; Shah et al.,
2005; Shah et al., 2007) and therefore do not include discussions with parents in their studies.
Although one study employed adolescents (ages 12-17), it did not include political
discussions with parents (Lee et al., 2013). A study by McDevitt (2006) showed that exposure
to political news motivated middle-school adolescents to initiate political discussions with
their parents; yet, political discussions with peers were not explicitly examined in his study.
Moreover, older adolescents might tend to address their parents less frequently than do
younger adolescents. Thus, the present study examines whether exposure to political news in
media longitudinally influences late adolescents’ involvement in political discussions with
parents, peers or both.
The present study
This study aims to investigate the effects of young people’s involvement in various
forms of political communication on their political participation. More specifically, our
overarching goal is to understand how first-time voters’ intentions to vote and actual voting
behaviours change during a pre-election period as a consequence of their political discussions
with parents, political discussions with peers, and exposure to political news in media. This
goal encompasses three specific questions. First, previous research has not conclusively
differentiated between the effects of parental and peer political discussions; therefore, we
examine whether changes in voting intention and voting are induced by political discussions
with parents or with peers (expecting a greater impact of peers). Second, the causal order
between political discussions in different contexts remains unclear; hence, we explore
whether political talk with parents precedes political talk with peers, or vice versa. Finally,
previous research has suggested that following politics in the media stimulates one’s greater
involvement in political talk, but political discussions with peers and parents have not been
differentiated from each other; therefore, we examine how exposure to political news affects
political talk when discussions with parents and peers are considered separately.
Method
Participants and procedure
We employ a subset of data from a larger panel study conducted in the Czech
Republic. This study was originally designed to examine changes in late adolescents’
sociopolitical attitudes, Internet use and personal relationships over a one-year period.
Participants completed online questionnaires five times within one year (i.e., February, May,
June, October, and December, 2010). Since the national parliamentary elections were held at
the end of May 2010, the present study uses pre-election data from February (T1) and May
(T2) as well as post-election data from June (T3).
Recruited participants were in their third and fourth years of Czech secondary
education, which usually lasts four years and is completed at the age of 19. Because Czech
citizens gain the right to vote in national elections at the age of 18, our participants were
potential first-time voters. A stratified multistage random sampling method (i.e., districts,
schools, classes) was used to obtain e-mail contacts of 1,000 young people living in the South
Moravian region. After being contacted via e-mail, 657 persons confirmed their willingness to
participate in the study, and 479 participated in the first wave of data collection. Out of these
participants, 21% lived in a big city (with a population about 400,000), while 79% lived in
smaller towns and villages; 52% students were from academically oriented grammar schools,
while 48% were from vocationally oriented secondary schools. Overall, the data collection
was affected by sample attrition as some participants dropped out the study or did not
participate in all waves of data collection (290 students participated in the second wave; 276
in the third). Ten participants younger than 18 were excluded from all analyses.
For the present study, analyses were conducted based on 223 young individuals (73%
women) who participated in the study in both T1 and T2. In these waves, they completed the
questionnaires regarding voting intention, political discussions with parents, political
discussions with peers, and exposure to political news in media. Out of the 223 participants,
182 young people reported information about their actual voting behaviour at T3.
However, if participants with some missing data are completely excluded from
analyses (i.e., listwise deletion), that can result in biased parameter estimates if the patterns of
missing data are not completely random (Enders and Bandalos, 2001). The above-mentioned
223 participants who were included in the main analyses did not differ significantly from the
256 participants who were excluded due to missing data in terms of their urban versus rural
backgrounds, but they did differ significantly in terms of their educational track (i.e., a greater
proportion of students from vocational schools had missing data). Therefore, we also
recomputed our final model using all available data from 454 participants (69% women) who
had participated in at least at one wave and completed at least some target measures (missing
data were treated by a full information approach; Enders and Bandalos, 2001). Results of all
available data are presented in brackets next to the main results.
Measures
Voting intention (T1 and T2). The strength of participants’ intention to vote was
measured by one question: “A parliamentary election will be held this year to decide on a new
government for our country. Are you going to take part in this election?” A three-point
response scale (0 = no; 1 = I’m not sure; 2 = yes) was used (MT1= 1.58, SDT1= 0.67; MT2=
1.67, SDT2= 0.66).
Political discussions with parents (T1 and T2). Participants were asked one
question: “In the last month, how often did you discuss political issues with your parents?” A
five-point response scale (0 = never; 1 = once a month; 2 = several times a month; 3 = several
times a week; 4 = daily or almost daily) was employed (MT1= 1.58, SDT1= 1.09; MT2= 1.86,
SDT2= 1.16).
Political discussions with peers (T1 and T2). Involvement in political discussions
with peers was measured by one question: “In the last month, how often did you discuss
political issues with your friends or classmates?” The response scale was the same as the one
used for parental discussions (MT1= 1.90, SDT1= 1.07; MT2= 2.11, SDT2= 1.10).
Exposure to political news in media (T1 and T2). The frequency of exposure to
political news was measured with regard to three types of media: 1) TV or radio, 2)
newspaper or journals, and 3) the Internet. Specifically, the question was as follows: “In the
last month, how often did you follow political news on TV or radio?” The same questions
were asked for newspapers or journals and for the Internet. Five-point response scales were
used (0 = never; 1 = less than once or twice a week; 2 = once or twice a week; 3 = three or
four times a week; 4 = daily). A total scale was computed by averaging all three items
(Cronbach alphas were .68 at T1 and .74 at T2; MT1= 2.15, SDT1= 0.85; MT2= 2.18, SDT2=
0.92).
Voting (T3). After the election, participants were asked a dichotomous question
about whether they actually voted: “A parliamentary election took place at the end of May.
Did you vote?” (0 = no; 1 = yes). Voting turnout in our sample was 80%.
Data analysis
In a preliminary analysis, we tested whether there were any associations between the
initial level of voting intention and political communication at T1 and voting at T3. Since the
outcome variable was binary, logistic regression models were used.
Next, to examine indirect effects from political communication to voting behaviour,
a cross-lagged model depicted in Figure 1 was estimated. Longitudinal paths of voting
intention, political discussions with parents, political discussions with peers, and exposure to
political news were estimated from T1 to T2. Relations of the same variables between T1 and
T2 were controlled as autoregressive paths (a), and relations among distinct variables from T1
to T2 were estimated as cross-lagged paths (b). In addition, longitudinal paths from the T2
variables to voting behaviour at T3 were estimated. Concurrent correlations among the
variables at T1 were freely estimated as well as residual correlations among the variables at
T2. Because one of our outcome variables – voting intention – was ordinal, we estimated
parameters using a polychoric correlation matrix, which is typically used for analyses with
ordinal outcomes (Kline, 2011; Muthén and Asparouhov, 2002). A weighted least square
estimator with mean- and variance-adjusted χ2
statistic (WLSMV) was used because of its
robustness to eventual non-normal distributions of the data.
--- Figure 1 ---
Results
Descriptive statistics
Bivariate correlations between study variables are presented at Table 1. Many
participants intended to vote during the pre-election periods and actually went to vote.
Specifically, at T1, 68% (n = 152) of participants intended to vote, 22% (n = 49) were
undecided, and 10% (n = 22) intended not to vote. At T2, 78% (n = 173) of participants
intended to vote, 12% (n = 26) were undecided, and 11% (n = 24) intended not to vote (5
participants did not answer this question at T2). At T3, 80% (n = 145) went to vote, whereas
20% (n = 37) did not.
Voting intentions between T1 and T2 were moderately correlated (r = .57). The
correlation between voting intentions at T1 and actual voting at T3 was also moderate (r =
.49), and the correlation between voting intentions at T2 and actual voting at T3 was strong (r
= .80).
As expected, correlation analyses also showed that voting intentions as well as actual
voting behaviour were positively linked to greater involvement in political discussions with
parents and with peers, and to a higher frequency of exposure to political news in media (see
Table 1).
--- Table 1 ---
Preliminary analysis
A greater likelihood of voting was strongly predicted by having a positive voting
intention at T1. Over and above the effect of voting intention, a greater likelihood of voting
was predicted by a greater frequency of discussions with peers, but not by discussions with
parents or exposure to political news (see Table 2). Hence, our results implied that some
forms of political communication (particularly political discussions with peers) were
positively associated with voting behaviour.
--- Table 2 ---
A cross-lagged model predicting voting intention and voting
The estimated model showed a good fit to the data (χ2
4 = 2.17, p = .70; CFI = 1.00;
RMSEA = 0.00 [χ2
4 = 2.71, p = .61; CFI = 1.00; RMSEA = 0.00]). As presented in Table 3,
autoregressive paths (a paths in Figure 1) of participants’ voting intentions, political
discussions with parents, discussions with peers and the frequency of exposure to political
news were all significant between T1 and T2. Their standardized coefficients were all
moderate to strong (i.e., ranged from .40 to .56 [from .39 to .56]), suggesting that these
variables were relatively stable from T1 to T2.
--- Table 3 ---
Regarding voting intentions and behaviour, changes in voting intentions at T2 were
positively predicted by the frequency of political discussions with peers at T1, but not by
political discussions with parents or by exposure to political news at T1 (see Table 3). Young
individual’s voting intentions at T2 strongly determined participants’ actual voting behaviour
at T3, but political discussions (both with parents and with peers) and exposure to political
news at T2 had no additional effect on voting behaviour at T3 (see Table 3). To assess
whether young people’s political communications at T1 had indirect effects through their
voting intentions at T2 on their actual voting behaviour at T3, 95% confidence intervals of
standardized indirect effects were examined. Regarding political discussions with peers, 95%
confidence intervals of standardized indirect effect were from 0.09 to 0.42 [from .11 to .40],
suggesting that political discussions with peers at T1 predicted increased voting intentions at
T2, which turned into actual voting at T3. In contrast, political discussions with parents (95%
standardized CI from -.03 to .31 [from -.02 to .29]) and exposure to political news (95%
standardized CI from -.07 to .25 [from -.10 to .21]) had no indirect effect on actual voting.
Relations among political discussions and exposure to political news
In the model, although involvement in political discussions with parents did not
induce changes in voting intentions, discussions with parents at T1 were found to predict
more frequent political discussions with peers at T2. Besides, political discussions with
parents at T1 marginally predicted a higher frequency of exposure to political news at T2. On
the other hand, political discussions with peers did not predict more frequent political
discussions with parents or more frequent exposure to political news at T2. Exposure to
political news did not predict political discussions with parents or peers (see Table 3).
Discussion
This study of first-time voters revealed three important findings. First, first-time
voters’ intentions to vote are predicted by political discussions with their peers but not with
their parents during an election campaign. Next, although political discussions with parents
are not directly related to changes in voting intention, they predict political discussions with
peers. Finally, exposure to political news in media does not predict changes in late
adolescents’ voting intention or in their political discussions during a pre-election period.
In line with our expectations, the results showed that political discussions with peers
before the election were associated with young people’s enhanced voting intentions, which in
turn increased their voting turnout, whereas political discussions with parents did not have
such effects. There are at least three complementary explanations for these discrepant findings
regarding peer and parental discussions. First, previous studies have suggested that political
discussions enhance individuals’ voting behaviours in multiple ways. For example, during
political discussions, people can directly persuade each other (e.g., Green et al., 2003; Gerber
and Green, 2000), perceive social support for their voting (Glasford, 2008; Glynn et al.,
2009), and increase their political knowledge (Eveland et al., 2005; Eveland and Thompson,
2006). Although all of these aspects appear in political discussions with both peers and
parents, transitions from adolescence to adulthood are characterized by the decreasing
influence of parents on adolescents’ socio-political orientations (Arnett, 2006; Vollebergh et
al., 2001). At the same time, peer relationships gain growing importance in this life period
(Fraley and Davis, 1997; Kobak et al., 2007). Thus, for late adolescents, peer discussions are
more likely than parental discussions to have substantial impacts on their voting behaviours.
Specifically, late adolescents may accept their friends’ political ideas more readily than their
parents’, and take political information from peers as more valid than such information from
their parents.
Second, the impact of peer political discussions might be because young individuals
prefer to discuss politics with peers who encourage their voting, rather than with peers who do
not. Although adolescents tend to make relationships with peers who are attitudinally similar
to themselves, they also prefer to establish contacts with peers from whom they can learn and
gain new insights (Kerr et al., 2003; LaFontana and Cillessen, 2002). Applying this idea to the
political domain, we suppose that young individuals may be likely to discuss politics with
those peers who have above-average positive attitudes regarding politics in order to engage
with more relevant and intriguing ideas. Such friends are, in turn, likely to give direct
persuasion, positive attitudes, and additional political information to young individuals – all
of which enhance voting. In contrast, since most adolescents have the same parental figures
throughout their lives and parents tend to have stable attitudes toward voting (Krosnick and
Alwin, 1989; Sears and Levy, 2003), no selection among alternative discussion partners is
possible and the influence of parents on adolescents’ voting may be quite stable.
Finally, political discussions with peers and parents might be characterized by
different prevailing communication patterns related to future political participation.
Specifically, previous studies on family communication have differentiated between socioand
concept-oriented communicational dimensions (e.g., Chaffe, McLeod and Atkin, 1971).
The former dimension entails sustaining harmony, deferring to authority, and avoiding
controversy. Due to the lack of free exploration, socio-oriented communication is not
expected to be an ideal circumstance for young people to increase their interests in politics.
On the other hand, the latter dimension is characterized by free exchange of ideas and
exposure to controversy and disagreement, and is expected to be positively related to political
interest (Hively and Eveland, 2009). From this perspective, it is possible that political
discussions with peers entail more concept-oriented and less socio-oriented features than
political discussions with parents. In general, parent-adolescent interactions are characterized
by greater asymmetry and avoidance of controversy than peer-adolescents interactions. For
example, in comparison to peers, parents often prefer to explain their own views over
listening adolescent’s views (Hunter, 1985); conflicts with parents, compared to those with
peers, involve greater coercion and negative affect afterward (Adams and Laursen, 2001); and
adolescents avoid discussing controversial issues with their parents more than with their peers
(Moore and Rosenthal, 1991). Consequently, young people might perceive political
interactions with peers as more intriguing and stimulating, which can have positive impact on
their voting intention.
Although our study found no significant longitudinal effect of parental discussions
on young people’s voting intention during the three-month short-term pre-election period, it is
possible that political discussions with parents may in the long run still have effects on
adolescents’ voting intentions. In fact, we found positive correlations between these variables.
Previous studies have shown that young people who grow up in a politicized family
environment (i.e., in which parents engage in political activities and discuss politics in their
family) have a greater tendency to learn political values and habits that support political
participation (Plutzer, 2002; Zukin et al., 2006). However, these socialization processes
usually proceed gradually over a long period of time, not within short time periods, such as
the one captured in this study. In contrast to our findings on the effects of political discussions
with peers, family discussions may have a more gradual effect on young people’s voting
intentions, whereas peer discussions may have a more immediate effect.
Moreover, we found that political discussions with parents lead to more frequent
political discussions with peers. This finding supports the idea that political discussions in the
families serve as “playgrounds” where young people may develop their political opinions and
practice their discussion skills before initiating discussions in other contexts (Kiousis et al.,
2005; McDevitt and Chaffe, 2002). Moreover, family is a more protected place in which
parents play the role of a secure base from which their children confidently explore to learn
about the world (Bowlby, 1982). Therefore, through political talk with parents, young people
may develop self-confidence and political competences that are necessary in order to initiate
political communications outside the family, such as with peers. Thus, taken together,
political discussions with parents may not be the arenas where young people primarily
develop their participatory intentions, but parental discussions seem to serve as a safe realm
within which young people may develop the confidence to participate in more public political
discussions.
Surprisingly, our study found no effect of late adolescents’ exposure to political news
during the pre-election period on their voting behaviour. This finding seems to contradict the
findings of previous studies that a positive relation exists between exposure to political news
in various media and political participation among young people (Bakker and de Vreese,
2011; Pasek et al., 2006; Quintelier and Vissers, 2008). However, these studies were crosssectional;
hence, they could suggest only associations but not causal effects between exposure
to political news and participation. In contrast, one recent longitudinal study showed no causal
effect between adolescents’ exposure to political news in newspapers, on television or online
and their political participation (Lee et al., 2013), which is consistent with our findings. Thus,
it is possible that both exposure to political news and political participation arise from some
common characteristics of adolescents, such as their political interest, but they do not affect
each other. Moreover, exposure to political news is a passive form of media use; however,
several studies have shown that more active media uses predict political participation in
young people. For instance, Quintelier and Vissers (2008) found that the strongest predictor of
adolescents’ political participation, among various online activities, was their forwarding of
political e-mails. Similarly, the above-mentioned study by Lee et al. (2013) showed that
adolescents’ online active political messaging (i.e., expressing their political views through emails,
text messages, Web links or social networking sites) predicted their political
participation. Hence, we believe that adolescents’ media use may have a causal link to their
voting intentions, but this effect applies to active rather than passive forms of media use.
Several limitations should be mentioned in regard to this study. First, voting
intentions and political communications were measured at only two time points before the
election. Taking measurements at more times would strengthen the evidence on the studied
effects and their directionality. Second, we measured only the frequency of political
discussions and political media use; however, other characteristics of these political
communications might be relevant too, such as who initiated the discussions, what
communication patterns dominated in the discussions, and whether political media were used
passively or actively. Third, our sample included only first-time voters who attended
secondary education; it did not include those who had not gone to secondary school and were
working or unemployed. Compared to students, young workers and the unemployed might
belong to and communicate with relatively different peer networks, and thus political
discussions with peers may have different effects on these groups. Therefore, our results
should be generalized only carefully to non-student youth.
The major strength of this study consists in its utilization of longitudinal data. Based
on this approach, we were able to find that political discussions with peers not only were
associated with voting intentions and behaviour but also predicted their changes during the
pre-election period. Likewise, we were able to identify that political discussions with parents
led to increased frequency of political discussion with peers. Compared to cross-sectional
designs, our longitudinal analyses provided much stronger evidence for causal links from
political discussions to voting behaviour. Yet, more precise factors of political discussions
that are supportive of voting intentions as well as broader community-related predictors of
political discussions still need to be explored in future studies. At the same time, our results
draw attention to the fact that the impact of exposure to political news in media on
adolescents’ political participation might be overestimated by studies that use only crosssectional
data. For further research in this area, employing longitudinal and
(quasi)experimental designs would be highly desirable.
References
Adams R and Laursen B (2001) The organization and dynamics of adolescent conflict with
parents and friends. Journal of Marriage and Family 63(1): 97–110.
Amnå E, Ekström M, Kerr M and Stattin H (2009) Political socialization and human agency.
The development of civic engagement from adolescence to adulthood. Statsvetenskaplig
Tidskrift, 111(1): 27-40.
Arnett JJ (2006) Socialization in emerging adulthood: From the family to the wider world,
from socialization to self-socialization. In Grusec JE and Hastings PD (eds.) Handbook
of socialization: Theory and research. New York, London: Guildford Press, pp. 208-
231.
Bakker TP and de Vreese CH (2011) Good news for the future? Young people, internet use,
and political participation. Communication Research 38(4): 451-470.
Beaumont E (2010) Political agency and empowerment: Pathways for developing a sense of
political efficacy in young adults. In Sherrod LR, Torney-Purta J and Flanagan CA
(eds.) Handbook of research on civic engagement in youth. Hoboken: Wiley, pp. 525-
558.
Bowlby J (1982) Attachment and loss. Vol.1: Attachment (2nd ed.). New York: Basic Books.
Cho J, Shah DV, McLeod JM, McLeod DM, Scholl RM and Gotlieb MR (2009) Campaigns,
reflection, and deliberation: Advancing an O-S-R-O-R model of communication effects.
Communication Theory 19(1): 66-88.
Ekström M and Östman J (2013) Family talk, peer talk and young people’s civic orientation.
European Journal of Communication 28(3): 294-308.
Enders CK and Bandalos DL (2001) The relative performance of full information maximum
likelihood estimation for missing data in structural equation models. Structural
Equation Modeling: A Multidisciplinary Journal 8(3): 430-457.
Eveland WP Jr and Thomson T (2006) Is it talking, thinking, or both? A lagged dependent
variable model of discussion effects on political knowledge. Journal of Communication
56(3): 523-542.
Eveland WP Jr, Hayes AF, Shah DV and Kwak N (2005) Understanding the relationship
between communication and political knowledge: A model comparison approach using
panel data. Political Communication 22(4): 423-446.
Flanagan CA (2003) Developmental roots of political engagement. Political Science &
Politics 36(2): 257-261.
Fraley RC and Davis KE (1997) Attachment formation and transfer in young adults’ close
friendships and romantic relationships. Personal Relationships 4(2): 131-144.
Gallatin J (1980) Political thinking in adolescence. In Adelson J (ed.) Handbook of adolescent
psychology. New York: Wiley, pp. 344-382.
Gerber AS and Green DP (2000) The effects of canvassing, telephone calls, and direct mail on
voter turnout: A field experiment. American Political Science Review 94(3): 653-663.
Glasford DE (2008) Predicting voting behavior of young adults: The importance of
information, motivation, and behavioral skills. Journal of Applied Social Psychology
38(11): 2648-2672.
Glynn CJ, Huge ME and Lunney CA (2009) The influence of perceived social norms on
college students’ intention to vote. Political Communication 26(1): 48-64.
Gordon HR and Taft JK (2011) Rethinking youth political socialization: Teenage activists
talk back. Youth & Society 43(4): 1499-1527.
Green DP, Gerber AS and Nickerson DW (2003) Getting out the vote in local elections:
Results from six door-to-door canvassing experiments. The Journal of Politics 65(4):
1083-1096.
Hively MH and Eveland WP Jr (2009) Contextual antecedents and political consequences of
adolescent political discussion, discussion elaboration, and network diversity. Political
Communication 26(1): 30-47.
Hunter FT (1985) Adolescents’ perceptions of discussions with parents and friends.
Developmental Psychology 21(3): 433-440.
Jennings MK (2004) American political participation viewed through the lens of the political
socialization project. In Hermann MG (ed.) Advances in political psychology.
Amsterdam: Elsevier, pp. 1-18.
Kerr D, Sturman L, Schulz W and Burge B (2010) ICCS 2009 European report. Amsterdam:
IEA.
Kerr M, Stattin H, Biesecker G and Ferrer-Wreder L (2003) Relationships with parents and
peers in adolescence. In Lerner RM, Easterbrooks MA and Mistry J (eds.) Handbook of
psychology. Vol. 6: Developmental psychology. Hoboken: Wiley, pp. 395-419.
Kiousis S, McDevitt M and Wu X (2005) The genesis of civic awareness: Agenda setting in
political socialization. Journal of Communication 55(4): 756-774.
Kline RB (2011) Principles and practice of structural equation modeling. New York,
London: Guildford Press.
Klofstad CA (2007) Talk leads to recruitment: How discussions about politics and current
events increase civic participation. Political Research Quarterly 60(2): 180-191.
Klofstad CA (2010) The lasting effect of civic talk on civic participation: Evidence from a
panel study. Social Forces 88(5): 2353-2375.
Kobak R, Rosenthal NL, Zajac K and Madsen SD (2007) Adolescent attachment hierarchies
and the search for an adult pair‐bond. New Directions for Child and Adolescent
Development 117: 57-72.
Krosnick JA and Alwin DF (1989) Aging and susceptibility to attitude change. Journal of
Personality and Social Psychology 57(3): 416-425.
LaFontana KM and Cillessen AHN (2002) Children’s perceptions of popular and unpopular
peers: A multimethod assessment. Developmental Psychology 38(5): 635-647.
Lee N-J, Shah DV and McLeod JM (2013) Processes of political socialization: A
communication mediation approach to youth civic engagement. Communication
Research 40(5): 669-697.
McDevitt M and Chaffe S (2002) From top-down to trickle-up influence: Revisiting
assumptions about the family in political socialization. Political Communication 19(3):
281-301.
McDevitt M and Chaffee S (2000) Closing gaps in political communication and knowledge:
Effects of a school intervention. Communication Research 27(3): 259-292.
McDevitt M (2006) The partisan child: Developmental provocation as a model of political
socialization. International Journal of Public Opinion Research 18(1): 67-88.
McDevitt M and Kiousis S (2007) The red and blue of adolescence: Origins of the compliant
voter and the defiant activist. American Behavioral Scientist 50(9): 1214-1230.
McIntosh H, Hart D and Youniss J (2007) The influence of family political discussion on
youth civic development: Which parent qualities matter? Political Science and Politics
40(3): 495-499.
McLeod JM and Shah DV (2009) Communication and political socialization: Challenges and
opportunities for research. Political Communication 26(1): 1-10.
McLeod JM (2000) Media and civic socialization of youth. Journal of Adolescent Health 27
(2): 45-51.
Chaffee SH, McLeod JM and Atkin CK (1971) Parental influences on adolescent media use.
American Behavioral Scientist 14(3): 323-340.
Moore S and Rosenthal D (1991) Adolescents’ perceptions of friends’ and parents’ attitudes
to sex and sexual risk-taking. Journal of Community & Applied Social Psychology 1(3):
189-200.
Muthén BO and Asparouhov T (2002) Latent variable analysis with categorical outcomes:
Multiple-group and growth modeling in Mplus. Available at:
https://www.statmodel.com/download/webnotes/CatMGLong.pdf.
Norris P (2003) Young people & political activism: From the politics of loyalties to the
politics of choice? Available at:
http://www.hks.harvard.edu/fs/pnorris/Acrobat/COE%20Young%20People%20and%20
Political%20Activism.pdf.
Pasek J, Kenski K, Romer D and Jamieson KH (2006) America’s youth and community
engagement: How use of mass media is related to civic activity and political awareness
in 14- to 22-year-olds. Communication Research 33(3): 115-135.
Piaget J (1965) The moral judgment of the child. New York: Free Press.
Plutzer E (2002) Becoming a habitual voter: Inertia, resources, and growth in young
adulthood. American Political Science Review 96(1): 41-56.
Putnam RD (2000) Bowling alone: The collapse and revival of American community. New
York: Simon and Schuster.
Quintelier E and Vissers S (2008) The effect of internet use on political participation: An
analysis of survey results for 16-year-olds in Belgium. Social Science Computer Review
26(4): 411-427.
Richardson WK (2003) Connecting political discussion to civic engagement: The role of civic
knowledge, efficacy and context for adolescents, PhD dissertation, College Park,
University of Maryland.
Saphir MN and Chaffee SH (2002) Adolescents’ contributions to family communication
patterns. Human Communication Research 28(1):, 86-108.
Sears DO and Levy S (2003) Childhood and adult political development. In Sears DO, Huddy
L and Jervis R (eds.) Oxford handbook of political psychology Oxford, New York:
Oxford University Press, pp. 60-109.
Sears DO (1975) Political socialization. In Greenstein FI and Polsby NW (eds.) Handbook of
political science. Vol. 2. Reading: Addison-Wesley, pp. 93-153.
Shah DV et al. (2007) Campaign ads, online messaging, and participation: Extending the
communication mediation model. Journal of Communication 57(4): 676-703.
Shah DV, Cho J, Eveland WP Jr and Kwak N (2005) Information and expression in a digital
age: Modeling internet effects on civic participation. Communication Research 32(5):
531-565.
Smetana JG (2011) Adolescents, families, and social development: How teens construct their
worlds. Chichester, Malden, MA: Wiley-Blackwell.
Teorell J (2003) Linking social capital to political participation: Voluntary associations and
networks of recruitment in Sweden. Scandinavian Political Studies 26(1): 49-66.
Valentino NA and Sears DO (1998) Event-driven political communication and the preadult
socialization of partisanship. Political Behavior 20(2): 127-154.
Vollebergh WAM, Iedema J and Raaijmakers QAW (2001) Intergenerational transmission
and the formation of cultural orientations in adolescence and young adulthood. Journal
of Marriage and Family, 63(4), 1185-1198.
Zukin C, Keeter S, Andolina M, Jenkins K, Delli Carpini MX (2006) A new engagement?
Political participation, civic life, and the changing American citizen. Oxford, New
York: Oxford University Press.
Tables and figures
Table 1. Bivariate correlations.
1. 2. 3. 4. 5 6. 7. 8.
1. Voting intention T1
2. Discussions with parents T1 .24
3. Discussions with peers T1 .23 .37
4. Exposure to political news T1 .24 .39 .47
5. Voting intention T2 .57 .30 .33 .27
6. Discussions with parents T2 .30 .46 .22 .23 .25
7. Discussions with peers T2 .26 .34 .49 .27 .30 .53
8. Exposure to political news T2 .35 .37 .34 .65 .35 .42 .43
9. Voting T3 .49 .21 .34 .24 .80 .20 .32 .27
Note. Ns = 223, except Voting T3 (N = 182). All correlations were significant at the .01 level.
Table 2. Logistic regression analysis predicting voting at T3.
Predictors B SE OR 95% CI for OR
Voting intention T1 1.46 0.31 4.32** 2.37; 7.86
Discussions with parents T1 0.13 0.23 1.14 0.73; 1.78
Discussions with peers T1 0.68 0.26 1.97** 1.19; 3.27
Exposure to political news T1 0.00 0.33 1.00 0.52; 1.93
Note. N = 182. Pseudo R2
= .24 (Cox & Snell), .38 (Nagelkerke).
Table 3. Predictors of voting intention, discussions with parents, discussions with peers, and
exposure to political news at T2, and voting at T3.
B SE β
Voting intention T2
Voting intention T1 0.97[0.98] 0.16[0.15] .47**[.49**]
Discussions with parents T1 0.18[0.18] 0.11[0.11] .14[.14]
Discussions with peers T1 0.33[0.33] 0.12[0.11] .26**[.27**]
Exposure to political news T1 0.14[0.09] 0.14[0.13] .09[.06]
Discussions with parents T2
Voting intention T1 0.33[0.32] 0.11[0.10] .19**[.19**]
Discussions with parents T1 0.43[0.40] 0.06[0.06] .41**[.39**]
Discussions with peers T1 0.02[0.05] 0.07[0.07] .02[.05]
Exposure to political news T1 0.02[-0.01] 0.08[0.08] .02[.00]
Discussions with peers T2
Voting intention T1 0.22[0.29] 0.09[0.10] .13*[.18**]
Discussions with parents T1 0.16[0.15] 0.06[0.06] .16**[.15*]
Discussions with peers T1 0.41[0.41] 0.06[0.06] .40**[.41**]
Exposure to political news T1 -0.01[0.01] 0.08[0.09] -.01[.00]
Exposure to political news T2
Voting intention T1 0.27[0.27] 0.08[0.08] .19**[.20**]
Discussions with parents T1 0.09[0.05] 0.04[0.04] .10*[.06]
Discussions with peers T1 0.00[0.05] 0.05[0.06] .00[.06]
Exposure to political news T1 0.61[0.60] 0.06[0.06] .56**[.56**]
Voting T3
Voting intention T2 2.77[2.31] 0.95[0.56] .98**[.95**]
Discussions with parents T2 -0.47[-0.14] 0.45[0.32] -.14[-.05]
Discussions with peers T2 0.48[0.29] 0.37[0.29] .14[.10]
Exposure to political news T2 -0.30[-0.19] 0.35[0.31] -.07[-.05]
Note. ** p < .01. * p < .05. Coefficients predicting voting intention T2 and voting T3 are
probit regression coefficients.
Figure 1. Longitudinal path model predicting voting.