Adolescence is a critical developmental period for the formation of political identities and behaviors (Erikson, 1968; Flanagan & Gallay, 2014; Krosnick & Duane, 1989; Sears & Levy, 2003), including radical beliefs and actions (Barracosa & Cherney, 2025; Beelmann, 2020; 2026; Carlsson et al., 2020). However, the drivers of youth political radicalism remain insufficiently understood. Existing research focuses largely on adults, relies heavily on cross-sectional designs, and uses attitudinal rather than behavioral measures. Moreover, relatively few studies examine political experiences, such as political alienation, as antecedents of radicalization (Campelo et al., 2018; Jahnke et al., 2021; Vergani et al., 2018). <\/p>\n\n\n\n
Yet, recent decades have seen a global rise in political alienation, characterized by political powerlessness, dissatisfaction, and perceptions that elected officials cannot be trusted and do not care about citizens\u2019 views (Pew Research Center, 2024; Public Agenda, 2022; Valgar\u00f0sson et al., 2025). These trends are particularly pronounced among young people, who increasingly report feeling disconnected from political systems and lacking meaningful influence within them (De\u017eelan, 2023; Gallup, 2023; Henn et al., 2005). For example, 68% of Swedish youth born between 1996 and 2010 believe society is heading in the wrong direction and report declining political interest (Christensen, 2024). Such discontent may translate into support for radical forms of political engagement: 47% of UK youth endorse revolutionary societal change (Channel 4 & Craft, 2024), and 55% of individuals aged 18-34 globally approve of tactics such as violence, property damage, or misinformation as tools for change (Edelman Trust Barometer, 2025). <\/p>\n\n\n\n
The present study seeks to address these gaps by building on political grievance and alienation perspectives as well as Significance Quest Theory, which converge in highlighting how experiences of political alienation, that is political distrust, feelings of powerlessness, and dissatisfaction with political processes, can motivate radicalism. Within these frameworks, we examine how political alienation contributes to the development of youth political radicalism, focusing on the longitudinal effects of political distrust, powerlessness, and dissatisfaction over a five-year period among Swedish youth.<\/p>\n\n\n\n
The Need for Significance and Political Grievance As Drivers of Radicalism<\/h2>\n\n\n\n
Radical political behavior refers to actions that challenge societal norms regarding both appropriate political goals and acceptable means of pursuing them (Kruglanski et al., 2018). Political radicalism is understood as a continuum of political action that varies in both goals and means. At one end are normative, conventional, mainstream and institutionally endorsed behaviors, including voting or signing a petition; at the other are violent, extreme actions, including terrorism. Between these endpoints lie various forms of non-normative political behavior that challenge prevailing political norms, rules, or authorities, often through illegal or confrontational means such as unauthorized demonstrations, property damage, or clashes with political opponents or the police. What unites these behaviors is not necessarily their level of violence, but their rejection of established boundaries regarding how political change should be pursued. Many radical actors advocate systemic change, seeking to disrupt and replace existing political, economic, social, and cultural structures and norms (McCauley & Moskalenko, 2008). While these various forms of radical political behaviors may be related, they are not equivalent. The present study focuses on non-normative political behavior, which may represent a precursor or correlate of violent radicalization but should not be interpreted as synonymous with it.<\/p>\n\n\n\n
Political radicalism has been theorized to be driven by the frustration of unmet psychological needs, which can motivate individuals to adopt ideologies and behaviors that deviate from societal norms (Echelmeyer et al., 2023). According to Significance Quest Theory (SQT), one of the most fundamental psychological needs is the need for significance, that is, to be respected, valued, to have dignity, and to matter to others (Kruglanski et al., 2014, 2022). Experiences of being disregarded, devalued, or wronged, whether politically, socially or personally, can frustrate this need, foster a sense of insignificance, and, in turn, drive individuals to adopt radical ideologies or behaviors as means to regain significance. Participation in radical political actions may enhance a sense of agency and value. While research has offered empirical support for SQT (Jasko et al., 2019; Miklikowska et al., 2022; Webber et al., 2018), prior work has focused primarily on social and personal sources of significance loss, with less attention to political experiences, particularly during adolescence.<\/p>\n\n\n\n
Grievance-based and alienation perspectives offer a complementary lens by emphasizing dissatisfaction with political institutions, perceived injustice, and lack of voice as key drivers of radicalization (Ajil, 2022; McCauley & Moskalenko, 2011; Ratelle & Souleimanov, 2017; Rydgren, 2007), including during adolescence (Beelmann, 2020; 2026). These perspectives highlight how perceptions of unfair treatment, exclusion from decision-making processes, and institutional unresponsiveness can generate grievances toward the political system. Such grievances, in turn, are reflected in individuals\u2019 broader orientations toward politics, including political alienation, that is distrust toward institutions, feelings of powerlessness, and dissatisfaction with how the system operates (Finifter, 1970; Olsen, 1969; Schwartz, 1976; Seeman, 1959). When conventional political channels are perceived as ineffective, these orientations may increase the appeal of alternative or even violent forms of political action (Van Buuren & De Graaf, 2014). <\/p>\n\n\n\n
Integrating these theoretical approaches, grievance-based and alienation perspectives identify the structural and institutional sources of discontent, while Significance Quest Theory provides a motivational account of how experiences of discontent may translate into action by undermining individuals\u2019 sense of significance. <\/p>\n\n\n\n
The Role of Political Alienation in Radicalization<\/h2>\n\n\n\n
Political alienation is often defined as a sense of having no meaningful role, voice, or influence in the conventional political process (Finifter, 1970; Olsen, 1969; Schwartz, 1976; Seeman, 1959). It is conceptualized here as a multidimensional construct comprising related but distinct components: political distrust, powerlessness, and dissatisfaction. Political distrust reflects a lack of confidence in political institutions and actors; powerlessness captures the perceived inability to influence political processes; and dissatisfaction refers to negative evaluations of how the political system performs. These dimensions represent distinct yet complementary ways in which individuals may experience estrangement from the political system and do not need to be simultaneously present. <\/p>\n\n\n\n
From the grievance perspectives, political alienation can be understood as reflecting how individuals experience systemic injustice, exclusion, and lack of voice within the political system. It captures the subjective manifestation of political grievances, translating structural and institutional conditions into individual-level perceptions. From the perspective of Significance Quest Theory, such experiences may signal a diminished sense of personal significance. When prescriptively prosocial political institutions, such as government agencies, elected officials, or other public authorities, are experienced as inaccessible, dishonest, unreliable, or unfair, individuals may come to view their beliefs and actions as inconsequential within the existing political system. This can foster a perception that they do not matter and motivate seeking alternative avenues to restore a sense of significance. <\/p>\n\n\n\n
Research on adults provides support for the relevance of the three dimensions of political alienation in radicalization. Experiences of disregard, indignation, or unfair treatment by political institutions and public authorities have been linked to decreases of political trust (Bennett et al., 2013; Fairbrother et al., 2022), higher powerlessness (Soss, 1999), and lower legitimacy beliefs (Sunshine & Tyler, 2003; Tyler & Huo, 2002). In turn, political distrust has been associated with greater acceptance of violence and illegal political behaviors (Marien & Hooghe, 2011; Villagran, 2025), as well as with engagement in illegal activities (Jackson et al., 2012; Tyler, 2006) and radical political behaviors (OECD, 2021). Similarly, political dissatisfaction has been linked to non-normative political participation (van Stekelenburg & Klandermans, 2013) and disengagement from conventional participation, such as voting (McKay et al., 2021; Olsen, 1969; Southwell, 2003; Southwell & Everest, 1998). In addition, political powerlessness has been related to protest voting as a means of expressing opposition to political elites (Horton & Thompson, 1962), anti-establishment attitudes (Gest et al., 2017), and greater approval of violence as a necessary means of achieving justice (Ransford, 1968). However, experimental manipulation of political powerlessness has been found to increase the legitimation of existing authorities (van der Toorn et al., 2014). This apparent contradiction suggests that powerlessness may give rise to two adaptive coping responses to significance loss: one characterized by opposition and non-normative action, and another characterized by system justification, where individuals legitimize the status quo to restore meaning (Jost, 2020).<\/p>\n\n\n\n
Adolescence As a Critical Formative Period for Alienation and Radicalization<\/h2>\n\n\n\n
The radicalizing potential of experiences of political distrust, powerlessness, or dissatisfaction may be particularly pronounced during adolescence. Political socialization and developmental theories identify adolescence as a critical period for the formation of political identity and behaviors, during which political orientations are especially malleable and sensitive to contextual influences (Erikson, 1968; Krosnick & Duane, 1989). Consistent with this heightened malleability, adolescents\u2019 approval of radical political action fluctuates more substantially over time than adults\u2019, whose approval tends to be relatively stable (Rink & Sharma, 2016; Weber, 2018). Adolescence is also a period in which individuals learn to navigate authority, institutions, and social hierarchies, actively constructing beliefs about how power operates and whether authorities are legitimate and responsive (Flanagan & Gallay, 2014; Flanagan & Tucker, 1999). Negative encounters with political institutions or public authorities during this formative phase may undermine emerging beliefs about political legitimacy and efficacy, possibly leading to an enduring sense of alienation (Citrin et al., 1975; Dermody et al., 2010; Fox, 2015; Neundorf & Smets, 2017). Moreover, ongoing development of cognitive control<\/a> and emotional regulation during adolescence is associated with increased risk-taking<\/a> and heightened emotional responses to frustration or perceived injustice (Casey & Caudle, 2013; Silvers et al., 2012), factors that have been found to be related to higher engagement in activities that seek to challenge existing political structures (e.g., protesting) (Oosterhoff & Wray-Lake, 2020). In line with this developmental vulnerability, adolescents have been shown to express higher approval of radical political action than adults (Watts, 1999). Together, this combination of developmental malleability, early encounters with political systems, and heightened reactivity suggests that negative political experiences in adolescence may have enduring consequences, potentially increasing vulnerability to radical alternatives (Barracosa & Cherney, 2025; Beelmann, 2020; Beelmann, 2026; Carlsson et al., 2020).<\/p>\n\n\n\n Empirical research on the relationship between political alienation and various forms and stages of radicalism among youth remains limited but offers important insights. A cross-sectional study of Dahl et al. (2017) found a positive link between political alienation and radical activism among youth in the UK, although this was not observed in seven other EU countries. Other studies have linked specific facets of alienation, particularly low institutional trust, to more radical forms of political participation among adults (Dekker et al., 1997) and to more favorable attitudes toward radical activism among youth over time (\u0160erek et al., 2018). Political dissatisfaction has been linked to a higher likelihood of law-breaking behavior (Abdelzadeh et al., 2015), and a meta-analytic review reported small but significant associations between dissatisfaction with political actors, police, and institutions and radical outcomes (Jahnke et al., 2021). Political distrust has also been associated with youth engagement in radical environmental activism (Cologna et al., 2021). In contrast, political alienation has shown only weak or inconsistent associations with normative political disengagement, such as non-voting (Dahl et al., 2017) and has been found to be unrelated to normative civic engagement among ethnic minority youth (Hope & Jagers, 2014). The perceptions of institutional discrimination or victimization have shown positive associations among ethnic minority youth (Hope & Jagers, 2014; Oosterhoff & Wray-Lake, 2020; see also Pinedo et al., 2024).<\/p>\n\n\n\n In sum, although political alienation may motivate radical forms of political expression, its role in radicalization remains underexplored, particularly during adolescence, a critical period for radicalization, and from a longitudinal perspective. Longitudinal designs offer a unique opportunity to examine the dynamic associations between alienation and radicalism over time. They also allow for the simultaneous investigation of individual- and contextual-level processes, capturing how both personal political experiences and exposure to climates of shared disillusionment in immediate social environments (e.g., classrooms) may contribute to radical political behavior. In the present study, we focus on non-normative political behaviors that challenge legal or institutional boundaries, which we conceptualize as occupying an intermediate position on a continuum ranging from conventional participation to violent extremism.<\/p>\n\n\n\n The present five-wave panel study seeks to advance understanding of youth political radicalism by investigating the role of political alienation in radical behavior across adolescence and by separating within- from between-person and between-classroom processes. We hypothesized that during years in which adolescents experienced higher political alienation than their own average, they would contemporaneously engage in more radical political behaviors than usual (effects on the within-person level). We further hypothesized that adolescents with higher overall levels of political alienation would exhibit higher overall levels of radicalism and a slower decline in radicalism over time compared to peers with lower alienation (effects on the between-person level). Finally, we tested the role of a classroom climate of political alienation (effects on the between-classroom level). Higher classroom-level alienation may be associated with greater individual radicalism through the normalization and reinforcement of shared grievances. However, radicalization may be driven primarily by direct personal experiences of alienation, such that only personally felt institutional disregard activates radical behavior (i.e., you care when it happens to you). <\/p>\n\n\n\n This study is part of a broader research initiative on youth development, conducted within the Political Socialization Program. The project explores a broad range of political and social attitudes and behaviors, along with contextual influences. The data collection was overseen by professors Erik Amn\u00e5, Mats Ekstr\u00f6m, Margaret Kerr, and H\u00e5kan Stattin, with financial support from Riksbankens Jubileumsfond. Ethical approval was obtained from the Regional Ethical Review Board in Uppsala (Dnr 2010\/115). Data were gathered annually between 2010 and 2014 in Sweden\u2019s seventh-largest city, which closely reflects national averages in terms of income levels, unemployment rates, and ethnic diversity (Statistiska Centralbyr\u00e5n, 2016). Ten schools were selected from neighbourhoods characterized by varying ethnic and socioeconomic compositions. Each participating class received a 100 EUR incentive. Data collection took place during ordinary school hours, in the presence of trained research assistants. Adolescents were informed about the voluntary nature of their participation, the confidentiality of their responses, and the content of the survey. Parents received questionnaires via mail, with a response rate of 65%. Approximately 3.8% of parents declined consent for their child\u2019s involvement.<\/p>\n\n\n\n The initial sample included 946 adolescents (M<\/em>age at T1 = 13.41 years, SD<\/em> = 0.53, 50.7% Female) from diverse migration and socioeconomic backgrounds (M<\/em> = 3.01, SD<\/em> = 0.80 for perceived financial situation of the family). Within the sample, 27.1% had a non-Nordic immigrant background (defined as having at least one parent born outside the Nordic region) and 71.5% identified as having Nordic background. Among those with non-Nordic immigrant background, 71.0% had at least one parent born outside of Europe (non-European immigrant background). Participants were clustered in 38 classrooms, which remained stable between T1 and T3. Three of these classrooms had fewer than eight participants (N<\/em> = 12). They were removed to minimize potential biases in individual and class-level data. Adolescents who switched classrooms once (N<\/em> = 31) were categorized based on where they spent most of their time, whereas those who switched classrooms twice (N<\/em> = 42) were removed from the analyses. By T4, students transitioned to new schools, joining various new classrooms alongside peers from outside the original sample. The final dataset included 892 adolescents (51.1% female) in 35 classrooms, which falls within the range commonly used in applied multilevel research (e.g., Maas & Hox, 2005). Adolescents were 13-14 years old at T1 (M<\/em> = 13.41, SD<\/em> = 0.52) and 17\u201318 years old at T5 (M<\/em> = 17.32, SD<\/em> = 0.47). Attrition analysis indicated no major systematic differences between participants who remained in the study and those who dropped out and a measurement invariance analysis supported metric invariance based on changes in approximate fit indices (Putnick & Bornstein, 2017), suggesting that the construct was measured consistently across time. The supplementary materials detail dropout and measurement invariance analyses. The number of units at the highest level in the current manuscript was N = 35, which exceeds commonly cited recommendations for three-level models (e.g., N = 30; Kreft & de Leeuw, 1998).<\/p>\n\n\n\n At every wave, youth reported the frequency of their engagement in radical political behaviors over the preceding 12 months using five items rated on a three-point Likert scale (1 = No<\/em>, 2 = Once or twice<\/em>, 3 = Yes, several times<\/em>): \u201cPainted political messages or graffiti on walls\u201d, \u201cTaken part in an illegal action\/demonstration or occupation\u201d, \u201dTaken part in a political event where property was destroyed\u201d, \u201cTaken part in a political event where there was a confrontation with political opponents or the police\u201d, and \u201cBroken the law on political grounds\u201d. A scale score was computed by averaging the five items. Similar items have been used before to tap political radicalism (e.g., Moskalenko & McCauley, 2009), including in adolescent samples (Mirazchiyski & Mirazchiyski, 2025; Pfundmair, Paulus, & Wagner, 2021) . The internal reliabilities at T1-T5 were good: .93, .93, .84, .82, and .89.<\/p>\n\n\n\n Political Distrust<\/em>. At every measurement occasion, youth reported on how much trust they have towards six authorities in the political system: the courts, the police, the parliament, political parties, the government, and the EU by rating them on a four-point Likert scale (from 1 = A lot of trust<\/em> to 4 = No trust at all<\/em>). A scale score was computed by averaging the six items. Similar items have been used to tap political distrust (Bertsou, 2019; European Social Survey, 2002; OECD, 2021; Valgar\u00f0sson et al., 2025). The internal reliabilities were good: .91, .92, .92, .90, and .90 for T1-T5, respectively. Political Powerlessness<\/em>. <\/strong>At every measurement occasion, youth indicated their political powerlessness by rating three items: \u201cOrdinary people lack the opportunity to affect political decisions\u201d (reversed), \u201cThose in power in our society lack interest in how people like me have it in life\u201d (reversed), and \u201cThose in power make decisions without finding out what people like me think\u201d (reversed) on a four-point Likert scale (from 1 = Applies perfectly<\/em> to 4 = Doesn\u2019t apply at all<\/em>). A scale score was computed by averaging the three items. Similar items have been used to tap political powerlessness (Horton & Thompson, 1962; Public Agenda, 2022; Ransford, 1968). The internal reliabilities were acceptable-to-good: .71, .75, .78, .79, and .81 for T1-T5, respectively. <\/strong>Political Dissatisfaction.<\/em> <\/strong>At all time-points adolescents reported on how satisfied they are with the way in which the sitting government in Sweden is handling its tasks on a four-point Likert scale (ranging from 1 = Very satisfied<\/em> to 4 = Not satisfied at all<\/em>). Similar items have been used to tap political dissatisfaction (European Social Survey, 2002)<\/p>\n\n\n\n Gender<\/em> was coded such that 1 indicated female and 2 indicated male. Immigrant background <\/em>was coded with <\/em>1 for youth with immigrant background (i.e. with at least one parent born outside of Nordic countries) and 0<\/em> for youth with no immigrant background (i.e, both parents born in Nordic countries). This definition distinguishes between youth with and without a non-Nordic immigrant background, reflecting meaningful differences in migration histories and social integration in Sweden. As robustness checks, we also considered a non-Swedish immigrant background (i.e, <\/em>at least one parent born outside of Sweden) and non-European immigrant background (i.e., at least one parent born outside of Europe) (see Supplementary Materials). SES<\/em>. Parents provided information on their household monthly income using a seven-point scale (1 = 1-10,000 SEK<\/em> to 7 = above 60,001 SEK<\/em>; M<\/em> = 5.06, SD<\/em> = 1.49), as well as their highest level of education on a five-point scale (1 = less than 9 years of education<\/em> to 5 = university college\/university<\/em>; M<\/em> = 4.31, SD<\/em> = 0.87). Parental responses were available for 60% of youth. Perceived financial situation of the family.<\/em> Adolescents rated one item \u201cWhat are your family finances like?\u201d on a four-point scale from 1 = \u201cMy parents always complain that they don\u2019t have enough money<\/em>\u201d to 4 = \u201cMy parents never complain about being short of money\u201d<\/em> (M<\/em> = 3.01, SD<\/em> = 0.78). <\/p>\n\n\n\n Correlations, means, and standard deviations between study variables are reported in Table 1. Correlations between the study and control variables are presented in supplementary materials. Adolescents\u2019 political alienation (distrust, powerlessness, dissatisfaction) was positively, although weakly, related to their radical behavior measured concurrently and across time. <\/p>\n\n\n\n Table 1<\/strong> A series of t-tests revealed that reported higher powerlessness than girls at T1, t<\/em>(796) = 2.38, p<\/em> = .018, mean difference = 0.07, 95% CI [0.013, 0.137]. They also reported higher distrust at T3, t<\/em>(669.44) = 2.34, p<\/em> = .020, mean difference = 0.12, 95% CI [0.020, 0.228]. Boys also reported higher dissatisfaction than girls at T2, t<\/em>(648.04) = 2.00, p<\/em> = .046, mean difference = 0.10, 95% CI [0.002, 0.203], T3, t(642.77) = 2.61, p <\/em>= .009, mean difference = 0.14, 95% CI [0.036, 0.254], and T4, t<\/em>(520.17) = 2.04, p <\/em>= .042, mean difference = 0.11, 95% CI [0.004, 0.220]. Regarding radicalism, boys reported higher levels of radicalism than girls at T1, t(<\/em>612.58) = 4.38, p<\/em> < .001, mean difference = 0.11, 95% CI [0.059, 0.154], T3, t<\/em>(347.36) = 5.15, p<\/em> < .001, mean difference = 0.07, 95% CI [0.046, 0.104], T4, t<\/em>(304.42) = 3.91, p<\/em> < .001, mean difference = 0.05, 95% CI [0.027, 0.080], and T5, t<\/em>(444.82) = 2.65, p<\/em> = .008, mean difference = 0.04, 95% CI [0.011, 0.072].<\/p>\n\n\n\n Non-Nordic immigrant youth reported significantly higher powerlessness than youth without a non-Nordic immigrant background at T5, t<\/em>(610) = 2.24, p<\/em> = .025, mean difference = 0.12, 95% CI [0.014, 0.218]. They also reported higher distrust at T2, t<\/em>(732) = 2.68, p<\/em> = .008, mean difference = 0.16, 95% CI [0.044, 0.284], T3, t<\/em>(247.87) = 2.05, p<\/em> = .041, mean difference = 0.14, 95% CI [0.005, 0.270], and T5, t(203.31) = 2.88, p<\/em> = .004, mean difference = 0.19, 95% CI [0.060, 0.319]. There were no significant differences in dissatisfaction. Regarding radicalism, non-Nordic immigrant youth showed higher levels than youth without a non-Nordic immigrant background at T1, t<\/em>(278.24) = 3.10, p<\/em> = .002, mean difference = 0.10, 95% CI [0.036, 0.160], T2, t<\/em>(225.76) = 3.52, p<\/em> = .001, mean difference = 0.12, 95% CI [0.051, 0.182]. The supplementary materials detail results for alternative operationalizations of immigrant background. <\/p>\n\n\n\n We used Mplus 8 (Muth\u00e9n & Muth\u00e9n, 1998-2018) and multilevel modelling to examine whether political alienation predicted changes in youth political radicalism both within individuals and between individuals over time. Within-person effects capture the relationship between fluctuations in alienation across the five waves (i.e., deviations from own average alienation score) and fluctuations in political radical behavior (i.e., deviations from own average radicalism score) across the same waves. These effects show if periods of heightened experiences of alienation were associated with higher-than-usual radical behaviors for the same individual.<\/p>\n\n\n\n In contrast, between-person effects refer to interpersonal differences, assessing how youth with different levels of political alienation vary in their engagement in radical political behavior. To control for individual differences in baseline levels, time-invariant predictors were adjusted for the overall sample average across all five time points (i.e., grand-mean centered) and time-varying predictors were adjusted for each individual\u2019s mean across the five time points (i.e., person-mean centered).<\/p>\n\n\n\n Initially, average change was modelled across three consecutive models. Model 1 specified a random intercept that decomposed variance in youth radicalism into within-person, between-person, and between-classroom components. Model 2 incorporated fixed linear and quadratic slopes representing the average trajectory over time. Given that adolescence is characterized by rapid and potentially non-linear developmental changes, we examine both linear and quadratic time trends to capture potential acceleration or deceleration in the development of political alienation and radicalism over time. Model 3 added a random slope to account for variability in change across individuals and classrooms. Subsequently, predictors of the random intercepts and slopes were added (i.e. political alienation at the between- and within-person levels) (Model 4A-C). Deviance test (\u0394-2LL) and AIC criterion were used to assess the nested models. Kreft and de Leeuw (1998) suggest that the model with lower deviance and AIC is the better-fitting model.<\/p>\n\n\n\n A random-intercept model (Model 1) was estimated to examine trajectories of youth radical political activism over time. Intraclass correlations showed that 22% of the variance was between adolescents and 0.5% between classrooms, with the remainder at the within-person level. Including linear and quadratic fixed effects in Model 2 improved model fit (\u0394-2LL = 46.11 (2), p<\/em> < .001; \u0394AIC = 42.11). The negative linear (B<\/em> = -0.04, 95% CI [-0.06, -0.03], p =<\/em> .001) and positive quadratic (B<\/em> = 0.01, 95% CI [0.00, 0.01], p =<\/em> .004) terms indicated a modest decline followed by stabilization. Allowing the linear slope to vary across individuals in Model 3 further improved model fit (\u0394-2LL = 185.26 (1), p<\/em> < .001; \u0394AIC = 254.84), although variance in slopes at the between-person level was negligible (\u03c3\u00b2 = 0.00, 95% CI [0.00, 0.00], p =<\/em> .001). No significant classroom-level slope variation was detected (B<\/em> = 0.00, 95% CI [0.00, 0.00], p =<\/em> .176) and this parameter was constrained to zero. Detailed results are presented in Table 2. In the next steps (Models 4A-C), political distrust, powerlessness, and dissatisfaction were added as predictors of youth radical political activism at the within-person and between-person levels.<\/p>\n\n\n\n Table 2<\/strong> Model<\/p>\n<\/td>\n Alienation Effect Models<\/p>\n<\/td>\n<\/tr>\n Fit <\/p>\n LL (df)<\/p>\n<\/td>\n Change <\/p>\n -2LL (df)<\/p>\n<\/td>\n L1 <\/p>\n Within <\/p>\n Person <\/p>\n<\/td>\n L2 <\/p>\n Between Person <\/p>\n<\/td>\n L3 <\/p>\n Between Classroom <\/p>\n<\/td>\n Variance<\/p>\n <\/p>\n<\/td>\n<\/tr>\n Model 1 <\/p>\n Unconditional<\/p>\n<\/td>\n -173.26 (4) <\/p>\n AIC=354.53<\/p>\n<\/td>\n –<\/p>\n<\/td>\n –<\/p>\n<\/td>\n –<\/p>\n<\/td>\n –<\/p>\n<\/td>\n .053 L1 .015 L2<\/p>\n Total:<\/p>\n .068<\/p>\n <\/p>\n<\/td>\n<\/tr>\n Model 2 <\/p>\n Fixed Linear and Slope L1<\/p>\n<\/td>\n -150.21 (6)<\/p>\n AIC=312.42 <\/p>\n <\/p>\n<\/td>\n 46.1 (2)<\/p>\n p<\/i> = .001<\/p>\n<\/td>\n B<\/i>lin<\/sub> = p =<\/i> .001<\/p>\n B<\/i>qd<\/sub> = p =<\/i> .004<\/p>\n<\/td>\n –<\/p>\n<\/td>\n –<\/p>\n<\/td>\n .053 L1<\/p>\n .015 L2<\/p>\n .000 L3<\/p>\n Total:<\/p>\n .068<\/p>\n <\/p>\n<\/td>\n<\/tr>\n Model 3 Random Linear Slope L2 <\/p>\n<\/td>\n -57.58 (7)<\/p>\n AIC=129.17<\/p>\n<\/td>\n 185.26 (1)<\/p>\n p<\/i> = .001<\/p>\n<\/td>\n B<\/i>qd<\/sub> = p =<\/i> .001<\/p>\n<\/td>\n –<\/p>\n<\/td>\n –<\/p>\n<\/td>\n .047 L1<\/p>\n .053 L2<\/p>\n .000 L3<\/p>\n Total:<\/p>\n .100<\/p>\n <\/p>\n<\/td>\n<\/tr>\n Model 4A<\/p>\n Distrust<\/p>\n at L1, L2, L3<\/p>\n<\/td>\n 146.98 (11)<\/p>\n AIC= -271.96 <\/p>\n<\/td>\n 178.79 (4)<\/p>\n p<\/i> = .001<\/p>\n<\/td>\n B<\/i>qd<\/sub> = p =<\/i> .007<\/p>\n B <\/i>= 0.033<\/p>\n p =<\/i> .001<\/p>\n<\/td>\n B<\/i>Level<\/sub> = 0.079<\/p>\n p<\/i> = .001<\/p>\n B<\/i>Slope<\/sub> = -0.007<\/p>\n p<\/i> = .245<\/p>\n<\/td>\n B<\/i>Level<\/sub> = <\/p>\n 0.005<\/p>\n p<\/i> = .878<\/p>\n <\/p>\n<\/td>\n .044 L1<\/p>\n .037 L2<\/p>\n .000 L3<\/p>\n Total: .081<\/p>\n<\/td>\n<\/tr>\n Model 4B<\/p>\n Powerlessness<\/p>\n at L1, L2, L3<\/p>\n<\/td>\n 84.07 (10)<\/p>\n AIC= -148.13 <\/p>\n<\/td>\n 53.06 (3)<\/p>\n p<\/i> = .001<\/p>\n<\/td>\n B<\/i>qd<\/sub> = p =<\/i> .001<\/p>\n B <\/i>= 0.022<\/p>\n p =<\/i> .009<\/p>\n<\/td>\n B<\/i>Level<\/sub> = 0.035<\/p>\n p<\/i> = .022<\/p>\n BSlope<\/sub> = B<\/i>Level<\/sub> = 0.088<\/p>\n p<\/i> = .171<\/p>\n<\/td>\n .045 L1<\/p>\n .041 L2<\/p>\n .000 L3<\/p>\n Total: .086<\/p>\n<\/td>\n<\/tr>\n Model 4C<\/p>\n Dissatisfactionat L1, L2, L3<\/p>\n<\/td>\n 139.58 (11)<\/p>\n AIC= -257.15 <\/p>\n<\/td>\n 163.98 (4)<\/p>\n p<\/i> = .001<\/p>\n<\/td>\n Bqd<\/sub> <\/i>= -0.003 <\/p>\n p =<\/i> .001<\/p>\n B <\/i>= 0.019<\/p>\n p =<\/i> .018<\/p>\n<\/td>\n B<\/i>Level<\/sub> = 0.050<\/p>\n p<\/i> = .013<\/p>\n B<\/i>Slope<\/sub> = 0.003<\/p>\n p<\/i> = .692<\/p>\n<\/td>\n B<\/i>Level<\/sub> = <\/p>\n 0.024<\/p>\n p<\/i> = .620<\/p>\n<\/td>\n .043 L1<\/p>\n .042 L2<\/p>\n .000 L3<\/p>\n Total:<\/p>\n .085<\/p>\n<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/figure>\n\n\n\n Model 4A included political distrust as a predictor of within-person change, the between-person random intercept and slope, and the between-classroom random intercept. This significantly improved model fit compared to the model without predictors (\u0394-2LL = 178.79 (4), p<\/em> < .001; \u0394AIC = 401.13).<\/p>\n\n\n\n At the within-person level, higher-than-usual distrust was associated with higher levels of radicalism (B <\/em>= 0.03, p<\/em> = .001, 95% CI [0.020, 0.047]), indicating that adolescents reported more radical behavior during periods of elevated distrust. <\/p>\n\n\n\n At the between-person level, adolescents with higher average distrust also reported higher levels of radicalism (B <\/em>= 0.08, 95% CI [0.049, 0.108], p<\/em> = .001). There were no effects of distrust on the slope of radicalism (B<\/em> = -0.01, 95% CI [-0.017, 0.003], p<\/em> = .245). Overall, distrust accounted for 19% of the variance in radicalism. At the between-classroom level, no significant association was found (B<\/em> = 0.00, 95% CI [-0.048, 0.058], p<\/em> = .878). See Figure 1A for illustration.<\/p>\n\n\n\n Figure 1<\/strong> To supplement these multilevel analyses, we examined the direction and timing of within-person associations by estimating a random-intercept cross-lagged panel model (RI-CLPM; Hamaker et al., 2015) (see Supplementary Materials for the model specification). The results of this model told a similar story to multilevel analyses. At the between-person level, distrust and radicalism were significantly correlated (r<\/em> = .35, p<\/em> = .001), indicating that adolescents who reported higher distrust across the five waves also reported more radicalism. The within-person associations were small and primarily concurrent, with a significant within-time association at T1 (r<\/em> = .02, p<\/em> = .047) and correlated changes between distrust and radicalism (r<\/em> = .02, p<\/em> = .047 at T1; r<\/em> = .05, p<\/em> = .045 and r<\/em> = .06, p<\/em> = .044 at T2\u2013T4) but no significant lagged effects, indicating that within-person increases in distrust coincided with simultaneous increases in radicalism.<\/p>\n\n\n\n The effects of distrust did not change substantially when controlling for parental education, household income, and youth perceived financial situation of the family, which had no significant effects (B<\/em> = -.0.00, 95% CI [-0.010, -0.004], p<\/em> = .878; B<\/em> = -0.00, 95% CI [-0.009, 0.003], p<\/em> = .405; B<\/em> = 0.00, 95% CI [-0.006, 0.016], p<\/em> = .456, respectively). Moderation by immigrant background and gender was also examined. A model including a freely estimated path from the interaction between distrust and immigrant background to youth political radicalism did not significantly differ from a model in which this path was fixed to zero (\u0394\u22122LL(1) = 0.26, p<\/em> = .613, \u0394AIC = 1.74), indicating that the effect of distrust was not different for non-Nordic immigrant youth and youth without a non-Nordic immigrant background. The main effect of immigrant background was not significant (B<\/em> = 0.01, [-0.004, 0.035], p<\/em> = .182). The results were substantially the same for alternative operationalizations of immigrant background (see Supplementary Materials). However, gender moderated the effects of distrust (\u0394-2LL = 20.142 (1), p<\/em> < .001; \u0394AIC = 18.14). A significant interaction between distrust and gender was found for the level of radicalism (B<\/em> = 0.08, 95% CI [0.052, 0.113], p<\/em> = .001), indicating that the effect of distrust was stronger among boys. The main effect of gender was also significant (B<\/em> = 0.04, 95% CI [0.027, 0.059], p<\/em> = .001), with boys exhibiting higher levels of radicalism than girls (see Figure 2A).<\/p>\n\n\n\n Overall, higher-than-usual levels of political distrust within individuals were associated with increases in radicalism. Adolescents who reported more distrust than their peers also reported higher levels of political radicalism. No significant classroom-level effects were observed.<\/p>\n\n\n\n Figure 2<\/strong> Model 4B included political powerlessness as a predictor of within-person change, the between-person random intercept and slope, and the between-classroom random intercept. This significantly improved model fit compared to the model without predictors (\u0394-2LL = 53.06 (3), p<\/em> < .001; \u0394AIC = 275.40).<\/p>\n\n\n\n At the within-person level, higher-than-usual powerlessness was associated with higher levels of radicalism (B<\/em> = 0.02, p<\/em> = .009, 95% CI [0.008, 0.035]), indicating that adolescents reported more radical behavior during periods of elevated powerlessness.<\/p>\n\n\n\n At the between-person level, powerlessness was not associated with differences in the slope of radicalism (B<\/em> = -0.00, 95% CI [-0.020\u20130.013], p<\/em> = .754). When this non-significant effect was constrained to zero, powerlessness significantly predicted the level of radicalism (B<\/em> = 0.03, 95% CI [0.010, 0.061], p<\/em> = .022). Constraining the slope did not significantly affect model fit (\u0394-2LL = 0.10, p<\/em> = .752) and did not alter other parameter estimates. Overall, powerlessness accounted for 14% of the variance in radicalism. At the between-classroom level, powerlessness did not significantly predict radicalism (B<\/em> = 0.09, 95% CI [-0.018, 0.195], p<\/em> = .171). See Figure 1B for illustration.<\/p>\n\n\n\n The random-intercept cross-lagged panel model told a similar story to multilevel analyses also for political powerlessness (see Supplementary Materials). Powerlessness and radicalism were correlated at the between-person level (r<\/em> = .16, p<\/em> = .047), whereas the within-person estimates were only borderline: a within-time association at T1 (r<\/em> = .02, p<\/em> = .056) and correlated changes across T2\u2013T5 (r<\/em> = .03, p<\/em> = .055; r<\/em> = .05, p<\/em> = .054; r<\/em> = .06, p<\/em> = .055; r<\/em> = .05, p<\/em> = .054), again with no significant lagged effects. Within-person increases in powerlessness were therefore weakly and concurrently linked to increases in radicalism.<\/p>\n\n\n\n The effects of powerlessness did not change substantially when controlling for the non-significant effect of parental education (B<\/em> = -0.00, 95% CI [-0.018, 0.195], p<\/em> = .384). Controlling for the non-significant effect of household income (B<\/em> = -0.00, 95% CI [-0.012, 0.000], p<\/em> = 126) rendered the effect of powerlessness on the level of radicalism non-significant (B<\/em> = 0.03, 95% CI [-0.003, 0.058], p<\/em> = .132), likely reflecting a loss of power due to reduced sample size when including this control variable or shared variance between household income and powerlessness. Model controlling for youth perceived financial situation of the family did not converge. Moderation by immigrant background and gender was also examined. A model including a freely estimated path from the interaction between powerlessness and immigrant background to youth political radicalism did not significantly differ from a model in which this path was fixed to zero (\u0394\u22122LL(1) = 0.03,The Present Study<\/h2>\n\n\n\n
Method<\/h2>\n\n\n\n
Data and Participants<\/h3>\n\n\n\n
Measures<\/h3>\n\n\n\n
Radical Political Behaviors<\/h4>\n\n\n\n
Political Alienation<\/h4>\n\n\n\n
Control Variables<\/h4>\n\n\n\n
Initial Analyses<\/h3>\n\n\n\n
Means, Standard Deviations, and Correlations Between Variables.<\/em><\/p>\n\n\n\nVariable<\/td> M<\/em><\/em><\/td> SD<\/em><\/em><\/td> 1<\/td> 2<\/td> 3<\/td> 4<\/td> 5<\/td> 6<\/td> 7<\/td> 8<\/td> 9<\/td> 10<\/td> 11<\/td> 12<\/td> 13<\/td> 14<\/td> 15<\/td> 16<\/td> 17<\/td> 18<\/td> 19<\/td> 20<\/td><\/tr> 1. Political Radicalism T1<\/td> 1.11<\/td> 0.34<\/td> –<\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td><\/tr> 2. Political Radicalism T2<\/td> 1.08<\/td> 0.31<\/td> .14***<\/td> –<\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td><\/tr> 3. Political Radicalism T3<\/td> 1.04<\/td> 0.18<\/td> .12**<\/td> .19***<\/td> –<\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td><\/tr> 4. Political Radicalism T4<\/td> 1.03<\/td> 0.15<\/td> .14***<\/td> .14***<\/td> .40***<\/td> –<\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td><\/tr> 5. Political Radicalism T5<\/td> 1.04<\/td> 0.18<\/td> .10**<\/td> .10*<\/td> .32***<\/td> .51***<\/td> –<\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td><\/tr> 6. Powerlessness T1<\/td> 1.52<\/td> 0.6<\/td> 0.06<\/td> 0.05<\/td> 0.01<\/td> 0.02<\/td> -0.03<\/td> –<\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td><\/tr> 7. Powerlessness T2<\/td> 2.43<\/td> 0.58<\/td> -0.02<\/td> 0.03<\/td> 0.06<\/td> 0.06<\/td> -0.01<\/td> .21***<\/td> –<\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td><\/tr> 8. Powerlessness T3<\/td> 2.46<\/td> 0.62<\/td> 0.05<\/td> 0.03<\/td> 0.06<\/td> 0.02<\/td> -0.03<\/td> .15***<\/td> .28***<\/td> –<\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td><\/tr> 9. Powerlessness T4<\/td> 2.4<\/td> 0.63<\/td> 0<\/td> 0.01<\/td> .10*<\/td> .09*<\/td> 0.07<\/td> .18***<\/td> .18***<\/td> .33***<\/td> –<\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td><\/tr> 10. Powerlessness T5<\/td> 2.41<\/td> 0.62<\/td> 0.02<\/td> 0<\/td> 0<\/td> .10*<\/td> 0.05<\/td> .09*<\/td> .18***<\/td> .25***<\/td> .38***<\/td> –<\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td><\/tr> 11. Distrust T1<\/td> 2.69<\/td> 0.72<\/td> .12**<\/td> .10*<\/td> .12**<\/td> 0.03<\/td> 0<\/td> .15***<\/td> .13**<\/td> .16***<\/td> .22***<\/td> 0.07<\/td> –<\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td><\/tr> 12. Distrust T2<\/td> 2.56<\/td> 0.71<\/td> .11**<\/td> .13**<\/td> .11**<\/td> .08*<\/td> 0.04<\/td> .08*<\/td> .14***<\/td> .18***<\/td> .20***<\/td> .14**<\/td> .55***<\/td> –<\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td><\/tr> 13. Distrust T3<\/td> 2.36<\/td> 0.7<\/td> 0.05<\/td> .08*<\/td> .10**<\/td> .19***<\/td> .13**<\/td> .13**<\/td> .13**<\/td> .20***<\/td> .24***<\/td> .20***<\/td> .40***<\/td> .51***<\/td> –<\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td><\/tr> 14. Distrust T4<\/td> 2.25<\/td> 0.64<\/td> .12**<\/td> 0.07<\/td> .12**<\/td> .22***<\/td> .10*<\/td> .10*<\/td> .09*<\/td> .16***<\/td> .32***<\/td> .19***<\/td> .34***<\/td> .43***<\/td> .57***<\/td> –<\/td> <\/td> <\/td> <\/td> <\/td> <\/td> <\/td><\/tr> 15. Distrust T5<\/td> 2.38<\/td> 0.63<\/td> .13**<\/td> 0.07<\/td> .11**<\/td> .19***<\/td> .15***<\/td> -0.01<\/td> .15**<\/td> .14**<\/td> .21***<\/td> .25***<\/td> .30***<\/td> .38***<\/td> .48***<\/td> . 56***<\/td> –<\/td> <\/td> <\/td> <\/td> <\/td> <\/td><\/tr> 16. Dissatisfaction T1<\/td> 2.14<\/td> 0.61<\/td> 0.05<\/td> 0.06<\/td> .09*<\/td> -0.01<\/td> 0.06<\/td> .11**<\/td> 0.06<\/td> .12**<\/td> .12**<\/td> .12**<\/td> .32***<\/td> .22***<\/td> .21***<\/td> .19***<\/td> .10*<\/td> –<\/td> <\/td> <\/td> <\/td> <\/td><\/tr> 17. Dissatisfaction T2<\/td> 2.11<\/td> 0.67<\/td> .12**<\/td> .15***<\/td> .11**<\/td> 0.04<\/td> .11*<\/td> .14***<\/td> .11**<\/td> .17***<\/td> .18***<\/td> .17***<\/td> .28***<\/td> .42***<\/td> .41***<\/td> .36***<\/td> .26***<\/td> .30***<\/td> –<\/td> <\/td> <\/td> <\/td><\/tr> 18. Dissatisfaction T3<\/td> 2.17<\/td> 0.72<\/td> -0.03<\/td> .09*<\/td> .15***<\/td> .20***<\/td> .10*<\/td> .09*<\/td> .14***<\/td> .16***<\/td> .20***<\/td> .19***<\/td> .22***<\/td> .36***<\/td> .54***<\/td> .43***<\/td> .34***<\/td> .25***<\/td> .52***<\/td> –<\/td> <\/td> <\/td><\/tr> 19. Dissatisfaction T4<\/td> 2.14<\/td> 0.68<\/td> 0.00<\/td> 0.06<\/td> .13**<\/td> .17***<\/td> .09*<\/td> 0.08<\/td> .13**<\/td> .18***<\/td> .30***<\/td> .24***<\/td> .17***<\/td> .25***<\/td> .38***<\/td> .54***<\/td> .37***<\/td> .25***<\/td> .42***<\/td> .45***<\/td> –<\/td> <\/td><\/tr> 20. Dissatisfaction T5<\/td> 2.39<\/td> 0.74<\/td> 0.00<\/td> 0.05<\/td> .08*<\/td> .16***<\/td> .15***<\/td> 0.05<\/td> .11*<\/td> .13**<\/td> .20***<\/td> .25***<\/td> .09*<\/td> .14**<\/td> .23***<\/td> .27***<\/td> .43***<\/td> .14**<\/td> .31***<\/td> .36***<\/td> .41***<\/td> –<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n Main Analyses<\/h3>\n\n\n\n
Results<\/h2>\n\n\n\n
Models to Account for the Level and Change in Adolescents\u2019 Political Radicalism.<\/em><\/p>\n\n\n\n\n\n
\n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n
\n Quadratic<\/p>\n\n \n \n
\n -0.043 <\/p>\n
\n 0.007 <\/p>\n\n \n \n \n \n \n \n \n
\n -0.003 <\/p>\n\n \n \n \n \n \n \n \n
\n -0.002 <\/p>\n\n \n \n \n \n \n \n \n
\n -0.003 <\/p>\n\n
\n @0<\/p>\n<\/td>\n\n \n \n \n \n \n \n \n \n \n Political Distrust As Predictor of Youth Radicalism<\/h3>\n\n\n\n
The effects of political distrust (A), powerlessness (B), and dissatisfaction (C) on the level of political radicalism across adolescence (i.e., from age 13 to 17). The figures present adjusted predictions (marginal effects) with 95% confidence intervals at one standard deviations.<\/em><\/p>\n\n\n\n![\"The](\"https:\/\/advances.in\/psychology\/wp-content\/uploads\/aip00061_figure1.svg\")
<\/figure>\n\n\n\n
The effects of political distrust (A), powerlessness (B), and dissatisfaction (C) on the level of political radicalism depending on the gender (1 = Female, 2 = Male). The figures present adjusted predictions with 95% confidence intervals at one standard deviations.<\/em><\/p>\n\n\n\n![\"The](\"https:\/\/advances.in\/psychology\/wp-content\/uploads\/aip00061_figure2.svg\")
<\/figure>\n\n\n\nPolitical Powerlessness As Predictor of Youth Radicalism<\/h3>\n\n\n\n