Self-Selection or Socialization? The Longitudinal Relation Between Civic Engagement and Political Orientations Among Adolescents

Abstract

Social capital theory states that civic engagement generates positive outcomes, such as social trust and political interest. Likewise, studies show that those involved in civic engagement generally report higher levels of social trust and political interest. It is still unclear, however, whether these differences are the result of socialization or selection. We used between-effects and fixed-effects regressions to examine the development of political orientations in a three-wave longitudinal sample of 1,050 adolescents. From our results, volunteering seemed to have no socialization effect whatsoever on political interest and potentially a weak enhancing effect on social trust. Associational membership did not predict social trust over time, but it seemed to socialize members into increased political interest over time. The results are discussed in light of the social capital debate about how civic engagemend in associational life and volunteering do – or do not – function as schools of democracy.

Keywords

civic engagement political interest social trust selection socialization adolescents

Citizens’ involvement in civic associations is believed to contribute to one’s development and learning of a variety of vital attitudes, norms, and values crucial to stable and effective democratic governance, and to be vital for citizens’ democratic participation in public affairs (Brehm & Rahn, 1997; Putnam, 2000). From this perspective, these associations are often viewed as schools of democracy, or even as an all-purposive “elixir” for society’s ills (Uslaner & Dekker, 2001, p. 177). As such, scholars suggest that civic associations will make their members more trustful of each other and of political institutions (Brehm & Rahn, 1997; Putnam, 1993), more politically active in society (Quintelier, 2008; Verba, Schlozman, Brady, & Nie, 1993), more interested in social and political issues (Verba, Schlozman, & Brady, 1995; Warren, 2001), better in their physical and mental health (Wilson, 2000), more politically efficacious (Almond & Verba, 1963), and more tolerant of minorities (Jennings & Stoker, 2004). Furthermore, youths’ involvement in associations has been shown to have long-lasting effects on their civic and political engagement. Using longitudinal data, scholars have shown that membership in civic associations and volunteer work at an early age have an impact on adult civic engagement and political participation (Feinstein, Bynner, & Duckworth, 2006; Hanks, 1981; Jennings & Stoker, 2004; McFarland & Thomas, 2006). In short, research on political socialization and social capital has repeatedly shown positive links between membership in associations and various political attitudes vital to the acceptance and functioning of democracy and society at large.

Given all these positive outcomes, civil society organizations and associations have attracted much attention over the past two decades (van Deth, Maraffi, & Newton, 1999; Warren, 1999). Yet, despite intensifying theoretical debate and many empirical insights into the role of associations and volunteer work in a democracy, it is still uncertain how civic engagement stimulates this large number of positive outcomes. Or, to put it differently, it is not clear whether it is civic engagement that makes citizens internalize certain democratic attitudes and values, or whether citizens who are already democratically-minded choose to engage in associational membership and volunteer work. The purpose of this study, therefore, is to contribute to a better understanding of the relation between civic engagement – which we define as associational membership and volunteer work – on one hand, and citizens’ development of social trust and political interest on the other. We will explore this link during adolescence because it is a time in peoples’ lives that is a formative and critical period for the emergence and development of political attitudes (Hatemi et al., 2009; Hooghe & Wilkenfeld, 2008). Consequently, the main question of this study is as follows:

Research Question 1: Does civic engagement affect youths’ political interest and social trust, or do civic engagement activities attract youths who are already politically interested or trustful?

Socialization and Self-Selection Effects

Overall, two main theoretical perspectives dominate current research on the role of civic engagement in democracy: socialization (i.e., civic engagement leads to the development of democratic political orientations) and self-selection (i.e., already democratically minded people are more likely to be involved in civic engagement).

The current study focuses on two vital democratic orientations: political interest and social trust. Political interest, that is, a citizen’s “intrinsic motivation to engage in politics” (Shani, 2009, p. 2), has been suggested to be a central element in young people’s political development, and it seems to play an important role in citizens’ involvement in politics and civic life (Best & Krueger, 2005; Fieldhouse, Tranmer, & Russell, 2007; Verba et al., 1995). At the same time, we know from previous research that political interest is extremely stable in adulthood, which suggests that longitudinal research would fruitfully focus on adolescence (Prior, 2010). Thus, by turning our attention to the development of political interest in adolescence, a time when changes are more likely to occur, we pave the way for a more comprehensive understanding of how it is related to civic engagement. Our interest in studying social trust stems from the fact that it is a central concept in social capital theory and, in research, on the role that associations and voluntary work play in generating political attitudes (Newton, 1999; Putnam, 2000).

According to the socialization perspective, dating back to de Tocqueville (1835/1969), positive face-to-face interactions taking place within associations (irrespective of whether they are political or nonpolitical organizations) between people of diverse backgrounds and characteristics (e.g., religion, ethnicity, political preferences) will foster a number of democratic and cooperative values and norms (Brehm & Rahn, 1997; Putnam, 1993, 2000). Hence, from this point of view, associations function as “learning schools for democracy” with the power to exert influence on their members’ political orientations. As mentioned above, several empirical studies have found support for the socialization perspective. Previous research on adults has shown that, compared with nonmembers and nonvolunteers, members of associations and people performing volunteer work are more supportive of democratic norms and values, more interested in politics and knowledgeable about political affairs, more politically efficacious, more tolerant and trustful, and more actively involved in politics (Almond & Verba, 1963; Brehm & Rahn, 1997; Delhey & Newton, 2003; Hooghe, 2003; van Deth, 2006; van Ingen & Bekkers, 2015; Verba et al., 1995; Wollebæk & Selle, 2002). In brief, according to the socialization perspective, we learn “the habits of the heart” (Bellah, 1985) regarding important values and norms, such as trust, reciprocity, and cooperation, by having close contact with people of diverse backgrounds through associational membership and voluntarism.

Over the years, however, the socialization perspective has been widely questioned, and from different angles. One of the difficulties concerns the direction of causality between civic engagement and the development of political attitudes and orientations. Let us illustrate with the relation between associational membership and social trust. Whereas Putnam (1995) stresses that “the causation flows from joining to trusting” (p. 666), others have maintained that the flow may be in the other direction (cf. Newton, 1999); that is, associations attract already trustful people. Furthermore, some scholars have argued that there may be “a dark side of civic engagement” (Fiorina, 1999), and that some kinds of civic involvement may in fact contribute to the development of “unsocial capital” (Keane, 1998; Levi, 1996). In addition, when it comes to the relation between associational membership and social trust, empirical findings vary vastly and are sometimes inconsistent with theoretical expectations from the socialization perspective. In some studies, for example, the relation is not significant, or rather limited, after controlling for relevant covariates (e.g., Badescu & Neller, 2007; Knack & Keefer, 1997; Nannestad, 2008; Quintelier, 2012; Whiteley, 1999). Consequently, the core assumptions underlying the socialization perspective within research on civic engagement have been questioned and challenged over the last two decades (Uslaner, 2000).

Another line of research has switched attention from socialization to self-selection effects (Newton, 1999; Stolle, 1998; Whiteley, 1999). According to proponents of this opposite line of thinking, adult citizens who already have relatively high levels of democratic orientations and attitudes are more likely to join associations and volunteer in the first place (Sønderskov, 2011; Stoll, 2001; Stolle, 1998; Uslaner & Brown, 2005; van Ingen & Bekkers, 2015).

In line with the self-selection perspective, studies have shown that being a member of an association does not produce social trust; instead, people who already have a certain level of social trust choose to join associations (Jennings & Stoker, 2004; Stolle, 1998). With regard to political interest, previous research using cross-sectional data has shown that citizens with greater interest in politics are more likely to volunteer and be members of voluntary associations (Bekkers, 2005). Similarly, using a panel of 16- and 18-year-old Belgian students, Quintelier (2012) has shown that some degrees of both self-selection and socialization seem to be at work with regard to political interest. In other words, youths already interested in politics seem to be more likely to become members of associations, such as school councils and political parties (Quintelier, 2012), within which they then develop an even greater political interest. To sum up, studies supporting the self-selection approach support the idea that citizens self-select into associations and voluntarism rather than associations and volunteer work socializing them into certain norms and attitudes. Thus, these studies’ positive correlations between civic engagement and various democratic orientations are explained by the presence of self-selection.

Limitations of Previous Research

Although previous studies provide a range of interesting theoretical and empirical insights into the relation between association membership and democratic orientations, they suffer from two important limitations. First, despite a large number of studies, the conflict between advocates of the socialization and self-selection perspectives seems to remain unsolved. The main reason is that many empirical studies, especially in the social capital tradition, have based their findings and conclusions on cross-sectional, rather than longitudinal, data (for exceptions, see Bekkers, 2012; Quintelier, 2012; van Ingen & Bekkers, 2015). Due to their cross-sectional nature, many studies have not been able to provide reliable answers to the question of the direction of causality, that is, whether civic engagement generates various democratic attitudes or whether people self-select into civic engagement based on their already present democratic characteristics.

This is not to say, however, that there are no longitudinal studies at all. For example, within both positive youth development research and extracurricular activity research, there is a long tradition of using panel data to study the long-term effects of prior civic engagement. Studies within these two research fields clearly show increases in positive outcomes and decreases in negative outcomes in young adulthood, as a consequence of participation in civic associations in adolescence. For example, these bodies of knowledge show that voluntary work and participation in extracurricular school activities are linked to lower depression (Fredricks & Eccles, 2005), higher levels of self-esteem (Coleman, 1961; Holland & Andre, 1987), less drug use (Darling, 2005), lower levels of delinquency (Holland & Andre, 1987; Landers & Landers, 1978), and higher academic achievement (Darling, 2005; Eidsmoe, 1964; Holland & Andre, 1987). Participation in various voluntary activities in adolescence has also been shown to increase civic engagement and political participation in adulthood (Fredricks & Eccles, 2006; Hanks, 1981; Holland & Andre, 1987). The focus of these studies, however, is often on the long-term effects of civic engagement, and their major drawback is that they seldom take into account self-selection effects. In sum, studies examining the socializing effect of civic engagement either have not considered self-selection or have tried to disentangle the socialization–selection conundrum using cross-sectional data. Hence, studies are needed that adequately separate the influence of socialization from self-selection.

A second limitation of previous research on this topic is its strong focus on adult cohorts (Stolle & Hooghe, 2004). There are, however, good reasons to focus on adolescent cohorts. An increasing number of studies show that political orientations, as well as behaviors, are shaped early in life (Flanagan & Sherrod, 1998; Jennings & Niemi, 1981; Torney-Purta, Lehmann, Oswald, & Schulz, 2001). For example, using longitudinal data on twins, recent genetic research has shown that political attitudes accumulate markedly between the ages of 9 and 17 (Hatemi et al., 2009). Furthermore, other studies have shown that, during adolescence, people actively develop their political orientations (Abramson, 1972; Hooghe & Wilkenfeld, 2008), and that many political orientations, especially political interest and social trust, are highly stable among adults (Krosnick & Alwin, 1989; Prior, 2010; Sherrod, Torney-Purta, & Flanagan, 2010; Stolle & Hooghe, 2004) and also among older youth (Hooghe & Wilkenfeld, 2008; Shani, 2009; Stolle & Hooghe, 2004). In sum, during adolescence, many critical changes occur in individual’s political development. Therefore, research on social capital explicitly calls for more research on youth samples (cf. Stolle & Hooghe, 2004; Uslaner, 2002). All in all, to gain better insights into how various political orientations develop and change, research should pay greater attention to the study of adolescents.

The Current Study

The current study makes use of three waves of data to try to overcome some of the limitations of previous studies and to contribute to the research on civic engagement and political attitudes. First, by using a longitudinal design, we are able to analyze empirically both socialization and self-selection effects. By examining the effects of associational membership and voluntary action on social trust and political interest in both, between-effects and fixed-effects regressions, our analyses will provide us with longitudinally modeled effects that can be compared to deduce whether we are facing socialization effects, selection effects, or both.

Second, in contrast to many studies within this research field, this study focuses on adolescents. As noted earlier, adolescence is a unique phase of life, during which young people face many developmental challenges, and previous research indicates that growth of political competencies during adolescence may be important for the long-term development of adult political attitudes and participation (cf. Verba et al., 1995). By studying adolescents, rather than older people, we can examine socialization as well as self-selection effects during a period where many changes occur, especially with regard to adolescents’ political development (Niemi & Hepburn, 1995).

Furthermore, by jointly focusing on political interest and social trust, we will be enable to gain a better understanding of role of civic engagement for two relatively distinct types of civic competencies. To put it differently, from a theoretical point of view, it will be interesting to investigate whether associational membership and voluntarism are distinctly related to social trust and political interest, or whether the mechanism (self-selection or socialization) is at work irrespectively of if examining social trust or political interest. In addition, and related to that point, both political interest and social trust are very essential preconditions for social activity and the functioning of democracy at large. All in all, examining the role of civic engagement in relation to political interest and social trust will add useful insights to knowledge about the development of political orientations.

Method

Participants and Procedure

The current study makes use of three waves of questionnaire data from Political Socialization Program (PSP), a longitudinal project on political socialization (Amnå, Ekström, Kerr, & Stattin, 2009). This community-based and cohort-sequential project is being carried out in a medium-sized Swedish city with a total population of about 135,000. The motive for choosing this specific city is that it is quire representative of national demographic characteristics. According to official statistics (Statistics Sweden, 2012), in 2010, when the first data collection was performed, the city was similar to the national average in annual mean income (234,058 Swedish Crowns/person compared with 237,186 Swedish Crowns/person for the whole country), the rate of unemployment (8.6% compared with 8.0% for the whole country), and the percentage of foreign-born persons (see Table 1).

Table 1.

Comparing the Share of Foreign Respondents.

	Sweden(15-24 years of age)^a	Sampled city(15-24 years of age)^a	Analytical sample
Share of respondents with foreign background	20%	24%	20.1%

Collected from Statistics Sweden (2012).

Thirteen schools within the city were selected for participation in the study: 10 of the city’s 14 junior high schools and three of its seven senior high schools. The 13 schools were strategically selected to represent both public and private schools and students’ diverse socioeconomic and ethnic backgrounds. All seventh graders in the junior high schools and first year students in the senior high schools were targeted in the study. The target sample included approximately 2,000 students at each time point. The participation rate ranged between 85% and 89% across the three time points. In the analytical sample, we only included respondents for whom we had data on associational membership on all three measurement occasions. The final analytical sample, then, included 1,050 participants (48.7% girls; M_age = 14.8 at the first time point).

The respondents filled out the questionnaires, approved in advance by the Regional Ethics Board in Uppsala, during regular school hours in their classrooms. To meet the possible objection that the presence of teachers might influence the responses of students, we ensured that no teachers were in the classrooms during data collection. Trained research assistants distributed the questionnaires and informed the students that participation in the study was voluntary; the students were also assured of the confidentiality of their responses. No student was paid for participating in the study, but each class received a contribution to their class fund.

Attrition Analyses

To establish whether dropouts differed from those who participated, we applied logistic regression analysis. The missing from T1 to T2 and from T2 to T3 were analyzed using all study variables (i.e., gender, age, socioeconomic status [SES], parents’ immigrant status, civic engagement, social trust, and political interest) from each previous time point as predictors. From T1 to T2, the only significant predictor of adolescent attrition was parents’ immigrant status. Youths with missing data at Time 2 were less likely to be from Swedish/Nordic backgrounds (odds ratio [OR] = 0.76, p < .05). Moreover, youths dropping out from T2 to T3 were less likely to be from Swedish/Nordic backgrounds (OR = 0.68, p < .05) and more likely to be younger (OR = 0.88, p < .05). However, the Nagelkerke R² for the whole block of variables were .02 and .01, respectively, meaning that participants and dropouts were not substantially different and indicating that attrition did not have a substantial impact on the results.

Measures

Political interest

In the current study, political interest was measured using two items: “How interested are you in politics?” and “How interested are you in what is happening in society?” (Amnå & Ekman, 2014), with the response scale ranging from 1 (not at all interested) to 5 (very interested). The Pearson’s correlations between these two items were .53, .62, and .67, at time points 1, 2, and 3, respectively. A mean index was created with higher scores indicating more interest.

Generalized social trust

Social trust was measured using two items: “Most people can be trusted” and “Most people are fair and do not take advantage of you” (Flanagan & Stout, 2010; Flanagan, Syvertsen, & Stout, 2007). Participants responded to the statements on a five-point scale, ranging from 1 (do not agree at all) to 5 (completely agree). The Pearson’s correlations between these two items were .68, .67, and .70, at time points 1, 2, and 3, respectively. A mean index was created with higher scores indicating more trust.

Membership of associations

This study assessed membership in four politically and semipolitically oriented associations: political associations, associations for peace or human rights (such as the Peace Movement, Amnesty International, or the like), environmental associations, and student councils (cf. Wollebæk & Selle, 2002). These associations are characterized by being nonprofit organizations that people join or leave voluntarily. Respondents who indicated that they were a member of at least one association within these categories were classified as members. Those who indicated that they had not been a member of any association of this kind at any time point were labeled nonmembers.

Voluntarism

We measured voluntarism by asking the respondents to report on their annual involvement in voluntary activities. The question was “Have you done any of the following during the last 12 months?” followed by “Worked voluntarily for a good cause?” The response options were “no,” “yes, occasionally,” and “yes, several times.” Positive responses were, thereafter, collapsed so that involvement was coded 1 and noninvolvement was coded 0.

Control variables

To take into account the possible effects of some other important factors, we included a number of conventionally used control variables: age, gender, parents’ immigrant status, and subjective SES. Gender was coded as 1 (boys) and 0 (girls). Parents’ immigrant status was measured as a dichotomous variable (1 = immigrant—neither of the respondent’s parents was born in Sweden/other Nordic countries; 0 = Swedish—at least one of the respondent’s parents was born in Sweden). Subjective SES was measured by the following five items: “If you want things that cost a lot of money (for example, a computer, skateboard, cell phone), can your parents afford to buy them if you want them?” “If you compare yourself with the others in your class, do you have more or less money to buy things?” “Does your family have more or less money than other families where you are living?” “How often do you and your family go on vacation?” and “How is personal finance in your family?” Cronbach’s alpha reliabilities for this scale were .80, .83, and .83, at time points 1, 2 and 3, respectively.

Analytical Strategy

We used several statistical methods to address our research questions. First, we performed a series of t tests to examine the differences between members/volunteers and nonmembers/nonvolunteers regarding their levels of political interest and generalized social trust.

Second, to examine our competing hypotheses about self-selection and socialization effects, we examined between-effects regressions as well as fixed-effects and first-difference (change scores) regressions. The between-effects regression averages respondents’ scores for both dependent variables (social trust and political interest) and independent variables (e.g., civic engagement) across the three time points. These scores are used to examine civic engagement effects on social trust and political interest based on between-person variation only. The effects from this way of procedure can thereby be directly compared with the cross-sectional correlations from previous research.

From prior research, we know that outcomes of cross-sectional data might be affected by self-selection. As mentioned above, if, for example, associational membership is the cause of already high social trust, then failure to control for such fixed background characteristics would result in biased estimates. But as a technique that is exclusively built on within-person variation, fixed-effects regressions can, to some extent, alleviate some of this problem (Allison, 2009). What the fixed-effects models do is to take into account time-invariant unobserved heterogeneity that can be correlated with observed variables. That is, the fixed-effects regressions account for variances from unobserved variables and remove biased inferences from correlation in the error terms that does not change over time (Wooldridge, 2000). Stated differently, the coefficients from a fixed regression are controlled for time-invariant selection effects without measuring them. To take selection effects into account and at the same time to use all three waves of data in this study, the average score of each respondent’s score is subtracted from each respondent’s score at each time point (for all respondents and for all variables in the model). With these “within transformations,” the data now measure how respondents deviate from their own average in all waves, and subtracting respondents’ averages for each variable implicitly does the same subtraction to the error term. In addition, by adding a dummy variable for each wave to the equation, the model captures the average effects of variables that do change over time. In sum, fixed-effects regressions allow us to better understand changes over time in our dependent variables and also the impact of phenomena that are usually hard to measure (e.g., selection).

The first-difference regressions are also based on within-person variation. When using this technique, (following the procedure of van Ingen and Bekkers (2015)), we examined transitions in associational membership and voluntarism by relating respondents joining an association or volunteering first at a subsequent time point to those who remained uninvolved at the previous point in time. The fixed-effects regression examined the relation between changes in civic engagement and changes in social trust and political interest. For example, the fixed-effects regressions can answer whether we can show that, on average, joining an association is related to increases in social trust when compared with not joining an association. Given their intuitive character, the first-difference regressions offer ideal tests of socialization. However, as they model only (a) those who joined associations or started volunteering and (b) uninvolved or nonvolunteers (and not those who exit associations or quit volunteering and those who remain involved), this is a less powerful technique compared with the fixed-effects regressions. We examined our models using Stata 13.0. The number of schools from which the respondents were sampled is rather small. Hence, to take into account potential autocorrelation, we used a school fixed-effects approach to adjust the standard errors.

Beyond associational membership and volunteering, factors such as gender, age, SES, and immigrant status have also been shown to be important for adolescents’ development of social trust and political interest (e.g., Quintelier, 2012; Wollebæk & Selle, 2002). We, therefore, controlled for these background characteristics in all our regressions.

Results

Descriptive Statistics

Before proceeding to the main analyses, we report the descriptive statistics on associational voluntarism and membership, and also the means and standard deviations of political interest and generalized social trust (Table 2). As can be seen, the highest frequency of members can be found in student councils, followed by political associations. The overall percentage of membership in any of the four associations was 20.1%, 13.2%, and 15.3% in 2010, 2011, and 2012, respectively. A further examination of Table 2 shows that the adolescents became slightly more interested in politics, but less trustful over the 3 years.

Table 2.

Percentages and Frequencies of Civic Engagement Over Time, and Means and Standards Deviations of Political Interest and Social Trust Over Time.

	T1 (2010)	T2 (2011)	T3 (2012)
Civic engagement
Voluntarism	54.6 (569)	46.6 (478)	51.3 (536)
Associational membership
Political association	2.6 (27)	4.0 (42)	4.0 (42)
Association for peace or human rights	3.0 (32)	1.4 (15)	1.7 (18)
Environmental association	1.6 (17)	1.2 (13)	1.3 (14)
Student council	15.0 (157)	13.5 (142)	10.6 (111)
Member of any of the types of associations above	19.1 (201)	17.5 (184)	14.6 (153)
Dependent variables, M (SD)
Political interest	2.92 (0.94)	3.03 (0.99)	3.03 (1.00)
Generalized social trust	3.19 (0.86)	3.13 (0.86)	2.92 (0.94)

Is Civic Engagement Associated With Higher Levels of Social Trust and Political Interest?

To reiterate, the main question in this study is as follows: Does civic engagement affect youths’ political interest and social trust, or does civic engagement attract youths who are already politically interested or trustful? However, before proceeding to this question, we need to examine one of the main assumptions underlying the social capital approach, namely, that association members and those involved in volunteering should be more politically interested and socially trustful than nonmembers and those not involved. As can be seen in Table 3, across the three time points, members and involved youths were found to be significantly more interested in politics than nonmembers and non-involved youths. The Cohen’s d indicated medium to large differences (Cohen, 1992). These results are in line with the findings of previous studies. Based on similar results, many previous studies have drawn the conclusion that associations generate political interest, among other outcomes. Yet, it is important to underline that these results do not per se indicate the direction of the relation between civic engagement and political interest. To examine the direction of the relation and to address this study’s specific research question, we have to examine the socialization–selection puzzle further.

Table 3.

Standardized Mean Scores on Political Interest and Generalized Social Trust According to Membership and Voluntarism.

			Nonmember/not involved	Member/involved	t value	Cohen’s d
Political interest	2010 (T1)	Associational membership	−.07	.28	4.47***	.28
	2010 (T1)	Voluntarism	−.20	.17	6.23***	.37
	2011 (T2)	Associational membership	−.08	.36	5.40***	.33
	2011 (T2)	Voluntarism	−.18	.22	6.65***	.42
	2012 (T3)	Associational membership	−.08	.45	6.07***	.37
	2012 (T3)	Voluntarism	−.15	.14	4.82***	.30
Generalized social trust	2010 (T1)	Associational membership	−.03	.11	1.69	.10
	2010 (T1)	Voluntarism	−.05	.05	1.60	.10
	2011 (T2)	Associational membership	−.01	.04	0.59	.04
	2011 (T2)	Voluntarism	−.09	.09	2.88**	.18
	2012 (T3)	Associational membership	−.01	.08	1.13	.07
	2012 (T3)	Voluntarism	−.05	.04	1.32	.08

p < .01. ***p < .001.

With regard to social trust, we found only one significant difference indicating that involvement in civic engagement is associated with higher values of social trust. This result seems to run counter to the presumption of socialization theorists that members or volunteers should display higher levels of social trust compared with nonmembers and non-involved. However, these results are much in line with previous studies reporting a nonsignificant or only a weak relation between social trust and civic engagement (e.g., Badescu & Neller, 2007; Brehm & Rahn, 1997; Claibourn & Martin, 2000; Knack & Keefer, 1997; Newton, 1999; Wollebæk & Strømsnes, 2008). In sum, our initial analyses show that civic engagement seems to be connected to higher political interest. However, with regard to generalized social trust, this link seems to be nonexistent or weak at best.

An additional way to examine differences between respondents is provided by the between-effects regressions (See Tables 4 and 5). However, by controlling for gender, age, SES, and immigrant status, the between-effects regressions account for a stricter test than the initial t tests. Without the controls (not included in the tables), we found a positive effect of volunteering on political interest (b = 0.15, p < .001) but not on social trust (b = 0.07, p > .05). For membership, there was also a positive effect on political interest (b = 0.22, p < .001) but again, no effect on social trust (b = 0.06, p > .05). These results were in line with our initial findings. Including the control variables generated stronger predictions of political interest from both modes of civic engagement. When predicting social trust, the inclusion of the controls made the effect from volunteering statistically significant, whereas the membership remained unable to predict social trust. Despite these controls, it is still possible that adolescents with higher values on social trust and political interest have self-selected into civic engagement. It is of course possible to control for further variables. However, as previously noted by van Ingen and Bekkers (2015), because all relevant control variables are unlikely to be in the data, “[a] more efficient method is to use fixed-effects regressions, which also control for unobserved heterogeneity caused by time-invariant factors” (p. 286).

Table 4.

Regression of Social Trust on Volunteering and Membership.

	Social trust (volunteering)			Social trust (membership)
	Between-effects regression	Fixed-effects regression	First-difference regression	Between-effects regression	Fixed-effects regression	First-difference regression
Volunteering	0.12*	0.19*
Membership				0.03	0.03
Participation transition
Remain uninvolved (ref)			0			0
Entry/start			0.08			0.05
Exit/quit			0.02			0.05
Remain involved			−0.02			−0.06
Gender	0.15**	—	−0.06	0.14*	—	−0.06
Age	−0.06	0.03	−0.01	−0.06	0.03	−0.01
SES	0.12***	−0.06	−0.04	0.12***	−0.06	−0.04
Immigrant status	0.35***	−0.08	—	0.34***	−0.07	—

Note. The fixed-effects regressions were controlled for year of measurement (dummy variables). The first-difference regression was controlled for “equation number” (e.g., controlled for the time variable). SES = socioeconomic status.

p < .05. **p < .01. ***p < .001.

Table 5.

Regression of Political Interest on Volunteering and Membership.

	Political interest (volunteering)			Political interest (membership)
	Between-effects regression	Fixed-effects regression	First-difference regression	Between-effects regression	Fixed-effects regression	First-difference regression
Volunteering	0.38***	0.12
Membership				0.37***	0.19
Participation transition
Remain uninvolved (ref)			0			0
Entry/start			0.11			0.21*
Exit/quit			−0.08			−0.26*
Remain involved			−0.03			−0.08
Gender	0.14*	—	0.03	0.10	—	0.02
Age	−0.01	−0.01	0.01	−0.02	−0.01	0.01
SES	0.08*	0.00	0.02	0.07	0.01	0.01
Immigrant status	−0.18**	0.26	—	−0.19*	0.23	—

p < .05. **p < .01. ***p < .001.

The Potentially Socializing Effect of Civic Engagement on Political Interest and Social Trust

We first turn to the fixed-effects regressions. For social trust (Table 4), fixed-effects regression generated similar effects of civic engagement as the between-effects regressions. Keeping in mind that our initial analysis did not show much differences on social trust when comparing members and nonmembers, and volunteers and nonvolunteers, these results still lend support to a situation in which socialization comes into play as the mechanism explaining the small gain for members and volunteers. Moreover, the observed associations between volunteering and membership, on one hand, and political interest, on the other, in our initial analyses (using t tests and between-effects regression) were not significant in the fixed-effect regression framework (Table 5). Hence, based on the findings of the fixed-effects regressions on political interest, the fact that those involved in civic engagement are high in political interest seems foremost to be the result of self-selection.

Turning to our strictest test of the hypothesis that civic engagement enhances social trust and political interest, the first-difference regressions seemed to go somewhat in different directions compared with the findings of the fixed-effects regressions: In contrast to the positive effect of voluntarism found in the fixed-effects model on social trust, the first-difference model did not show that changes in civic engagement predicted changes in social trust. In addition, just as in the fixed-effects regression, there was no significant effect of voluntarism on political interest (b = 0.11, p = .14). With regard to membership in associations, this mode of civic engagement had no effect on social trust (b = 0.05, p = .58) but it was positively linked to political interest (b = 0.21, p = .02). Hence, the indication from the fixed-effects regression that civic engagement may socialize people into more trustful individuals was not supported by the first-difference regression. However, the positive effect of associational membership on political interest indicated that associational membership may have a socializing effect on adolescents’ political interest.

Discussion

The aim of the current study was to gain a better understanding of the relation between two measures of civic engagement (associational membership and volunteering) and two important political orientations: political interest and social trust. The main research question we attempted to address was whether civic engagement affect youths’ political interest and social trust, or does civic engagement attract youths who are already politically interested or socially trustful?

The findings from this study corroborate with previous studies on social capital suggesting a nonsignificant or limited relation between civic engagement and social trust (e.g., Nannestad, 2008; Quintelier, 2012). Associational membership seemed not to contribute anything to higher levels of social trust. Voluntarism, a more demanding mode of civic engagement, demonstrated a weak and inconclusive effect on social trust. Thus, it seems unlikely, although it cannot be ruled out, that adolescents become more trustful as a consequence of volunteering. All in all, these results suggest that we have to direct our attention to and further analyze other factors in order to gain a better understanding of how social trust develops during adolescence.

With regard to political interest, the socialization–selection conundrum seems to be contingent on the particular mode of civic engagement in which adolescents choose to engage: for volunteering, the cross-sectional effects of the between-effects regression diminished in the fixed-effects regression. Also, there was no positive effect of voluntarism on the first-difference framework. Together, these results suggested that there was very little support for an overtime socializing effect of voluntarism on political interest. Instead, the differences between involvement in civic engagement and not in our initial analysis were likely the result of self-selection. However, although the associations from the between-effects regressions were not present in the fixed-effects framework, we saw that our strictest test of the socialization hypothesis—the first-difference regression model—provided support for an influence effect. Nonetheless, as the socialization effect of associational membership on political interest evident in the first-difference model was not not present in the fixed-effects model, a socialization effect is not likely to entirely explain the higher levels of political interest among those involved in civic engagement. Therefore, part of the explanation here seems to be self-selection.

As such, our results correspond to the findings of prior research (e.g., Quintelier, 2012) where both self-selection and socialization seem to play a role with regard to political interest. As mentioned in the “Introduction” section, the idea that civic engagement has advantageous outcomes for those involved is not new. However, the presence of a self-selective mechanism harmonizes with an important message often overlooked in previous research: Young people are not completely passive recipients and objects of socializing agents and institutions. Rather, they seem also to be active agents, meaning that they are actively involved in their surrounding worlds and their own political development (cf. Amnå, Ekström, Kerr, & Stattin, 2009; McDevitt & Chaffee, 2002). A consequence of this particular finding may be that the socialization perspective that currently dominates the debate over social capital development is more frequently combined with an active agent perspective. Such a way ahead may provide a more accurate understanding of how citizens’ involvement in societal and associational life corresponds to their development of political orientations.

Together, the results from analyzing these longitudinal data contribute to our understanding of young people’s development of political attitudes by challenging the assumption that associations function as schools of democracy. Our aim was to better understand whether socialization effects or selection effects account for observed differences among those who are involved in civic engagement and those who are not. Some of our analyses leave us with findings that depict a relation between civic engagement and the development of political orientations, which is still very much unclear. Given that our strictest test of the socialization hypothesis provided support for an enhancing effect of associational membership on political interest, it may be that there is a dual process present for political interest, both socialization and selection. What seems much clearer from our analyses, though, is that associational membership has no enhancing effect on social trust and that volunteering has no enhancing effect on political interest. Hence, we can conclude that the mode of civic engagement is important for how different political orientations develop.

The current study has some limitations and strengths that warrant attention. A first potential limitation concerns whether findings from our sample can be generalized to a wider population. As noted above, the data used in this study come from a community-based sample of adolescents in a medium-sized Swedish city. It might, therefore, be argued that findings cannot be generalized from this particular context. Ideally, randomly selected samples drawn from the general population are preferable. Nonetheless, at a macrolevel, there are similar patterns in the demographic characteristics used here and the data found at the national level (e.g., income, parents’ immigrant status, and unemployment). At the same time, the research project from which the data in the current study were obtained is based on a community-based approach because it is more suitable and practically feasible given the longitudinal design of the project. Bearing this in mind, the data used in the study seem to be representative of the general population, and have adequate validity. Related to that contention, it might also be argued that the findings cannot be generalized to other democratic countries because of the specific institutional arrangements and contextual settings particular to Sweden. However, even if the purely empirical findings in the current study may differ from those in other countries, it is reasonable to presume that the theoretical contributions are not as context specific and dependent. To clarify further, even if, for example, levels of social trust, political interest, or percentages of people involved in associations were to vary across democratic societies, there is good reason to believe that the patterns of results regarding the self-selection and socializations effects might be similar in various political settings. Still, to ensure that our findings are generalizable to other political settings, comparative studies are needed. Second, this study has focused on political and semi-public associations rather than other types. This choice may constitute a limitation because it is possible that the findings differ for other types of associations. However, given the fact that, in the current study, we were interested in the relation of membership in associations to political interest and social trust, the selection of associations seems justifiable.

The study has, at the same time, several strengths. First and foremost, by adopting a 3-wave longitudinal design, separated at 1-year intervals, we were able to take into account both socialization and self-selection effects. Additionally, we employed fixed-effects regression which thereby could alleviate some of the difficulties of statistically controlling for phenomenon that cannot be observed (Allison, 2009). Consequently, this study has generated further insights into the ongoing debate about civic engagement and development of political orientations within social capital literature, and thereby exceeds research findings in the field based solely on cross-sectional data. A second strength concerns the focus on adolescents. As noted previously, research in developmental psychology as well as in political science has demonstrated that the attitudinal patterns of adolescents remain relatively stable over the life course. Therefore, adolescence seems to be a good starting point for an exploration of how membership of associations relates to political interest and social trust. Taken as a whole, despite its limitations, the study presents important insights into research on social capital. It provides evidence that young people, even at an early age, self-select into associations, and that social trust is not related to membership in associations during adolescence.

Much remains to be done to establish a more comprehensive understanding of the role of associations in young people’s development of democratic orientations. Further research may benefit from embracing a more agency-oriented approach to young people by using research designs that allow different directions of effects to be examined. To embrace such an approach is likely to generate important information about youth’s participation in civic engagement and more accurate knowledge for authorities formulating youth policies related to political and democratic involvement.

Footnotes

Authors’ Note

This study was made possible by access to data from the Political Socialization Program, a longitudinal research program at Youth & Society (YeS) at Örebro University, Sweden. Professors Erik Amnå, Mats Ekström, Margaret Kerr, and Håkan Stattin were responsible for the planning, implementation, and financing of the collection of data.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The data collection and the study were supported by grants from Riksbankens Jubileumsfond.

Author Biographies

Viktor Dahl earned his PhD in Political Science at the Örebro University, Sweden. His research concerns political socialization and the development of young people’s political values, attitudes, and behavior. Specifically, he is interested in the potential of social networks on its members’ political orientations.

Ali Abdelzadeh defended his PhD thesis “Transcending Dichotomies: The Many Faces of Youth Dissatisfaction in Democracy” in 2014. Since then, he has continued researching how young people organize around political issues, often with the aim of understanding their political participation and other democratic competencies.

References

Abramson

P. R.

(1972). Political efficacy and political trust among Black schoolchildren: Two explanations. Journal of Politics, 34, 1243-1269.

Allison

P. D.

(2009). Fixed effects regression models. Los Angeles, CA: Sage.

Almond

G. A.

Verba

(1963). The civic culture: Political attitudes and democracy in five Nations. Princeton, NJ: Princeton University Press.

Amnå

Ekman

(2014). Standby citizens: Diverse faces of political passivity. European Political Science Review, 6, 261-281.

Amnå

Ekström

Kerr

Stattin

(2009). Political socialization and human agency: The development of civic engagement from adolescence to adulthood. Statsvetenskaplig Tidskrift, 111, 27-40.

Badescu

Neller

(2007). Explaining associational involvement. In van Deth

J. W.

Montero Jose

Westholm

(Eds.), Citizenship and involvement in European democracies: A comparative analysis (pp. 158-187). New York, NY: Routledge.

Bekkers

(2005). Participation in voluntary associations: Relations with resources, personality, and political values. Political Psychology, 26, 439-454.

Bekkers

(2012). Trust and volunteering: Selection or causation? Evidence from a 4 year panel study. Political Behavior, 34, 225-247.

Bellah

R. N.

(1985). Habits of the heart: Individualism and commitment in American life. Oakland: University of California Press.

10.

Best

S. J.

Krueger

B. S.

(2005). Analyzing the representativeness of Internet political participation. Political Behavior, 27, 183-216.

11.

Brehm

Rahn

(1997). Individual-level evidence for the causes and consequences of social capital. American Journal of Political Science, 41, 999-1023.

12.

Claibourn

M. P.

Martin

P. S.

(2000). Trusting and joining? An empirical test of the reciprocal nature of social capital. Political Behavior, 22, 267-291.

13.

Cohen

(1992). A power primer. Psychological bulletin, 112, 155.

14.

Coleman

J. S.

(1961). The adolescent society: The social life of the teenager and its impact on education. New York, NY: Free Press of Glencoe.

15.

Darling

(2005). Participation in extracurricular activities and adolescent adjustment: Cross-sectional and longitudinal findings. Journal of Youth and Adolescence, 34, 493-505.

16.

Delhey

Newton

(2003). Who trusts? The origins of social trust in seven societies. European Societies, 5, 93-137.

17.

de Tocqueville

. (1969). Democracy in America. Garden City: Doubleday. (Original work published 1835)

18.

Eidsmoe

R. M.

(1964). High school athletes are brighter. Journal of Health, Physical Education, Recreation, 35(5), 53-54.

19.

Feinstein

Bynner

Duckworth

(2006). Young people’s leisure contexts and their relation to adult outcomes. Journal of Youth Studies, 9, 305-327.

20.

Fieldhouse

Tranmer

Russell

(2007). Something about young people or something about elections? Electoral participation of young people in Europe: Evidence from a multilevel analysis of the European social survey. European Journal of Political Research, 46, 797-822.

21.

Fiorina

M. P.

(1999). Extreme voices: A dark side of civic engagement. In Skocpol

Fiorina

M. P.

(Eds.), Civic engagement in American democracy (pp. 395-426). Washington, DC: Brookings Institution Press.

22.

Flanagan

C. A.

Sherrod

L. R.

(1998). Youth political development: An introduction. Journal of Social Issues, 54, 447-456.

23.

Flanagan

C. A.

Stout

(2010). Developmental patterns of social trust between early and late adolescence: Age and school climate effects. Journal of Research on Adolescence, 20, 748-773.

24.

Flanagan

C. A.

Syvertsen

A. K.

Stout

(2007). Civic measurement models: Tapping adolescents’ civic engagement. Washington, DC: Center for Information and Research on Civic Learning and Engagement.

25.

Fredricks

J. A.

Eccles

J. S.

(2005). Developmental benefits of extracurricular involvement: Do peer characteristics mediate the link between activities and youth outcomes? Journal of Youth and Adolescence, 34, 507-520.

26.

Fredricks

J. A.

Eccles

J. S.

(2006). Is extracurricular participation associated with beneficial outcomes? Concurrent and longitudinal relations. Developmental Psychology, 42, 698-713.

27.

Hanks

(1981). Youth, voluntary associations and political socialization. Social Forces, 60, 211-223.

28.

Hatemi

P. K.

Funk

C. L.

Medland

S. E.

Maes

H. M.

Silberg

J. L.

Martin

N. G.

Eaves

L. J.

(2009). Genetic and environmental transmission of political attitudes over a life time. The Journal of Politics, 71, 1141-1156.

29.

Holland

Andre

(1987). Participation in extracurricular activities in secondary school: What is known, what needs to be known? Review of Educational Research, 57, 437-466.

30.

Hooghe

(2003). Value congruence and convergence within voluntary associations: Ethnocentrism in Belgian organizations. Political Behavior, 25, 151-175.

31.

Hooghe

Wilkenfeld

(2008). The stability of political attitudes and behaviors across adolescence and early adulthood: A comparison of survey data on adolescents and young adults in eight countries. Journal of Youth and Adolescence, 37, 155-167.

32.

Jennings

M. K.

Niemi

R. G.

(1981). Generations and politics: A panel study of youth adults and their parents. Princeton, NJ: Princeton University Press.

33.

Jennings

M. K.

Stoker

(2004). Social trust and civic engagement across time and generations. Acta Politica, 39, 342-379.

34.

Keane

(1998). Civil society: Old images, new visions. Cambridge: Polity.

35.

Knack

Keefer

(1997). Does social capital have an economic payoff? A cross-country investigation. The Quarterly Journal of Economics, 112, 1251-1288.

36.

Krosnick

J. A.

Alwin

D. F.

(1989). Aging and susceptibility to attitude change. Journal of Personality and Social Psychology, 57, 416-425.

37.

Landers

D. M.

Landers

D. M.

(1978). Socialization via interscholastic athletics: Its effects on delinquency. Sociology of Education, 51, 299-303.

38.

Levi

(1996). Social and unsocial capital: A review essay of Robert Putnam’s Making Democracy Work. Politics & Society, 24, 45-55.

39.

McDevitt

Chaffee

(2002). From top-down to trickle-up influence: Revisiting assumptions about the family in political socialization. Political Communication, 19, 281-301.

40.

McFarland

D. A.

Thomas

R. J.

(2006). Bowling young: How youth voluntary associations influence adult political participation. American Sociological Review, 71, 401-425.

41.

Nannestad

(2008). What have we learned about generalized trust, if anything? Annual Review of Political Science, 11, 413-436.

42.

Newton

(1999). Social and political trust in established democracies. In Norris

(Ed.), Critical citizens: Global support for democratic government (pp. 169-187). Oxford, UK: Oxford University Press.

43.

Niemi

R. G.

Hepburn

M. A.

(1995). The rebirth of political socialization. Perspectives on Political Science, 24, 7-16.

44.

Prior

(2010). You’ve either got it or you don’t? The stability of political interest over the life cycle. The Journal of Politics, 72, 747-766.

45.

Putnam

R. D.

(1993). Making democracy work: Civic traditions in modern Italy. Princeton, NJ: Princeton University Press.

46.

Putnam

R. D.

(1995). Bowling alone: America’s declining social capital. Journal of Democracy, 6(1), 65-78.

47.

Putnam

R. D.

(2000). Bowling alone: The collapse and revival of American community. New York, NY: Simon & Schuster.

48.

Quintelier

(2008). Who is politically active: The athlete, the scout member or the environmental activist? Young people, voluntary engagement and political participation. Acta Sociologica, 51, 355-370.

49.

Quintelier

(2012). Socialization or self-selection? Membership in deliberative associations and political attitudes. Nonprofit and Voluntary Sector Quarterly, 42, 174-192.

50.

Shani

(2009). On the origins of political interest (Doctoral dissertation). Princeton, NY: Princeton University.

51.

Sherrod

L. R.

Torney-Purta

Flanagan

C. A.

(Eds.). (2010). Handbook of research on civic engagement in youth. Hoboken, NJ: John Wiley.

52.

Sønderskov

K. M

. (2011). Does generalized social trust lead to associational membership? Unravelling a bowl of well-tossed spaghetti. European Sociological Review, 27, 419-434.

53.

Statistics Sweden. (2012). Antal personer med utländsk eller svensk bakgrund efter region [Number of persons with foreign or Swedish background by region]. Available from http://www.scb.se.

54.

Stoll

M. A.

(2001). Race, neighborhood poverty, and participation in voluntary associations. Sociological Forum, 16, 529-557.

55.

Stolle

(1998). Bowling together, bowling alone: The development of generalized trust in voluntary associations. Political Psychology, 19, 497-525.

56.

Stolle

Hooghe

(2004). The roots of social capital: Attitudinal and network mechanisms in the relation between youth and adult indicators of social capital. Acta Politica, 39, 422-441.

57.

Torney-Purta

Lehmann

Oswald

Schulz

(2001). Citizenship and education in twenty-eight countries: Civic knowledge and engagement at age fourteen. Amsterdam: International Association for the Evaluation of Educational Achievement (IEA) International.

58.

Uslaner

E. M.

(2000). Producing and Consuming Trust. Political Science Quarterly, 115, 569-590.

59.

Uslaner

E. M.

(2002). The moral foundations of trust. New York, NY: Cambridge University Press.

60.

Uslaner

E. M.

Brown

(2005). Inequality, trust, and civic engagement. American Politics Research, 33, 868-894.

61.

Uslaner

E. M.

Dekker

(2001). The “social” in social capital. In Dekker

Uslaner

E. M.

(Eds.), Social capital and participation everyday life. London, England: Routledge.

62.

van Deth

J. W

. (2006). Democracy and involvement: The benevolent aspects of social participation. In Torcal

Montero Jose

(Eds.), Political disaffection in contemporary democracies: Social capital, institutions and politics (pp. 101-129). London, England: Routledge.

63.

van Deth

J. W.

Maraffi

Newton

. (1999). Social capital and European democracy. London, England: Routledge.

64.

van Ingen

Bekkers

. (2015). Generalized trust through civic engagement? Evidence from five national panel studies. Political Psychology, 36, 277-294.

65.

Verba

Schlozman

K. L.

Brady

H. E.

(1995). Voice and equality: Civic voluntarism in American politics. Cambridge, MA: Harvard University Press.

66.

Verba

Schlozman

K. L.

Brady

H. E.

Nie

N. H.

(1993). Citizen activity: Who participates? What do they say? American Political Science Review, 87, 303-318.

67.

Warren

M. E.

(1999). Democracy and trust. Cambridge, UK: Cambridge University Press.

68.

Warren

M. E.

(2001). Democracy and association. Princeton, NJ: Princeton University Press.

69.

Whiteley

P. F.

(1999). The origins of social capital. In van Deth

J. W.

(Ed.), Social capital and European democracy (pp. 25-44). London, England: Routledge.

70.

Wilson

(2000). Volunteering. Annual review of sociology, 26, 215-240.

71.

Wollebæk

Selle

(2002). Does participation in voluntary associations contribute to social capital? The impact of intensity, scope, and type. Nonprofit and Voluntary Sector Quarterly, 31, 32-61.

72.

Wollebæk

Strømsnes

(2008). Voluntary associations, trust, and civic engagement: A multilevel approach. Nonprofit and Voluntary Sector Quarterly, 37, 249-263.

73.

Wooldridge

(2000). Introductory econometrics, a modern approach. Cincinnati, OH: South-Western Thompson Learning.