Although radicalism, holding an extreme and possibly violent stance for change, is plausibly preventable by law, the prevention is uncertain and requires empirical research (Ferguson & McAuley, 2020; Ludigdo & Mashuric, 2021). Such plausibility stems from the awareness of the law, including its enforcement and abidance, premised on deterrence and norm theories. Deterrence theory proposes that law enforcement deters radicalism when criminal or illegal (Hirtenlehner & Schulz, 2021; Johnston & Bose, 2020). Likewise, norm theory suggests that realizing and internalizing the social norm, which means the common practice of people in society, for law abidance proscribe radicalism when it is criminal or illegal (Johnston & Bose, 2020). Nevertheless, deterrence has appeared ineffective or counterproductive in preventing crime or violence (Drinkard et al., 2017). Such counter-production suggests that deterrence invokes radicalism (Tang & Chung, 2022). Whereas the antisocial social norm has induced crime or violence, the social norm for law abidance has not exhibited its preventive effect on radicalism (Mulla et al., 2019). The effects of deterrence and the law-abiding norm on radicalism are thus uncertain and require empirical clarification. Notably, radicalism, law non-abidance, and antisocial norms prevail more in the youth than older people (Farrington, 2018; Liu et al., 2023;

Vanclair & Fischer, 2011). Such youth characteristics are worth addressing (Botticher, 2017). Meanwhile, deterrence and norm theories may lapse when youth do not regard radicalism as criminal or illegal (Ludigdo & Mashuric, 2021). Such a possible lapse prompts the present study to clarify the effectiveness of the social influence of deterrence and the law-abiding social norm for preventing youth radicalism. This clarification thus informs the use of social influence for that prevention.

Radicalism is prevalent in youth to be socially disruptive, possibly causing violence and death (Ludigdo & Mashuric, 2021). Radicalism includes the practice or support for crime, extremism, illegality, and terrorism to achieve social change (Harpviken, 2021). It is adversarial, bumping, coercive, destructive, resistant, threatening, and not merely compromising, peaceful, silent, and sedentary, as in civil disobedience (Sedgwick, 2010; Wong et al., 2019). These features, which can be dehumanizing, plaguing, and polarizing, make radicalism noxious and objectionable (Berger, 2018). Moreover, radicalism can arise from conspiracism, prejudice, racism, and xenophobia, which are socially undesirable (Liu et al., 2023). Hence, radicalism necessitates prevention, including by law, notably for public order or national security (Sjoen & Jore, 2019).

The national security law applies to tackle dehumanizing, terrorism, political crime, and thus their grounds in radicalism against national governance or welfare (Sheikh, 2014). By imposing certain, severe, and swift punishment, this law has deterred aggression, deviance, and opposition and thus alleviated distress and ill-being (Sabia & Bass, 2017). This law is draconic and thus deterrent and threatening because it likens radicalism or violence to imperiling national security (Li et al., 2022; Sheikh, 2014). For such deterrence, the national security law started to apply to Hong Kong in mid-2020 (Baehr, 2022; Lee & Chan, 2023). The effectiveness of this law is a notable concern because the central government of China newly imposed the law on the regional administration of Hong Kong, where freedom from charges about national security had prevailed before (Baehr, 2022). Hong Kong youth are particularly fond of such freedom and are resistant to the national security law (Wong et al., 2022).

Meanwhile, their radicalism rests on localism, which glorifies Hong Kong and its autonomy, independence, and separation from China (Zamecki, 2018). Such resistance to and separation from China casts doubt on Hong Kong youth’s acceptance of and compliance with the national security law and, thus, its effectiveness (Wong et al., 2022). Notably, such resistance persists after national security law enforcement to fuel radicalism, such as incitement and preparation to attack the government or police with firearms for the separatist cause (Leung, 2022; Wong & Cheung, 2021). Nevertheless, radicalism diminishes in intensity, scope, and publicity after the enforcement (Lo & Hung, 2022; Zheng, 2020).

Awareness of national security law enforcement, including prosecution and sentencing, necessarily registers the effectiveness and, thus, the consequences of the law. Generally, awareness is a precondition for attitude and practice (Tan et al., 2022). Resultant attitudes include acceptance, belief, interest, and norm-setting, whereas resultant practices include civic, political, and social participation and transgression (Jovarauskaitė et al., 2020). According to a review of research in diverse contexts, awareness of pandemic law has fostered legally compliant, preventive, and protective practices (Bish & Michie, 2010). Besides its instrumental functions, awareness is commendable for realizing people’s capability, enlightenment, freedom, and rationality (Jin & Chen, 2020). Accordingly, the law is effective because of its awareness to uphold capable, enlightened, rational, and voluntary compliance.

Agreement on compliance or abidance by law is instrumental to securing a healthy and lawful instead of a risky life (Drinkard et al., 2017). Besides, such agreement is valuable for realizing the social contract, convention, rationality, and dutifulness and fairness underlying citizenship (Bertea, 2019). This agreement represents a personal asset or strength for development (Drinkard et al., 2017).

Awareness of people’s support for law abidance registers the corresponding social norm, which means an appropriate and widespread practice. Such awareness is valuable for showing or suggesting citizenship, knowledgeability, learning, morality, prosociality, social capital, and socialization (Bicchieri, 2017). Moreover, awareness of the legal or prosocial norm is instrumental to academic achievement and moral and socially responsible practices (Yang, 2017). This awareness is also a developmental asset (Benson et al., 2006). Furthermore, this awareness is natural and ubiquitous for socialization or social influence through social comparison or reference to prevent deviance (Mayet et al., 2012).

Deterrent and Normative Prevention of Radicalism

Radicalism is preventable by the social influence of awareness of national security law enforcement and the social norm for law abidance, according to deterrence and norm theories. Deterrence theory holds that awareness of punishment or related enforcement for criminal or illegal activity certainly, severely, and swiftly generates fear or a felt threat to deter the activity (Hirtenlehner & Schulz, 2021). This theory rests on the supposed rational calculation of the cost of punishment or related enforcement. The theory has expounded deterrence against illegal activity in general (Sweeten et al., 2013). In this study, such awareness concerns national security law enforcement. Meanwhile, norm theory states that awareness of the social norm or custom engenders conformity to the norm (Waterman et al., 2022). This theory presumes social sensitivity to and assimilation of the social norm of collective consciousness. The theory has warranted the effects of illegal and prosocial social norms on illegal and prosocial activity, respectively (Canlas & Molino-Magtolis, 2023). In this study, such awareness concerns people’s support for law abidance. Hence, deterrence and norm theories and effects are distinguishable and independent. However, awareness of deterrence and the social norm concern law, which can be a common denominator confounding the effects of deterrence and the norm. In other words, deterrence and the norm can be correlated (Merhi & Ahluwalia, 2019). To avoid such confounding, the study needs to examine deterrence and norm effects simultaneously with the following hypotheses.

Awareness of national law enforcement is likely to deter radicalism based on deterrence theory. Such deterrence has arisen from such enforcement as legal education and intervention (Amjad, 2009). Moreover, radicalism has declined with awareness of its cost (Ebers & Stephan, 2022). Similarly, violence as a feature of radicalism has diminished with disciplining or punishment (Levitt & Lochner, 2001). Violence has also declined with awareness of arrest (Berger, 2005). More broadly, illegal activity as a feature of radicalism has eroded with such enforcement as incarceration and other punishment and their risk or threat (Hirtenlehner & Schulz, 2021).

Awareness of people’s support for law abidance is likely to discourage radicalism based on norm theory. Typically, radicalism has risen with awareness of social norms for radicalism, including martyrdom and violence (Ferguson & McAuley, 2020; Ismail et al., 2022; Miconi et al., 2020; Tang et al., 2020). Violence has similarly evolved with awareness of social norms for delinquency, violence, or even victimization (Palacios et al., 2019). More generally, illegal activity has sprung with awareness of social norms for antisociality, crime, delinquency, deviance, violence, or illegal activity (Hirtenlehner & Schulz, 2021). Conversely, illegal activity has shrunk with awareness of the social norm for morality (Brugman et al., 2003).

These two hypotheses require testing, considering their empirical uncertainty. Notably, deterrence theory has yet to secure clear empirical support (Dolling et al., 2009). Accordingly, radicalism has risen with awareness of imprisonment, repression, and threat as ways of deterrence (Ferguson & McAuley, 2020; Pfundmair & Mahr, 2023). Violence has also escalated with arrest, conviction, disciplining, and punishment as deterrence processes (Antunes & Ahlin, 2014). Likewise, illegal activity has increased with awareness of arrest, incarceration, involvement in or intervention from the justice system or police, legal sanctions, and penalties as deterrence processes (Hutcherson, 2012). Radicalism can also arise without considering its cost (Mukhitov et al., 2022). Hence, the punishment received has raised retaliation and violence (Antunes & Ahlin, 2014). Moreover, the expectation for punishment has not affected violence and illegal activity (van Rooij et al., 2017). Therefore, national security law enforcement awareness may instigate radicalism as a backfire (Lee & Chan, 2023; Kobayashi et al., 2021; Vickers & Morris, 2022).

Testing the hypotheses needs to control for agreement on law abidance and background characteristics, possibly confounding the hypothesized relationships. Such agreement or support for the law has impeded radicalism, illegal activity, and violence (Muthuswamy, 2022). The agreement may also raise awareness through selective attention (Sohn et al., 2012). Meanwhile, radicalism has been higher in the youth who has been irreligious, locally born, male, or younger than in the other (Harpviken, 2021). Additionally, illegal activity has been less frequent in the married youth than in the other (Miconi et al., 2020). Awareness has been higher in the youth with higher education, male gender, or native status than in the other (Ten Dam et al., 2020). Hence, these background characteristics may exert common influences on the awareness, agreement, and radicalism, thus confounding relationships among the latter.

Situational Deterrence

The deterrent effect is likely to be weaker in the presence of other factors, according to situational (action) theory (Hirtenlehner & Schulz, 2021). This theory reasons that the other factors interfere with the situational calculation of punishment cost. Hence, self-control and friends’ morality have diminished the deterrent effect of expected punishment (Hirtenlehner & Schulz, 2021). This theory thus regards the deterrent effect as situational. Based on this theory, awareness of national security law enforcement is less likely to prevent radicalism when agreement on law abidance or awareness of support for law abidance is higher. In other words, the inverse effect of awareness of the enforcement is likely to be weaker when the agreement or awareness of the support is higher. Nevertheless, such situational deterrence requires empirical examination, considering the contradictory findings available (Pauwels et al., 2011).

The examination in Hong Kong is felicitous for its bridging role in exchanging upgraded knowledge between the East and West (Lui, 2014). This role builds on cultural, economic, and political compatibility between Hong Kong and other places because of its geography and history as the entrance of China and East Asia (Zamecki, 2018). Meanwhile, Hong Kong integrates administrative and cultural features from other places, making it distinctive and crucial for testing and generalizing knowledge developed from others. Such distinctiveness lies in emerging radicalism in massive protests featuring arson, blocking, vandalism, and violence fueled by localism, nativism, and media connectivity (Tang et al., 2020; Zamecki, 2018). Radicalism is glaring and nascent as it defies harmony, lawfulness, and peacefulness conventionally practiced according to Chinese or Confucian ethos in Hong Kong (Lee & Chiu, 2018). The national security law and its enforcement are also new to Hong Kong since its return to China (Baehr, 2022; Lee & Chan, 2023). Such changes in radicalism and law amplify the significance of building new knowledge by scrutinizing the relationship between radicalism and law abidance, notably regarding national security.

To recapitulate, the following two hypotheses about the social influence on the youth’s radicalism derive from deterrence and norm theories.

  1. 1.

    Radicalism declines with awareness of national security law enforcement.

  2. 2.

    Radicalism declines with awareness of support for law abidance.

Method

In the fourth quarter of 2022, a random-sample telephone survey of 883 Chinese youths (aged 18–24 years) in Hong Kong provided data for statistical analysis. Such sampling started with a random extraction of telephone numbers from the territory-wide residential telephone directory database, followed by randomly selecting a youth, if available, in each residence contacted through the telephone numbers. The household telephone survey has been more representative and workable than the mobile telephone survey, which has been biased, interrupted, costly, and inadvisable (Badcock et al., 2017; He et al., 2018). A trained survey interviewer ensured the anonymity, confidentiality, quality, and voluntariness of the selected youths’ informed consent and responses to the survey, according to the procedure approved by a research ethics committee. The survey contacted 1,499 residences with Chinese youths on weekday evenings and weekend daytime and evenings to attain a response rate of 58.9%. This rate was higher than the typical rate of 25%, suggesting youth’s higher cooperativeness (Keeter et al., 2006). The resultant sample with complete data for analysis afforded to test a considerably weak effect (|r| > 0.094) with 95% statistical confidence and 80% statistical power.

Participants

The youths averaged 20.9 years in age and 15.0 years in educational grade (see Table 1). Among them, 55.0% were female, 81.6% were locally born, 13.8% were born in Mainland China, 85.6% were irreligious, 10.9% lived with their spouses, and 3.0% lived alone. These characteristics functioned as control variables for testing the hypothesized effects.

Table 1 Means/percentages and standard deviations (N = 883)
Full size table

Measurement

The survey measured each youth’s radicalism, agreement on law abidance, awareness of support for law abidance, and awareness of national security law enforcement with multiple rating items (see Table 2). The items appeared randomly within sections to minimize biases due to similar preceding items (Tourangeau et al., 2000). Before factor analysis, the items scored on a 0-100 scale, with 0 for the lowest level of “none” and 100 for the highest level of ‘very much.” Remarkably, self-reported radicalism, law-abidance, and antisocial and prosocial norms have been valid for predictive analysis (Kennedy et al., 2018; Shahzalal & Adnan, 2022; Vitaro et al., 2015; Wong et al., 2019).

Table 2 Standardized factor loadings
Full size table

Radicalism in the previous month, as the outcome of the analysis, combined six items, such as the acts of “supporting radical actions” and “bumping in support of demonstrations” (Wong et al., 2019). The internal consistency reliability (α) was 0.911.

Awareness of national security law enforcement in the previous semi-year as the predictor of radicalism combined five items, such as those concerning awareness of “education for the national security law” and “prosecution under the national security law” (Shek et al., 2022a). This measure thus covered the implementation and consequences of the national security law in Hong Kong. The internal consistency reliability (α) was 0.814.

Awareness of support for law abidance in the previous semi-year as the predictor of radicalism combined four items, such as those concerning people’s support for “law abidance being fair” and “law abidance being national” (Akers, 1998). The internal consistency reliability (α) was 0.797.

Agreement on law abidance in the previous semi-year as the predictor of radicalism combined three items, such as those concerning “law abidance being fair” and “obeying the law to avoid confusion” (Akers, 1998). The internal consistency reliability (α) was 0.657.

Analysis

Testing the hypotheses required measures free of the method artifact of acquiescent rating (Podsakoff et al., 2003). Such measures were the trait factors distilled from confirmatory factor analysis to be distinct from the method artifact factor (using robust maximum likelihood estimation via Mplus, Muthen & Muthen, 2006). This identification thus verified the factorial validity of the measures, with the convergence of items to their respective discriminable factors.

Hypothesis testing using the trait factors tackled the potential endogeneity (i.e., intercorrelation) problem among the factors by two-stage least squares (2SLS) estimation (Jaeger, 2008). This estimation required the first stage to use background characteristics as instrumental variables to generate predicted trait factors for estimation in the second stage, focusing on predicting radicalism as the latest outcome.

Assessing the stability of the hypothesized effects across conditions engaged interactions between awareness of national law enforcement and support for law abidance and others. Each interaction was a product of two predictors, based on their standard scores, to minimize the problem of multicollinearity in estimation (Dunlap & Kemery, 1987).

Results

Before factor analysis, the youth’s radicalism, on average, was not low (M = 35.5 on a 0-100 scale, see Table 1). On average, the youth’s agreement on law abidance and awareness of support for law abidance and national security law enforcement were modest (M = 46.5–50.6). Hence, there was room for attenuating the youth’s radicalism and raising the youth’s law abidance.

The trait factors of radicalism, agreement on law abidance, and awareness of support for law abidance and national security law enforcement were identifiable from confirmatory factor analysis. This analysis showed the convergence of items on their respective factors substantially constrained to be independent of the method factor (λ = 0.649-0.706 for radicalism, λ = 0.351-0.521 for awareness of national law enforcement, λ = 0.315-0.452 for awareness of support for law abidance, and λ = 0.594-0.606 for agreement on law abidance, see Table 2). Such factorial validity stood on the good fit of the analysis (L2(718) = 1,869, SRMR = 0.040, RMSEA = 0.027, CFI = 0.944, Marsh et al., 2004).

In the first stage preceding hypothesis testing, background characteristics were generally significantly predictive of agreement on law abidance and awareness of support for law abidance and national security law enforcement. Most notable was the lower agreement (β = − 0.170, see Table 3) and awareness about law abidance (β = − 0.154) and the higher awareness of the enactment (β = 0.155) in the locally born youth than the other. Additionally, the agreement was higher in the youth living with the spouse or alone, younger, higher in education, or not locally born. Awareness of support for law abidance was lower in the youth who was locally born, irreligious, or not living alone. By contrast, awareness of national security law enforcement was higher in the youth who was locally born, religious, higher in education, or younger.

Table 3 Standardized effects on agreement and awareness
Full size table

Hypothesis testing supported Hypotheses 1 and 2 concerning the decline of rationalism with awareness of national security law enforcement and support for law abidance. Supporting Hypothesis 1, awareness of the enforcement in the previous semi-year displayed a significant strong inverse effect on radicalism in the previous month (β = − 0.492 and − 0.496 based on OLS and 2SLS estimations, see Table 4). Similarly, supporting Hypothesis 2, awareness of support for law abidance in the semi-year manifested a significant strong inverse effect on radicalism in the previous month (β = − 0.367 and − 0.572 based on OLS and 2SLS estimations). These effects were net of the significant moderate inverse effect of the agreement on law abidance in the previous semi-year on radicalism in the previous month (β = − 0.265 and − 0.351 based on OLS and 2SLS estimations). Additionally, the hypothesized effects were net of those of background characteristics. Without controlling the agreement and awareness, radicalism was significantly higher in the youth with lower education, older age, native status, male gender, not having religious faith, or not living with a spouse or living alone (see Table 4).

Table 4 Standardized effects on radicalism
Full size table

The hypothesized effects hardly varied substantially across conditions, even though some variations were statistically significant. Only two variations or interaction effects displayed adequate statistical power (β > |0.068|). One was the positive effect of the interaction between agreement on law abidance and national security law enforcement awareness on radicalism (β = − 0.085, see Table 5). The effect suggested that the agreement weakens the inverse effect of that awareness on radicalism. Similarly, the effect of the interaction between awareness of support for law abidance and awareness of national security law enforcement on radicalism was significantly positive (β = 0.054). That is, awareness of the support weakened the inverse effect of awareness of the enforcement. Another sufficiently powerful effect was the positive effect of the interaction between education and enforcement awareness on radicalism (β = 0.084). This interaction effect suggested that education weakened the inverse effect of awareness of the enforcement.

Table 5 Additional alternate standardized interaction effects on radicalism
Full size table

Discussion

The statistical analysis supported the two hypotheses that radicalism declines with awareness of national security law enforcement and law abidance. Additionally, agreement on law abidance increased with awareness of national security law enforcement, given the control for background characteristics. Moreover, these characteristics did not substantially alter hypothesized relationships. These findings emerge from Hong Kong youth having modest levels of awareness of national law enforcement and law abidance and agreement on law abidance and a lower but nonnegligible level of radicalism on average (see Table 1). Such levels echo similar findings and reflect their radical bases (Tang & Chung, 2022; Yu et al., 2023; Zheng, 2020). For instance, the locally-born youth reported higher radicalism and awareness of national security law enforcement and lower awareness of and agreement on law abidance than the other youth (see Tables 3 and 4). These differences reflect the influences of localism, which evolves from locally-born status (Tang & Chung, 2022; Zamecki, 2018).

More specifically, hypotheses that radicalism declines with awareness of national security law enforcement and people’s support for law abidance attained support. The hypothesized inverse effects were strong (|β| > 0.4) even adjusted for endogeneity and considerably stable across conditions (i.e., absolute standardized interaction effect < 0.1), thus endorsing deterrence and norm theories underlying the hypotheses. Accordingly, deterrence theory and norm theory, respectively, warrant the inverse effects of awareness of national security law enforcement and awareness of support for law abidance on radicalism. In other words, youth radicalism is preventable by legal deterrence and proscription (Streitwieser et al., 2019).

Awareness Effects

For deterrence theory, awareness of the enforcement is likely to affirm the certainty, severity, and swiftness of the national security law to deter radicalism through threatening (Hirtenlehner & Schulz, 2021). Accordingly, the national security law imposes severe punishment after sure and swift arrests to curb further insecurity (Lee & Chan, 2023; Tung et al., 2021). The law is also an appropriate response to radicalism for quenching its continuation or escalation (Koehler, 2017). The deterrent effect also reflects the youth’s tendency to use the information for hedonistic choices to obtain pleasure and avoid pain (Cote, 2014). Specifically, the deterrent effect reveals the Hong Kong youth’s valuation of enjoyment and support for law enforcement (Wu et al., 2013). According to deterrence theory, these youth’s inclinations operate in deterrence (Pauwels et al., 2011; Schulz, 2014).

The norm effect registers the youth’s gregariousness and relational sensitivity (Cote, 2014). Precisely, the norm effect echoes the Hong Kong youth’s interdependence, social sensitivity, and susceptibility to media influence (Wong & Au-Yeung, 2018). Such interdependence, sensitivity, and susceptibility establish peer and media influences as norms to propagate and raise radicalism (Tang et al., 2020; Zheng, 2020). According to norm theory, these youth’s tendencies underlie the norm effect (Ramiro et al., 2019). Additionally, the norm for law abidance is likely to be normative because the law is typically normative, universally applicable, and forceful (Bertea, 2019). To the youth, the social norm also represents an asset conducive to development and socialization, thus meriting the youth’s conformity (Mayet et al., 2012). The youth cares about such conformity to achieve social approval (Massengale et al., 2017). While social norms for deviance influence the youth, social norms for morality or prosociality are at least equally influential because their conformity draws social support to become normative (Bryan et al., 2014).

Albeit weak, the statistically significant variations or the interaction effects illustrate situational deterrence, ensuing from the elaboration of deterrence theory by situational theory. The significant reductions in the deterrence effect due to education, agreement on law abidance, and awareness of support for law abidance indicate interference with the situational consideration of deterrence based on awareness of national security law enforcement. These reductions indicate that education, the agreement, and the social norm for law abidance preempted the deterrence effect. Such preemption echoes those driven by morality and self-control covered in situational theory (Schepers & Reinecke, 2018). Accordingly, education and law abidance are commensurate or compatible with morality and self-control (Drinkard et al., 2017; Hudson & Brandenberger, 2022).

Among other significant but weak variations or interaction effects, the strengthened effect of the interaction or combination of agreement on law abidance and awareness of support for law abidance notably exhibits the advantage of the combination. Accordingly, the attitude and norm are complementary to each other. The complementarity effect grows with compatibility between the attitude and norm (Ajzen & Fishbein, 2014). More generally, this effect escalates with the fit between the person and environment, respectively corresponding to the attitude and norm (Kwon, 2019). As such, the norm and attitude or agreement reinforce the effect of one another.

Background Effects

Significant but slight variations in the effects of the agreement and awareness on radicalism due to the birthplace (i.e., being locally born vs. others) or native status are noteworthy. Accordingly, native status strengthened the effects of the agreement on law abidance and awareness of national security law enforcement and weakened the effect of awareness of support for law abidance. Such variations reflect the specific influence of Hong Kong due to Westernization (Wang & Chang, 2010). This influence enhances the individualistic effects and erodes the collectivistic effect (Ma, 2019; Narvaez, 2014). Hence, the individualistic effects of agreement and deterrence strengthened with native status, and the collectivistic effect of the social norm weakened with native status. Additionally, agreement and awareness about law abidance were significantly lower, and national security law enforcement awareness was significantly higher in the native than the others. Such differences are again attributable to the Westernization of Hong Kong natives to avoid legal control (Narvaez, 2014).

Limitations and Future Research

Limitations in the cross-sectional design, self-report measurement, and sampling in a Chinese Hong Kong society happen to the study. This design could not ensure the temporal order for prediction, and the measurement might not be objective enough. Hence, the effects of the agreement and awareness on radicalism might occur concurrently, even with the adjustment for endogeneity, and the measurement best reflected perceptions. Moreover, the sample in Hong Kong might be distinctive in its incorporation of Western culture into Chinese tradition. As such, the hypothesized relationships exhibited slight differentials between the native and other, suggesting the enhanced individualistic effect and abated collectivistic effect due to Westernization. These limitations and possibilities require future research to address. Future research can employ experimental and panel designs to strengthen prediction and causal inference, multiple measurements to strengthen objectivity, and samples from diverse sociocultural contexts to strengthen generalizability. The experimental design can communicate support for law abidance and national security law enforcement and induce agreement on law abidance, typically in a controlled or laboratory setting to assess the change in radicalism. Alternatively, the panel survey can measure the agreement, awareness, and radicalism repeatedly to gauge the change in radicalism in a real-life setting. The measurement can combine multiple sources, including the youth and his or her associates, to boost objectivity. Notably, measuring the impact of the law-abidance norm directly can supplement the statistical analysis of the impact. Future research can utilize samples from diverse contexts to scrutinize the generality and specificity of findings and their variations due to contextual factors, including culture and Westernization.

Theoretical Implications

The inverse effects of awareness of national law enforcement and law abidance on radicalism warrant the relevance of deterrence and norm theories to deradicalization in youth. Such relevance rests on the rational calculation and social sensitivity and responsivity underlying the theories, respectively (Hirtenlehner & Schulz, 2021; Waterman et al., 2022). This relevance also dismisses contradicting views about the counterproductive effects of deterrence and the legal norm. Moreover, the weakened deterrence effect of awareness of national security law enforcement due to law abidance and its social norm supports the situational formulation of deterrence theory. This formulation reveals that norm conformity partly preempts the rational calculation underly deterrence. Additionally, both deterrence and norm effects of the law benefit from mediation by agreement on the law. These moderation and mediation effects thus enrich the generating mechanisms of deterrence and norm theories.

Practical Implications

For preventing youth radicalism, law abidance, its social norm, and national security law enforcement, including education, prosecution, and sentencing, are recommendable for promotion. Such promotion builds on deterrence and norm theories to establish the deterrent and normative effects of law abidance and national security law enforcement. Accordingly, the enforcement needs to be certain, severe, and swift, and law abidance needs to prevail. Crucially, the enforcement and norm need to be the focus of youth’s awareness, which, in turn, needs to be the focus of promotion. Specifically, promoting enforcement awareness is more preventive for youth with lower education, agreement on law abidance, or awareness of support for law abidance. This promotion builds on situational deterrence informed by situational theory. Establishing the law, including its enforcement, abidance, and corresponding social awareness, is the foundation for preventing radicalism (Zhou, 2017). Additionally, such prevention hinges on establishing the law as socially normative (Clubb & O’Connor, 2019).

Education is helpful for deradicalization as education has proven effective in raising abidance by the national security law (Shek et al., 2022). Such education can facilitate rational calculation and social sensitivity to learn about the merit and norm or popularity of the national security law to capitalize on deterrence and norm effects. This facilitation benefits from discussion, explanation, sharing, and support among educators and youth (Grygiel, 2013; Seal & Harris, 2016; Shek et al., 2022). Accordingly, discussion, explanation, and sharing about the merit and norm of the national security law and its abidance among supportive people helps nurture the abidance. This nurturing aims to advance intelligence and wisdom, including the social aspects of responsiveness and responsibility (Grygiel, 2013; Sternberg, 2003). Hence, such advancement is to make the abidance rationally and socially viable.

Preventing radicalism is more needed for youths who are older, lower in education, locally born, male, irreligious, or not living with spouses or living alone. This need is observable elsewhere (Del Pino-Brunet et al., 2023; Harpviken, 2021; Jakubowska et al., 2021; Miconi et al., 2020). Such prevention requires promoting youth’s agreement on law abidance, awareness of support for law abidance, and awareness of national security law enforcement. Promoting the agreement is higher in need for youths who are older, locally born, lower in education, or not living with spouses or not living alone. This need also happens elsewhere (Dong & Zeb, 2022; Drinkard et al., 2017). Promoting awareness of support for law abidance is more needed for youths locally born, irreligious, or not living alone. Likewise, promoting awareness of national security law enforcement is higher in need for youths who are irreligious, older, lower in education, or not locally born.