What predicts university students’ compliance with the law: Perceived legitimacy or dull compulsion?

Abstract Studies on police legitimacy, particularly from the developed western societies have shown that compliance with the law is primarily shaped by feelings that laws and legal authorities are legitimate and should be obeyed. However, we are unsure whether such findings would be consistent if tested among university students in a transitional society like South Africa where police-students relationship is driven by conflict, and where police misconduct is relatively high. That is, do university students comply with the law in South Africa because they perceived the police as legitimate or due to feelings of pervasive helplessness or what scholars refer to as ‘dull compulsion”. This article explores the exact factor that shapes compliance with the law among South African university students. Findings indicate that police effectiveness, predatory policing and experience of police abuse predicted voluntary compliance with the law among university students. Additionally, neither perceived legitimacy nor dull compulsion shaped compliance. The implications of the findings are discussed.


PUBLIC INTEREST STATEMENT
In most western societies, compliance with the law and legal authorities is shaped by normative feelings that legal authorities are legitimate and should be obeyed. But in transitional African societies where public-police relationship is driven by suspicion and conflict, compliance may be shaped by other factors than perceived legitimacy -particularly by feelings of endemic powerlessness or dull compulsion. In this article, we focused on university students, and determine the exact factor that shapecompliance with the law among this specific population component. Although the relationship between the South African Police Service (SAPS) and university students has not really been fantastic in recent times, but we are unsure whether such incompatibility could shape the latter's compliance with the law. From our survey, neither perceived legitimacy nor dull compulsion shaped compliance. Rather, compliance was shaped by other exogenous factors. We recommend that for compliance to be enhanced and sustained, the SAPS officers need to desist from unethical practices, and adopt permissive methods in dealing with university students.

Introduction
The importance of legal authorities' legitimacy in shaping acquiescence to the law has spurred exponential research and investigations into why people comply with the law (see Hinds & Murphy, 2007;Murphy et al., 2009;Reisig et al., 2014). While most of these studies were carried-out in the developed western societies (e.g., Australia, United Kingdom (U.K.), United States of America (U.S.), and Canada, little is known about law-abiding attitudes towards legal authorities (especially the police) in transitional African societies. Such dearth of research limits the scope of generalizing findings and theories from most western studies across diverse cultures and contexts. Considering this dilemma, this present study attempts to redress such conspicuous gap in legal-criminological literature by explaining the basis for compliance with the law in a transitional African society like South Africa.
Studies have shown that compliance with the law is based on peoples' persuasion that it is morally right to do so, and because legal authorities are considered to be legitimate (Tyler, 1990;Sunshine & Tyler, 2003). Contrastively, it is also based on peoples' prudent assessments of both the incentives and repercussion of aberrancy or noncompliance (see Nagin, 1998). Although the Tylerian explanations on compliance revolve around normative feelings that legal authorities should be obeyed because they are legitimate (Tyler, 2006(Tyler, , 1990, but such submission might have also sprout out of feelings of endemic powerlessness or what scholars have described as "dull compulsion" (see Akinlabi & Murphy, 2018;Carrabine, 2005;Tankebe, 2009). Notably, dull compulsion has only received a single testing in criminological research in Africa (see Akinlabi & Murphy, 2018); but no study has ever tested such concept in South Africa.
From a study conducted in Ghana in 2009, Tankebe contends that in post-colonial or transitional societies where there are widespread cases of police corruption, brutality and abuse, citizens' lawful behaviour or adherence to the law (i.e. compliance with the law), should not be misconstrued as an indication of police legitimacy. Rather, he avers that such acquiescence to the law, and cooperation with the police might have sprout out of a consciousness of "dull compulsion" (see Tankebe, 2009). Bolstering his argument, Tankebe explains that dull compulsion is inevitable under certain circumstances where people believe they are incapable, powerless, or unable to challenge police corruption, abuses, and other auxiliary misconducts (Tankebe, 2009). Additionally, they could helplessly cooperate and obey the directives of the police, and by extension comply with the law, to avoid police retaliation (Tankebe, 2009). In other words, people may choose to conform to police directives and voluntarily comply with the law, not necessarily out of perceived legitimacy, but because they are afraid of the consequences of not complying. Such circumstances often result in quasi-conformity, or what Akinlabi and Murphy described as "a pseudo conformity" (Akinlabi & Murphy, 2018; also see Carrabine, 2005, for a similar perspective on the corrections).
Although Tankebe and Carrabine provide plausible explanations on the nature of compliance, but the phrase "dull compulsion" itself was never subjected to empirical scrutiny or tested in their studies. However, Akinlabi tested dull compulsion in public compliance with the law, in a study that was conducted in Nigeria in 2016 (for detail see Akinlabi, 2016). Nevertheless, his study focused broadly on adult population with little consideration for specific component of the populationone of which is university students. Moreover, we are unsure whether the findings of his study will be consistent if tested in another jurisdiction, especially in a multi-racial, complex, and more divided society like South Africa. Hence, the present study represents the first attempt to empirically test "dull compulsion" as a precondition for compliance with the law in South Africa. However, unlike the previous study that focused on the general population, the present study focuses on specific component of the population-university students. The present study is driven by the need to know the exact factors that shape compliance with the law among university students.
Similar to the reports from western jurisdictions (see Avdija, 2010;Chow, 2012), university students in South Africa are often subjected to higher police scrutiny than adult population (see Bello & John-Langba, 2020b). Such scrutiny is notably driven in recent times by security exigencies -owing majorly to incessant students' protests over increment in school fees (#feesmustfall) and other ancillary student-related matters (Lange, 2017). Moreover, university students are usually audacious and dogmatic, and more inclined to pursuing their grievances in a narrow-minded manner without civilizing influences (Murphy, 2015;Murphy & Gaylor, 2010). They are recalcitrant to constituted authority, obstinately unruly, flout traffic rules, congregate in public space, and engage in risky activities that contravene the law, to mention a few (Murphy, 2015;Bello & John-Langba, 2020b). These acts constitute public nuisance and are more likely to attract spontaneous police attention (Akinlabi, 2017b;Barclay et al., 2007). It may also account for why young people are more pessimistic about the police than older population (see Akinlabi, 2017a;Hinds, 2009;Murphy, 2015).
Nevertheless, we cannot always conclude that university students will not comply with the law owing to their frequent clashes with the police. For example, in a study on student perception of the police in Ghana, Boateng (2016) found that students still have confidence in the Ghanaian police (and possibly cooperate or comply with the law) despite their involvement in risky activities. Similarly, unlike previous studies, we are not sure if compliance with the law among university students will be shaped by feelings of endemic powerlessness or perceived legitimacy. Hence, this article aims to determine the exact factor that shape compliance with the law among South African university students. Specifically, we want to determine whether compliance with the law among university students is predicted by perceived legitimacy or dull compulsion?
Additionally, we also consider whether in a multiracial society like South Africa, precursors like police corruption, procedural justice policing, police effectiveness, predatory policing and abuses, are linked with university students' perception of police legitimacy, compliance with the law, and expression of dull compulsion. However, before explaining the study methodology, it is appropriate to provide a concise information about the complexities of policing students in South Africa, followed by a review of literature on the major variables considered in the study. These will be followed by explanations on research methodology, findings, discussions, and conclusion.

Policing students in modern South Africa: complexities and disgruntlements
Generally, policing in South Africa has been troublesome in recent years (Bello & John-Langba, 2020b). Reports of people having negative experiences with the South African Police Service (SAPS) officials are bothersome, albeit not new. More worrisome is the relationship between the police and young people, especially university students (see Bello & John-Langba, 2020a). Although people of different socioeconomic strata have reported unsatisfactory experiences, or encounters with the SAPS (Vetten, 2011), but the views of specific component of the population are blurred in such reports. However, this present study intends to cast an illuminating light on an underexplored context-that is, the views of university students about the police. This study only focuses on the South African Police Service (SAPS).
Broadly, the relationships between the police and university students in South Africa have been characterised by suspicion and conflict, culminating in low level of compliance and confidence in the police (see Bello & John-Langba, 2020a). Although studies from western societies show similar findings (see Avdija, 2010;Chow, 2012;Murphy & Gaylor, 2010), but there may be divergences in terms of the nature and extent of such relationships when compared to those held in South Africa.
Besides, the context may also differ. Notably, in recent times, South Africa universities have become confusable with protest actions and compartments of illegitimate activities, regardless of the academic and democratic values that guide their operations (see Bello & Steyn, 2019). These protests actions have frequently resulted in wanton destruction of university properties and by extension public facilities. Such scenarios have often attracted police attention. The police have sometimes teargassed student protestors in order to disperse them. Unfortunately, such approach has occasionally escalated the problem-so much that ordinary protest actions have sometimes snowballed into full-blown police-students faceoffs (see Bello & John-Langba, 2020b).
Moreover, there is high discontentment among people (including university students) with the prevailing policing style in contemporary South Africa. The officials of the SAPS have been entangled in censurable misconducts, corruption, predatory policing and abuse of power (Burger, 2011;Faull, 2010). Besides the incessant daily reports of unpleasant public personal and vicarious experiences with the SAPS, there are also notable scandals and corrupt cases linked to police officers across the country (see Burger, 2011;Faull, 2010). The entrenchment in the public domain that the SAPS are corrupt is therefore not surprising (Faull, 2010;Newham & Faull, 2011). Despite the publicised reforms in the police institution, we are yet to witness durable transformation in the SAPS in real sense of it. Police officers are still involved in petty roadside corruption-demanding for inducement to look-away from offenses, or sanction offenders (Faull, 2010). It is also a commonplace in South Africa to see people grease the palms of police officers to avoid unwarranted interrogations or unnecessary delays when stopped in the street.
There are also concerns about police use of force in the public, especially to compel obedience during public protest and demonstrations (Bello & Steyn, 2019). While there may be need for the police to use of force in certain circumstances, it should be applied appositely without violating the rules of engagement. When police officers indiscriminately open fire on defenseless protestants, their legitimacy becomes questionable. Students' compliance with the law may also be jettisoned. The same is applicable in situations where people (including university students) experience predatory policing-that is, policing by suppression (see Gerber & Mendelson, 2008). Therefore, considering the prevailing policing situation in the country, determining whether university students will consider the police as worthy of their deference remains a puzzle for empirical scrutiny.
Indisputably, certain factors have been found to shape peoples' (including university students) willingness to either comply or be intransigent to the law. These factors include but are limited to: perception of police legitimacy, assessment of police procedural justice, experiences of police corruption, police effectiveness in controlling crime, predatory policing and experiences of police abuse (see Akinlabi & Murphy, 2018;Murphy, 2015;Sunshine & Tyler, 2003;J. Tankebe, 2010). This present study explores the relationship between these variables, and their effects on perception of police legitimacy or dull compulsion, and compliance with the law among university students. For instance, previous studies especially from western societies found that when students are treated fairly, they tend to perceive the police as legitimate (see Avdija, 2010;Chow, 2012;Williams & Nofziger, 2003). Moreover, researchers have contested that findings from western-oriented studies may have jurisdictional and contextual limitations, especially when tested in non-western societies (see Akinlabi, 2017a;Akinlabi & Murphy, 2018;Bottoms & Tankebe, 2012;Murphy & Cherney, 2011;Tankebe, 2009). For example, unlike the dominant findings in western jurisdictions, Akinlabi and Murphy (2018) found no correlation between police corruption and citizens' compliance with the law in Nigeria. However, since Akinlabi and Murphy's study was based on general population, we uncertain of the consistency of their results if tested in another jurisdiction, and in another context (i.e. among university students). Hence, the essence of this present study.

Procedural justice policing, police legitimacy, police effectiveness and compliance
The importance of perception of the police as a precondition for compliance with the law has propelled exponential research on procedural justice policing, police effectiveness and police legitimacy, in legal-criminological field. Such upsurge in interest might have emanated from the understanding that public-police relationship and cooperation are fundamental requirements for effective policing. There is also the realization that law-abiding attitudes are shaped by a mixture of normative factors (Akinlabi & Murphy, 2018;Hough et al., 2013). Other views emphasized the indisputability of legitimacy for legal authorities to garner public support and cooperation (see Beetham, 1991;Bradford et al., 2014).
In the developed western societies, research has shown that compliance with law is based on judgements about police legitimacy (Tyler & Fagan, 2008). It implies that people will likely comply with or acquiesce to the law and cooperate with the police when they perceive the police as legitimate (Hinds, 2009;Murphy, 2015). Fundamentally, studies have shown that judgement about police legitimacy is largely influenced by two main factors-i.e. procedural justice and police effectiveness (Jackson et al., 2012;Tyler, 1997;Sunshine & Tyler, 2003). That is, the fairness of procedures through which the police make decisions and treat people, and the effectiveness of the police in carrying out their constitutional responsibilities (police effectiveness) (see Jackson et al., 2012).
While procedural justice policing is measured by factors such as, the fairness of procedures, quality of decision-making and treatment from the police; police effectiveness generally typifies the influential impetus to legitimize the police. Generally, police effectiveness is about how well the police carryout their constitutional duties. That is, how good they are in controlling crime, detecting and deterring criminals, and in sanctioning lawlessness. Therefore, the underlying assumption is that police legitimacy will be assessed based on peoples' judgement about the fairness of treatment they receive from the police. That is, whether people are fairly or justly treated by the police or not, and whether the police are effective in keeping their communities safe and in controlling crime in their neighbourhoods, or not (Sunshine & Tyler, 2003). Such assessment whether positive or negative will influence public compliance with the law.
Moreover, there is the possibility that in a country like South Africa where police-public relations are marked by violence, abusive and predatory form of policing, and corruption, people may not consider the police as legitimate (see Burger, 2011;Faull, 2010;Gerber & Mendelson, 2008). If that is the case, will people not obey the police or acquiesce to the law because they failed to consider the SAPS as legitimate? This study proposes that the public might acquiesce to the law due to other reasons. We also argue that police legitimacy may not offer an all-inclusive explanation for compliance with the law, especially in transitional African societies. And as Tankebe (2009) opined, feelings of dull compulsion might be a major contributory factor in explaining compliance behaviour. That is, people might comply with the law, not necessarily out of perceived legitimacy of the police, but because they fear police reprisals for not complying.
Importantly, there is currently no notable scientific study that precisely examine or theoretically substantiate the motivational ground for compliance with the law in South Africa. Conjectures about the subject, if mentioned at all, are anecdotes from media chats or reports, deliberations, or in pastime jokes, not based on empirical evidence, let alone sound one. Investigations into why people choose to comply with or disobey the law may passably auspicate the weakness of the law or legal system. Nonetheless, it is just a slight indication of a larger possibility. There are other justifiable reasons people comply or become adamant towards the law that transcend the weakness of the law or legal system.
As earlier explicated, previous studies on the nexus between procedural justice policing, police legitimacy, and police effectiveness, and compliance with the law focused on the general population with little consideration for specific component of the population. With the incessant policestudents' confrontations in South Africa, we are unsure if such conflicting relationships will affect students' perception of police legitimacy, and negatively affect compliance with the law. Although previous studies from the general population infer that the pendulum could swing either way (Akinlabi & Murphy, 2018;Tankebe, 2009), but we are unsure whether similar outcome will be generated or not, if tested among university students. Moreover, we cannot always conclude that students' negative experiences with the police will culminate in negative perceptions of the police without subjecting it to empirical scrutiny. That is, do university students comply with the law because they see the police as legitimate and worthy of being deferred to? Or do they comply with the law out of a sense of powerlessness or dull compulsion? In specific terms, this study tests the exact predictors of compliance with the law among university students in South Africa.

Participants selection for data collection
Data for this study were delineated to a research conducted in 2018 from a sample of students at the University of KwaZulu-Natal, South Africa. We were able to obtain a sample of 482 students from the institution through a convenience sampling technique. The sampling technique was adopted due to tight students' schedules during the survey period. Nevertheless, we endeavored to secure a sample that relatively reflect the university's student demographic composition and diversity.
The first author worked with some research assistants to administer questionnaires to students in lecture theatres after due permissions were obtained from the authorities of the institution. The aim of the study was conveyed to students, and comprehensive explanations were given on ethical issues before consent forms were signed and questionnaires administered. A total number of 600 hard copy questionnaires were administered, but only 482 copies were completed and retrieved, indicating a response rate of 80.3%. The socio-demographic compositions of the study participants are presented in Table 1.

Measures
A couple of instruments was used to test the study hypotheses. The study essentially has three outcome variables, i.e. police legitimacy, dull-compulsion and compliance with the law, and were measured as composite. The composite measures of five concepts reputed to shape self-reported compliance with the law were computed on one hand and its predictions of police legitimacy or dull compulsion were constructed on the other hand. All instruments were measured on a five-point Likert scale ranging from 1-"strongly disagree" to 5-"strongly agree". All the scales were adapted from previous studies, such as Akinlabi and Murphy (2018) Tyler (2006), and Carrabine (2004). The significance of these instruments to the present study are explained below:

Police legitimacy
It assessed participants' perspectives on their commitment to obey and consider the police worthy of their deference. The Cronbach alpha result was (α = 0.87), Mean = 9.10; SD = 4.05. A high score on this scale implies a higher perception of police legitimacy.

Compliance with the law
It assessed the extent of participants' willingness to comply with the law. The Cronbach alpha was α = 0.90; Mean = 9.03; SD = 4.22. A high score on this scale indicates a higher level of compliance with the law.

Dull compulsion
It assessed participants' perceived feelings of whether rebuffing police officers' directives would be detrimental to their life. The Cronbach alpha was α = 0.87; Mean = 9.11; SD = 4.10. A high score on this scale implies a high level of dull compulsion.

Procedural justice
It assessed participants' perceived fairness or justness of police procedures. The Cronbach alpha was α = 0.83; Mean = 13.56; SD = 5.25. A high score on this scale implies that the police often adopt fair procedures when dealing with people.

Police effectiveness
It assessed participants' perceived effectiveness of the police in controlling crime. The Cronbach alpha was α = 0.89; Mean = 11.40; SD = 4.99. A high score on this scale implies the police are doing well in discharging their duties.

Experience of police abuse
It assessed participants' experience of police abuse. It has a Cronbach alpha of α = 0.90; Mean = 9.04; SD = 4.22. A high score on this scale implies a higher experience of police abuse.

Police corruption
Assessed participants' experiences of police corruption. The Cronbach alpha was α = 0.89; Mean = 9.07; SD = 4.06. A high score on this scale implies higher perception of police corruption.

Predatory policing
Considered as policing by suppression (see Gerber & Mendelson, 2008). The scale assessed participants' perceived experiences of regular suppression by the police. The Cronbach alpha was α = 0.89; Mean = 9.14; SD = 4.04. A high score on this scale indicates a high level of police subjugation.

Result
Data for this study were analysed using IBM SPSS statistics 25. Three major analyses were conducted-Correlation, Ordinary Lead Square (OLS) regression, and Multiple regression analyses, besides the descriptive statistics. For correlation, a bivariate correlation was carried-out (i.e. Pearson correlation) to evaluate the relationship between the scale variables (see Table 2). The results show that there are significant correlations between the variables. While some correlated positively, others were negative. For instance, while compliance with the law was negatively correlated with police legitimacy (r = −.112, p <.01), police effectiveness was positive (r = .390, p < .01). It implies that while students who were unwillingly to comply with the law are less likely to view the SAPS as legitimate, those who considered the SAPS as effective in controlling crime are likely to perceive the SAPS as legitimate. Contrastively, police effectiveness and experience with the police correlated positively with police legitimacy.

Predictors of police legitimacy and dull compulsion
To address the aims of this study, OLS and multiple regression analyses were conducted. Tables 3  and 4 show the predictors of police legitimacy and dull compulsion, and predictors of university students' self-reported compliance with the law, respectively. From the first model (police legitimacy) as presented in Table 3, none of the demographic variables predicted police legitimacy. However, only one out of the five independent variables exerted significant effect on police legitimacy. Specifically, police effectiveness (β = .16; P < .00) was a significant predictor of police legitimacy. Unexpectedly, procedural justice did not predict police legitimacy among university students. It implies that if the police are effective in combating crime in the country, the higher the likelihood that university students will consider them as legitimate.
The above findings indicate that for students to defer to the police in South Africa, the effectiveness of the police in crime control and helping to protect lives and properties are fundamental. Importantly, this regression model accounts for only 9% of the overall perception of police legitimacy. The value is relatively small in comparison to the values reported in studies on students and the police in most western societies. The implication is that although perceived police effectiveness, exerted significant effect on police legitimacy, there are other factors that are likely to shape university students' perception of police legitimacy beyond police effectiveness.
In the second model (dull compulsion); similar to the first, none of the demographic variables predicted dull compulsion. Meanwhile two of independent variables-police abuse (β = .13; P < .05), and police corruption (β = .32; P < .001) predicted dull compulsion. It implies that students who viewed the police as abusive and corrupt show a greater tendency to rationally obey the police (i.e. an expression of dull compulsion due to endemic powerlessness). That is, students may comply with, or rationally acquiesce to the SAPS not because they perceived the SAPS as legitimate but due to their prudent calculations of the risk and consequences of disobeying police orders.
Turning to table 5, eleven (11) variables were computed to predict university students' voluntary compliance with the law in South Africa. In step1, four (4) demographic variables were computed; but none predicted voluntary compliance with the law. In step 2, five variables were added to the initial four demographic variables. In this model, a combination of gender (β = .07; P < .05), experience of police abuse (β = −.14; P < .05), police effectiveness (β = .24; P < .001) and predatory policing (β = −.14; P < .05) predicted compliance with the law. It implies that while male students who viewed the police as effective in crime control are more likely to comply with the law; those who had experienced police abuse and predatory policing are less likely to acquiesce to the police or comply with the law.
In step 3, police legitimacy and dull compulsion (though outcome variables) were mediated with other five (5) scales and four (4) demographic variables to test their effects on voluntary compliance with law. Of these variables, it is gender (β = .09; P < .05), experience of police abuse (β = −.16; P < .05), police effectiveness (β = .26; P < .001) and predatory policing (β = −.15; P < .05) that predicted self-reported compliance with the law. Notably, police legitimacy and dull compulsion did not make any important impact in the model. The results are similar to ones obtained in step 2. Summarily, it suggests that while male students and those who viewed the police as effective have greater tendency to comply with the law, those who had experienced police abuse and predatory policing are less likely to comply with the law in South Africa.

Discussion
This study aims to achieve two things: first, to determine the factors that explain university students' perception of police legitimacy and feelings of dull compulsion; and secondly, to determine the exact factors that predict university students' self-reported compliance with the law. We established earlier that western studies have consistently found police effectiveness and procedural justice as key determinants of police legitimacy, even in studies among students and young people (see Hinds, 2009;Hough et al., 2013;Murphy, 2015;Murphy & Gaylor, 2010). That is, when students perceive the police as effective in crime control and in keeping their communities safe, and when they view them as procedurally fair, chances are high that they will consider the police as legitimate. We also found procedural justice to be the most significant determinant of police legitimacy in previous studies. Judgements about police legitimacy were also found to the strongly linked with public self-reported compliance with the law (see Hough et al., 2013;Jackson et al., 2012).
In the current study, police effectiveness, not procedural justice policing, proved to be the greatest significant predictor of police legitimacy among university students in South Africa. Consistent with the findings of Akinlabi and Murphy (2018) and Tankebe (2009), we found that procedural justice might not always account for, and possibly be the most significant determinant of police legitimacy in all contexts and cultures (even in studies on university students). For instance, in comparison to the developed western societies, South Africa is a multiracial and complex society with high crime rate, exacerbated by increasing level of unemployment, inequality and poverty. Hence, the major concern of an average South African student will possibly be safety and security of lives and properties. Therefore, the capacity of the SAPS officers to effectively control crime and ensure communities are safe is important to South African university students. If these expectations are unmet, university students may likely view the SAPS as ineffective and question their legitimacy.
Additionally, we established earlier that we were unsure if perceived legitimacy or expression of dull compulsion would be associated with compliance with the law among university students in South Africa. Interestingly, in the current study, neither perceived legitimacy nor dull compulsion was associated with university students' compliance with the law. However, we do not envisage that perceived legitimacy would not predict compliance with the law. Previous findings from western studies (albeit among general population) found a strong linkage between perceived police legitimacy and compliance (see Hough et al., 2013;Reisig et al., 2014). Unfortunately, this is not the case among South African university students. Our findings also contradict what Tankebe (2009) found in his Ghanaian study that dull compulsion could explain why Ghanaians are acquiesced to the law.
If neither perceived police legitimacy nor dull compulsion predicted compliance, what then explained compliance with the law among university students in South Africa? From our findings, a variety of variables predicted compliance with the law; they include gender, experience of police abuse, perceived police effectiveness and predatory policing. Unexpectedly, male students showed greater tendency to comply with the law, than the female folks. Police effectiveness also exerted significant effect on students' self-reported compliance. That is, those who perceived the police as effective in their duties are more likely to comply with the law.
Contrastively, those who have experienced police abuse and predatory policing were less likely to comply with the law. Unlike Tankebe's (2009) findings where experiences of police abuse were positively associated with compliance (suggesting a sense of dull compulsion), in the present study, perceived experience of police abuse and predatory policing reduced compliance. It implies that when police abuse, and subjugate people (including university students), then university students are less likely to comply with the law in South Africa. Importantly, these findings indicate that there could be contextual variations in judgement about compliance with the law (although our study is limited to university students). Therefore, generalising some western theories and findings to transitional societies may not always suffice, particularly in studies on students' samples.

Recommendations and conclusion
This paper sets out to determine the factors that shape police legitimacy and dull compulsion, and the factors that predict compliance with the law among South African university students. This study has made significant contributions to the field of criminology by complementing and refuting previous findings especially from the developed western societies, and in a few African studies (in relation to predictors of police legitimacy among university students). It is also the first empirical study in South Africa (although we focused on university students) to test Tankebe's (2009), and Akinlabi and Murphy (2018) dull compulsion assumption as prerequisite for compliance with the law.
Importantly, while these findings have implications for effective policing in contemporary South Africa particularly in relation to university students, the study is however not immune to limitations. Considering its cross-sectional nature, it is difficult to make causal conclusions. As such, these findings should, be treated with caution, as they do not represent the views of the entire South African university student population. Additionally, since this study is limited to the views of university students, we cannot generalise the findings to the broader South African society. Therefore, future studies can focus on the broader South African population, i.e., both adult and younger population, and make both causative and comparative inferences. Nevertheless, the highlighted limitations cannot detract the relevance and contribution of this study to existing literature on compliance with the law among university students, especially in transitional societies. Future research can adopt a longitudinal survey approach, and explore factors not considered in this study.
Notably, this study has contributed to a very scarce literature in South Africa, and by extension, Africa. Specifically, it has shown that university students will only perceive the police as legitimate if police officers are not abusive and subjugating. And when they are effective in fighting crime in the country. Thus, when SAPS officers desist from unethical practices and adopt permissive methods in dealing with university students, compliance with the law and cooperation with the police among this population component will be enhanced and sustained.