The youth’s participation in community service is a viable but under-researched way to prevent youth violence (Ballard et al., 2015; Thomas & Sanderson, 2011). Violence is an intended act to devastate the life or property of other people, including those of public authorities and private spheres (Stoddard et al., 2011). Possibly because of their pubertal development, youth are likely to be perpetrators of violence (Farrington, 2007). Thus, youth violence is a concern for prevention (Luengo Kanacri et al., 2020). Meanwhile, the youth’s participation in community service such as helping disadvantaged people presumably plays a role in socializing, prevention, and development for the youth and society (Ballard et al., 2015). Such a role is a reasonable factor for preventing violence (Brezina & Agnew, 2013). Nevertheless, empirical and theoretical bases for violence prevention are yet uncertain, notably in view of findings contradicting the prevention (Drummond et al., 2011; Wright & Fitzpatrick, 2006). Such findings have indicated that community involvement foments violence perpetration. Dispelling the uncertainty requires the scrutiny of main and conditional effects of the participation on violent perpetration. These effects are explicable with empowerment theory, which posits that empowerment in the sense of erasing powerlessness is effective for violence prevention (Bittle, 2002; Krahe, 2014). Empowerment means a process to realize and enhance personal and/or collective power through participation and action (Buelens et al., 2017. Ekholm & Lindström Sol, 2019). Participation in community service is a form of empowerment (Buelens et al., 2017; Knapp et al., 2010). More specifically, the participation is empowerment when it addresses such relative powerlessness as residing in poor housing and female gender. For the scrutiny of the main and conditional preventive effects, the present study relies on panel data on randomly sampled youths in Hong Kong, China. The social ecology of Hong Kong features a dense structure of population, services, information, and their interactions and influences, which permeate inhabitants’ urban life (Lee & Ting, 2015). The study thereby seeks to illustrate the ecological validity (i.e., in real life) of the preventive effect of community service participation.

The study of youth violence prevention is vital and necessary. In the first place, violence needs prevention because of its severe harm to its victim and perpetrator, and society as well (Stoddard et al., 2011). That is, all such violence as assaulting, hitting, injuring, and vandalizing is harmful (Vitaro et al., 2013). Moreover, the perpetrator would feel shameful, guilty, depressed, rejected by others, and perform poorly (Uusitalo-Malmivaara, 2013). Violence also creates anxiety in society (Binken & Blokland, 2012). Moreover, youth violence tends to reflect deeper problems such as the criminal or gang activity (Binken & Blokland, 2012). The victims of youth violence are also more likely to be youth than are others (Hansen-Nord et al., 2014). For recent years in Hong Kong, for instance, youth violence has often targeted police during mass turbulence (Shek, 2020). Youth violence perpetration happens, according to empowerment theory, when power to control antisocial behavior or perform prosocial behavior is deficient (Diamond et al., 2018; Higgins et al., 2005). Public policy and practice are thus imperative to address the violence (Thomas & Sanderson, 2011). Herein, mobilizing youth’s community service is one supposed approach (Thomas & Sanderson, 2011). More specifically, the service organizer invites youth to buttress safety or prevent violence in the community (Ballard et al., 2015; Teixeira, 2016).

Participation in community service includes performing volunteer service, serving the community, advising on youth policy, preventing crimes in the community, and developing social capital in the community (Lanero et al., 2017; Nygreen et al., 2006). In other words, such performance characterizes empowerment with its assertion of ideas and realization of power to tackle community problems (Gullan et al., 2013; Zeldin et al., 2016). The participation is typically a response to invitation by community, government, or social service agencies through school and other channels (Marzana et al., 2012). Such participation has appeared to be intrinsically important and beneficial to the participant as well as the community (Teixeira, 2016). The intrinsic importance resides in the virtues of the participation in altruism, philanthropy, and social inclusion (Duffy & Raque-Bogdan, 2010). In addition, the participation is beneficial in making people socially responsible, connected, and integrated, thus keeping the community safe (Sherrod et al., 2002). Particularly, the benefits happen in consolidation of the participant’s identity, confidence, knowledge, and skills (Coulter-Kern et al., 2013). Such benefits also affirm the empowering role of the participation. The value of the participation thereby draws public policy support for its promotion, including the support in Hong Kong (Planty et al., 2006; Xu & Yang, 2018).

For promoting youth’s participation in community service, it is crucial to invite, mobilize, or engage youth in the participation (Carlson, 2006). The invitation can originate from educational, work, and other settings and through various media. In the urbanized and compact context of Hong Kong, government, social service, and educational agencies are keen to organize community service programs and invite youth to participate in them (Xu & Yang, 2018). Some youth respond to the invitation because they are students or social service users expected or required to participate. For minimizing the bias due to the youth's personal selectivity, such participation is important to demonstrate its contribution and promotability. An additional step to minimize the selectivity is the use of propensity weighting to give higher weights to cases with lower propensities of selectivity (Morgan & Winship, 2007).

Conditions for the Empowerment Effect

Participation in community service is likely to prevent violent perpetration in the youth according to empowerment theory. Empowerment theory is applicable when empowerment tackles the problems of alienation, oppression, marginalization, deprivation, and powerlessness (Christens, 2012). It concerns the presence of the problems or powerlessness to frustrate goal attainment and foment problem behavior (Chapin & Cox, 2001). This concern marks the Marxian origin of empowerment theory (Lee, 2001). That is, empowerment theory posits that empowerment is beneficial in mitigating powerlessness (Speer et al., 2013). To mitigate the frustration and agitation, empowerment functions to make people capable and skillful to solve problems, build relationships, and contribute to society (Langhout et al., 2013). Empowerment can operate in terms of training, enlightening, and opportunity provision for voicing, working, and/or serving (Ekholm & Lindström Sol, 2020). As such, one gains power through accomplishing, coordinating, leading, organizing, decision making, and working generally (Kang, 2016). Consequently, empowerment enables and consolidates self-determination and self-actualization (Gullan et al., 2013). Empowerment also raises power through securing achievement, independence, ownership, reputation, and various resources in the person (Chamodraka et al., 2016). As such, empowerment eventually serves to resist oppression, marginalization, alienation, and related problems (Huiskamp & Hartmann-Mahmud, 2006). As the youth is relatively powerless, his or her development benefits from empowerment (Berkowitz et al., 2006). Specifically, empowerment has made youth socially responsible, autonomous, efficacious, self-sufficient, and receiving social support (Gullan et al., 2013). Empowerment also underlies the development of abilities to solve problems, manage others, start businesses, and do well in life (Kwan & So, 2008). Holding empowerment as a general process to offset powerlessness, empowerment theory thereby explains the contributions of involvement in caring, consultation, critical thinking, decision making, leading, managing, moral education, ownership, power sharing, problem solving, self-determination, service-learning, therapy, and voicing (Ferguson et al., 2011; Gullan et al., 2013; Knapp et al., 2010; Langhout et al., 2013; Liang et al., 2021; McWhirter, 1994; Zeldin et al., 2014). For instance, empowerment theory explains the youth’s civicness with the youth’s awareness and understanding about inequity, criticalness, and having a systemic worldview (Moore et al., 2016).

Participation in community service is a key component of empowerment (Langhout et al., 2013). The participation is integral to the Marxian origin of empowerment in involvement in and building of the community (Reisch et al., 2013). Specifically, the participation realizes the principles of empowerment in social contribution or development, social integration, training, socialization, upskilling, and self-determination in community involvement (Lanero et al., 2017; Nygreen et al., 2006). Notably, the participation reflects empowerment in terms of learning, voicing, self-actualizing, and achieving (Haski-Leventhal et al., 2008). The participation thus heightens and demonstrates capability for prosocial behavior (Pascarella & Terenzini, 2005). At least, the participation reflects receiving the opportunity provided by empowerment to strengthen capability. The participation has manifested achievement in academic, career, and identity fields, building relationships, and making the participant prosocial and socially responsible (Kahne et al., 2013). Notably, the cases of such a prosocial contribution have also happened in Hong Kong (Xu & Yang, 2018). The prosocial practices have been contradictory to violent perpetration (Strohschein & Matthew, 2015). Conversely, participation in community service has prevented the participant from blaming others, being lonely, delinquency, playing truant, abusing drugs, and having relational problems (Gillath et al., 2005). Delinquency, drug abuse, and relational problems connect to violence (Petering et al., 2016).

Key to empowerment theory, empowerment is more effective for one who is more powerless (Chamodraka et al., 2016; Speer et al., 2013). In other words, the more powerless person is more receptive to and thus benefitting more from empowerment. As such, the preventive effect of community service participation is likely to be greater in the more powerless person. Generally, the youth tends to suffer from alienation and powerlessness, as subordinated in school, low occupational position, and insecure status (Cote, 2014; Hurrelmann & Quenzel, 2015; Zamecki, 2018). The suffering has delayed the youth’s achievement (Hurrelmann & Quenzel, 2015). More specifically, the youth’s female gender and living in poor housing are likely to be conditions for empowerment to realize its preventive effect. This relative powerlessness or deprivation stems from status inconsistency between the youth’s desired and given statuses (Hurrelmann & Quenzel, 2015). The female youth tends to suffer more gender discrimination, harassment, and exploitation as a sexual object, and thus deprivation than does the male counterpart (Bay-Cheng, 2015). Hence, the female youth has experienced more rejection, victimization, and stress than has the male one (Pedersen et al., 2017). For instance, the female youth’s relative powerlessness originates from childcare in Hong Kong (Lui & Choi, 2015). Living in poor housing, as an indicator of poverty, is an obvious disadvantage (Anthony & Robbins, 2013). Essentially, living in poor housing may be responsible for violence in the youth resident (Farrington, 2007). Living in poor housing may also impede academic achievement (Dooley & Stewart, 2004). More generally, poverty has been the origin of youth problems and disadvantage, including violence and inadequate social capital (Davis & Shlafer, 2016). For instance, living in such poor housing as a divided unit in an apartment characterizes poverty and disadvantage in Hong Kong (K. Cheung, 2015). Hence, female gender and poor housing, according to empowerment theory, are likely to be conditions that augment the preventive effect of community service participation on violent perpetration. Notably, the prevention is necessary for all youth, regardless of their gender and housing condition (Foshee & Matthew, 2007). Disadvantaged youth, nevertheless, are likely to benefit particularly from community service participation, according to empowerment theory.

Hypotheses

Empowerment theory and related research undergird hypotheses about the contributions of participation in community service to the youth due to invitation by a governmental or non-governmental organization.

  1. 1.

    Participation in community service reduces violent perpetration. That is, the participation has a negative, alleviating main effect on the perpetration.

  2. 2.

    Participation in community service reduces violent perpetration more in the female youth than in the male youth. That is, the coupling of the participation and female gender together has a negative, alleviating interaction effect on the perpetration.

  3. 3.

    Participation in community service reduces violent perpetration more when living in poor housing than in the other. That is, the coupling of the participation and living together has a negative, alleviating interaction effect on the perpetration.

For an adequate test of the effect of community service participation on violent perpetration, statistical controls or adjustments for prior, background, and response characteristics are imperative. Most evident is the need for controlling for prior violent perpetration, which possibly influences the participation and violent perpetration, thus confounding the effect of the participation on violent perpetration. The confounding stems from the stability and antisocial influences of violent perpetration (Vitaro et al., 2007; Vogel & Keith, 2015). Additionally, background and response characteristics are necessary for controlling. The background characteristics include the youth’s age, gender, education, and living in such poor housing as divided housing. On the one hand, these characteristics have affected participation in community service, such that the participation has been higher with higher age or education, or in the female youth and family with a better financial condition (Handy et al., 2016). On the other hand, such background characteristics have been useful for predicting violent perpetration, such that violent perpetration has been higher with lower age or education, higher poverty, or in the male youth than in the female one (Farrell et al., 2017; Vogel & Keith, 2015). The response characteristic of acquiescence or rating indiscriminately highly is certainly a source of method artifact in need of control (Kam et al., 2012). This response characteristic can shape responses to questions about violent perpetration and community service participation and thus confound their relationship. The influence of acquiescence reflects carelessness in rating all items, thus inflating their positive relationships. Another response characteristic in need of control was social desirability, which means the tendency to give socially desirable responses that may not reflect the real situation. Social desirability has affected reports of delinquency and other social practices (Cheung et al., 2001). In addition, a more crucially needed adjustment is for selectivity into the observed level of participation in community service. Such selectivity is identifiable as the match between the observed and predicted participation, with prior violent perpetration and background and response characteristics as predictors. Adjustment for the selectivity applies the counterfactual approach to causal analysis (Morgan & Winship, 2007). Herein, the counterfactual means the absence of selectivity, as identified by statistical adjustment. The analysis thus assigns greater weight to cases with lower selectivity into the participation or nonparticipation.

Hong Kong, as a Chinese metropolis, is suitable for the study for generating findings relevant to international advancement of knowledge. Such suitability rests on the compatibility of Hong Kong with most parts of the world, notably including the East and West. Hence, those issues of violence, community service, and empowerment prevail in Hong Kong as well as most other places (Adorjan & Chui, 2014). Youth violence and its associated worsening of quality of life is increasingly rampant there (Shek, 2020). More importantly, relative deprivation is prominent in youth in Hong Kong, as well as in other places (Cote, 2014; Hurrelmann & Quenzel, 2015; Shek, 2020; Zamecki, 2018). The compatibility stems from its intensive Westernization and globalization (K. Cheung, 2015). Consequently, Hong Kong becomes the hub for interchange of knowledge generated locally and elsewhere (Kuah-Pearce & Fong, 2010). The global flow of knowledge and people would also champion the worth of this Hong Kong study.

Method

Participants

Panel data for analysis emanated from a two-wave survey of 1,710 Chinese youths (aged 16–24 years) in Hong Kong. These youths constituted a random sample, with a telephone survey using randomly selected residential telephone numbers of the whole metropolis. The first wave of the survey conducted in early 2016 initially contacted 3,145 youths of the age range and successfully interviewed 1,733 of them. This Wave 1 survey attained a response rate of 51.1%. Six months later in the same year, Wave 2 survey successfully interviewed 1,710 of the initial sample, showing a retention rate of 98.7%. Notably, the residential telephone panel survey generally could have a high retention rate, despite a lower response rate initially, because of the residential stability, incentive offered, and legitimacy of the survey institution (Bonke & Fallesen, 2010; Stephens et al., 2007; Tourangeau & Ye, 2009). Accordingly, the respondent obtained a shopping coupon sent by mail as an incentive immediately after each response.

The two-wave sample eventually had an average of 20.3 years in age and 11.6 years in education (i.e., about Grade 11) (see Table 1). Among them, 49.6% were female and 6.9% dwelling in divided housing, which was poor rental housing in Hong Kong. Divided housing meant the division of an already small apartment into several compartments for poor people to rent.

Table 1 Means and standard deviations (N = 1,710)
Full size table

Measurement

Measurement of violent perpetration, participation in community participation, and social desirability relied on multiple items (see Table 2). Each item took a five-level scale to generate scores, with 0 for the lowest level, 25 for the second level, 50 for the mid-level, 75 for the fourth level, and 100 for the highest level. Such scoring facilitated interpretation and comparison without any distortion in linear analysis (Preston & Colman, 2000).

Table 2 Standardized factor loadings
Full size table

Violent perpetration in past three months was the average of six items, such as “injuring others” and “vandalism” (Vitaro et al., 2013). Such perpetration has been predictive of rejecting others and predictable by participation in unsupervised activity, friends’ rejection, and socioeconomic deprivation (Vitaro et al., 2007). It held an internal consistency reliability coefficient (α) of 0.699 and 0.726 in Wave 1 survey and Wave 2 survey respectively, based on confirmatory factor analysis.

Participation in community service due to invitation during the year before the survey was the average of five items, such as “serving the community because of invitation to do so,” “advising on youth policy because of invitation to do so,” and “performing volunteer service because of invitation to do so” (Bringle et al., 2004). Such participation illustrated empowerment in realizing power and its impacts (Buelens et al., 2017; Gullan et al., 2013; Zeldin et al., 2016). It has been predictive of and predictable by various forms of civic engagement, such as political participation (Marzana et al., 2012; Simons & Cleary, 2006). It held an internal consistency reliability coefficient (α) of 0.763 in Wave 2 survey, based on confirmatory factor analysis.

Social desirability in past three months was the average of four items, such as “being ready to help others” and “willing to admit mistakes” (Paulhus, 1991). It held an internal consistency reliability coefficient (α) of 0.659 in the Wave 2 survey.

Acquiescence or the tendency to rate everything indiscriminately highly, being a control factor to indicate the response set or method artifact, was the average of all rating items (Podsakoff et al., 2003; Zagorski, 1999). The items included those of violent perpetration, community service participation, and social desirability. Its internal consistency reliability (α) was 0.922 and 0.964 at Wave 1 survey and Wave 2 survey respectively.

Selectivity was the match between observed and predicted participation in community service participation. The predicted participation was that resulting from a linear regression analysis of the participation on prior violent perpetration and background and response characteristics (Morgan & Winship, 2007). Hence, the absolute residual or mismatch between the observed and predicted participation indicated non-selectivity. A lower mismatch, conversely, indicated higher selectivity into community service participation or nonparticipation.

Procedures

The survey proceeded with random sampling of all residential telephone numbers in Hong Kong to contact households for interviewing their Chinese youth members aged from 16 to 24 years. The contact could take an unlimited number of attempts until obtaining definite consent or refusal to the interviews. In case more than one youth in the age range lived in the residence, the youth with the latest birthday was the selected respondent. This counted as a way of random sampling within the household (Salmon & Nichols, 1983). The survey, conducted by seasoned interviewers, operated in the evenings of weekdays and the daytime and evenings of weekends.

Data analysis began with confirmatory factor analysis to verify the three trait factors of violent perpetration, participation in community service, and social desirability, given the method factor of acquiescence altogether (via Mplus, Muthen & Muthen, 2006; Podsakoff et al., 2003). That is, the analysis identified the trait factors by the respective indicators, which also loaded on the common method factor. The analysis aimed to find substantial and higher loadings on the trait factor than on those on the method factor to buttress factorial validity. This validity incorporated convergent validity into the trait factors and discriminant validity for distinguishing the various factors. To accommodate any distribution in the variables, the analysis employed robust maximum likelihood estimation (MLM) (Lei & Wu, 2012; Muthen & Muthen, 2006).

In a hierarchical way, linear regression analysis served to estimate the main and interaction effects of participation in community service, and the main effects of prior violent perpetration, background characteristics, and response characteristics. This identification also identified non-selectivity as the absolute difference between the actual and predicted participation. Such non-selectivity then served as a weighting factor in subsequent analysis to adjust for selectivity, according to counterfactual logic (Morgan & Winship, 2007). This adjustment aimed to prefer cases with high non-selectivity and downplay cases with high selectivity. That is, analysis based on weighted data estimated effects primarily based on cases not selected for community service participation or nonparticipation. This was to minimize the selectivity bias when estimating the effect of community service participation. Such estimation would be comparable to that based on a randomized-controlled experiment (Morgan & Winship, 2007).

Subsequently, linear regression analysis served to test the hypotheses and estimate the various effects on violent perpetration. Herein, participation in community service, prior violent perpetration, and background and response characteristics were basic predictors in the first step of the analysis. Controlling for the response characteristic of acquiescence was necessary to avoid spurious positive relationships among inflating ratings, such as that of violent perpetration. In addition, interactions between the participation with gender and living in divided housing were additional predictors in the second step of analysis. To minimize the problem of multicollinearity in the analysis, the interactions were the multiplicative products of the standard scores of the participation and conditions (Brambor et al., 2006). The analysis operated twice, with and without adjustment for selectivity by weighting cases according to non-selectivity.

Results

Confirmatory factor analysis attained a very good fit to warrant estimates that illustrated the factorial validity of the trait factors of violent perpetration, participation in community service due to invitation, and social desirability. The very good fit was evident in the low standardized root-mean-square residual (SRMR), the high comparative goodness of fit index (CFI) and other indicators (L2(73) = 258, SRMR = 0.031, RMSEA = 0.038, p-close = 1.000, CFI = 0.970; Marsh et al., 2004). More importantly, loadings on the trait factors were substantial and stronger than were those on the method factor of acquiescence (see Table 2). Accordingly, loadings ranged from 0.394 to 0.653 on the trait factor of violent perpetration; from 0.458 to 0.674 on the trait factor of the participation; and from 0.386 to 0.635 on the trait factor of social desirability. Meanwhile, some loadings on the method factor of acquiescence were also substantial (λ > 0.3). The trait factors were therefore suitable and the acquiescence factor was necessary for hypothesis testing through linear regression analysis.

Linear regression analysis showed that prior violent perpetration and almost all background and response characteristics were significantly predictive of participation in community service during Wave 2 survey. Specifically, the participation was significantly greater when age (β = 0.120), acquiescence (β = 0.179), or social desirability (β = 0.167) was higher (see Table 3). The participation was significantly lower when prior violent perpetration (β = -0.123) or education (β = -0.057) was higher. Moreover, the participation was significantly higher in the female youth than in the male one (β = 0.062). The analysis enabled the identification of selectivity and non-selectivity into the participation for weighting cases according to counterfactual logic.

Table 3 Standardized effects on community service participation due to invitation
Full size table

Linear regression analysis then proceeded with the analysis of violent perpetration with and without weighting cases for the selectivity adjustment. The analysis showed support for all the three hypotheses, with the control for prior, background, and response characteristics. First, Hypothesis 1 about the negative effect of participation in community service on violent perpetration found support from the corresponding significant negative effects (β = -0.081 & -0.152, see Table 4). Notably, the effect held with or without adjustment for selectivity. Second, Hypothesis 2 about the stronger negative effect of the participation on violent perpetration in the female than in the male one obtained support from the corresponding significant interaction effect (β = -0.059 & -0.061). That is, the participation was helpful to prevent more female perpetration than male perpetration. Third, Hypothesis 3 about the stronger negative effect of living in divided housing on violent perpetration attained support from the corresponding significant interaction effect (β = -0.063 & -0.069). That is, the participation was more helpful to prevent violent perpetration by the poorer youth.

Table 4 Standardized effects on violence, weighted and not weighted by non-selectivity into invited community service participation
Full size table

The effects and their related hypothesis testing were acceptable because the analysis controlled for prior violent perpetration and background and response characteristics. Clearly, prior violent perpetration during Wave 1 was a significant predictor of violent perpetration at Wave 2 (β = 0.253 & 0.241, see Table 4). Moreover, violent perpetration was significantly higher when acquiescence was higher (β = 0.486 & 0.461) and social desirability was lower (β = -0.108 & -0.079). Violent perpetration was also significantly higher in the youth living in divided housing than in the youth not living in the housing (β = 0.080 & 0.056). These significant effects indicated the need for controlling for prior violent perpetration, and background and response characteristics.

Discussion

Results support all the hypothesized main and interaction effects of participation in community service on violent perpetration, with the control for prior perpetration and adjustment for selectivity into the participation. This support illustrates the relevance of empowerment theory to explaining the prevention of violent perpetration through the participation. Essentially, the preventive effect is credible because it resulted from statistical control and adjustment. The results thus suggest that the participation represents empowerment to prevent violent perpetration in the youth, who suffers from relative deprivation, status inconsistency, or powerlessness. This suffering is particularly evident in the youth who was female or living in poor, divided housing. Moreover, the preventive effect was stronger with adjustment for selectivity than without the adjustment. That is, the youth with lower selectivity to participate benefited more from the participation. This benefit illustrates the contribution of empowerment to provide new knowledge and skills to the participant (Taylor, 2011). Overall, the results are important to refute the view that community service and empowerment are irrelevant to fomenting violence (Ballard et al., 2015; Drummond et al., 2011; Van Ryzin & Dishion, 2013). Instead, the results reveal that community service is a form of empowerment that is helpful to prevent youth violence.

The preventive effect of participation in community service is justifiable by the merit of community service as a way of empowerment. Accordingly, the participation is meaningful, fruitful, and rewarding to gather the strength in community service and violence prevention. That is, the participation is helpful to community service and thus demonstrates and hones the capability for the service (Nygreen et al., 2006). Community service is beneficial to the participant possibly because of the strengthening of social capital through the service (Larson et al., 2006). Social capital in turn is beneficial because of its reciprocity and stability (Ferguson, 2006). Reciprocity means that the participant receives help after helping others and stability means that the help is predictable and trustworthy. Both reciprocity and stability reflect the social nature and thus norms to maintain enduring social relationships (Piehler & Dishion, 2007). These properties also sustain the prosocial effect of community service and discourage the antisocial effect of violence. Accordingly, community service aims to enhance reciprocity, cohesion, mutual care, and conflict resolution in the community, as opposed to hurting others through violence (Bridger & Luloff, 2001). More specifically, community service has helped prevent youth violence (Payne, 2006). Because of the social nature of reciprocity inside the community, the prevention can also happen to the youth participant in community service. Hence, by participating in community service generally and violence prevention specifically, the youth can present his or her own violent perpetration.

The preventive effect of community service participation or empowerment is also justifiable with the powerless conditions found in the youth with female gender or living in divided housing. Accordingly, empowerment particularly applies to raising the power of the powerless youth. In this connection, powerlessness largely appears to represent relative deprivation or status inconsistency. Meanwhile, the youth aspires for agency, autonomy, and ideals (Cote, 2014). Hence, the youth is sensitive to relative deprivation or status inconsistency and thus suffers from powerlessness.

Limitations and Future Research

The study has an obvious limitation in reliance on a single Chinese metropolis. This limitation suggests that present findings best apply to places like Hong Kong in terms of urban and sociocultural conditions. Such conditions would heighten social concern, connection, and influence, such as mobilizing youth for community service (Cui et al., 2016). For instance, the dense and somewhat poor residential conditions in Hong Kong expedite social contact and communication (K. Cheung, 2015). Thus, places with such structure and orientation for community are likely to share the preventive effect of community service. Future research is thus necessary to clarify the contextual influence by incorporating data from multiple contexts for comparative, contextual analysis. That is, such research needs to gauge contextual factors from different contexts and scrutinize their influences on the preventive function of community service participation.

A fuller design in future research should incorporate measures of empowerment and other components of empowerment theory to substantiate its explanation for the preventive effect of participation in community service. The empowerment can cover provided opportunities to coordinate, lead, make decisions, organize, plan, and voice, as well as perform community service (Gullan et al., 2013; Roth et al., 1998; Zeldin et al., 2014, 2016). By mending the limitation in measuring empowerment in the present study, future research can demonstrate that the participation undergoes a process of empowerment to strengthen capability and successively contain violent perpetration. The process can include enlightening, facilitating, opportunity giving, and training to become a helper or leader (Ekholm & Lindström Sol, 2020). Future research also needs to demonstrate the essential premise of empowerment theory that empowerment is helpful particularly when offsetting powerlessness. This means that future research ought to show that female gender and living in poor housing engender powerlessness and successively enhance the preventive effect of empowerment.

Implications

Given youth violence as typically problematic and in need of prevention, inviting youth to participate in community service is a helpful means for the prevention. The invitation is justifiable because youth not inclined to participate can benefit more from the invited participation. Moreover, youth who are female or living in divided housing are particularly important targets for the invitation. With respect to empowerment theory, the invited participation needs to be meaningful, fruitful, and rewarding to upgrade power and remedy powerlessness in youth. This means that community service needs to emphasize empowerment or capacity building to enhance participants’ skills to prevent violent perpetration in the community and themselves (Payne, 2006). It also requires that community service is to resolve rather than provoke conflicts (Ohmer & Brooks, 2013). Such community service can develop social enterprises to enhance its work relevance and financial viability (Taylor, 2011). In this connection, a social enterprise means one organizing formal, committed, and financially sustainable work for addressing societal or community needs. Notably, the social enterprise for community service can seek to improve employment and housing (Connelly et al., 2011). Community service based on social enterprises would ensure its meaning and reward. Meanwhile, safeguarding the reciprocal and prosocial features of community service is also vital for capitalizing on its preventive benefit (Reisch et al., 2013). For such safeguarding, incorporating collaboration is helpful in community service (Hendricks & Rudich, 2000). In all, community service needs to cultivate the norms of reciprocity, collaboration, and prosociality (Gravenkepper, 2007).