1 Introduction

Along with imparting knowledge, schools are expected to educate children and youth toward their future role as engaged citizens of democratic societies and active participants in the political and social processes. The trajectory of political development is essentially a gradual acquisition of knowledge, attitudes and skills necessary for active participation in democratic life (Saha 2000; Schine 2001). Hence, it is the role of the school to instill attitudes and nurture behaviors deemed appropriate for democratic citizenship (Hurn 1985). In this capacity, schools contribute to students’ personal development as well as to societal cohesion and solidarity.

The challenge of educational policies, then, is to negotiate a fine line between academic achievement and the civic and political development of children and youth. In recent decades, the pendulum has shifted away from civic education toward academic learning and achievement, a change of emphasis partly attributable to globalization trends that are also reflected in the proliferation of international achievement tests (e.g., Callahan et al. 2010). This paper is premised on the understanding that the above shift does not in any way diminish the importance of the role of schools in citizenship education. Moreover, it offers a new perspective on the school’s mission in shaping students’ civic behavior.

Our basic assumption is that expressions of civic behavior in school are indicative of students’ civic norms and future conduct as democratic citizens. In this paper, we focus on four facets of students’ civic behavior, the first two reflecting disciplinary misdemeanors, while the other two—civic engagement: (a) academic dishonesty in school; (b) violence in school; (c) extra-curricular school activities; and (d) community voluntary action. These four behaviors were chosen for two main reasons. First, they reflect students’ actual behavior rather than declared future intentions. Second, they encompass both negative and positive conduct. Specifically, we investigate whether and how these four facets of civic behavior are associated with students’ sense of distributive and procedural justice in school.

The sense of justice in school has, in recent decades, attracted a growing academic interest, both theoretically and empirically. This subject has been studied in relation to the teacher-student personal interaction and as part of the effects of the school context (climate). As elaborated below, both these factors have been shown to play a role in students’ socialization, affecting a variety of educational outcomes.

Our investigation contributes to the current understanding of the subject by refining the distinctions among several salient domains. First, we differentiate between perception of procedural justice as an organizational (collective) feature of the school and personal perception of distributive justice. Within the latter, we differentiate between the subjective evaluation of the instrumental (grades distribution) as opposed to relational (teacher-student) domain. We investigate whether and how each of these three aspects of the sense of justice relates to the four kinds of students’ civic behavior enumerated above.

The study was carried out in Israel, a relatively new state characterized by the heterogeneity of its mostly immigrant population and by inner conflicts—national, ethnic and religious—which accentuate the role of schools in the civic socialization of the country’s young generation (Ichilov et al. 2005). The structure of the Israeli educational system, which is mostly public, is an outcome of political arrangements reflecting major societal rifts. Specifically, three separate sectors are operating under the umbrella of a central public education system: Jewish-general (secular), Jewish-religious, and Arab. These sectors differ significantly in respect of their ecological environment, access to resources, definition of the school’s mission, and curricular emphases. Such a variation may have a considerable effect on the relationship between students’ sense of justice and their civic behavior, as well as—directly—on the behavioral outcomes. Accordingly, in our investigation, educational sectors are regarded as contextual factors, on par with other such parameters.

As already stated, our study explores the effect of the sense of distributive and procedural justice on the four stated aspects of students’ attested civic behavior in the school setting, and the possible variation in these effects in the three systemic contexts. The following section lays out the theoretical background of our investigation and elaborates the concepts of (a) the sense of distributive and procedural justice in school; (b) disciplinary misconduct—dishonesty and violence; (c) civic engagement—extra-curricular activity and voluntary community action. Our hypotheses concerning the relationship between the sense of justice and civic behavior at school are formulated next. A section laying out the structure of the educational system in Israel ensues, followed by a description of the research design, including the sample, the variables and the method of analysis. The findings are presented in the next section, and are then summarized and discussed in the final section.

2 Theoretical framework

2.1 Sense of justice

Education is a distinct “sphere of justice,” inasmuch as it is an arena in which resources and rewards are constantly distributed (Connell 1993; Deutsch 1979; Resh & Sabbagh 2016; Walzer 1983). Within schools, teachers are a major agent of reward (and punishment) allocation. They evaluate students’ performance and learning behavior, sort students into classrooms, tracks and ability groups, assign grades, and award certificates. Teachers (and peers) also offer (or withhold) praise, support, help, encouragement, respect and esteem. The major objects of this distribution are students, who in turn evaluate the “fairness” or “unfairness” of their reward and punishment allocation based on their perceptions of deservedness. Thus, the justice aspect of teachers’ behavior and students’ evaluation of its fairness become an integral component of students’ educational experience (Dalbert 2004); they also constitute a hidden message about society and the world around. It is possible that, as such, these factors also contribute to the shaping of students’ attitudes and behavior.

The burgeoning research in this area has indicated that the sense of justice leads to a variety of outcomes. Thus, it affects students’ wellbeing and academic motivation (e.g., Dalbert and Maes 2002; Peter and Dalbert 2010), the sense of belonging and identification with the school (e.g., Berti et al. 2010; Resh & Sabbagh 2014), the perception of school as a fair place (e.g., Peter and Dalbert 2010; Gogard 2011, 2012), trust in teachers and in people at large (e.g., Gorard 2012; Resh & Sabbagh 2014), democratic attitudes (Resh & Sabbagh 2014) as well as the readiness to extend help, and school and community engagement (Berti et al. 2010; Saha 2012; Lenzi et al. 2014). It was also found to reduce alienation and resistance that are manifested in a variety of disciplinary problems, particularly violence (e.g., Arum and Velez 2012; Chory-Assad 2002; Chory-Assad and Paulsel 2004; Khoury-Kassabri et al. 2005; Blank and Shavit 2013).

In exploring the sense of justice, we distinguish between distributive justice, i.e., evaluations of the justness of instrumental (grades allocation) and relational (teacher-student relations) resource distribution, on the one hand, and procedural justice, i.e., evaluations regarding the justness of the processes (or means) by which these resources are distributed, on the other. In what follows, we briefly explain the meaning of these two aspects of the sense of justice.

2.1.1 Distributive justice

Justice can be conceived of as a social exchange in which students (and for that matter, all individuals) make contributions (e.g., study hard) on the expectation of a reward (e.g., high grades) (Törnblom 1992). In a number of studies, justice has been conceptualized in proportional terms, in the sense that the actual resources (rewards) that people receive are evaluated relative to the just or expected amount thereof. Jasso’s (1980, 2005) theory of distributive justice is based on a comprehensive model for determining degrees of perceived (in)justice. The general formula developed by Jasso to assess perceived (in)justice can be applied to a variety of resources, groups and contexts (Jasso 1980, p. 3; see also Sabbagh and Vanhuysse 2012):

$${\text{Justice evaluation}} = { \ln }\left( {{\text{Actual share}}/{\text{Just share}}} \right)$$

‘Actual share’ is the amount of a resource that is actually received. ‘Just share’ (or just reward) is the amount expected, which is based on the social norm, or entitlement, derived from comparisons with reference groups. When the ‘actual share’ does not match the ‘just share,’ the result is a perception of injustice (e.g., Berger et al. 1972; Martin 1981; Crosby 1982; Markovsky 1985; Törnblom 1992; Jasso 1980, 1993; Hegtvedt and Markovsky 1995). Specifically, a lower-than-expected actual reward is likely to elicit a sense of deprivation, or under-reward. Conversely, individuals whose actual reward exceeds their expectations are likely to feel over-rewarded.

It is important to keep in mind, however, that the sense of distributive justice is context-bound and complex, being subject to such factors as the type of the resource being distributed, as well as the social norms, or entitlement, that shape the perception of “justness” in different contexts (Walzer 1983; Sabbagh et al. 1994; Randall and Mueller 1995; Dar and Resh 2001). In this investigation, we are concerned with two types of resources that affect the sense of justice in the school context: (1) instrumental rewards, i.e., the grades distributed by teachers, and (2) relational rewards, i.e., students’ relations with teachers.

2.1.2 Procedural justice

This term refers to the perceived fairness of the processes (means) by which distributions, or the decisions about them, are carried out (Leventhal 1976; Lind and Tyler 1988). Components of fair procedure include consistency, universality, transparency in using the criteria for allocation, and having a “voice,” i.e., the legitimacy of appeal when the “fairness” of a procedure is in question. Hence, the procedures of resource distribution may, in themselves, constitute a source of the sense of (in)justice, affecting the perceived legitimacy of distribution outcomes and satisfaction with the latter (e.g., Lind and Tyler 1988; Tyler et al. 1996; Hegtvedt et al. 2003).

Criteria for procedural justice seem to reflect an organizational climate whereby the accepted rules and regulations (as perceived by the organisation’s members) are applied universally to all.Footnote 1 Thus, while the sense of distributive justice reflects the individual evaluation of a gap between the actual and the deserved reward, procedural justice is essentially an institutional feature (e.g., that of a school).

Empirical research into procedural justice has overwhelmingly focused on organizations, with hardly any investigations conducted in the educational field. The few studies on education that we have found do not distinguish between the distributive and the procedural justice, applying a single scale to both these variables (Berti et al. 2010).

2.2 Deviant behavior: academic dishonesty (cheating) and violence

Recent decades have been marked by a growing concern about the high rate of disorder and violence in schools, which has often been linked to the so-called “[global] crisis of moral authority” (e.g., Arum 2003; Benbenishti and Astor 2005; Agnich 2011). Disorder and violence render the school setting uncomfortable, chaotic and even frightening, to teachers as well as to students and their parents. Crucially, this is also considered to be a factor that disrupts orderly learning and academic achievement. Indeed, ample evidence attests to the relationship between disciplinary problems and students’ achievement (e.g., Coleman and Hoffer 1987; Lee and Bryk 1989; Finn 1989; Arum and Velez 2012; Shavit and Blank 2012; Blank and Shavit 2013).

Extending this rationale, we assume that orderly behavior and the ability to respect and abide by organizational rules are central to nurturing the civic consciousness of future citizens. Conversely, deviant behavior on the part of individual students and, for that matter, any disruption of school order in general—not only disturbs the daily learning process, but negatively affects students’ civic development and identity. In the current study, we examine two different, albeit interrelated, facets of deviant conduct in school which, for many years now, have been of great concern to parents, teachers, and educational policy makers:

  1. 1.

    Academic dishonesty—cheating in the learning process;

  2. 2.

    Violence—verbal and/or physical intimidation of peers and teachers.

2.2.1 Academic dishonesty

Using cheating as a ladder to academic success is a violation of the norm of honesty, a central element in the process of organized learning in school (or university). Conversely, abiding by accepted behavioral rules, and avoiding academic dishonesty in particular, is a desirable feature of a student’s civic identity. It may also be indicative of the student’s future behavior as an adult citizen.

Research on academic dishonesty focuses on background and personal factors that may predict or explain the inclination to cheat in the learning process. These have been shown to include gender (boys more than girls), lower SES and level of achievement, lower self-efficacy and reduced sense of attachment (e.g., Finn and Frone 2004; Anderman and Murdock 2007). Arum (2003) suggests that rule-breaking behavior (including violence), especially in the US, is a ramification of a single overall macro-factor: the breakdown of moral authority and its institutional practices. Investigations that relate more specifically to school (or classroom) factors that play a role in ‘encouraging’ this phenomenon focus mainly on the pedagogical practices, goal structure, testing policy and general disciplinary policy of a school (e.g., Murdock et al. 2007; Anderman 2007; Nicholas and Berliner 2007).

To the best of our knowledge, little scholarly attention has been accorded to the examination of the relationship between students’ sense of justice and academic dishonesty. It stands to reason, however, that if students feel that they are being treated fairly by their teachers, they will be more likely to avoid cheating, a sort of noblesse oblige towards both teachers and peers. Indeed, although the notions of ‘teachers’ fairness’ or ‘classroom fairness’ figure only in very few studies, and then only peripherally, these factors have been shown to be negatively related to academic dishonesty (e.g., Murdock et al. 2007).

2.2.2 Violence

Compared to cheating, violence—both verbal and physical—is a much stronger manifestation of ‘rule breaking’ and disorder in school. As mentioned, it is widely acknowledged that the recent prevalence of school violence has wreaked havoc in the day to day workings of schools and has heightened a sense of threat experienced by both students and teachers. Importantly, violence also disrupts the learning process, and thus negatively affects academic achievement. As mentioned above, the spread of school violence has been attributed to the general ‘breakdown of moral authority,’ an explanation that, in itself, has become a matter of great public concern (Arum 2003).

Investigations of personal “risk factors” have identified individual characteristics that may be related to deviant and violent behavior in school, among them gender (girls are less involved) (Jenkins 1995; Khoury-Kassabri et al. 2005); race (racial and ethnic minorities are more involved) (Farkas et al. 2002), and SES (the lower class and the poor are more involved) (Gregory et al. 2010; Jenkins 1995). In recent decades, greater attention has been drawn to the school’s role in either inciting or decreasing school disciplinary problems, especially violence. It has been pointed out that the degree of perceived teachers’ support correlates positively with students’ sense of belonging to the school, and an increase in the latter brings about a decrease in student involvement in violence (Goodenow 1993; Khoury-Kassabri et al. 2009). Another salient factor is the school’s disciplinary policy, which has been found to correspond to the level of discipline. Specifically, the perceived fairness or clarity of a school policy, as well as the perceived extent to which rules are enforced, were both singled out as factors that reduce student involvement in violent incidents (Esposito 1999; Arum 2003; Arum and Velez 2012; Khoury-Kassabri et al. 2005; Blank and Shavit 2013). The few studies that investigated the relationship between students’ justice perceptions and their deviant behavior did indeed find that the more students evaluate their teachers as fair, the more they tend to judge the institutional arrangements (at school) as fair, and the less they are involved in deviant behavior (Chory-Assad 2002; Chory-Assad 2004; Sanches et al. 2012; Blank and Shavit 2013).

It would thus be logical to assume that the sense of fairness at school—the perception that there are clear and fair disciplinary rules that teachers apply across the board and universally—will moderate personal tendencies toward deviant behavior, i.e., will decrease the chances of cheating and violence in school.

2.3 Civic engagement: extra-curricular and voluntary community activities

Education through schooling is an inherently collective endeavour: children (students) are learning in classes that are situated in schools; they meet daily and interact with groups of peers and with their teachers. Hence, social life and social activity are inseparable from the process of schooling. Moreover, this ‘informal’ part of schooling becomes an essential component of civic socialization: students are encouraged to engage, both internally (within the school) and externally (outside it, in the community), in voluntary social activity (Wilson 2000).

2.3.1 Extracurricular activity in school

In the context of this article, extracurricular activity is relegated to the ‘informal’ part of school curriculum that involves (1) social life—student council, school newspaper, academic help to other students etc., and (2) leisure activity—sports, art groups etc. At the most general level, participation in extracurricular activity is a function of school engagement. This means that, in relation to the educational process, an individual routinely experiences a sense of meaning and purpose, which may be reflected in a variety of educational outcomes (Mahoney et al. 2003). Moreover, this form of engagement is visible to school personnel (i.e., teachers, principals), who may hold higher expectations for extracurricular activity participants compared with other students (Van Matre et al. 2000).

Extracurricular engagement in school is widely assumed to contribute positively to students’ personal development and educational outcomes. Consequently, most of the related investigations have focused on the effects of extracurricular activity and their possible mediators or moderators. Overall, researchers concur that engagement in this form of informal activity has positive short- and long-term effects on both educational and personal-developmental outcomes. The former include grades, course choices, educational and occupational aspirations, and college application and participation, while the latter—identity formation, interpersonal competence, an increase in connections with peers and a concomitant decrease in alienation, and avoidance of risky behavior and substance abuse (e.g., Fredricks and Eccles 2006; Eccles and Tempelton 2002; Eccles et al. 2003; Mahoney et al. 2003; Marsh and Kleitman 2002). Evidence has also been adduced that disadvantaged students especially benefit from this kind of activities.

This study treats extracurricular activity as an outcome, an indicator of civic engagement. We hypothesize that, along with personal background variables, students’ sense of justice in school—i.e., their evaluation of teachers as fair—will be positively correlated with the amount of extracurricular activity in which they engage. Students who feel that the instrumental and relational rewards they receive are ‘just’ (i.e., they get what they are entitled to), and that the procedures through which these are allocated are fair, will be more connected to school, more involved in extracurricular activities, and more responsive to suggestions and requests for voluntary engagement within the school.

2.3.2 Voluntary activity in the community

Community volunteering has become a mainstay of education for citizenship in school, and in some cases part of the formal high school curriculum (Yates and Youniss 1999; Niemi et al. 2000; Wilson 2000). This development is premised on the understanding that social engagement is an integral part of civic education and that the strength and stability of democratic societies hinge on engaged and active citizens. Indeed, studies in Israel and in other countries have shown that programs of community volunteerism are widespread, especially in secondary schools, and that between 40 and 60% of all the students participate in such action.

As in the case of extracurricular activity, investigations of students’ community volunteering have mostly focused on its effects on educational outcomes. Research suggests that voluntary social activity is a positive experience that contributes to students’ sense of wellbeing and personal growth as well as to their development as democratic citizens. The two former aspects encompass academic motivation, self-image and confidence, and personal interaction skills (Wilson 2000; Howard 2003). The democratic citizenship outcome involves political knowledge, awareness of social problems, awareness and sensitivity to “others” in the community, and to injustice and inequality, social responsibility, and critical thinking (Youniss et al. 1997; Eccles et al. 2003; Ichilov 2003a, b; Rienders and Youniss 2006).

In parallel to extracurricular activity, we perceive community volunteering as an outcome, which is affected by the sense of justice in the school. Students who perceive their teachers’ behaviour as ‘just’ and the procedures through which the rewards are allocated as fair will be more trustful in general (Resh & Sabagh 2013), and hence more willing to commit to social engagement and voluntary activity (Saha 2012; Lenzi et al. 2014)

2.4 Hypotheses

In light of the above, we hypothesize that elements of students’ civic behavior will be affected by their sense of distributive and procedural justice in school. More specifically:

Hypothesis 1 (H1)

The sense of distributive justice will be negatively related to (a1) dishonesty and (a2) violence in school.

Hypothesis 2 (H2)

A positive evaluation of procedural justice will be negatively related to (b1) dishonesty and (b2) violence in school.

By the same token,

Hypothesis 3 (H3)

The sense of distributive justice will be positively related to (c1) extra-curricular activity in school and (c2) engagement in community volunteerism.

Hypothesis 4 (H4)

Procedural justice will be positively related to (d1) extra-curricular activity in school and (d2) engagement in community volunteerism.

2.5 Israel: social divides and sectorial educational system

School sectors in Israel reflect major socioeconomic and cultural divides. Israel is a relatively young state with a Jewish majority (about 80%)—mostly immigrants and descendants of immigrants of European, Middle Eastern and North African extraction—and an indigenous Arab minority (about 20%). Generally, the Arab minority is marginalized politically and deprived economically, as well as discriminated against formally and informally in many aspects of public life (Al Haj 2002). In the Jewish majority, a major cleavage is secular-religious.Footnote 2 The secular-oriented part of the Jewish majority, mostly of European-American extraction, is by and large more affluent. The religious group, with a greater proportion of Jews of Afro-Asian extraction, is less affluent; its aspiration has been to merge a religious Jewish lifestyle with the rather secularist principles of Zionism.

Israel’s educational system is almost entirely public, with free and compulsory schooling of one year of kindergarten and 12 years of primary (grades 1–6), junior-high (grades 7–9) and senior-high school (grades 10–12). The system is centrally administered by the Ministry of Education, which allocates and monitors educational resources, assigns new teachers and controls a range of educational issues, including the school curriculum (Gaziel 1996). However, under the umbrella of a unified national system operate three major semi-autonomous sectors. Differentiated along national-ethnic (Jews/Arabs) and religious (secular Jews/religious Jews) lines,Footnote 3 these sectors reflect the tripartite social structure. Contrary to the prevailing rhetoric of equal opportunity, the sectors differ in terms of the quantity and quality of school resources and the degree of autonomy granted to them, as well as the structure of their curriculum—especially the messages transmitted through the social sciences and civic studies. Owing to its political power as an essential member of the coalition, the religious Jewish sector enjoys richer resources and considerable administrative and pedagogical autonomy. The resources at the disposal of the Israeli Arab sector are much more limited. Moreover, their curriculum, while resonating to some extent the group heritage (Arab, Druze), is closely controlled by the Ministry of Education.Footnote 4

2.6 Hypotheses

We assume, albeit somewhat speculatively, that the different educational trajectories—separate school premises, differential curricular emphases, and separate community contexts (especially, for the Arabs)—should manifest in cross-sector differences in students’ civic behavior.

Hypothesis 5 (H5)

There will be a lower incidence of violence in the Jewish religious sector compared to the Jewish secular sector (e1). On the other hand (e2), the level of violence in the Arab sector will be higher compared to the Jewish sector. The rationale for this hypothesis is that most schools in the Jewish religious sector are uni-gender and seem to maintain harsher disciplinary standards. The Arab sector suffers from lack of resources, which is reflected in larger classes and lower teacher-student ratios. In these circumstances, violent behavior might be an expression of resistance and/or anomie.

Hypothesis 6 (H6)

Community volunteering is expected to be more prevalent in the Jewish religious sector (f1), reflecting a strong tradition of charity (tzedaka) promoted by the Jewish religion. Conversely, community volunteering will be more limited in the Arab sector (f2), where opportunities for such activities are scarce, and many of the informal practices are relegated to the extended family (hamula) (Zydan 2007).

2.7 School structure

Two additional variables, both relating to school structure, are considered as possible factors affecting students’ behavior:

2.7.1 School composition

On par with students’ personal background, the factor that renders the atmosphere in school either conducive or non-conducive to extracurricular activity and community volunteering is the student-body at the collective level. Here, we treat the educational background of the student-body as an indicator of culturally advantaged or disadvantaged upbringing. We therefore hypothesize that the higher the educational-background level of a student-body, the less violence and the more extracurricular activity and community volunteering the students will report.

2.7.2 School size

Big schools are more often than not crowded physically and encumbered by bureaucracy. Both these factors have been shown to contribute to student alienation, which fosters violence. On the other hand, bigger schools afford more opportunities for informal activities, thus warranting expectations for higher levels of both extracurricular activity and community volunteering.

3 Research design

3.1 Sample

The investigation was carried out on a national sample of 48 middle schools representing the three major sectors of the Israeli public school system: 24 schools in the general (secular) Jewish sector; 9 schools in the religious Jewish sector; and 15 schools in the Israeli-Arab sector. In each sector, schools were sampled countrywide to represent the educational composition (the disadvantage index)Footnote 5 and regions across the country. In each school, two 8th grade and two 9th grade classes were randomly selected.Footnote 6 In schools with ability-based classes, a “stronger” and a “weaker” class were selected at each grade level. The sample includes 5084 students in 187 classes: 2542 students in the secular Jewish sector, 826 in the religious Jewish sector, and 1716 in the Israeli-Arab sector. Data were collected during the 2010–2011 school year.

3.2 Research tool

An anonymous self-administered questionnaire was answered by all students in the selected classes during a class period in the presence of a teacher and a research assistant. Information about the schools—disadvantage index, sector, size and district—was obtained from Ministry of Education files.

3.3 Variables

Dependent Variables Four variables were constructed to represent students’ civic behavior at school: extra-curricular activity (within school), community volunteering (outside school), academic dishonesty and violence (both within school).

  1. (a)

    Extra-curricular activity (Sch.activ.): The students were asked whether they were participating in two sets of within-school extracurricular activities. The first targeted the school community, e.g., being a member of student committees, contributing to a school newsletter, mentoring younger students etc. The second set comprised individual extracurricular activities of students’ choice such as dance, choir, sports, chess, etc. The variable represents the number of activities counted for each student.

  2. (b)

    Community volunteering (Comm.activ.): The students were asked whether they were active in any out-of-school voluntary activity such as a youth movement, sports clubs, environmental action, helping in various community organizations, etc. The variable represents the number of activities counted.

  3. (c)

    Academic dishonesty was measured on a scale and computed as a mean value of four items related to dishonest behavior in school. The students were asked to assess, on a scale of 1-never to 7-frequently (α = 0.67), four behaviors: “I have copied in exams”, “I have copied homework from friends”, “I missed classes”,”I stole (took out) some staff from students’ school bags”.

  4. (d)

    Violence was measured by asking the students to assess, on a scale of 1-never to 7-frequently (α = 0.78), four items tapping the use of verbal and physical violence toward students and teachers, and computed as a mean of these values.

Independent Variables As specified below this type of variables involve both distributive and procedural aspects of justice.

  1. (a)

    Sense of distributive justice in school concerned grades (instrumental reward) and teacher-student interactions (relational reward). Following Jasso’s argument (1980, 2007) that under-reward is felt more keenly than over-reward, the sense of (in)justice was formulated as a logarithmic function of the ratio of actual and deserved (just) “good,” as follows:

$${\mathbf{J}} = {\mathbf{ln}}\left( {{\mathbf{Actual}}/{\mathbf{Just}}} \right),$$

where J denotes sense of (in)justice. The measure equals zero when the actual reward matches the student’s evaluation of the deserved reward; it is negative when the student thinks s/he is under-rewarded, and positive in the case of over-rewarding.

  1. (a1)

    Sense of distributive justice as regards grades (instrumental) For each of the four subjects—language (Hebrew or Arabic), math, English (as a foreign language) and history—students reported, on a scale of 1–100, the actual grade they received and the grade they thought they deserved (for example: “What was your final grade in math last year?”, “What is the grade in math you think you deserved?”). Sense of justice was calculated as described above; the measure is the mean of the four J’s (α = 0.72).

  2. (a2)

    Sense of distributive justice as regards teacher-student interactions (relational) In ten pairs of items, respondents evaluated, on a 7-point scale, the amount of a relational resource (e.g., praise, attention or help) they actually received from their teachers, and the amount of that resource they felt they ought to have received. Responses ranged from 1-very little to 7-very much. Sample items included: (1a) “My teachers help me in learning matters when I ask for help” and (1b) “My teachers ought to help me in learning matters when I ask for help”; (2a) “My teachers treat me respectfully even if the comments and ideas I express are critical” and (2b) “My teachers ought to treat me respectfully…” As with grade allocation, J was calculated for each pair of statements. Since factor analysis produced one principal component, the measure is a mean of the ten outcomes (α = 0.76).

  3. (b)

    Procedural justice (see “School-level variables” below).

Personal level controls We included four background variables that, as elaborated below, are likely to have an independent effect on outcomes and which should therefore be controlled in investigating the ramifications of the sense of justice.

  1. (a)

    Gender Female-0; male-1. There is recurring evidence that boys perceive greater injustice in school, especially in regard to grade allocation (e.g., Dar and Resh 2001; Jasso and Resh 2002; Resh and Dalbert 2007). Girls seem to adjust better to the student role, and hence may also conform more easily to school behavioral standards.

  2. (b)

    Parents’ education Set as a mean of the father’s and the mother’s education level, measured on a 5-point scale, ranging from 1—elementary school to 5—academic education. This variable served as a proxy for SES,

  3. (c)

    Academic self-image Students’ self-reported evaluations of their academic performance, ranging from 1—very weak to 6—excellent. This variable served as a proxy for achievement.

    By and large, students from advantaged backgrounds and those with higher achievements are also more adjusted to school life and to its academic, social and behavioral requirements. Accordingly, they are expected to perceive less injustice (Resh 2010), to participate to a greater extent in extra-curricular and community activity, and to exhibit less non-normative behavior (Benbenishti and Astor 2005; Gregory et al. 2010).

  4. (d)

    Justice sensitivity. Measured on a 6-item scale, based on items developed by Schmitt et al. (2005). Possible responses ranged from 1—not sensitive at all to 7—very sensitive (α = 0.79).

Justice sensitivity is seen as an indication of a personality disposition manifested in inter-individual differences in reaction to unfair situations. This trait heightens the experience of injustice towards oneself and accentuates the perception of unfair treatment of others (Schmitt et al. 2005). These and other facets of justice sensitivity have been used to predict a wide range of phenomena, including decisions in game paradigms and altruistic punishments (Schmitt et al. 2010). In the context of the current study, we assume that those who are more sensitive to justice will tend to participate in voluntary activity and to refrain from violent and dishonest behavior in school. Justice sensitivity was measured using the scale comprising such items as “When I encounter injustice it infuriates me”; “It is hard for me to become attached to someone who is insensitive to justice”; etc.

3.4 School-level variables

  1. (a)

    Procedural justice Following Tyler and Blader (2000), we constructed a scale of nine items tapping the various criteria according to which students are likely to assess the fairness of procedures (e.g., clear criteria for allocating rewards; universal, equal treatment of students; right to appeal in case of “unfair” treatment). Example of questions: “There’s no point in complaining about teachers in our school because nothing will be done about it anyway” (reversed); “School rules and regulations apply equally to all students”; “Teachers in my school allocate grades to students based on unified criteria.” Possible answers ranged from 1—definitely do not agree to 7—definitely agree (α = 0.69). The mean perception of procedural justice in a school indicates the collective sense regarding the “fairness” (or “unfairness”) of its procedures.

  2. (b)

    Sector Affiliation with the secular Jewish, religious Jewish or Arab educational sectors was converted into two dummy variables: religious Jewish—1, else—0; Arab—1, else—0. The secular Jewish sector served as the reference category.

  3. (c)

    Educational composition The mean level of parents’ education in a school.Footnote 7

  4. (d)

    School size The number of students in a school (retrieved from the Ministry of Education information).

4 Data analysis

Data were analyzed using the Hierarchical Linear Model (HLM), a statistical method that is especially appropriate for data of two or more nested levels of measurement (see Bryke and Raudenbush 1992; Raudenbush and Willms 1995), as it allows for the decomposition and simultaneous estimation of effects at individual and aggregate levels. Individual-level variables in our analysis were personal background, justice sensitivity and sense of distributive justice. School-level variables were procedural justice, sectorial affiliation, school composition and school size.

5 Findings

Table 1 below displays correlations among the four behavioral indicators (the dependent variables). It transpires that deviant behavior in school and pro-social activities, both in and outside the school, are two separate, largely unrelated domains. At the same time, a strong correlation is observed between the two social engagement indicators, extracurricular school activity and community volunteering (r = 0.51), and between the two deviant behaviors, dishonesty and violence (r = 0.56). However, the correlations between the variables across these pairs, although negative as expected, are weak and mostly insignificant. It is noteworthy, as well as logical, that the correlations with pro-social indicators are somewhat stronger for violence (r = −0.040 and −0.039) than for dishonesty (r = −0.034, and −0.012), which Israeli students do not generally regard as a great deviation from accepted norms of behavior.

Table 1 Correlations among civic behavior indicators
Full size table

We now move to our main analysis investigating the possible effects of students’ sense of distributive and procedural justice on their civic behavior in school and the role of school structural variables—sector affiliation, composition, size and– in shaping these behaviors. By and large, the hypothesized relationship between the sense of justice and civic behavior was confirmed. However, variation emerged in the various aspects of this relationship across the four civic behavior variables studied (Tables 2, 3).

Table 2 Academic dishonesty and violence (HLM)
Full size table
Table 3 School engagement and community volunteering (HLM)
Full size table

Not surprisingly, between-school variation in the prevalence of violence is much greater than in dishonesty (7% and 4% respectively). Moreover, the structural variables—composition, sector, size—do not appear to have any significant effect on the prevalence of disruptive behavior in school. However, students’ personal sense of justice, both instrumental (grades distribution) and relational (interactions with teachers), is significantly related to dishonest and violent behavior: Experiencing justice in school seems to be an effective deterrent to disruption of civic order in school. As concerns violent behavior, an additional factor in reducing it appears to be students’ positive judgment about a school’s procedural justice.

Of the personal controls, gender and academic position have consistent effects: Similar to what was found in other studies, boys and “weak” students reported involvement in dishonest and violent behavior to a greater extent.

In respect to students’ pro-social activity, the results are somewhat different. First, it is important to note that the between-school variation is more pronounced in both civic school activity and community volunteering (20% and 14% respectively). Personal controls—gender, parents’ education and academic self-image—have a significant effect only on community volunteering, and in the expected direction: Girls, higher parents’ education and a more positive academic image have all emerged as conducive to participation in voluntary out-of-school activity, but seem to affect within-school extra-curricular participation only slightly. The same is true for justice sensitivity, which positively affects community activity but hardly at all within-school activity.

Interestingly, neither is the sense of instrumental justice related to within-school activity. On the other hand, the interpersonal (relational) dimension of justice is positively related to community activity, and the sense of procedural justice (as a school variable) is significantly related to both within and out-of-school civic activity, reducing the unexplained between-school variance by 20.6% and 12.5% respectively.

Finally, both sector and school-size were found to affect civic activity. Students in the Arab sector reported greater involvement in extra-curricular school activities and community volunteering (compared to students in the Jewish secular sector).Within-school civic activity was shown to be more intensive in larger schools.

6 Summary and discussion

Recent global trends have placed greater focus on academic preparation, constantly challenging the traditional role of schooling as a vehicle for civic education within nation-states. Nonetheless, imparting democratic knowledge and instilling citizenship values and behaviors remains a central educational mission in preparing a new generation for life both within the nation state and in the global world.

The current investigation focused on four possible indicators of school students’ civic behavior: two positive (participation in within-school extracurricular activities and out-of-school community volunteering) and two negative (academic dishonesty and violence). Both the antecedents and the consequences of these behaviors have been extensively studied. Moreover, in recent decades, research on factors that may construct and shape these behaviors has moved beyond personal indices like gender, SES and self-image toward the role of schools and teachers in this respect (e.g., Aalsma and Lapsley 2001; Moberg et al. 2009). Our study adds to this body of research by focusing on the students’ sense of distributive and procedural justice in school as factors affecting four indicators of civic behavior. Most existing investigations into the effects of students’ sense of justice have used a generalized and rather vague measure termed “teachers’ fairness.” This study, on the other hand, distinguishes between the instrumental (grades distribution) and relational (teacher-student relations) distributive justice, on the personal level, and the sense of procedural justice as a school feature, on the collective level. This refined approach has enabled us to test the effect of each of these three aspects of the sense of justice on the four civic behaviors studied, as well as similarities and differences in the relations observed.

By and large, our findings have confirmed our overall assumption that the sense of justice in all its aspects has an independent positive effect on student outcomes (controlling for personal background factors). However, the effect of the three different dimensions of the sense of justice has been found to differ across the four behaviors studied.

More specifically:

6.1 Academic dishonesty

Relatively little between-school difference emerged with respect to dishonesty. As mentioned earlier, although dishonesty falls under the rubric of non-normative behavior, the kind of dishonesty measured here is not perceived as a serious deviation from accepted informal school norms—and in some cases is even condoned as innocuous mischief or showing off. It is therefore plausible to assume that such dishonesty will be present to a similar degree in all schools. Violence is a different matter altogether, and its prevalence can be expected to vary significantly across schools.

Still, on the personal level, both instrumental and relational aspects of the sense of distributive justice have been found to significantly reduce dishonest academic behavior. This observation suggests that ‘just’ behavior on the part of teachers may induce in students a reciprocal commitment to ‘fairness.’ On the other hand, none of the school variables—the procedural justice or the structural variables—were found to have any effect at all on dishonest behavior.

6.2 Violence

Violence seems to be significantly affected by both measures of distributive justice (grades distribution and relations with teachers) and, even to a greater degree, by the sense of procedural justice. The sense of justice can be conceived of a sort of contract, both formal and informal, between the teachers and the students; if it is broken through teachers’ unjust conduct, students no longer feel compelled to honor it either. Crucially, such a violation of the ‘contract’ may arouse feelings of frustration and anger that can easily translate into delinquent behavior.

For either dishonesty or violence we did not find any effect of school structural features (sectorial affiliation, student composition and school size). We found this result surprising: We had expected deviant behaviors to be more prevalent in large and ‘low’ composition schools. It may be that the introduction of sense of justice variables into the equation depressed these effects: when students feel that they are fairly treated, they ‘reciprocate’ by displaying more normative behavior irrespective of which school they happen to attend.

6.3 School engagement (extracurricular activity) and community volunteering

Extracurricular as well as community voluntary activity can both be considered as indicators of a social-capital development. The creation of, and participation in, social networks, social awareness and social responsibility are all behaviors that are essential to becoming an active citizen of a democratic society as well as of today’s global world (Coleman 1988; Cotterell 2007; Putnam 2002).

Extracurricular activities are organized by and within the school, and larger schools usually enjoy more resources and offer a wider range of extracurricular opportunities. It is therefore logical that school size was found to have a significant (positive) effect on extracurricular activity but not on out-of-school community volunteering.

Considerable between-school variance emerged in both extracurricular and voluntary community activity (20% and 14% respectively). Yet, contrary to our hypothesis, extracurricular activity was not related to either aspects of the sense of distributive justice, and voluntary activity was related to relational justice alone. At the same time, both these behavioral outcomes seem to depend on procedural justice, suggesting that students will be more inclined to engage in informal school activities in a climate of fairness.

Another unexpected—and interesting—finding is that students in the Arab sector are more involved in the various activities, both within and outside school. This could be attributable to the general context of these schools, the majority of which are located in rather secluded and mostly socio-economically disadvantaged all-Arab communities. Thus, it is up to the school to motivate a student to engage in extra-curricular activity, both in school and in the surrounding community.

The findings on civic behavior obtained in the current research align with our previous results regarding a strong relationship between the sense of distributive and procedural justice and students’ civic attitudes, specifically, democratic orientation, trust and sense of belonging to the school (Resh & Sabbagh 2013, 2014). Combined, these conclusions underscore the importance of the sense of justice in school as a facilitating factor in the development of an involved, active future citizen of a democratic society.

Main limitations of this study have to do with its cross-sectional design, which restricts causality inferences. In the relationship between justice and behavior demonstrated here, the causal precedence of the former can be established on theoretical grounds. Only a longitudinal study, however, would be able to substantiate such an inference.

It stands to reason that the sense of justice is a significant factor in the process of schooling not only in Israel but in other countries as well. It would, however, take a cross cultural investigation to determine the scope of this factor, as well as the type of variables that may moderate the relationship between the sense of justice and civic development.

Finally, an avenue for further research—, one related specifically to Israeli reality—has been opened by the unforeseen results regarding sectorial differences. A deeper investigation is needed to understand the mechanisms behind the relationship between the sense of justice and civic behavior in the Arab school sector in Israel.