Subjective Obligation to Obey the Police in Diverse Societies: A Social Resistance Perspective

Abstract A sense of obligation to obey the police is an important predictor of public cooperation and compliance with the law. Minorities tend to feel less obligated to obey the police than the majority. Previous work based on the social resistance framework shows that the experiences that shape the lives and attitudes of minorities may encourage them to actively engage in a variety of everyday resistance acts against the majority group, which may include high-risk and delinquent behaviours. The present study tests this framework for the first time concerning the self-perceived obligation to obey the police while also considering different minority groups who experience varying levels of marginalization. We use a representative sample of about 1,100 Israelis from four minority groups – Muslims, immigrants from the former Soviet Union, ultra-Orthodox Jews and Jews of Ethiopian origin – along with the Jewish majority group. The results show that Muslims feel the least obligation to obey the police, followed by Jews of Ethiopian origin. Social resistance was negatively related to the self-perceived obligation to obey the police among Muslims and ultra-Orthodox Jews while controlling for demographic characteristics and previous theoretical explanations, namely procedural justice, self-help and anger.


INTRODUCTION
Inquiry into the antecedents of people's voluntary compliant attitudes and behaviours towards the police and other legal authorities constitutes a significant research frontier in criminological studies (e.g.Mazerolle et al. 2013a, b;Tankebe 2013;Tyler 2004).A core objective of the police is to ensure citizens' compliance with and obedience to their directives and to the law more generally.Since police forces cannot necessarily be present at any given place and time, it is thanks to voluntary, everyday, law-abiding behaviour among most of the public that the police can operate effectively to maintain social order by focusing their efforts on those who are less compliant (Tyler 2004).The scientific literature suggests that citizens' perceptions of police legitimacy, and especially a sense of obligation to obey figures who represent the authority of the state, are the most prominent factors that can shape compliance and cooperation (Bolger and Walters 2019;Factor and Mehozay 2023;Reisig, Trinkner, and Sarpong 2023;Tyler and Nobo 2023;Worden and McLean 2017).
Empirical evidence suggests that in many societies, members of minority groups tend to hold more negative views towards the police compared to the majority group (Ben-Porat 2008;Factor, Castilo, and Rattner 2014).This is also reflected in a weaker sense of obligation to obey the police (Murphy and Cherney 2012).However, the empirical literature examining how people in a given society view the police typically focuses on disparities between the main majority and minority groups while paying less attention to heterogeneity between different minority groups who experience varying levels of marginalization (Sargeant, Davoren, and Murphy 2021;Unnever, Barnes, and Cullen 2016;Weitzer 2014).This practice of lumping together diverse minorities fails to account for their differing levels of conflict with the majority group and different levels of perceived discrimination and alienation, all of which may reflect particular historical experiences (Mentovich et al. 2020;Unnever and Gabbidon 2011).Thus, there is a need for a more critical and nuanced investigation that considers variations in attitudes and behaviours between different minority groups.This is the first motivating factor behind the present study.
Our second motivation relates to the reasons for differences in citizens' compliance with the law.Previous works have suggested several explanatory frameworks for why individuals and groups may differ in this respect, including procedural justice, self-help, and strain or anger (Barkworth and Murphy 2015;Tankebe and Asif 2016;Tyler and Nobo 2023;Weisburd 1988).However, while such propositions hold some promise for explaining group differences, they largely do not account for macro-level considerations, such as social class, power relations, discrimination and alienation from the state.Recently, the social resistance perspective (Factor, Kawachi, and Williams 2011) has been suggested to overcome limitations and bridge social structure and personal agency.This theory suggests that power relations and the position of a non-dominant minority group in society may promote resistant and disobedient behaviour on the part of group members.
A range of findings supports the social resistance framework concerning different behaviours and attitudes (Factor et al. 2013b;Letki and Kukołowicz 2020).However, the theory has not yet been tested in the context of the obligation to obey the police.Using a survey with a representative sample of 1,091 Israelis, the present study examines the perceived obligation to obey the police among five different social groups -Muslims, immigrants from the former Soviet Union (FSU), ultra-Orthodox Jews, Jews of Ethiopian origin and members of the Jewish majority group.Then, to better understand attitudes towards obeying the police among minority group members, we test the relationship between social resistance and perceived obligation to obey the police while controlling for previous alternative explanations and social-demographic circumstances.

Police Legitimacy and Citizens' Obedience
The voluminous empirical research on police legitimacy across multi-national settings shows that legitimacya citizen's feeling that "the authority or institution is entitled to be deferred to and obeyed" (Sunshine and Tyler 2003, 514) affects a range of desirable behaviours relating to the legal system (Bolger and Walters 2019;Walters and Bolger 2019).A greater perception of the police as legitimate may have a range of positive outcomes.For instance, people who view the police as legitimate are more willing to cooperate and assist in controlling crime (Murphy, Hinds, and Fleming 2008) and to comply more generally with the law in their everyday lives (Sunshine and Tyler 2003).During a citizen-police officer interaction, greater perceptions of police legitimacy make citizens more likely to defer to police authority, more willing to accept the decision made, and more satisfied with the interaction (Tyler 2004).During the COVID-19 pandemic, it was shown that legitimacy affected citizens' willingness to comply with social distancing restrictions (Murphy et al. 2020).
While there is an ongoing debate among academics about the components of institutional legitimacy, one well-recognized core element (along with trust) is an obligation to obey individuals and organizations in authority (Reisig et al. 2023;Tyler 2004;Tyler and Nobo 2023;Worden and McLean 2017).Tyler and Jackson (2014), for example, found in a random sample of American residents that obligation to obey is the most prominent component of legitimacy in shaping compliant behaviour.Some scholars even treat the obligation to obey as a standalone concept rather than an institutional legitimacy component (Tankebe 2013).Obligation to obey can be defined as citizens' internal normative belief that they should defer to and comply with authority because it is the right thing to do, irrespective of the likelihood of being rewarded or punished (Tyler 2004).This sense of obligation to obey develops during childhood and adolescence as part of a suite of beliefs and attitudes about the law and authorities, and reflects a process of "legal socialization" whereby children learn that compliance with the authorities should take precedence over their self-interest even when they view a law or order as wrong (Fagan and Tyler 2005).In the rest of this paper, the "obligation to obey" means this sense of internal normative commitment rather than objective legal duty.
Aside from the normative importance of citizens' obligation to obey the police and its value to police legitimacy, the obligation to obey the police is hypothesized to predict actual cooperation and compliance with the police.Although lately there has been some weakening of the consensus on this topic (Reisig et al. 2023), some studies have indeed found that obligation is correlated with actual obedience to police directives (Tyler and Jackson 2014;Wolfe and McLean 2021;Woo, Maguire, and Gau 2018).
In short, legitimacy, especially the obligation to obey, makes a significant impact on shaping cooperative behaviour and compliance with authorities.Yet different groups within society may have different attitudes toward the law, its enforcement agencies, and the state more generally.Hence, their obligation to obey the law and authorities may also differ.

Obligation to Obey the Police among Different Social Groups
The police are the most visible and immediate representative of the state.As such, in diverse societies, the police may be seen as embodying a superordinate national identity (Murphy, Sargeant, and Cherney 2015;Skogan and Frydl 2004).However, members of minority groups who are alienated from the state often tend to have a troubled relationship with law enforcement (Skogan and Frydl 2004).The decisions, policies and behaviours of both the police as an institution and the individual police officers who embody it can create a perception of boundaries between citizens who are entitled to protection, citizens who remain vulnerable, and those who are viewed as a threat (Ben-Porat and Ghanem 2017).Yet such perceptions are not shaped solely by what the police do or how they operate but rather by the interaction between three factors: institutional police policies and procedures; the behaviour of actual police officers on the ground; and the attributes and perceptions of citizens.These last include citizens' social identity and values, feelings of alienation from the state, the symbolic status of what the police represent, and whether the police are believed to represent a particular social group (Factor and Mehozay 2023;Hasisi and Weitzer 2007;Mentovich et al. 2020).
Numerous studies have found wide gaps between evaluations of the police by members of minority groups, whether racial, ethnic or religious, and the majority group in different societies (Peck 2015).Broadly speaking, members of minority groups tend to exhibit lower levels of trust in the police and to rate the police lower on measures of legitimacy (Ben-Porat and Yuval 2012;Mentovich et al. 2020;Murphy et al. 2015;Murphy and Cherney 2012).Such attitudes are often grounded in a perception that police treatment of minority group members is driven by bias (e.g.see Weitzer and Tuch 1999).In practice, they typically lead to lower willingness to cooperate, lower obedience and less compliant attitudes toward the police (Murphy and Cherney 2012;Sargeant et al. 2021).
Yet findings in some societies have been mixed, suggesting that not all minority groups hold unfavourable attitudes toward the police.For instance, in Australia, compared with the Vietnamese minority, Indians and Arabic speakers are more likely to comply with the police (Sargeant et al. 2021) and to cooperate in crime control and counter-terrorism efforts (Murphy et al. 2018).In Israel, the Druze community is a subgroup within the Arab minority that holds similar political orientations to the Jewish majority.Their perceptions of trust in the police are generally positive and similar to those of the Jewish majority (Hasisi 2007).
These mixed results in the policing literature chime with the argument of Unnever and Gabbidon (2011) that criminological research should move beyond uncovering disparities in offending and its antecedents solely between the majority and minority groups.Rather, we must acknowledge that the disposition to obey varies across social groups, and there is a need for further, more nuanced examination of the factors that might account for these variations (Unnever et al. 2016).These may include different historical relations with the majority, with consequent differences in levels of discrimination, social alienation and conflict with national institutions such as the police (Factor et al. 2011).

Previous Explanations for the Obligation to Obey the Police
Previous criminological research has identified several possible explanations for why people might act unlawfully and display incompliant behaviour and disobedience towards police forces.The prominent theoretical frameworks suggested are procedural justice, self-help and strain theory.Taking a procedural justice perspective, Tyler's process-based model of legitimacy holds that citizens' satisfaction and legitimacy perceptions toward authorities, such as the police, depend more on whether the authorities exercise their power through just procedures than on instrumental concerns about whether the outcome of a given case was favourable (Tyler and Nobo 2023).Over time, stronger perceptions of legitimacy improve citizens' willingness to comply and cooperate with the authority (Tyler 2004).Conversely, police forces deemed to operate unjustly and unfairly will be viewed as less legitimate, reducing citizens' compliance and willingness to obey.The procedural justice model for police legitimacy and compliance has been subject to a great deal of empirical inquiry, with results supporting the core argument that procedural justice can influence police legitimacy and, consequently, citizens' willingness to cooperate and obey (Factor et al. 2014;Jonathan-Zamir and Weisburd 2011;Mazerolle et al. 2013b;Sargeant et al. 2021).For example, in a sample of Ghanaian immigrants in the United States, Pryce, Johnson, and Maguire (2017) showed that perceived procedural justice significantly increases feelings of obligation to obey police directives.
The second explanation for uncompliant behaviour can be found in self-help or vigilantismthe phenomenon whereby private citizens take the law into their own hands (Brown 1975;Little and Sheffield 1983).Black (1983) suggested that vigilante behaviour is a means by which people or groups express their grievances towards the conduct of others and thereby engage in a form of social control rather than depending upon a third party such as the police.Examining the vigilante phenomenon in Israel, Weisburd (1988) found support for the idea that those who participate in vigilante violence are indeed fulfilling a role of community social control.Studies have also shown that public support for self-help via violent vigilantism is related to broader concerns with the rule of law and correlated with judgments of police trustworthiness and obligation to obey the police (Jackson et al. 2013;Tankebe 2009;Tankebe and Asif 2016;Yagil and Rattner 2002).However, we must note that the association between self-help and the obligation to obey the police might work in both directions.
Last, the General Strain Theory (Agnew 1992) has been used to explain involvement in delinquent behaviour at the individual level.Agnew's theory extends classical strain theories, focusing mainly on monetary success, to other stressors, such as negative relationships.Each of these relationships, or "strain stimuli", can increase the likelihood of experiencing negative emotions like disappointment, depression, and, more importantly, anger.Anger is said to be the prime factor that reduces inhibitions and leads the individual toward a desire to take action (e.g. to avenge a wrong) in the form of delinquent behaviour (Agnew and Brezina 2010).Accordingly, Aseltine, Gore, and Gordon (2000) showed that anger mediates the relationship between strain and adolescent misconduct, including violent and aggressive acts.Some studies have linked procedural justice and strain frameworks by suggesting that procedural injustice constitutes a source of strain.Indeed, research in legal settings has found that negative emotions, like anger, mediate the negative effect of perceived procedural injustice (seen as an expression of strain) on subsequent tax compliance behaviour (Murphy and Tyler 2008) and obedience to police directives (Barkworth and Murphy 2015).Building on these findings, one can speculate that members of non-dominant minorities suffering different strains in life originating from their social status, like discrimination, might develop feelings of resentment and anger towards the larger society, which is seen as failing to protect and advance the minorities' interests.This might then weaken non-dominant group members' sense of obligation to the police in its role as an immediate representative of the state and the majority group.
While these propositions show some promise in uncovering the mechanism(s) that might drive an individual to disobey the rule of law and its enforcers, they appear to be general models seeking to explain the genesis of individual deviant and criminal acts on the micro level, at the expense of the role of social context and its impact, i.e. macro-level explanations (Matsueda 2017).More specifically, the assumptions derived from the procedural justice, self-help and strain-anger theories do not directly consider power relations or social class and so fail to address crucial differences between members of different groups that might affect their involvement in risky or delinquent behaviour and their obedience to police authorities.
Considering these limitations, there is growing interest in integrating insights across different levels of explanations in more recent criminological literature.These newer perspectives combine macro-level theories, which focus on how structural conditions such as institutional discrimination or economic deprivation affect behaviour (e.g.see Bui 2009;Cockerham 2005;Hipp 2011;Krivo and Peterson 2000;Phillips and Bowling 2003;Sampson 1987;Williams and Collins 1995), with microlevel theories, which focus on personal agency and individual characteristics that shape behaviour, like self-control, self-efficacy, anger or rational choice (e.g.see Agnew 2016;Burt, Lei, and Simons 2017;Kirk and Matsuda 2011;Williams and Mohammed 2009).Since individual behaviour always takes place within some context (Cockerham 2005;Factor et al. 2011;House and Mortimer 1990), this integrative approach allows for a more comprehensive understanding of core criminological questions, such as the differential tendencies of different people and groups to be involved in delinquent and criminal behaviour (Baumer and Arnio 2015).As such, this model should also be beneficial for explaining differences between groups in levels of obedience to the police.

The Social Resistance Framework
A more recent critical-theoretical explanation for the higher involvement of members of minority groups in risky and delinquent behaviours and, by extension, lower obligation to obey the police was suggested by Factor et al. (2011).Incorporating both macro-structural inequalities and micro-individual agency, the social resistance framework takes a further step and adds an active component by seeing non-dominant minority groups as actively resisting the dominant group.In its basic formulation, the theory suggests that non-dominant minority groups within a given society tend to engage more often in unlawful and harmful behaviours 1 as active displays of everyday opposition to the dominant group and its rules of conduct.More precisely, it suggests that social resistance among nondominant minority groups may emerge from and express: (a) alienation from the dominant group and lack of attachment to the country, with concomitant lack of commitment to the law; and (b) rejection of the dominant group's identity, culture and behaviours, while reinforcing the minority group's own collective identity in opposition to that of the dominant group (Ewick and Silbey 2003;Scott 1985).For example, if healthy behaviours are perceived as associated with the dominant group, non-dominant minority group members may engage deliberately (consciously or not) in harmful behaviours like smoking, unhealthy eating habits or lower physical activity.In this way, marginalized group members aim to change the social order and signal to the dominant group that its power is limited.Factor et al. (2011) further argue that acting unlawfully is also encouraged through positive reinforcement, as it provides immediate gratification, while the negative outcomes of such acts are perceived as distant or irrelevant.
The social resistance framework, therefore, adds to previous explanations and differs from well-known theories in several main respects.As described above, it integrates macro-structural and micro-agentic approaches, and it directly explains the involvement of members of non-dominant minorities in risky and delinquent behaviours (in contrast to previous general models which explain criminal acts and obedience, such as the procedural justice model) while adding an active component.More specifically, in contrast with the general strain theory (Agnew 1992), which explains delinquent behaviour as a way to cope with negative emotions by attaining success and material goods, the social resistance framework sees criminal behaviour as an active expression of resistance and a means to cope with discrimination.Moreover, self-help theory (and, similarly, defiance theory; Sherman 1993) explains deviance after the emergence of criminal behaviour.By contrast, the social resistance framework intends to explain the origins of the criminal behaviour itself (Factor et al. 2011;Itskovich and Factor 2023).
The social resistance theory has been tested empirically across a variety of nondominant minority groups in the United States (Factor, Williams, and Kawachi 2013c;Haddad et al. 2023), Israel (Factor et al. 2013b;Itskovich and Factor 2023;Savaya et al. 2023), and Central and Eastern Europe (Langley et al. 2021;Letki and Kukołowicz 2020), and findings show general support for its theoretical propositions.Examining the theory in the context of traffic violations, Factor et al. (2013b) found that social resistance had a direct and much greater impact on non-Jewish minority drivers in Israel compared with the Jewish majority group, while for the latter, the main antecedents of delinquent behaviour were procedurally unjust treatment by the police and non-commitment to the law.Building on 1 There is empirical evidence that non-dominant minorities have higher rates of involvement in high-risk and delinquent behaviour compared to the majority group in various societies (e.g.see Blom and Jennissen 2014;Burt and Simons 2015;Factor 2018;Friese and Grube 2008;Gofen, Cohen-Blankshtain, and Ibraheem 2021;Marshall 1997;Sitney, Caldwell, and Caldwell 2016;Stucky 2012;Veen et al. 2011;Wormith, Hogg, and Guzzo 2015).However, although this pattern appears to be widespread, there are notable exceptions in some societies and non-dominant groups (e.g.see Factor et al. 2011;Ujcic-Voortman et al. 2010).

International Annals of Criminology
hypotheses suggested by the framework, Letki and Kukołowicz (2020) showed that group alienation and discrimination increase uncooperative attitudes in areas such as tax morality or "green" behaviour when non-dominant group members are spatially clustered.Similar evidence of this clustering effect was shown by Haddad et al. (2023), who examined the higher involvement in crashes of Black pedestrians in a US city.Social resistance has also been found to help explain the positive effect of social alienation on psychological distress and sleeping problems among minority-group adolescents (Savaya et al. 2023).

The Current Study
As we have seen, the social resistance framework has been shown to provide a comprehensive explanation for minority groups' lower adherence to the rule of law while complementing previous theories, emphasizing both macro and micro factors and considering the group's social status.However, the framework has yet to be tested thus far concerning the obligation to obey a specific authority, such as the police.The present study addresses this gap.
Extending the social resistance framework, we propose that members of nondominant minority groups who perceive themselves as discriminated against and who feel alienated from and lack of attachment to the country may actively express their rejection of the dominant group and the state's institutions by disobeying the police, which is seen as an arm of the state.This resistance is not necessarily related directly to police conduct but to macro minority-majority relations and the position of non-dominant minority groups in society.It should be noted here that members of non-dominant minority groups may choose to express resistance by defying directives or orders from police officers even if they generally respect and comply with the law (just as one might express resistance by smoking or engaging in other lawful but risky behaviour).Thus, disobeying the police, under this framework, is simply another way to express one's objection to existing social structures.
We test the relationship between social resistance and obligation to obey the police among five social groups in Israel that vary in their social alienation and dominance/marginalization while controlling for previous explanations and sociodemographic variables.Israel is an ideal setting for the present study as a diverse multi-ethnic society.According to Israel's Central Bureau of Statistics (2022a), around 74% of the Israeli population is Jewish, and the rest is mainly Arab.Within these broad categories, Israel consists of various social groups distinct from one another regarding their common history, identity, language, social norms, values and socio-economic circumstances.Considering this diversity and the challenges it provokes, Israel is a "deeply divided society" (Hasisi 2007) that constitutes a useful case for the study of intergroup relations and policing in diverse democratic societies (Mentovich et al. 2020;Perry and Jonathan-Zamir 2014).For this study, we focus on four visible minority groups in Israel -Muslims, ultra-Orthodox Jews, Jewish citizens of Ethiopian origin and Jewish immigrants from the FSUwhich were found in previous studies to hold different sets of values and attitudes (Factor et al. 2014;Hasisi and Weitzer 2007;Kimmerling 2004;Shafir and Peled 2002;Yuval 2021).We use the Jewish population to compare, excluding the three Jewish minority groups studied (referred to hereafter as the Jewish majority group).
Muslims comprise around 86% of Israeli Arabs, who together make up around 21% of the Israeli population (non-Muslim Israeli Arabs include Christians, Druze, Bedouins and Circassians) (Central Bureau of Statistics 2022a, b).Israeli Muslims are a non-assimilating minority with a distinct culture and language.Members of this group suffer from higher rates of poverty and discrimination relative to the majority, even though they have full rights under the law (Factor 2019).Muslims in Israel generally identify as Palestinians rather than Israelis, rely more heavily on traditional informal forms of social control (Sorek 2011) and exhibit a lack of willingness to be in contact with the police (Hasisi 2007).Generally, their perceptions of the police are more negative and critical compared to those of the Jewish majority and even those of other Arab groups in Israel (Hasisi and Weitzer 2007;Weitzer and Hasisi 2008;Zureik, Moughrabi, and Sacco 1993).
Ultra-Orthodox Jews are a Jewish religious minority that comprises nearly 10% of the country's population (Ben-Porat and Yuval 2019).Despite their shared religious/ethnic identity, the ultra-Orthodox community stands out from the mainstream Jewish majority in their attitudes and behaviours.Historically, ultra-Orthodox Jews have expressed a general distrust, refusal to cooperate, and antagonism towards Israeli society and authority (Yagil and Rattner 2002;Yogev 2022a).These views have been reflected in its complex relations with the Israeli police, which at times have been expressed in violent clashes over social issues such as gay pride parades, swimwear advertisements (Brewer et al. 1996) and, more recently, enforcement of COVID-19 regulations (Gilman 2021).However, since the mid-1990s, the ultra-Orthodox community has improved its communication with the Israeli authorities, which has resulted in more positive views of the police and its legitimacy and a higher willingness to cooperate with police directives.It has been suggested that this positive trend is due to a convergence between the community's right-wing ideology and the state's governing leadership ideology, which has moved over the years more to the right (Yogev 2022b).
Jews of Ethiopian origin arrived in Israel in two main waves of immigration, in 1984-5 and 1991.They constitute 1.7% of the Israeli population (Central Bureau of Statistics 2022c).The Ethiopian-origin community is one of the most disadvantaged segments of Israeli society in terms of employment rates and levels of education (Offer 2004).Moreover, until recently, some religious circles in Israel refused to recognize Ethiopian immigrants as Jews (Rabinowitz 2020).The results of that can be seen in studies reporting on Ethiopians' feelings of alienation, marginalization and "otherness" in their own country (Ben-Eliezer 2008).Jews of Ethiopian origin tend to view the police in a negative light, reflecting an experience of police discrimination and mistreatment (Abu and Ben-Porat 2021) and overrepresentation of the community in national crime statistics (Shouach and Ben-Eliezer 2022).However, some studies report high levels of trust in the police among citizens of Ethiopian descent despite perceived discrimination, a phenomenon that may reflect a desire for integration and inclusion as equal members of the Israeli state (Abu, Yuval, and Ben-Porat 2017).
Lastly, immigrants from the FSU arrived in Israel during and following the collapse of the Soviet Union in 1989 and now comprise about 11% of the Israeli population (Central Bureau of Statistics 2016).This mass immigration of over 1 million people increased the Israeli Jewish population by 18% and triggered significant changes in Israeli society, economics and politics (Remennick 2017).FSU immigrants were generally secular, with high education and human capital levels, and white-collar occupational backgrounds (Amit 2012).Struggles with integration and feelings of rejection from Israeli society (Edelstein and Bar-Hamburger 2007) led FSU immigrants to consolidate their status as a separate community in Israeli society, whose members tend to hold on to the "Russian" components of their identity (Amit 2012;Leshem 2012;Shechory and Ben-David 2010).Nevertheless, the similarities between the FSU and Israel as modern societies with high levels of education and literacy made the cultural gap between the immigrants and the absorbing society relatively small (Walsh, Fogel-Grinvald, and Shneider 2015).Hence, with time, the FSU immigrants could integrate into the employment market (though at the price of occupational downgrading) and the educational system (Leshem 2012).In this light, some scholars even consider the FSU immigrants as "regular Israelis with an accent" (Remennick 2017).
The current study aims to explore variations in the obligation to obey the police between different minority groups that experience varying levels of marginalization while using the social resistance perspective to explain these differences.Specifically, we hypothesize that the association between social resistance and obligation to obey the police will differ across the examined minority groups, controlling for previous explanations and sociodemographic variables.

Data
The research was designed as an observational study based on a national randomdigit telephone survey.The sample included 1,091 Israelis -257 Muslims, 244 immigrants from the FSU, 88 respondents of Ethiopian origin and 241 ultra-Orthodox Jews, along with 261 Jewish Israelis who did not fall into any of the last three categories (the majority group).To ensure an adequate sample size for each subgroup (i.e. to increase the statistical power of our analyses and reduce standard errors), each minority subgroup was boosted, or oversampled, beyond their actual proportion in the Israeli population (Weisburd and Britt 2014).Each group was randomly sampled to achieve a representative sample of the group.
The interviews took place in early September 2015.Trained interviewers from the University of Haifa's survey institute, who were bilingual where necessary, conducted the interviews using the participants' language of preference (Hebrew, Arabic or Russian).To ensure adequate representation of the subsamples and a high response rate, up to 10 contacts were attempted for each sampled household on different days and hours.In cases where a refusal was encountered, the household was contacted again by an experienced interviewer.
The total response rate was 46% and ranged from 66% for the Jewish majority subsample to 26% among the respondents of Ethiopian origin.The total cooperation rate was 62%, ranging from 82% for the Jewish majority subsample to 50% among respondents of Ethiopian origin (for a description of the rate calculations, see American Association for Public Opinion Research 2016).These rates are comparable to those found in other large telephone surveys (e.g.Hasisi and Weitzer 2007;Lee et al. 2009;Schneider et al. 2012).The five subsamples were weighted separately by gender and age to make each subsample similar to the distribution of the corresponding subpopulation according to national data (Central Bureau of Statistics 2017).Table 1 shows the descriptive statistics for each of the five groups.

Research Tool
The questionnaire was based on the UNREST and DRQ questionnaires, which have been previously validated (Factor, Kawachi, andWilliams 2013a, Factor et al. 2013b).In addition, the questionnaire included scales for estimating respondents' obligation to obey the police and perceptions of procedural justice (Jackson et al. 2011;Jonathan-Zamir and Weisburd 2011;Mehozay and Factor 2017;Tyler and Jackson 2014), self-help (vigilantism; Tankebe 2009) and anger (Spielberger et al. 1985).The questionnaire was translated and back-translated to Arabic and Russian and was tested in a small pilot sample of 15 respondents before being administered to the entire sample.

Variables
Our dependent variable is "obligation to obey the police", measured in the questionnaire with three items (see Table 2 for the wording of the items).The main independent variable is "social resistance", measured with three items."Procedural justice" was measured with four items, "self-help" or vigilantism with two items, and "anger" with six items.Each item was measured on a six-point Likert scale ranging from 1 (strongly disagree) to 6 (strongly agree).
To test the construct validity of the scales, we conducted confirmatory factor analysis for all items (see Table 2).All the items are significant (p < 0.001), and the fit indices (comparative fit index = 0.955; root mean square error of approximation = 0.046) indicate a good fit of the model (Cheung and Rensvold 2002;Hair et al. 2006;Schumacker and Lomax 1996).Cronbach's α values for the scales, also presented in Table 2, are higher than the traditional cut-off of 0.7 (DeVellis 2003), which suggests that the scales have internal reliability.The single exception is the self-help scale, which has just two items.In cases where scales have a small number of items, which may produce small Cronbach's α values, it is recommended to calculate the mean inter-item correlation for the items in the scale.As shown in Table 2, the mean inter-item correlation for self-help was 0.23, within the recommended range of 0.2-0.4(Briggs and Cheek 1986;Pallant 2007).

Data Analysis
As a first step in analysing the data, we compared the means of the main research variables across the five groups using analysis of variance (ANOVA).Then, we performed multivariate linear regressions, predicting the obligation to obey the police from social resistance and the control variables for each of our five groups.
We analysed each group separately, rather than one model for the full sample, while including interaction terms between group affiliation and social resistance (for other examples of this approach, see Hasisi and Weitzer 2007;Murphy et al. 2018;Yuval 2021).We chose this approach for several reasons.First, our main interest in the current study is to better understand the characteristics of the individual minority groups and to identify any differences between them.Second, there are good reasons The police generally treat citizens with respect 0.85 The police generally treat citizens in a fair manner 0.88 The police generally explain their actions to people who contact them 0.79 The police allow citizens to express their opinions before making a decision regarding their case 0.71 Self-help (r = 0.23) International Annals of Criminology to believe that group membership interacts with other variables in the model.Thus, using one model would require adding many interaction terms, making the model less efficient.In this vein, using one model would also make it difficult to simultaneously test the alternative explanations previously offered in the literature (procedural justice, self-help and anger), as this would require the inclusion of still more interaction terms.Third, using one model based on a representative sample of the Israeli population would require weighting the groups according to their size in the population, thus making the sample size for some groups very small (for example, only about 22 Ethiopian Jews).
Before proceeding, we performed two tests to check the regression models.First, we calculated variance inflation factor (VIF) values for the independent variables to test for multicollinearity.The VIF values ranged from 1.05 to 1.46, with a mean of 1.16, indicating that the regressions do not raise concerns about multicollinearity.Second, to explore the potential effect of missing cases on the results, we tested whether the probability of missing cases in our independent variables is associated with the dependent variable.We found no significant difference (p = 0.118) in the means of obligation to obey the police between the observations with and without missing data.These findings indicate that missing cases are random (i.e.not associated with the dependent variable) and, therefore, do not affect the results (Allison 2001).
To illustrate the main findings, the regression coefficients were used to calculate the marginal effect displays of the predicted obligation to obey the police across social resistance levels and the five social groups, controlling for the other variables in the model by setting them to their means (Fox 2008).Because there are differences in the levels and range of social resistance between the groups, to compare the effects of social resistance graphically, we first calculated each group's average level of social resistance and one standard deviation above and below this mean (Weisburd and Britt 2014).We then calculated each group's corresponding obligation to obey prediction and plotted these results.

RESULTS
We start by comparing the main research variables across the five social groups using ANOVA.As can be seen from Table 3, there are significant differences between the groups in all of the main research variables.Most notably, Muslims (3.59) and Jews of Ethiopian origin (4.02) have the lowest levels of obligation to obey the police.Ultra-Orthodox Jews (2.58) and Muslims (2.56) have the highest levels of social resistance.Jews of Ethiopian origin (2.67) and ultra-Orthodox Jews (2.81) score lowest in perceived procedural justice, while Muslims (3.21) and ultra-Orthodox Jews (3.09) score highest in self-help.Muslims (2.76) and Jews of Ethiopian origin (2.32) report the highest levels of anger.
Next, using a set of multivariate linear regressions, we test the association between social resistance and obligation to obey the police within each of the five social groups, controlling for alternative explanations, previous contact with the police, and sociodemographic variables.Table 4 presents the regressions.As can be seen, among Muslims (b = -0.18,p = 0.039) and ultra-Orthodox Jews (b = -0.20,p = 0.007), there is a significant negative association between social resistance and obligation to obey the police when controlling the other variables in the model.Among the other two marginalized minority groups, namely Jews of Ethiopian origin (b = -0.25,p = 0.136) and immigrants from the FSU (b = -0.27,p = 0.316), we could not find a significant association.However, it is important to note that for both groups, the association is negative and that the Ethiopian-origin sample is small, which might make it difficult to find significant effects (Weisburd and Britt 2014).

International Annals of Criminology
Figure 1 presents the predicted obligation to obey the police by social resistance at the average and one standard deviation above and below the average across the five social groups while controlling the other variables in the model, which are set to their means.Although the figure is based on different models, it illustrates the predicted effect of social resistance separately for each group after controlling for the same variables.In addition, the fact that for each group, the effects are presented at both the average level of social resistance and one standard deviation above and below this mean further adjusts for differences across the groups.
The figure clearly shows the effects described above.First, for all the groups studied, there is a negative association between social resistance and the obligation to obey the police.Second, we can see (as also shown in Table 3) that there are differences across the groups in the average level of social resistance, presented as the middle point in each line, and the range, which can be seen from the length of each line.Third, looking at the relative positions of the lines, we can see that Muslims have the lowest levels of obligation to obey (over the different levels of social resistance), followed by the Jewish FSU and Ethiopian-origin groups, the Jewish majority and ultra-Orthodox Jews.Finally, we can see from Figure 1 that the strongest effect is found among ultra-Orthodox Jews and Muslims.For instance, when the other variables in the model were fixed to their means, ultra-Orthodox Jews with social resistance levels one standard deviation below the mean scored, on average, 4.67 in obligation to obey, while for those with social resistance levels one standard deviation above the mean, this score dropped to 4.06about a 13% decrease.Similarly, Muslims with social resistance levels one standard deviation below the mean scored 3.92 in obligation to obey, but among those with social resistance levels one standard deviation above the mean, this score dropped by about 12% to 3.45.
Regarding the previous explanations, Table 4 shows, not surprisingly, that procedural justice is positively and significantly associated with the obligation to obey the police in all the social groups studied.However, self-help and anger only somewhat affect the responsibility to obey in any of the groups.

DISCUSSION
In many societies, alienated minorities constitute a substantial proportion of the population.Reluctance to identify with the state may cause members of these groups to view police authorities, which represent the state, as illegitimate (Murphy et al. 2015).This, in turn, results in lower feelings of obligation to obey police directives and to cooperate with police officers on the ground.These circumstances are challenging both for the police as an institution and for overall civic health.The police depend on widespread legitimacy and compliance to avoid confrontations and operate more effectively (Tyler 2004).From the perspective of civic health, the obligation to obey the police is an important value in democratic societies and a predictor of public cooperation and compliance with the law (Reisig et al. 2023).
The present study was motivated by two main lacunae in the literature.First, while the fact that minority group members tend to hold more negative feelings towards the police compared to the majority is well established (Peck 2015), the potential heterogeneity between minority groups has received less scholarly attention (Peck 2015;Unnever et al. 2016;Unnever and Gabbidon 2011).Second, the social resistance framework (Factor et al. 2013b, c;Letki and Kukołowicz 2020) holds that the experiences that shape the lives and attitudes of minorities may encourage them to actively engage in various everyday resistance acts against the majority group.Yet, though this framework has been tested and supported in the realm of high-risk and delinquent behaviours, it has not been examined in the context of the obligation to obey the police.The current study addresses both issues: (a) by examining the obligation to obey the police in Israel among different minority groups who experience varying levels of marginalization; and (b) by testing the association of social resistance with these attitudes while controlling for socioeconomic circumstances and three alternative explanations, namely procedural justice, self-help and anger.Towards these ends, we surveyed a representative sample of Israelis from five distinct social groups that differ in their social standing and circumstances: Muslims, immigrants from the FSU, Jews of Ethiopian origin, ultra-Orthodox Jews and members of the Jewish majority group who do not fall into any of the categories above.
Our results reveal significant differences among the minority groups studied in their attitudes regarding the obligation to obey the police, social resistance, procedural justice, self-help and anger.Two of the four minority groups -Muslims and Jews of Ethiopian originfeel relatively less obliged to obey the police, while Muslims and ultra-Orthodox Jews have the highest levels of social resistance.In addition, we found a significant negative association between social resistance and obligation to obey among Muslims and ultra-Orthodox Jews, but not among Jews from the FSU and those of Ethiopian descent, where the association existed but did not reach significance.At a basic level, these findings confirm that, as assumed, minority groups differ from each other in their views and attitudes towards authorities and the police.
Delving more deeply, the findings force us to ask why certain groups within different societiesin our case, Muslims and ultra-Orthodox Jewshold more negative views towards the state than others or are more likely than others to engage in social resistance.In our case, these findings probably reflect the fact that both Muslims and ultra-Orthodox Jews identify in ways that explicitly set them apart from mainstream Israeli society.As described above, Muslims in Israel tend to view themselves more as Palestinians than Israelis (Sorek 2011), and their religious affiliation inherently excludes them from identifying with Israel as a Jewish state.This is consistent with previous findings showing that around 70% of Israel's Arab minority do not recognize the state's right to maintain a Jewish majority (Smooha 2013).As for ultra-Orthodox Jews, while they identify with the Jewish affiliation of the Israeli state, they often express their wish for the state to rely on Jewish religious law rather than on democratic values (Stern et al. 2021).
Moreover, both Muslims and ultra-Orthodox Jews in Israel tend to live (by choice) apart from other Israeli communities, concentrated in their cities or neighbourhoods (Cahaner and Malach 2021;Central Bureau of Statistics 2022b).Previous findings show that social resistance is a more potent force in communities that tend to cluster spatially (Haddad et al. 2023;Letki and Kukołowicz 2020).This may be partly because such clustering reduces these groups' exposure to different values and social norms (including police legitimacy), which can perpetuate social alienation and affect the obligation to obey and compliant behaviour.
Immigrants from the FSU are substantially more integrated within the majority society in Israel and share a relatively similar mentality (Walsh et al. 2015), which may explain their lower social resistance and higher obligation to obey the police.Our findings regarding the Ethiopian Jews (i.e. the lack of a significant negative association between social resistance and obligation to obey) are somewhat surprising, given this community's long history of a conflicted relationship with the police and feelings of alienation in Israeli society (Abu and Ben-Porat 2021;Ben-Eliezer 2008).It should be considered that the Ethiopian sample was relatively small, which can reduce the probability of finding a significant effect (Weisburd and Britt 2014).On the other hand, the community today has a strong desire to be included in Israeli society as equal members (Abu et al. 2017), and this aspiration may be reflected in the responses to our survey questions on identification with the state and social resistance.
More generally, our results raise interesting questions about the mechanisms underlying the observed patterns.For instance, our findings hint that social resistance may be a stronger predictor of obligation to obey the police among nondominant minority groups which are more marginalized (Muslims and ultra-Orthodox Jews in the present sample).However, it may be that the effects of social resistance are similar across the groups and that groups with higher levels of social resistance also have lower levels of obligation.These questions are beyond the scope of the present study and must await future research.
Another notable finding in our research is that procedural justice was correlated significantly and positively with the obligation to obey the police among all social groups in our study.This adds to the well-established literature on the process-based model of legitimacy, which holds that how police officers treat the public and display their authority have a substantive impact on how citizens of all social groups view the police and their compliance-related attitudes and behaviours (Mazerolle et al. 2013b;Tyler and Nobo 2023).Interestingly, controlling the other variables in the model, anger was not found to be related to the obligation to obey the police in any of the five groups.That result contradicts previous findings that show that anger has an impact on compliance with the law (Aseltine et al. 2000;Barkworth and Murphy 2015;Murphy and Tyler 2008).This discrepancy may reflect the fact that we measured anger as a general trait that is not specifically related to the context of policing or alienation from the state, as has been done in some previous studies (e.g.Barkworth and Murphy 2015).We presume that asking people of different social groups about their anger, specifically towards the police, might result in different outcomes than those we found.We encourage future research to address these issues.
Concerning self-help, while theoretically, this explanation can be applied in the context of the obligation to obey, its lack of significance in our results is consistent with previous empirical findings (Tankebe and Asif 2016).Acts of self-help and vigilante behaviour are more common in places where the low sense of obligation to obey the police is accompanied by palpable neglect by the police of their duty to prevent crime and keep citizens safe (Black 1983;Tankebe and Asif 2016).Applying that argument to Israel, we can assume that the insignificance in our results among all social groups hints that Israel is generally a law-abiding society where citizens largely respect the police and obey police orders.In addition, the association between self-help and the obligation to obey the police might work in the other direction, such that attitudes toward self-help are an outcome of the obligation to obey.Future studies should explore this based on longitudinal or experimental research designs that can test for causality.
It is also interesting to note that although our regression models included these four theoretical explanations, namely social resistance, procedural justice, self-help and anger, they explain only about 12 to 17% of the variance in obligation to obey the police across the five groups studied.This may suggest that criminologists should continue their efforts to find suitable explanations for this phenomenon.
Our study is not devoid of limitations.First, the data in the study are crosssectionali.e. the dependent and independent variables were measured at the same time.This limits the conclusions that might be drawn about the causal links between our variables.Future studies should apply other techniques, such as randomized controlled experiments, to identify causal associations between social resistance and the obligation to obey the police.Second, the study is based on self-reported data, with all their well-known potential limitations.We encourage other scholars to validate our results using a range of qualitative and quantitative methodologies.Third, the survey was conducted in 2015.Although it is reasonable to assume that the general attitudes and social mechanisms explored in the current study have not changed dramatically since then, future studies should be conducted to provide updated results.Fourth, the sample size of Jews of Ethiopian origin is rather small.Additional studies are required to validate the current results regarding this group with a larger sample.Finally, as mentioned earlier, it might be the case that some members of non-dominant minorities resist the police but nonetheless abide by the law.Future studies should explore this tension and seek to elucidate the conditions under which members of non-dominant minority groups choose one behaviour over another to express their social resistance.
From a practical perspective, the current study offers insights into issues that have yet to be fully explored in the Israeli context or other diverse societies with notable policy implications.Primarily, our results strengthen the notion that securing public legitimacy should be a central goal of the police, as an institution, and state officials.On the state's part, this can be achieved by enhancing social inclusion and identification with the state while putting effort into reducing discrimination and racismactual or perceived.For the police, this may be best accomplished by focusing on procedural justice policing, which has been found numerous times to enhance both compliant attitudes and actual compliance (Mazerolle et al. 2013b).In this respect, our findings suggest that while procedural justice is, of course, important for all citizens, some social groups need more focused attention if they are to be persuaded that their communities are being treated by tenets of procedural justice.Thus, states and police forces should adopt an evidencebased approach and focus on social groups whose integration remains incomplete.Based on the results of our study, in the Israeli context, these would be Muslims and ultra-Orthodox Jews.
In conclusion, the current study explored, for the first time, the effect of social resistance on the obligation to obey the police among five distinct social groups in Israel, a diverse and multi-ethnic society, while controlling for alternative explanations.The study expanded the empirical literature in several ways.First, levels of obligation to obey the police among the three Jewish subgroups in our studyultra-Orthodox Jews, Jews of Ethiopian descent and immigrants from the FSUwere unknown until now.Second, the social resistance framework has not yet been tested on the obligation to obey the police.Third, the study allowed us to refine our understanding of majority-minority relations by analysingto the best of our knowledgethe largest number of groups in a given society yet examined at one time in the context of the social resistance framework.Finally, we strengthened empirical support for the social resistance framework by controlling for other theoretical propositions that might explain differences in compliant behaviour and obligation to obey.The results indicate that among Muslims and ultra-Orthodox Jews, there is a significant association between social resistance and obligation to obey the police, while no significant association between these two concepts was found among the two immigrant groups studied -Jews of Ethiopian descent and immigrants from the FSU.In addition, among the three explanations for the obligation to obey the police that were offered previously in the literature (procedural justice, self-help and anger), we found a significant effect only for procedural justice.
Our findings add to the growing body of work on the social value of identification with the statesomething increasingly salient as societies in the twenty-first century continue to diversify through social change (Murphy et al. 2015).

Figure 1 .
Figure 1.Predicted obligation to obey the police by social resistance at the average and one standard deviation above and below the average for the five subpopulations.FSU, former Soviet Union.

Table 1 .
Descriptive Statistics of the Research Subsamples

Table 2 .
Confirmatory Factor Analysis for the Main Research Items (Standardized Coefficients) and Cronbach's α Values of the Factors

Table 3 .
Descriptive Statistics of the Main Research Variables and Analysis of Variance by Subpopulation Groups FSU, former Soviet Union.

Table 4 .
Linear Regression of Obligation to Obey the Police, Social Resistance and Control Variables by Subpopulation Groups