Skip to main content
SearchLoginLogin or Signup

Why do People Cooperate with the Police and Criminal Courts? A Test of Procedural Justice Theory in 30 Countries

This paper presents a cross-national test of the portability of procedural justice theory. Analysis of sample survey data from 30 diverse social, political and legal contexts across Europe and beyond finds qualified support for the theory across national borders.

Published onJul 01, 2024
Why do People Cooperate with the Police and Criminal Courts? A Test of Procedural Justice Theory in 30 Countries
·

Abstract

This paper presents a cross-national test of the portability of procedural justice theory. Analysis of nationally representative sample survey data from 30 diverse social, political and legal contexts across Europe and beyond finds qualified support for the theory across national borders. First, in most countries a normative account of public cooperation with the police (based on procedural justice and legitimacy) has greater empirical purchase than an instrumental account (based on effectiveness and fear of crime). Second, procedural justice is the strongest predictor of legitimacy in most of the 30 countries, but it is an especially important predictor of legitimacy and cooperation in countries where the police seem to be a positive group authority (using national levels of procedural justice as a proxy measure). We finish with a call for more methodologically equivalent research into why procedural justice theory works better in some social, political and legal contexts than it does in others.

Academics, policy makers and justice officials have long known the importance of public cooperation with the police and criminal courts (Meares, 2017; Tyler, 2017; Frydl & Skogan, 2004). Law-breaking goes unnoticed if nobody reports it; crimes go unsolved if the public do not come forward with vital information; trials are ineffective if nobody gives testimony in court. Policy makers and practitioners need to understand how to encourage public cooperation, and when designing policing policy and practice, two ideal types often guide discussion (Tyler, 2011a, 2011b). On the one hand, an instrumental account puts crime-fighting and risk-reduction center stage. If people are more willing to help a police force that they view as effective at deterring crime, protecting communities and exerting authority on the streets, then police need to demonstrate their crime-fighting skills to the public. On the other hand, a normative account prioritises legitimacy and fair process in police-citizen interactions. If people are more willing to help police that they believe are moral, just and appropriate, then police need to prioritise the rightful exercise of power and authority, especially fair interpersonal treatment and decision-making during interactions with citizens.

In this paper, we present the fullest comparative cross-national test to date of procedural justice theory (PJT, Tyler, 2006a, 2006b; Sunshine & Tyler, 2003). A good deal of evidence from the US (Tyler & Fagan, 2008; Tyler & Jackson, 2014), UK (Jackson et al., 2013; Huq et al., 2017), Australia (Murphy & Cherney, 2012; Mazerolle et al., 2013) and Israel (Jonathan-Zamir & Weisburd, 2013; Metcalfe et al., 2016) supports two key propositions of PJT: first, that procedural justice is a stronger predictor of legitimacy than distributive justice and effectiveness; and second, that procedural justice and legitimacy are stronger predictors of public cooperation with the police, compared to factors like effectiveness, perceptual deterrence and fear of crime (for reviews of the literature, see: Jackson, 2018; Bolger & Walters, 2019). Studies thus far support the idea that legal authorities should—as much as politically and practically possible—privilege consensual models of crime-control that emphasise fair process, group inclusion and legitimacy, over more pro-active, aggressive forms of policing that emphasise strength, power and deterrence (Tyler, 2003).

Yet despite the popularity of PJT in an increasing number of countries across the world, the evidence base is limited in at least two respects.1 First, the geographical coverage needs to be expanded to cover a wider range of social, political and legal contexts. Second, there is little commitment in the literature to methodological equivalence in sampling, conceptualisation, measurement and modelling. If we are to test the portability of PJT across national borders, we need comparable cross-national data. If we are to develop a properly comparative cross-national body of evidence, we need a body of work that (a) covers a larger number of countries and (b) utilizes (at the bare minimum) some kind of agreed basis on which to define, measure and model concepts in consistent ways in different national settings. All this needs to be done within a theoretical framework that has definitional clarity and analytical power.

By way of contribution, we present findings from a methodologically equivalent 30 country study. We link nationally representative sample survey data from Round 5 of the European Social Survey2 (ESS) to two matching representative-sample surveys of the US and South Africa.3 First, we model the predictors of legitimacy. We treat the value-content of legitimating norms—how police need to be seen to act if they are to be judged as legitimate holders of power in a given national context—as an empirical question, to be assessed using nationally representative sample surveys rather than imposed top-down by researchers (Jackson & Bradford, 2019; Trinkner, 2019). Second, we model the predictors of cooperation. We assess and compare the motivational bases driving people’s willingness to cooperate with the police and criminal courts in each country. Third, we assess whether the potential sources and consequences of police legitimacy vary according to the average national levels of procedural justice.

Overall, our study not only tests PJT in 30 countries, it also narrows the gap in the literature on what it is about the social, political and legal jurisdiction that gives the theory particular purchase. PJT predicts that when officers send messages of group inclusion and status to most citizens—by acting in procedurally just ways—they show to those they police that they are a valued part of the group that legal authorities represent (a group that one could conceptualise as mainstream law-abiding society). Building on the idea that the police represent a stronger and more identity-relevant group authority in countries with high levels of enacted procedural justice towards citizens, we examine whether people are more likely to orient themselves towards legal authority in normative (rather than instrumental ways) in countries with high aggregate levels of procedural justice. We find that (a) procedural justice is a more important predictor of legitimacy than distributive justice, effectiveness and fear of crime in countries where the police are routinely experienced to behave in procedurally just ways, and (b) procedural justice and legitimacy are more important predictors of cooperation than effectiveness and fear of crime in those same countries. We conclude with the idea that high national-levels of procedural justice bring with it an increased salience of—and connections to—the group that the police represent, thereby enhancing the normative nature of police-citizen relations in those contexts (Trinkner, 2019).

The rest of the paper proceeds as follows. After reviewing PJT, we outline the theoretical model that we test in each country. We then set out an explanation for why procedural justice and legitimacy may be more important in some countries than in others. After describing the fieldwork, measures and analytical strategy, we report the findings of country-specific analyses and the cross-national comparisons. Our conclusions (a) position the contribution within the existing literature, (b) discuss study limitations, and (c) suggest some future lines of enquiry.

A CROSS-NATIONAL PERSPECTIVE ON PROCEDURAL JUSTICE THEORY

What legitimates the police?

PJT is a popular explanatory framework for understanding what legal authorities can do to secure legitimacy in the eyes of citizens and, relatedly, how justice officials can encourage people to cooperate with the criminal justice system and comply with the law (Sunshine & Tyler, 2003; Tyler & Huo, 2002). A starting premise of PJT is that, when legal officials are seen to wield their power and authority in normatively appropriate ways, those who are subject to (and beneficiaries of) that power acknowledge their rightful authority to govern. PJT posits that procedural justice is the single biggest factor underpinning the normative justification of police power, especially in contexts where individuals identify with the social group that the authority represents (Tyler & Lind, 1992; Lind & Tyler, 1988; Tyler, 1997; Trinkner 2019).

Why is fair interpersonal treatment and decision-making in police-citizen interactions so fundamental to the legitimation of the police? First, there is a commonly held normative expectation that legal officials should treat citizens with respect and dignity, should give people a sense of voice and transparency in the decision-making process, should act impartially, and should convey trustworthy motives. By conforming to normative expectations about how power should be exercised in police-citizen interactions, procedural justice helps to (a) generate the belief that the institution is normatively valid and (b) strengthen the internalized moral value that one should voluntarily defer to police authority (Jackson, 2018; Jackson & Bradford, 2019). Second, Tyler’s relational framework of procedural justice and legitimacy stresses the importance of group identities to police-citizen relations (Tyler, 1997). When authority figures wield power in ways that accord with principles of fair process, this sends signals of value, status and inclusion to individuals (Lind & Tyler, 1988; Tyler & Blader, 2003a, 2003b) within the social categories the police represent (categories that are usually conceptualised and operationalized in terms of law-abiding national, community or citizenship identities).4 Especially in contexts in which the police already represent an important superordinate group in society, procedural justice helps build group solidarity and bonds; people internalize the goals, values and motivations of the group—and this includes recognizing the legitimacy of group authorities (Tyler, 1997; Trinkner, 2019).

Looking across the available evidence (Jackson, 2018), procedural justice does seem to matter the most when police power is being exercised. Studies in the US, UK, Australia and Israel routinely find that being seen to act in procedurally fair ways helps to legitimate the police. Importantly, the effectiveness of the police and whether police allocate outcomes—such as arrests, citations, protection and service—and finite resources fairly across aggregate social groups (i.e. distributive justice) both seem to be less important to the generation and maintenance of legitimacy.5 Yet, while the evidence for the importance of procedural justice to police legitimacy is relatively strong in the US, Australia, UK and Israel, studies set in South Africa, China, Pakistan and Ghana have found that effectiveness and/or freedom from corruption are at least as important to legitimacy as fair interpersonal treatment and decision-making (Bradford et al., 2014; Boateng et al., 2022; Sun et al., 2017; Jackson et al., 2014; Tankebe, 2009; although also see Wu & Liu, 2023). Intriguingly, these are countries that have more authoritarian approaches to policing and governance, and/or have yet to establish a baseline level of effectiveness and safeguards against corruption, and/or have police that are historically rooted in colonialism. In such contexts, it may be that people do not identify very strongly with the social group that the police represent (Trinkner, 2019). This would partly explain why legitimacy seemed to depend less on procedural justice and more on factors such as effectiveness and distributive justice (Tyler, 1997).

Public cooperation with the police and criminal courts

In addition to referencing the criteria that people use to judge the rightful authority of police, PJT is also interested in understanding why people cooperate with the police, comply with the law, and support the empowerment of legal authorities and (reasonable) use of force. In particular, PJT predicts that cooperation (the focus of the current study) is encouraged through procedural-justice based approaches to rule enforcement that create legitimacy and motivate people to help group authorities. Tests of PJT in the US, UK, Australia and Israel routinely find that, when people view the police as a moral, just and appropriate institution that is entitled to dictate appropriate behavior in certain circumstances, they also tend to think that it is the right thing to do to support their goals and function (e.g. Sunshine & Tyler, 2003; Tyler & Fagan, 2008; Jackson et al., 2013; Jonathan-Zamir & Weisburd, 2013; Mazerolle et al., 2013; White et al., 2016).6 Notably, Bolger & Walter’s (2019) meta-analysis of the literature found not only that procedural justice was a common indirect predictor of cooperation through legitimacy; it was also a direct predictor of cooperation (bypassing legitimacy). Procedural justice—and the identification processes it strengthens—means that people internalise the goals, values and motivations of the group (Trinkner, 2019).7 Leaving aside the role that legitimacy plays, this strengthens the willingness of people to come forward voluntarily to report a crime, give vital information to officers to aid their investigation, and give evidence in court.

Outside of the US, UK, Australia and Israel, the evidence is patchier. Kochel et al. (2013) found that procedural justice and legitimacy were positive predictors of the decision of victims to report the crime to the police in Trinidad & Tobago. By contrast, Tankebe (2009) showed that effectiveness was a stronger predictor of people’s willingness to cooperate with the police than legitimacy in Ghana. Work from Japan found that neither duty to obey nor normative alignment with the police predicted general willingness to report a crime and give information to officers (Tsushima & Hamai 2015). In the latter study, most research participants expressed a strong willingness to cooperate—perhaps the norm to cooperate was generally strong in Japan, meaning that there was less ‘space’ within which legitimacy can operate (cf. Jackson et al., 2021)?

THE CURRENT STUDY: A CROSS-NATIONAL TEST OF PROCEDURAL JUSTICE THEORY

Overall, PJT is receiving an increasing amount of attention in a growing number of countries across the world. Studies have explored the norms that legitimate and different motivations to cooperate, the evidence generally supports key PJT predictions. Yet, international enthusiasm has outpaced commitment to cross-national coverage and comparability. Studies have tended to cluster in the US, UK, Australia and Israel, and even when the net has been cast wider, the lack of methodological equivalence—the directly incomparable approaches to sampling, measurement and modelling—means that it is difficult to make formal comparisons across studies. We do not know whether PJT works better in some countries more than other countries because we lack methodologically equivalent nationally representative data from a large number of countries. Consider, for example, the lack of agreement on how to define, measure and model legitimacy. Kochel et al. (2013) captured legitimacy using indicators of obligation to obey the police and law; Tankebe (2009) measured legitimacy using items about trust in the police and obligation to obey the police; and Tsushima & Hamai (2015) tapped into legitimacy using indicators of normative alignment and duty to obey. Such a lack of methodological equivalence, as well as inconsistency in the methods of modelling used, means that it is difficult to (i) compare findings and (ii) know precisely why PJT works better in some countries than it does in other countries. For the evidence base to develop, we need properly comparative, cross-national work.

This is the goal of the current paper. In our test of the portability of PJT across 30 countries, the methodologically equivalent data cover the US, Northern Europe (e.g. Denmark, UK and Ireland), Western Europe (e.g. France, Germany and Netherlands), Eastern Europe (e.g. Hungary, Ukraine and Russian Federation), Southern Europe (e.g. Portugal, Spain and Greece), down to Israel and South Africa. Figure 1 details the theoretical model that we test in each country. We motivate this formulation of the theoretical model in the following section.8

FIGURE 1 A procedural justice model of public cooperation with the police

Legitimation and legitimacy

We define legitimacy along two connected lines: normative alignment and duty to obey (Jackson et al., 2012, 2013). First, normative alignment is the belief that officers generally act in ways that align with their values and normative expectations. Officers embody the institution (they wield its power). They implicitly and explicitly claim authority on this basis and, when citizens generally believe that officers wield the power of the institution in normatively appropriate ways (i.e. in ways that accord with values and norms dictating normatively appropriate use and direction of power), they recognize the institution’s moral right to power. Second, duty to obey is the consent part of legitimacy—this is the recognition of the authority’s right to govern (for discussion, see Posch et al., 2021; Hamm et al., 2022). When people internalize the overarching moral value that they should willingly obey an authority because of the source (because they’re the police) not the content of its instructions or orders, they allow that authority to dictate appropriate behavior (Tyler & Jackson, 2013).9 They feel a moral obligation to obey rules and orders (and willingly accept decisions) that emanate from an institution that they willingly assent to (Tyler, 2006a, 2006b).

Normative alignment was captured by asking research participants whether they think the police generally act in ways that align with their own values and normative expectations (duty to obey was also measured using content-independent items). This puts clear methodological water between measuring legitimacy and measuring public attitudes towards specific forms of police conduct, specifically procedural justice, distributive justice, effectiveness and lawfulness. By modelling the predictors of normative alignment and duty to obey (using measures of procedural justice, distributive justice, effectiveness and lawfulness), it becomes an empirical question which normative expectations legitimate the police in each of the 30 countries. Research participants are dictating (at scale) the content of appropriate use of power, not the researcher (for discussion, see Jackson & Bradford, 2019; Trinkner, 2019). If, for instance, procedural justice perceptions emerge as the key predictor of legitimacy in a given context, then we can infer that procedural justice is a basic expectation about the legitimate use of power in that jurisdiction.10

So, what criteria do people in different countries use to judge the legitimacy of the police (Figure 1)? Procedural justice is the first potential predictor of normative alignment and duty to obey. A positive link between procedural justice and normative alignment would reflect the idea that citizens partly judge the normative appropriateness of the institution (whether power is appropriately used) on the basis of whether officers seem fair, respectful, neutral and transparent in their interpersonal treatment and decision-making during police-citizen interactions. If procedural justice emerges as the strongest predictor of normative alignment in a given country, then procedural justice is the strongest legitimating norm (with regards to police behavior) in that particular context. Similarly, a positive link between procedural justice and duty to obey would indicate that citizens are judging, to some degree, the rightful authority of the institution to dictate appropriate behavior (i.e. whether they feel it is their moral duty to defer to officers) on the basis of whether officers seem to respect the importance of fair process.

The second potential predictor is distributive justice. Distributive justice is the extent to which outcomes (the ‘goods’ and ‘bads’ of policing) are distributed fairly across aggregate social groups in society. To measure perceptions of distributive justice, research participants were asked whether they thought police officers treated different groups (rich vs poor, ethnic majority vs ethnicity minority) equally.11 If perceived distributive justice is a positive predictor of normative alignment, then people are judging the normative appropriateness of the institution to some degree on whether officers do not discriminate against certain groups in society in their treatment of citizens. If perceived distributive justice is a positive predictor of duty to obey, then people are judging the rightful authority of the institution to some degree on the same basis.

The third potential predictor is lawfulness. In terms of normative alignment, lawfulness may be an important normative expectation about the appropriate use of power. People may expect that officers follow their own rules as well as the rules that govern other members of a society and judge their moral right to power on whether they actually do or not. In terms of duty to obey, lawfulness may partly shape the rightful authority to dictate appropriate behavior. Deference may depend to some degree on the belief that officers themselves ‘stick to the rules’. Feeling a duty to obey officers may thus depend to some degree on reciprocity—if they do not follow rules and commands, why should I?

The fourth potential predictor is effectiveness. If procedural justice, distributive justice and lawfulness reference particular types of fairness in police action, effectiveness is about the outcomes they produce in relation to crime, risk and safety. Legitimacy could depend on their ability to affect these outcomes – after all, the police are given power to enforce the law and respond to problems. If effectiveness is a positive predictor of normative alignment, then being able to visibly fight crime is a legitimating societal norm regarding the appropriate use of power. If effectiveness is a positive predictor of duty to obey, then people are more likely to recognize the rightful authority of the police to dictate appropriate behavior when they view the police as strong in the fight against crime.

The final potential predictor is fear of crime. Things here are a little more speculative, since there is less evidence on the link between fear of crime and legitimacy. The relationship could be negative. People who worry about their own personal risk of criminal victimization may question the appropriateness and authority of the police. They might believe that a key function of the police is to assert social control and to make people feel safe, and fear of crime may encourage them to question the legitimacy of the police. The relationship could also be positive. When people worry about becoming a victim of crime, they may to some degree feel dependent on the outcomes of the police. They may feel motivated to grant them legitimacy in order to manage threats and reduce uncertainty (Van Der Toorn et al., 2011; Kochel, 2018).

Why do people cooperate with the police and criminal courts?

On the right-hand side of Figure 1 is willingness to cooperate with the police. To measure this construct, we capture instances of behavior that are of perhaps most immediate benefit to the administration of justice—the calls that individuals make to the police to report crimes or anti-social behavior, the information that they give to the police to aid their investigations, and their willingness to get involved in legal trials and court proceedings. While individuals come into contact with the police for many different reasons, acts such as these involve matters of personal concern to those involved. They are of great practical benefit to the police and criminal courts, as well as an active recognition of the propriety of the police remit over matters of crime and disorder.

The first two potential predictors of cooperation are normative alignment and duty to obey. If normative alignment is a positive predictor of willingness to cooperate, then a certain type of reciprocity may be at play: believing that the police act in normative appropriate ways encourages research participants to act in normatively appropriate ways, one of which is to help police and courts fight crime. The role that duty to obey plays in willingness to cooperate would, by contrast, be more to do with respect for a position of authority; people defer to the police’s implicit expectation in a given society that citizens should proactively help them fight crime.12

Another potential predictor of cooperation is procedural justice. One reason why people might be willing to call the police to report a crime, give information important to some potential police investigation, and provide evidence in court, is that they believe that officers are generally respectful, professional and objective (and might therefore treat them in such ways in an encounter). Some studies have found that procedural justice predicts cooperation, adjusting for legitimacy. Reisig & Lloyd (2009) found that obligation was not a significant predictor of willingness to cooperate with the police, drawing on data from a sample of adolescents in Jamaica, but procedural justice was a significant predictor. Kochel et al.’s (2013) study in Trinidad & Tobago also found that obligation was a weak (and statistically insignificant) predictor of people’s willingness to report a crime to the police, and that procedural justice was more important. As already mentioned, Bulger & Walters (2019) found in their meta-analysis evidence for a direct link between procedural justice and cooperation.

These three potential predictors reflect normative motivations to cooperate. The final two reflect instrumental motivations. The fourth potential predictor is the belief that the police are effective at fighting crime and managing people’s sense of risk. People may be more likely to cooperate when they believe that the police are effective at catching criminals, deterring crime, and responding quickly in an emergency. The goal of the police is to reduce crime and risk and they think helping an effective police force would be beneficial in that regard. Conversely, if the police seem to be ineffective in fighting crime, then people may conclude what’s the point…what’s in it for me? As with procedural justice, this could be separate from the role that legitimacy may play.

The final potential predictor of cooperation is fear of crime. As with the link between fear of crime and legitimacy, our position is speculative. On the one hand, people who worry about the risk of being burgled and physically attacked in the street may be more likely to call and assist the police and the courts. They may be motivated to report crimes and give information to the police because they are concerned about their own personal risk. Correspondingly, people who are not worried about criminal victimization may be less likely to cooperate with legal authorities because they do not feel that they need legal authorities to reduce their material risk. On the other hand, people who worry about crime may be less willing to help the police, possibly because they think that there are potentially negative personal consequences from identifying the culprit to the police and aiding investigation and prosecution.

Cross-national variation in strengths of the relationships predicting legitimacy and cooperation

In addition to testing the model in Figure 1 in each country separately, we also examine whether the theory appears to work better in some countries than it does in others. As Tyler (1997: 337) emphasizes: “It is important to recognize that the relational model may not provide a good description of authority relations in all settings. In particular, people should care more about relational issues when they identify with the authority and the group that authority represents.” While the norms dictating appropriate use of police power may vary according to varying social, political and institutional practices, arrangements and histories (Jackson & Bradford, 2019; Trinkner, 2019), we test whether procedural justice is an especially important individual-level predictor of legitimacy in countries with a high level of procedural justice—and conversely, whether procedural is less important as an individual-level predictor of legitimacy in countries with low average levels of procedural justice. Following Trinkner (2019), we assume that the routine signalling of inclusion, worth and value by police to citizens through widespread enactment of procedurally just principles means that people in those countries are more likely to (a) view the police as a positive group authority and (b) identify as a law-abiding citizen of the nation state (Murphy et al., 2022).13 This would also imply that, in such contexts, normative (relational) motivations to cooperate will be more important than instrumental motivations.

We test whether procedural justice and legitimacy (both relational factors) are more important individual-level predictors of cooperation than instrumental factors, in countries in which the police represent an identity-relevant legal authority. Using a country-level proxy for identity-relevance and superordinate group inclusion, we assess whether police-citizen relations are more normative than instrumental in contexts with higher national levels of perceived procedural justice, while in countries with lower national-levels of perceived procedural justice, they are more instrumental than normative. The aggregate level of procedural justice in a given country is an adequate proxy for identity-relevance and superordinate group inclusion, not least because individual level studies have consistently identified a strong association between perceived procedural justice and salient social categories, such as nation, state and the notion of the law-abiding citizen (for a recent review see Chan et al. 2023).

We predict that in countries with high aggregate levels of procedural justice, procedural justice will be a stronger individual level predictor of legitimacy, and effectiveness and fear of crime will be weaker predictors of legitimacy. We also predict that in countries with high levels of procedural justice, procedural justice and legitimacy will be stronger individual level predictors of cooperation, with effectiveness being a weaker predictor. In other words, in countries where we assume (through our proxy) that the police are a prototypical authority figure of a powerful social group that people are more likely to identify with, individuals draw from their perceptions of procedural justice more identity-relevant positive or negative information that then informs legitimacy and cooperation, compared to countries where the police are not a prototypical authority figure of a powerful social group. By contrast, in those countries where people at the aggregate level do not tend to believe that police are procedurally just, other factors (such as effectiveness and fears of crime) will be more important predictors of legitimacy and cooperation.14

THE STUDY

Overall, we make three related contributions. First, we test PJT in a comparative cross-national context using nationally representative data from 30 countries. We use structural equation modelling (SEM) to fit the model represented in Figure 1 for each of these countries. We examine (a) the predictors of legitimacy in different countries and (b) the factors that explain variation in behavioral intentions to cooperate with the police and criminal courts. Second, because we use data that were collected with an unusual amount of attention to methodological equivalence, and because we fit the same statistical models in each country, we can directly compare the strengths of these associations across the countries. This also provides future single-country studies with a methodological benchmark to facilitate the careful accumulation of comparative evidence on both fronts, i.e. the locally specific preconditions of legitimacy and the role that legitimacy may or may not play in motivating cooperation. Third, we estimate mean country levels of perceived procedural justice, and we examine how the associations from the model vary by these national-level measures of procedural justice. The goal is to kick-start further work on the context-level factors that may help explain why PJT works better in some countries than in other countries.

Data and methodological equivalence

Established in 2001, the European Social Survey (ESS) is an academically driven, face-to-face interview survey that measures a wide range of attitudes and beliefs (Jowell et al., 2007).15 The ESS runs every two years, charting interactions between Europe's changing institutions and the attitudes, beliefs and behavior patterns of its diverse populations. The survey is directed by a core scientific team and is funded by the European Commission as a European Research Infrastructure Consortium. Each participating country contributes to the central funding and covers the costs of its own fieldwork and national coordination.

The questionnaire comprises an invariant core of questions asked of all respondents in each round, and a series of rotating modules which change from round to round. Rotating modules focus in considerable detail on a particular issue. They are selected through a competitive process, where teams of researchers are invited to submit proposals for the modules in each round. The core data for our analysis come from a 45-question rotating module on ‘trust in justice’ which was included in Round 5 of the ESS (see European Social Survey, 2011a; Jackson et al., 2011; Hough et al., 2013a, 2013b). The survey was conducted in 2010/11, with 28 countries taking part: Austria, Belgium, Bulgaria, Croatia, Cyprus, Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Ireland, Israel, Lithuania, Netherlands, Norway, Poland, Portugal, Russian Federation, Slovakia, Slovenia, Spain, Sweden, Switzerland, UK, and Ukraine.

All the countries used some form of random probability sampling, in most cases multistage stratified sampling. The mode of data collection was computer-assisted personal interviewing in around half of the countries, and paper-assisted personal interviewing in half. While the target response rate was 70%, realised response rates for Round 5 ranged from 30.5% (Germany) to 81.4% (Bulgaria). The average response rate was 60.2% (standard deviation 11%). The total realised sample size was n=58,838, and the sample sizes for individual countries ranged from 1,083 for Cyprus to 3,031 for Germany.

We also use data from two further countries, South Africa and the US. Both of them fielded some version of the ESS module. The South African data come from the 2012 round of South African Social Attitudes Survey (SASAS), an annual national survey conducted by the Human Sciences Research Council (HSRC), South Africa’s statutory research agency. This survey round consisted of a nationally representative sample of South African adults aged 16 years and over living in private households. SASAS employs a multistage probability sampling design, where the first-stage sampling units are 500 Population Census enumeration areas (EAs), stratified by province, geographical sub-type and majority population group. Interviews were conducted face-to-face. The realised sample size is n=2,518.

The US data come from a random sample of individuals drawn from a GFK Knowledge Networks research panel of U.S. adults (see Tyler & Jackson, 2014). Knowledge Networks uses random digit dialing and address-based sampling methods to construct and maintain the panel. Some 2,561 respondents were initially selected from the larger panel and invited to take part in this survey. A total of 1,603 individuals completed the survey online. The survey was fielded in August and September of 2012.

Variables and measures

The analysis involves variables that represent the eight concepts that are shown in Figure 1. These are perceived procedural justice, distributive justice, lawfulness and effectiveness of the police, respondents’ fear of crime (worry about criminal victimization), normative alignment with the police, duty to obey the police, and willingness to cooperate with the police and courts. The supplementary materials (section S1) describe the survey questions (“items”) and response options that were used as measures of these concepts. The same items were mostly used in the ESS and the surveys in South Africa and the US, apart from small differences in some items as shown in supplementary section S1. The questions are listed using their wordings in the English-language questionnaires. For the ESS they were translated to the other languages in the survey by the ESS team, using established procedures that aim to achieve stimulus equivalence of the translated questions. The items were also fielded in multiple languages in South Africa, and in English and Spanish in the US.,

In addition to these measures of the main concepts of interest, three background characteristics of respondents are also used in the analysis as controls: age (in years), gender (as male or female), and highest level of educational qualification (in three categories: less than upper secondary, upper secondary, or more than upper secondary education).

Statistical modelling

The results of the analysis are estimated regression coefficients from linear structural equation models (SEMs) corresponding to the model that is represented in Figure 1. They were obtained in two steps. First, factor analysis measurement models for each of the five latent variables in the model were estimated using pooled data for all the countries together. Second, the model in Figure 1 (the “structural model”) was then estimated separately for each country. Mean values of all the variables in each country were also estimated at the same time. The motivation and details of these steps are described in Section S2 of the supplementary materials.

By fitting the measurement models on the pooled data, we make the assumption of formal measurement equivalence across the countries, meaning that the parameters of the measurement models do not depend on the country. This means that we impose on each latent variable an operational definition which is the same in each country and which is determined by the measurement models that are estimated for all the countries together. The estimated coefficients of all the variables then refer to the variables thus defined and are comparable across the countries in this sense.

The estimation was carried out in the R language (R Core Team, 2022), using the lavaan package (Rosseel, 2012) and some additional packages as described in supplementary section S2. The R code that was used for the analysis is included in supplementary materials (section S4).

RESULTS

Measurement modelling

The measurement models were fitted first, separately for each of the five latent variables (procedural justice, effectiveness, normative alignment, obligation to obey, and willingness to cooperate). This used the sample of respondents from all the countries together, including all respondents who answered at least one measurement item for a given latent variable. The resulting sample sizes were between 56,615 and 58,060 for the five models.

Estimated parameters for these models are shown in the supplementary materials (Table S3a for the loading parameters, supplement S4 for the full results). The estimated loading parameters (λkl\lambda_{kl} in the notation of supplementary section S2) are all positive. This implies that higher values of each latent variable indicate higher levels of it in the direction of the naming of the variable (e.g. higher willingness to cooperate). The estimated measurement parameters (τkl\tau_{kl}, λkl\lambda_{kl} and θkl2\theta_{kl}^{2}) from this step were then taken forward as fixed values to the next step of estimation.

Fitting the structural model in each country

The structural equation models for the latent variables were then estimated in each country. In the rest of this section, we examine and compare the estimated regression coefficients for these models. These coefficients are presented in a graphical form in figures below, and in table form in supplementary Tables S3c-d (full output for the models is also included in supplementary section S4) The sample sizes for each country in this step are shown in Table S3b of the supplementary materials, together with R2 statistics for each country. Briefly, R2 was typically higher for normative alignment (ranging from 0.28 to 0.70) than for obligation to obey (ranging from 0.05 to 0.32) and cooperation (ranging from 0.03 to 0.14).

The estimated coefficients of the regression models are displayed in Figures 2-4. In each figure the response variable is the same: normative alignment with the police in Figure 2, duty to obey the police in Figure 3, and cooperation with the police in Figure 4. Each of the plots within a figure shows the estimated coefficients (and their 95% confidence intervals) of one predictor of that response variable. For example, the plot at the top left of Figure 2 shows the coefficients of procedural justice as a predictor of normative alignment, for each of the 30 countries. The order of the countries in Figure 2 is obtained by calculating for each country the averages of the absolute values of the coefficients of the normative predictors of normative alignment (procedural justice, distributive justice and lawfulness) and of the coefficients of its instrumental predictors (effectiveness and worry about crime) and taking the ratio of these two averages. The countries are shown in ascending order (from the bottom) of this ratio, so that the country at the top of the plot (Austria) is the one for which the normative predictors are the strongest, and the country at the bottom (South Africa) the weakest, relative to the instrumental predictors. A similar rule of ordering is used in Figures 3 and 4, using values of the predictors of obligation to obey and cooperation respectively (and including obligation to obey and moral alignment among the normative predictors in Figure 4). All of the variables were defined in such a way that their standard deviations across all the countries are 1. This means that every regression coefficient β in every figure is expressed in a “fully standardised” form where a difference of 1 unit of this cross-national standard deviation in the corresponding explanatory variable is associated with an expected difference of β standard deviation units in the response variable.

Starting with normative alignment in Figure 2, we find that the strongest predictor in nearly all of the countries is procedural justice (the exceptions are South Africa and Slovakia, where effectiveness has larger estimated coefficients than procedural justice). The standardized regression coefficients ranged from 0.24 (in South Africa) to 0.76 (in Spain) and all were statistically significant (p<.001). In all of countries, apart from Austria, effectiveness is a positive and significant predictor. The standardized regression coefficients are generally smaller than procedural justice, ranging from 0.05 (in Austria) to 0.40 (in France). Of note is that the ratio between procedural justice and effectiveness is larger than one for all countries but South Africa and Slovakia, ranging from 1.21 (Russian Federation) to 10.00 (Austria), with a fair number of countries hovering somewhere between 1.5 and 4. The other three predictors are smaller in size and much more variable in their statistical significance. For distributive justice (statistically significant at the 5% level in 13 countries), standardized regression coefficients range from -0.03 (in Sweden) to 0.19 (in Austria). For lawfulness (statistically significant in 15 countries), standardized regression coefficients range from -0.16 (in Ukraine) to 0.21 (in Israel). For fear of crime (statistically significant in 9 countries), standardized regression coefficients range from -0.07 (in Bulgaria) to 0.13 (in France).

Figure 3 shows the coefficients of the predictors of duty to obey. The conclusions are pretty much the same as for normative alignment. The coefficients for procedural justice are again positive and significant in all countries, and largest in all but South Africa and Slovakia, where effectiveness is largest. For procedural justice, standardized regression coefficients range from 0.05 (in South Africa) to 0.44 (in Slovenia). For effectiveness, standardized regression coefficients range from -0.02 (in Israel) to 0.30 (in South Africa). The absolute value for the ratio between procedural justice and effectiveness is larger than one for all countries but South Africa and Slovakia, ranging from 1.09 (Russian Federation) to 22.79 (USA), with a fair number of countries hovering somewhere between 1.25 and 2.75. As with normative alignment, the other three predictors are generally smaller in size and much more variable in their statistical significance. For distributive justice (statistically significant in 8 countries), standardized regression coefficients range from -0.11 (in Cyprus) to 0.14 (in Ukraine). For lawfulness (statistically significant in 9 countries), standardized regression coefficients range from -0.08 (in Ukraine) to 0.19 (in Croatia). For fear of crime (statistically significant in 10 countries), standardized regression coefficients range from -0.11 (in Slovenia) to 0.12 (in Poland).

Figure 4 presents the coefficients for the final part of the model in Figure 1, the model for willingness to cooperate with the police and courts. Overall, normative motivations to cooperate seem to be stronger than instrumental motivations to cooperate, with procedural justice, normative alignment and obligation to obey being stronger positive predictors of cooperation than effectiveness and fear of crime—indeed, the latter two tend not to be statistically significant and/or tend to be negative predictors. However, compared to the predictors of normative alignment and obligation to obey, the size of the regression coefficients is generally smaller. For procedural justice, standardized regression coefficients range from -0.09 (in Lithuania) to 0.30 (in Austria) and are statistically significant in 19 countries. For normative alignment, standardized regression coefficients range from -0.09 (in Cyprus) to 0.21 (in USA) and are statistically significant in 13 countries. For obligation to obey, standardized regression coefficients range from -0.07 (in Slovenia) to 0.16 (in Israel) and are statistically significant in 9 countries. Notably, with the exceptions of Finland and Hungary, if legitimacy is a significant predictor of cooperation, it is either normative alignment or obligation to obey. For effectiveness, standardized regression coefficients range from -0.22 (in Austria) to 0.16 (in Lithuania) and are statistically significant in 10 countries. For fear of crime, standardized regression coefficients range from -0.14 (in Sweden) to 0.10 (in Israel) and are statistically significant in 11 countries.

FIGURE 2 Regression coefficients predicting normative alignment with the police

FIGURE 3 Regression coefficients predicting duty to obey the police

FIGURE 4 Regression coefficients predicting cooperation with the police

Thus far we have found that procedural justice is the strongest positive predictor of both legitimacy and cooperation in nearly all of the countries. Moreover, in 19 countries at least one aspect of legitimacy is a significant positive predictor of cooperation. This supports the portability of PJT across many of these national borders. But does PJT work ‘better’ in some countries than in others? This would be an instance of a `cross-level interaction’ where the strength of an association at an individual level (between an individual’s perception of the procedural justice of the police and their willingness to cooperate with the police, say) may vary according to a country-level contextual variable (here the country-level average of perceived procedural justice). The final part of our analysis examines this question. To set the scene, Figure 5 shows the estimated averages of the main variables in each country. In these plots, the countries are shown in ascending order of average levels of procedural justice, so that the country at the bottom of the plot (Ukraine) is the one for which national estimated levels of procedural justice are the lowest, and the country at the top (Denmark) the one where national estimated levels of procedural justice are the highest. The countries with higher levels of procedural justice are generally in Northern and Western Europe while the countries with lower levels of procedural justice are generally in Eastern Europe (with the addition of Greece, Israel and South Africa). Notably, the general pattern for procedural justice also plays out for normative alignment and effectiveness, and to some degree distributive justice, but not for fear of crime and obligation to obey.16

Figures 6 and 7 summarize the results from the cross-level interactions. Figure 6 shows the estimated country means of perceived procedural justice (on the X-axis of the plots) against regression coefficients in the countries (on the Y-axis) predicting normative alignment and obligation to obey, while Figure 7 shows the estimated country means of perceived procedural justice (on the X-axis of the plots) against regression coefficients in the countries (on the Y-axis) predicting cooperation. Generally speaking, normative predictors are stronger and instrumental predictors are weaker in countries with higher national levels of procedural justice.

FIGURE 5 Estimates of the population averages of the main variables considered in the analysis, separately for each country. The countries are in the same order in each plot, by ascending order of the average of perceived procedural justice of the police

FIGURE 6 Scatterplots of estimated country means of perceived procedural justice (on the X-axis of the plots) against regression coefficients in the countries (on the Y-axis). The top two plots show the coefficients of procedural justice predicting normative alignment with the police and obligation to obey the police, and the bottom two plots show the coefficients for perceived police effectiveness predicting normative alignment with the police and obligation to obey the police

FIGURE 7 Scatterplots of estimated country means of perceived procedural justice (on the X-axis of the plots) against regression coefficients in the countries (on the Y-axis). The plots show the coefficients of five explanatory variables (procedural justice, normative alignment with the police, obligation to obey the police, police effectiveness and worry about crime) predicting willingness to cooperate with the police

DISCUSSION

Procedural justice theory (PJT) originated in the US (Tyler and Huo, 2002; Sunshine & Tyler, 2003; Tyler, 2006a, 2011a; Papachristos et al., 2012) but has since attracted considerable international attention. Studies have so far been conducted the UK, Australia, Israel, South Africa, China, Taiwan, South Korea, Nigeria, Ghana, Sierra Leone, Pakistan, Trinidad & Tobago, Jamaica, Chile and Brazil. Yet until now, our ability to systematically accumulate comparable and extensive cross-national evidence has been rather limited.17 Most of the studies that have tested a PJT account of public cooperation with the police and criminal courts have been conducted in the US, UK, Australia and Israel. Moreover, scholars have tended to use different forms of sampling, measurement strategies, and approaches to modelling the predictors and potential outcomes of legitimacy (Jackson, 2018).

We have presented findings from the most comprehensive formal assessment of the portability of PJT across diverse social, legal and political contexts to date. Our 30 country dataset enjoys an unusually high level of methodological equivalence due to the European Social Survey’s (ESS) commitment to cross-country equivalence of sampling, mode of interviewing, concepts and measures, and the fielding of the same module in representative sample surveys in the US and South Africa. Our analysis has found that, in most of the countries under investigation, people were more likely to recognize the legitimacy of an institution when they believed that the police made fair decisions and interacted respectfully with citizens. Crucially, procedural justice tended to be a stronger predictor of police legitimacy than distributive justice, lawfulness, fear of crime and the effectiveness of the police in crime-control—although the relative magnitude of statistical effects of effectiveness on legitimacy was typically larger than found in much prior research. The only two countries where effectiveness was a more important predictor of legitimacy than procedural justice were South Africa and Slovakia. Moreover, when officers were seen to be legitimate and procedurally just holders of legitimate power, people were more likely to report being willing to cooperate with the police and criminal courts. In most countries—the exceptions were Belgium, Greece, Croatia, Lithuania, Poland and Portugal, where only one normative factor was a significant predictor—people’s willingness to cooperate was predicted by at least two out of the three normative factors under investigation, i.e. procedural justice, obligation to obey and/or normative alignment with the police. Only in Greece and Lithuania were instrumental factors (police effectiveness and fear of crime) more important predictors of cooperation than normative factors (procedural justice, normative alignment and duty to obey).

Overall, we found that PJT worked in most countries. But we also found that it tends to work a little bit better in countries with high national levels of fair interpersonal treatment and decision-making. When ranked according to country-level average levels of procedural justice, Eastern European countries (especially Russian Federation, Ukraine, Bulgaria and Lithuania), Israel and South Africa generally had lower scores, while Northern and Western European countries generally had higher scores. Looking at the correlations between (a) national-level estimates of levels of procedural justice and (b) coefficients from the models for legitimacy and cooperation, people living in countries with generally high levels of procedural justice (e.g. ones in Northern and Western Europe) tended to be more oriented to legal authority in normative rather than instrumental ways. In particular, procedural justice was a stronger positive predictor of legitimacy among countries with higher aggregate levels of procedural justice, and effectiveness was a weaker positive predictor of legitimacy in the very same countries.18 The best explanation for this is that, when people in a given country collectively believe that the police treat citizens with respect and dignity and make fair decisions, the police represent a positive authority of the superordinate group (typically conceptualized as the law-abiding member of the community). Heightened identity relevance of the police then strengthens the role that procedural justice and legitimacy play in pro-active public cooperation (Tyler, 1997; Trinkner, 2019; Bradford & Jackson, 2024).

Limitations

We should mention three key limitations. First, the data are from 2010 to 2012. The national level pictures we have presented in this paper may now look different—due for instance, to increasing political polarisation and popularism, decreasing levels of institutional trust, and the rise of the far right. Second, we have used observational (non-experimental) survey data. We have estimated national-level conditional correlations between public perceptions of police activity, perceptions of legitimacy and willingness to cooperate, in a way that allows for formal country level comparisons. It is important to do so, but we cannot draw causal conclusions from our findings. Third, we did not measure actual (behavioral) cooperation. Saying one will cooperate in a hypothetical scenario may be an easy and socially acceptable response in a survey, but actually doing so might be more costly and complicated. But it remains valuable to examine stated willingness to cooperate. To say that one would or would not cooperate is a more concrete type of stated attitude than, say, general confidence in the police. It is one that recognizes a particular role and function of the police in certain situations, as well as positive intentions and a willingness to act in supportive ways. In an ideal world, we would have captured both intentions to cooperate and past cooperative acts. But space restrictions in the survey precluded this, indeed doing so would have limited the number of people available for part of the analysis—only a proportion of a representative sample would have been in a recent situation in which they had made a choice to cooperate or not with legal authorities.

Where next?

Our findings raise a number of issues for future research. From a technical perspective, we need more methodologically equivalent cross-national studies if they are to properly compare what the police need to do if they are going to be recognised as legitimate holders of power by those they protect and serve. This means measuring and modelling legitimacy in comparable ways. It also means including a full range of police behavior—procedural justice, distributive justice, effectiveness, boundaries, and lawfulness. From a theoretical perspective, we need to understand why things differ from one national context to the next. We have examined one potential explanation for the varying importance of procedural justice to legitimacy and of procedural justice and legitimacy to cooperation. We have found that people relate to legal authorities in more normative than instrumental ways when they live in a country where the police represent an important superordinate group that people are more likely to identify with. We have used national levels of procedural justice perceptions as a proxy, which is defensible. But it is important that future cross-national studies directly measure the identity-relevance of the police.

Future research should focus on people’s direct and indirect encounters with officers, as well as ‘focusing events’ such as the police killing of George Floyd (see Fine et al., forthcoming). It should also address why people obey the law. Unfortunately, while Round 5 of the ESS contained measures of people’s compliance with the law, the items did not really work. Very low numbers admitted to having made an exaggerated or false insurance claim or buying something that might be stolen, 97% and 92% respectively, and the question about committing a traffic offence may have largely captured bicycle offences. We recommend future cross-national studies (a) use better measures of self-reported offending behavior and (b) model various normative and instrumental factors, including in their framework perceptual deterrence, legal legitimacy and personal morality, alongside perceptions of the police (Trinkner et al., 2018).

To close

Without public reporting, crimes go unnoticed. Without information given, crimes remain unsolved. Without courtroom testimony, trials fail. Drawing on nationally representative data from 30 countries across Europe and beyond, we have found that the most important thing that police can do to encourage public cooperation is to act in procedurally just ways. When people see police as legitimate and fair, they are more likely to help the police and criminal courts (Tyler, 2011a, 2011b). This is true not just in the US, UK and Australia, but also in countries like France, Greece, Spain, Norway, Hungary and Ukraine.

Research shows that effective policing is not just about enforcing the law in an uncompromising way—it is about being fair too. These goals go hand in hand (Weisburd et al., 2022). Police can confidently assert their power without misusing their authority, both in how they interact with people and how they make their decisions. While aggressive policing is sometimes needed, real success relies on more than just force and deterrence. It is also about upholding the moral and social foundations of police authority (Tyler, 2023).

FUNDING

This study was funded by UK ESRC grant reference: ES/L011611/1, EU FP7 Eurojustis grant reference 217311, EU FP7 Fiducia grant reference 290563. The US survey was supported by grants from Yale Law School and New York University. The SASAS survey was supported by a grant from the Open Society Foundation for South Africa (OSF-SA).

REFERENCES

Baćak, V., & Apel, R. (2020). The thin blue line of health: Police contact and wellbeing in Europe. Social Science & Medicine, 267, 112404. https://doi.org/10.1016/j.socscimed.2019.112404

Boateng, F. D. (2017). Police legitimacy in Africa: A multilevel multinational analysis. Policing and Society, 27(8), 917-933. https://doi.org/10.1080/10439463.2016.1251439

Boateng, F. D., & Buckner, Z. N. (2017). Police legitimacy in Asia: Findings from a multilevel hierarchical non-linear analysis. Policing and Society, 27(4), 341-358. https://doi.org/10.1080/10439463.2017.1365915

Boateng, F. D., Pryce, D. K., & Abess, G. (2022). Legitimacy and cooperation with the police: Examining empirical relationship using data from Africa. Policing and Society, 32(3), 411-433. https://doi.org/10.1080/10439463.2022.2037554

Bolger, P. C., & Walters, G. D. (2019). The relationship between police procedural justice, police legitimacy, and people's willingness to cooperate with law enforcement: A meta-analysis. Journal of Criminal Justice, 60, 93-99. https://doi.org/10.1016/j.jcrimjus.2019.01.001

Bottoms, A., & Tankebe, J. (2012). Beyond procedural justice: A dialogic approach to legitimacy in criminal justice. Journal of Criminal Law and Criminology, 102(1), 119-170.

Bottoms, Anthony E., and Tankebe, Justice. 2021. Procedural justice, legitimacy and social contexts. In Procedural Justice and Relational Theory, edited by Meyerson, D., Mackenzie, C., and MacDermott, T., 85–110. London: Routledge.

Bradford, B. (2014). Policing and social identity: Procedural justice, inclusion and cooperation between police and public. Policing & Society, 24(1), 22-43. https://doi.org/10.1080/10439463.2012.724068

Bradford, B., & Jackson, J. (2018). Police legitimacy among immigrants in Europe: Institutional frames and group position. European Journal of Criminology, 15(5), 567-588. https://doi.org/10.1177/1477370817749496

Bradford, B., Huq, A., Jackson, J., & Roberts, B. (2014). What price fairness when security is at stake? Police legitimacy in South Africa. Regulation and Governance, 8(2), 246–268. https://doi.org/10.1111/rego.12012

Bradford, B. & Jackson, J. (2024). Legitimacy, relational Norms and reciprocity. In Hollander-Blumoff, R. (ed.) Research Handbook in Law and Psychology. Elgar Press.

Bradford, B., Jackson, J., & Stanko, E. (2009). Contact and confidence: Revisiting the impact of public encounters with the police. Policing & Society, 19(1), 20-46. https://doi.org/10.1080/10439460802457594

Brouwer, J., Van Der Woude, M., & Van Der Leun, J. (2017). Border policing, procedural justice and belonging: The legitimacy of (cr)immigration controls in border areas. British Journal of Criminology. https://doi.org/10.1093/bjc/azx050

Chan, A., Bradford, B., & Stott, C. (2023). A systematic review and meta-analysis of procedural justice and legitimacy in policing: the effect of social identity and social contexts. Journal of Experimental Criminology, 1-58. https://doi.org/10.1007/s11292-023-09595-5

Elliott, I., Thomas, S., & Ogloff, J. (2011). Procedural justice in contacts with the police: Testing a relational model of authority in a mixed methods study. Psychology, Public Policy, and Law, 17, 592-610. https://doi.org/10.1037/a0024212

European Social Survey. (2011a). Round 5 Module on Trust in the Police & Courts – Final Question Design Template. London: Centre for Comparative Social Surveys: City University London. Retrieved from http://www.academia.edu/4160691/ESS_Round_5_Design_Template_for_Trust_in_Justice_Module

European Social Survey. (2011b). Trust in justice: Topline findings from the European Social Survey. ESS Topline Results Series Issue 1. Jackson, J., Hough, M., Bradford, B., Pooler, T. M., Hohl, K., & Kuha, J. Retrieved from https://www.europeansocialsurvey.org/sites/default/files/2023-06/ESS5_toplines_issue_1_trust_in_justice.pdf

European Social Survey. (2012). Policing by consent: Understanding the dynamics of police power and legitimacy. ESS Country Specific Topline Results Series Issue 1 (UK). Jackson, J., Hough, M., Bradford, B., Hohl, K., & Kuha, J. Retrieved from https://www.europeansocialsurvey.org/sites/default/files/2023-06/ESS5_gb_toplines_policing_by_consent.pdf

Factor, R., Mahalel, D., Rafaeli, A., & Williams, D. R. (2013). A social resistance perspective for delinquent behavior among non-dominant minority groups. British Journal of Criminology, 53(5), 784-804. https://doi.org/10.1093/bjc/azt035

Fine, A., Oliveira, T., Jackson, J., Bradford, B., Pósch, K. & Trinkner, R. (forthcoming). Did the murder of George Floyd damage public perceptions of police and law in the United States? Journal of Research in Crime & Delinquency.

Frydl, K., & Skogan, W. (Eds.). (2004). Fairness and Effectiveness in Policing: The Evidence. National Academies Press.

Hamm, J. A., Wolfe, S. E., Cavanagh, C., & Lee, S. (2022). (re)organizing legitimacy theory. Legal and Criminological Psychology. 27(2), 129-146. https://doi.org/10.1111/lcrp.12199

Hasisi, B., & Weisburd, D. (2011). Going beyond ascribed identities: The importance of procedural justice in airport security screening in Israel. Law & Society Review, 45(4), 867-892. https://doi.org/10.1111/j.1540-5893.2011.00459.x

Hough, M., Jackson, J., & Bradford, B. (2013a). Legitimacy, trust and compliance: An empirical test of procedural justice theory using the European Social Survey. In J. Tankebe & A. Liebling (Eds.), Legitimacy and Criminal Justice: An International Exploration (pp. 326-352). Oxford: Oxford University Press.

Hough, M., Jackson, J., & Bradford, B. (2013b). The governance of criminal justice, legitimacy and trust. In S. Body-Gendrot, R. Lévy, M. Hough, S. Snacken, & K. Kerezsi (Eds.), The Routledge Handbook of European Criminology (pp. 243-265). Oxon: Routledge.

Huq, A., Jackson, J. & Trinkner, R. (2017). Legitimating practices: Revisiting the predicates of police legitimacy. British Journal of Criminology, 57, 5, 1101-1122, doi: 10.1093/bjc/azw037

Jackson, J. (2018). Norms, normativity and the legitimacy of legal authorities: International perspectives. Annual Review of Law and Social Science, 14, 145-165. https://doi.org/10.1146/annurev-lawsocsci-110316-113734

Jackson, J., & Bradford, B. (2019). Blurring the distinction between empirical and normative legitimacy? A methodological commentary on “Police legitimacy and citizen cooperation in China”. Asian Journal of Criminology, 14(4), 265-289. https://doi.org/10.1007/s11417-019-09289-w

Jackson, J., & Pósch, K. (2019). New directions for research into fairness and legal authority: A focus on causal mechanisms. In E. A. Lind (Ed.), Social Psychology and Justice (Frontiers of Social Psychology Series), pp. 181-212. New York: Routledge.

Jackson, J., Asif, M., Bradford, B., & Zakar, M. Z. (2014). Corruption and police legitimacy in Lahore, Pakistan. British Journal of Criminology, 54(6), 1067-1088. https://doi.org/10.1093/bjc/azu069

Jackson, J., Bradford, B., Hough, M., Kuha, J., Stares, S. R., Widdop, S., Fitzgerald, R., Yordanova, M., & Galev, T. (2011). Developing European indicators of trust in justice. European Journal of Criminology, 8(4), 267-285. https://doi.org/10.1177/1477370811411458

Jackson, J., Bradford, B., Hough, M., Myhill, A., Quinton, P., & Tyler, T. R. (2012). Why do people comply with the law? Legitimacy and the influence of legal institutions. British Journal of Criminology, 52(6), 1051-1071. https://doi.org/10.1093/bjc/azs032

Jackson, J., Bradford, B., Stanko, E. A., & Hohl, K. (2013). Just Authority? Trust in the Police in England & Wales. Oxon: Routledge.

Jackson, J., Brunton-Smith, I., Bradford, B., Rodriguez-Oliveira, T., Pósch, K., & Sturgis, P. (2021). Police legitimacy and the norm to cooperate: Using a mixed effects location-scale model to estimate social norms at a small spatial scale. Journal of Quantitative Criminology, 37(2), 547-572. https://doi.org/10.1007/s10940-020-09467-5

Jackson, J., Bradford, B., Giacomantonio, C., & Mugford, R. (2023a). Developing core national indicators of public attitudes towards the police in Canada. Policing and Society, 33(3), 276-295. https://doi.org/10.1080/10439463.2022.2102757

Jackson, J., Fine, A., Bradford, B., & Trinkner, R. (2023b). Social identity and support for defunding the police in the aftermath of George Floyd. Group Processes & Intergroup Relations, 26(4), 833-858. https://doi.org/10.1177/13684302221128230

Jackson, J., McKay, T., Cheliotis, L., Bradford, B., Fine, A. & Trinkner, R. (2023c). Centering race in procedural justice theory: Systemic racism and the under-policing and over-policing of Black communities. Law & Human Behavior, 47, 1, 68–82, https://doi.org/10.1037/lhb0000524

Jonathan-Zamir, T., & Weisburd, D. (2013). The effects of security threats on antecedents of police legitimacy: Findings from a quasi-experiment in Israel. [insert] Journal of Research in Crime and Delinquency, 50(1), 3-32. https://doi.org/10.1177/002242781141800

Jowell, R., Roberts, C., Fitzgerald, R., & Eva, G. (2007). European Social Survey. Measuring Attitudes Cross-Nationally: Lessons from the European Social Survey. London: Sage.

Kochel, T. R. (2018). Police legitimacy and resident cooperation in crime hotspots: Effects of victimisation risk and collective efficacy. Policing and Society, 28(3), 251-270. https://doi.org/10.1080/10439463.2016.1174235

Kochel, T. R., Parks, R., & Mastrofski, S. (2013). Examining police effectiveness as a precursor to legitimacy and cooperation with police. Justice Quarterly. https://doi.org/10.1080/07418825.2011.633544

Kyprianides, A., Bradford, B., Jackson, J., Yesberg, J. A., Stott, C., & Radburn, M. (2021). Identity, legitimacy and cooperation with the police: Comparing general-population and street-population samples in London. Psychology, Public Policy, and Law, 27(4), 492-508. https://doi.org/10.1037/law0000312

Lind, E. A., & Tyler, T. R. (1988). The Social Psychology of Procedural Justice. New York: Plenum.

Mazerolle, L., Antrobus, E., Bennett, S., & Tyler, T. R. (2013). Shaping citizen perceptions of police legitimacy: A randomized field trial of procedural justice. Criminology, 51(1), 33-63. https://doi.org/10.1111/j.1745-9125.2012.00289.x

Meares, T. (2017). Policing and procedural justice: Shaping citizens' identities to increase democratic participation. Northwestern University Law Review, 111(6), 1525-1536. Retrieved from https://www.proquest.com/scholarly-journals/policing-procedural-justice-shaping-citizens/docview/1946276272/se-2

Metcalfe, C., Wolfe, S. E., Gertz, E., & Gertz, M. (2016). They protect our homeland but neglect our community: Homeland security overemphasis, legitimacy, and public cooperation in Israel. Journal of Research in Crime and Delinquency, 53(6), 814-839. https://doi.org/10.1177/002242781665738

Moravcová, E. (2016). Willingness to cooperate with the police in four central European countries. European Journal on Criminal Policy and Research, 22(1), 171-187. https://doi.org/10.1007/s10610-015-9271-0

Murphy, K., & Cherney, A. (2012). Understanding cooperation with police in a diverse society. British Journal of Criminology, 52(1), 181-201. https://doi.org/10.1093/bjc/azr065

Murphy, K., Bradford, B., & Jackson, J. (2016). Motivating compliance behavior among offenders: Procedural justice or deterrence? Criminal Justice and Behavior, 43(1), 102-118. https://doi.org/10.1177/0093854815611166

Murphy, K., Bradford, B., Sargeant, E., & Cherney, A. (2022). Building immigrants’ solidarity with police: Procedural justice, identity and immigrants’ willingness to cooperate with police. British Journal of Criminology, 62(2), 299-319. https://doi.org/10.1093/bjc/azab052

Murphy, K., Tyler, T. R., & Curtis, A. (2009). Nurturing regulatory compliance: Is procedural justice effective when people question the legitimacy of the law? Regulation & Governance, 3(1), 1-26. https://doi.org/10.1111/j.1748-5991.2009.01043.x

Nagin, D. S., & Telep, C. W. (2017a). Procedural justice and legal compliance. Annual Review of Law and Social Science, 13, 5-28. https://doi.org/10.1146/annurev-lawsocsci-110316-113310

Nagin, D. S., & Telep, C. W. (2017b). Response to “Procedural justice and policing: A rush to judgment?”. Annual Review of Law and Social Science, 13, 55-58. https://doi.org/10.1146/annurev-lawsocsci-120516-024409

Oliveira, T. R., Jackson, J., Murphy, K. & Bradford, B. (2021). Are trustworthiness and legitimacy “hard to win, easy to lose”? A longitudinal test of the asymmetry thesis of police-citizen contact. Journal of Quantitative Criminology, 37(4), 1003-1045, https://doi.org/10.1007/s10940-020-09478-2

Papachristos, A. V., Meares, T. L., & Fagan, J. (2012). Why do criminals obey the law? The influence of legitimacy and social networks on active gun offenders. Journal of Criminal Law and Criminology, 102(2), 397-440.

Pósch, K., Jackson, J., Bradford, B., & MacQueen, S. (2021). “Truly free consent”? Clarifying the nature of police legitimacy using causal mediation analysis. Journal of Experimental Criminology, 17(4), 563-595. https://doi.org/10.1007/s11292-020-09426-x

R Core Team. (2022). R: A Language and Environment for Statistical Computing. Vienna: R Foundation for Statistical Computing.

Radburn, M., Stott, C., Bradford, B., & Robinson, M. (2018). When is policing fair? Groups, identity and judgements of the procedural justice of coercive crowd policing. Policing and Society, 28(6), 647-664. https://doi.org/10.1080/10439463.2016.1234470

Reisig, M. D., & Lloyd, C. (2009). Procedural justice, police legitimacy, and helping the police fight crime: Results from a survey of Jamaican adolescents. Police Quarterly, 12(1), 42-62. https://doi.org/10.1177/1098611108327311

Reisig, M. D., Bratton, J., & Gertz, M. G. (2007). The construct validity and refinement of process-based policing measures. Criminal Justice and Behavior, 34(8), 1005-1028. https://doi.org/10.1177/0093854807301275

Rosseel, Y. (2012). lavaan: An R package for structural equation modeling. Journal of Statistical Software, 48(1), 1-36. 10.18637/jss.v048.i02

Sargeant, E., Murphy, K., & Cherney, A. (2014). Ethnicity, trust and cooperation with police: Testing the dominance of the process-based model. European Journal of Criminology, 11(4), 500-524. https://doi.org/10.1177/1477370813511386

Skogan, W. G. (2006). Asymmetry in the impact of encounters with the police. Policing and Society, 16(2), 99-126. https://doi.org/10.1080/10439460600662098

Skogan, W. G. (2012). Assessing asymmetry: The life course of a research project. Policing and Society, 22(3), 270-279. https://doi.org/10.1080/10439463.2012.704035

Sun, I. Y., Wu, Y., Hu, R., & Farmer, A. K. (2017). Procedural justice, legitimacy, and public cooperation with police: Does Western wisdom hold in China? Journal of Research in Crime and Delinquency, 54(4), 454-478. https://doi.org/10.1177/0022427816638705

Sunshine, J., & Tyler, T. R. (2003). The role of procedural justice and legitimacy in shaping public support for policing. Law and Society Review, 37(3), 555-589. https://doi.org/10.1111/1540-5893.3703002

Tankebe, J. (2009). Public cooperation with the police in Ghana: Does procedural fairness matter? Criminology, 47(4), 1265-1293. https://doi.org/10.1111/j.1745-9125.2009.00175.x

Tankebe, J. (2013). Viewing things differently: The dimensions of public perceptions of police legitimacy. Criminology. [insert] https://doi.org/10.1111/j.1745-9125.2012.00291.x

Trinkner, R. (2019). Clarifying the contours of the police legitimacy measurement debate: A response to Cao and Graham. Asian Journal of Criminology, 14(4), 309-335. https://doi.org/10.1007/s11417-019-09300-4

Trinkner, R., & Tyler, T. R. (2016). Legal socialization: Coercion versus consent in an era of mistrust. Annual Review of Law and Social Science, 12, 417-439. https://doi.org/10.1146/annurev-lawsocsci-110615-085141

Trinkner, R., Jackson, J., & Tyler, T. R. (2018). Bounded authority: Expanding “appropriate” police behavior beyond procedural justice. Law & Human Behavior, 42(3), 280-293. https://doi.org/10.1037/lhb0000285

Tsushima, M., & Hamai, K. (2015). Public cooperation with the police in Japan: Testing the legitimacy model. Journal of Contemporary Criminal Justice, 31(2), 212-228. https://doi.org/10.1177/1043986214568836

Tyler, T. R. (1988). What is procedural justice? Criteria used by citizens to assess the fairness of legal procedures. Law & Society Review, 22(1), 103-135. Retrieved from https://www.proquest.com/scholarly-journals/what-is-procedural-justice-criteria-used-citizens/docview/1297907919/se-2

Tyler, T. R. (1989). The psychology of procedural justice: A test of the group-value model. Journal of Personality and Social Psychology, 57(5), 830-838. https://doi.org/10.1037/0022-3514.57.5.830

Tyler, T. R. (1994). Psychological models of the justice motive. Journal of Personality and Social Psychology, 67(5), 850-863. https://doi.org/10.1037/0022-3514.67.5.850

Tyler, T. R. (1997). The psychology of legitimacy: A relational perspective on voluntary deference to authorities. Personality and Social Psychology Review, 1(4), 323-344. https://doi.org/10.1207/s15327957pspr0104_4

Tyler, T. R. (2003). Procedural justice, legitimacy, and the effective rule of law. Crime and Justice, 30, 283-357.

Tyler, T. R. (2006a). Why People Obey the Law. Princeton University Press.

Tyler, T. R. (2006b). Psychological perspectives on legitimacy and legitimation. Annual Review of Psychology, 57, 375-400. https://doi.org/10.1146/annurev.psych.57.102904.190038

Tyler, T. R. (2011a). Trust and legitimacy: Policing in the USA and Europe. European Journal of Criminology, 8(4), 254-266. https://doi.org/10.1177/147737081141146

Tyler, T. R. (2011b). Why People Cooperate: The Role of Social Motivations. Princeton University Press.

Tyler, T. R. (2017). Procedural justice and policing: A rush to judgment? Annual Review of Law and Social Science, 13, 29-53. https://doi.org/10.1146/annurev-lawsocsci-110316-113318

Tyler, T. R. (2023). Whither legitimacy? Legal authority in the twenty-first century. Annual Review of Law and Social Science, 19(1), 1-17. https://doi.org/10.1146/annurev-lawsocsci-110722-074236

Tyler, T. R., & Blader, S. L. (2003). The group engagement model: Procedural justice, social identity, and cooperative behavior. Personality and Social Psychology Review, 7(4), 349-361. https://doi.org/10.1207/S15327957PSPR0704_07

Tyler, T. R., & Fagan, J. (2008). Legitimacy and cooperation: Why do people help the police fight crime in their communities? Ohio State Journal of Criminal Law, 6, 231-275.

Tyler, T. R., & Huo, Y. J. (2002). Trust in the Law. Russell-Sage.

Tyler, T. R., & Jackson, J. (2014). Popular legitimacy and the exercise of legal authority: Motivating compliance, cooperation and engagement. Psychology, Public Policy, and Law, 20(1), 78-95. https://doi.org/10.1037/a0034514

Tyler, T. R., & Lind, E. A. (1992). A relational model of authority in groups. In M. Zanna (Ed.), Advances in Experimental Social Psychology (Vol. 25, pp. 115-191). Academic Press.

Tyler, T. R., & Trinkner, R. (2018). Why Children Follow Rules: Legal Socialization and the Development of Legitimacy. New York: Oxford University Press.

Van Damme, A., Pauwels, L., & Svensson, R. (2015). Why do Swedes cooperate with the police? A SEM analysis of Tyler’s procedural justice model. European Journal on Criminal Policy and Research, 21(1), 15-33. https://doi.org/10.1007/s10610-013-9224-4

Van der Toorn, J., Tyler, T. R., & Jost, J. T. (2011). More than fair: Outcome dependence, system justification, and the perceived legitimacy of authority figures. Journal of Experimental Social Psychology, 47(1), 127-138. https://doi.org/10.1016/j.jesp.2010.09.003

Weisburd, D., Telep, C. W., Vovak, H., Zastrow, T., Braga, A. A., & Turchan, B. (2022). Reforming the police through procedural justice training: A multicity randomized trial at crime hot spots. Proceedings of the National Academy of Sciences, 119(14), e2118780119. https://doi.org/10.1073/pnas.2118780119

White, M. D., Mulvey, P., & Dario, L. M. (2016). Arrestees’ perceptions of the police: Exploring procedural justice, legitimacy, and willingness to cooperate with police across offender types. Criminal Justice and Behavior, 43(3), 343-364. https://doi.org/10.1177/0093854815602501

Williams, B. (2005). In the Beginning was the Deed: Realism and Moralism in Political Argument. Princeton, NJ: Princeton University Press.

Wu, G., & Liu, Y. (2023). Extending procedural justice theory to the Chinese context: The role of collective efficacy. British Journal of Criminology, 63(1), 40-58. https://doi.org/10.1093/bjc/azab115

SUPPORTING INFORMATION

Supplement S1. Survey measures

Table S1. Full wording

ESS

South Africa

US

Perceived police procedural justice:

Now some questions about when the police deal with crimes like house burglary and physical assault.

Now some questions about when the police deal with crimes like house burglary and physical assault.

Thinking about the police in your community

[1] Based on what you have heard or your own experience how often would you say the police generally treat people in [country] with respect?

[1] Based on what you have heard or your own experience how often would you say the police generally treat people in South Africa with respect?

[1] How often do the police treat the people with dignity and respect?

[2] About how often would you say that the police make fair, impartial decisions in the cases they deal with?

[2] About how often would you say that the police make fair, impartial decisions in the cases they deal with?

[2] How often do the police make fair and impartial decisions in the cases they deal with?

[3] And when dealing with people in [country], how often would you say the police generally explain their decisions and actions when asked to do so?

[3] And when dealing with people in South Africa, how often would you say the police generally explain their decisions and actions when asked to do so?

[3] How often do the police explain their decisions and actions in ways that people can understand?

Options: Not at all often (1), Not very often (2), Often (3), Very often (4)

Options: Not at all often (1), Not very often (2), Often (3), Very often (4)

Options: Almost never (1), Sometimes (2), Often (3), Almost always (4)

[1], [2], [3] used as items in factor analysis measurement model of perceived procedural justice.

Perceived police distributive justice:

Now some questions about whether or not the police in [country] treat victims of crime equally. Please answer based on what you have heard or your own experience.

Now some questions about whether or not the police in South Africa treat victims of crime equally. Please answer based on what you have heard or your own experience.

…“your community” instead of name of the respondent’s country

[1] When victims report crimes, do you think the police treat rich people worse, poor people worse, or are rich and poor treated equally?

[1] When victims report crimes, do you think the police treat rich people worse, poor people worse, or are rich and poor treated equally?

[1] When victims report crimes do you think that the police…

Treat rich people worse than others (0), Treat rich and poor people equally (1), Treat rich people better than others (0)

Options for [1]: rich people treated worse (0), poor people treated worse (0), rich and poor treated equally (1)

Options for [1]: rich people treated worse (0), poor people treated worse (0), rich and poor treated equally (1)

[2] And when victims report crimes, do you think the police treat some people worse because of their race or ethnic group or is everyone treated equally?

[2] And when victims report crimes, do you think the police treat some people worse because of their race or ethnic group or is everyone treated equally?

[2] When victims report crimes do you think that the police…

Treat white people worse than minorities (0), Treat people of different ethnicities equally (1), Treat white people worse than minorities (0)

Options for [2]: People from a different race or ethnic group than most [country] people treated worse (0), People from the same race or ethnic group as most [country] people treated worse (0), Everyone treated equally regardless of their race or ethnic group (1)

Options for [2]: White, Indian and Coloured South Africans are treated worse than black South Africans (0), Black South Africans are treated worse than other race groups (0), [equally, as in ESS] (1)

Combined measure of distributive justice: [1]+[2]

Perceived police lawfulness:

Now one last question about the police and things they may or may not do.

[No introduction]

Do you disagree or agree that:

[1] How often would you say that the police in [country] take bribes?

[1] How often would you say that the police in South Africa take bribes?

[1] The police take bribes

Options for [1]: 11-point scale, from Never (5) to Always (1)

Options for [1]: 11-point scale, from Never (5) to Always (1)

Options for [1]: Disagree strongly (5), Disagree (4), Neither (3), Agree (2), Agree strongly (1)

[2] … please say to what extent you agree or disagree with each of the following statements about the police in [country]… The decisions and actions of the police are unduly influenced by pressure from political parties and politicians.

[2] … please say to what extent you agree or disagree with each of the following statements about the police in South Africa… The decisions and actions of the police are unduly influenced by pressure from political parties and politicians.

[2] The decisions and actions of the police are unduly influenced by pressure from political parties and politicians.

Options for [2]: Agree strongly (1), Agree (2), Neither agree nor disagree (3), Disagree (4), Disagree strongly (5)

Options for [2]: Agree strongly (1), Agree (2), Neither agree nor disagree (3), Disagree (4), Disagree strongly (5)

Options for [2]: Disagree strongly (5), Disagree (4), Neither (3), Agree (2), Agree strongly (1)

Combined measure of lawfulness: average of [1] and [2]

Perceived police effectiveness:

[1] Based on what you have heard or your own experience how successful do you think the police are at preventing crimes in [country] where violence is used or threatened?

[1] Based on what you have heard or your own experience how successful do you think the police are at preventing crimes in South Africa where violence is used or threatened?

[1] How successful do you think the police are at preventing crimes where violence is used or threatened in your community?

[2] And how successful do you think the police are at catching people who commit house burglaries in [country]?

[2] And how successful do you think the police are at catching people who commit house burglaries in South Africa?

[2] And, how successful do you think the police are at catching people who commit house burglaries?

Options for [1] and [2]: 11-point scale, from Extremely unsuccessful (0) to Extremely successful (10)

Options for [1] and [2]: 11-point scale, from Extremely unsuccessful (0) to Extremely successful (10)

Options for [1] and [2]: 11-point scale, from Extremely unsuccessful (0) to Extremely successful (10)

[3] If a violent crime were to occur near to where you live and the police were called, how slowly or quickly do you think they would arrive at the scene?

[3] If a violent crime or house burglary were to occur near to where you live and the police were called how slowly or quickly do you think they would arrive at the scene?

[3] If a violent crime were to occur near to where you live and the police were called, how soon do you think they would arrive at the scene?

Options for [3]: 11-point scale, from Extremely slowly (0) to Extremely quickly (10)

Options for [3]: 11-point scale, from Extremely slowly (0) to Extremely quickly (10)

Options for [3]: 11-point scale, from Extremely slowly (0) to Extremely quickly (10)

[1], [2], [3] used as items in factor analysis measurement model of perceived effectiveness.

Worry about criminal victimization:

Now I would like to ask you about how much (if at all) you worry about

specific crimes. How worried are you about …

[1] How often, if at all, do you worry about your home being burgled?

[1] How often do you worry about your home being burgled?

[1] Having your home broken into and something stolen

[2] How often, if at all, do you worry about becoming a victim of violent crime?

[2] How often do you worry about becoming a victim of violent crime?

Being mugged or robbed

Options: All or most of the time (4), Some of the time (3), Just occasionally (2), Never (1)

Options: All or most of the time (4), Some of the time (3), Just occasionally (2), Never (1)

Options: Very worried (4), Fairly worried (3), Not very worried (2), Not at all worried (1)

Combined measure of worry about victimization: average of [1] and [2]

Normative alignment with the police:

…please say to what extent you agree or disagree with each of the following statements about the police in [country].

…please say to what extent you agree or disagree with each of the following statements about the police in South Africa.

Do you disagree or agree that:

[1] The police generally have the same sense of right and wrong as I do.

[1] The police generally have the same sense of right and wrong as I do.

[1] The police generally have the same sense of right and wrong as I do.

[2] The police stand up for values that are important to people like me.

[2] The police stand up for values that are important to people like me.

[2] The police stand up for values that are important to people like me.

[3] I generally support how the police usually act.

[3] I generally support how the police usually act.

[3] You generally support how the police act in your community.

Options: Agree strongly (5), Agree (4), Neither agree nor disagree (3), Disagree (2), Disagree strongly (1)

Options: Agree strongly (5), Agree (4), Neither agree nor disagree (3), Disagree (2), Disagree strongly (1)

Options: Agree strongly (5), Agree (4), Neither agree nor disagree (3), Disagree (2), Disagree strongly (1)

[1], [2], [3] used as items in factor analysis measurement model of normative alignment.

Duty to obey the police:

Now some questions about your duty towards the police in [country]… To what extent is it your duty to…

Now some questions about your duty towards the police in South Africa… To what extent is it your duty to…

Now some questions about the police in your community. Do you disagree or agree that it is your responsibility that:

[1] back the decisions made by the police even when you disagree with them?

[1] was not included in the survey

[1] You should support the decisions made by police officers even when you disagree with them.

[2] do what the police tell you even if you don’t understand or agree with the reasons?

[2] do what the police tell you even if you don’t understand or agree with the reasons?

[2] do what the police tell you even if you don’t understand or agree with the reasons?

[3] do what the police tell you to do, even if you don’t like how they treat you?

[3] do what the police tell you to do, even if you don’t like how they treat you?

[3] do what the police tell you to do, even if you don’t like how they treat you?

Options: 11-point scale, from Not at all my duty (0) to Completely my duty (10)

Options: 11-point scale, from Not at all my duty (0) to Completely my duty (10)

Options: Disagree strongly (0), Disagree (2.5), Neither (5), Agree (7.5), Agree strongly (10)

[1], [2], [3] used as items in factor analysis measurement model of obligation to obey.

Willingness to cooperate with the police and courts:

Now some questions about what you would do if you were the only witness to a crime.

Now some questions about what you would do if you were the only witness to a crime.

Imagine that you were the only witness to a crime.

[1] Imagine that you were out and saw someone push a man to the ground and steal his wallet. How likely would you be to call the police? Would you be…

[1] Imagine that you were out and saw someone push a man to the ground and steal his wallet. How likely would you be to call the police? Would you be…

[1] If you saw someone push a person to the ground and steal their purse or wallet how likely would you be to call the police?

Options for [1]: Not at all likely (1), Not very likely (2), Likely (3), Very likely (4)

Options for [1]: Not at all likely (1), Not very likely (2), Likely (3), Very likely (4)

Options for [1]: Very unlikely (1), Unlikely (2), Likely (3), Very likely (4)

[2] How willing would you be to identify the person who had done it? Would you be…

[2] How willing would you be to identify the person who had done it? Would you be…

[2] How willing would you be to identify the person who had committed the crime?

[3] And how willing would you be to give evidence in court against the accused? Would you be…

[3] And how willing would you be to give evidence in court against the accused? Would you be…

[3] And how willing would you be to give evidence in court against the accused? Would you be…

Options for [2] and [3]: Not at all willing (1), Not very willing (2), Willing (3), Very willing (4)

Options for [2] and [3]: Not at all willing (1), Not very willing (2), Willing (3), Very willing (4)

Options for [2] and [3]: Same as for [1], even though question stem has “willing” rather than “likely”.

[1], [2], [3] used as items in factor analysis measurement model of willingness to cooperate.

The first procedural justice item captures treatment with respect and dignity. When people feel demeaned or subjected to negative stereotypes, they view themselves as diminished as people and disrespected beyond what is appropriate when dealing with the law. Conversely, acknowledging people’s rights and acting with courtesy leads them to feel fairly treated. The second item captures neutrality, which refers to making decisions based on the consistent application of rules based on proper procedure rather than on personal opinions or prejudices. The third item captures voice, which means providing opportunities for citizens to participate in decision making processes. Opportunities for voice need not involve a formal or elaborate mechanism; studies of police street stops, for example, indicate that when officers provide people an opportunity to tell their side of the story before they take action, people are more likely to feel fairly treated.

Distributive justice refers to the just distribution of finite resources—how the benefits and burdens of policing are distributed—across aggregate groups (e.g. race, gender and social class). In the development stage of the module, it was clear that respondents understood more clearly the notion of fair (and equal) treatment (rich vs poor, majority vs minority ethnic groups) rather than fair (and equal) outcome, and that fair treatment was a powerful way in which fair outcomes were achieved. The answers to these two items were combined into one measure of perceived distributive justice, which was the number of items to which the respondent answered that different groups are treated equally (with values 0, 1, or 2; treated as missing if both questions were not answered).

Perceived lawfulness was measured by asking respondents how often they thought that the police in their country take bribes, and to what extent they thought that the decisions and actions of the police are unduly influenced by pressure from political parties and politicians. The average of these two items was used as a combined measure of perceived lawfulness, with values from 1 for lowest to 5 highest level of lawfulness (and treated as missing if both questions were not answered).

The perceived police effectiveness indicators reference key aspects of crime detection, reduction and protection, while also covering both property and violent crime. Fear of crime was measured by asking respondents how often they worry about their home being burgled and about becoming a victim of violent crime. The average of these two items was used as a combined measure of worry about victimization, with values from 1 for lowest to 4 highest level of worry (and treated as missing if both questions were not answered).

Legitimacy is defined along two positively correlated dimensions (Jackson et al., 2012, 2023). The first dimension is normative alignment with the police. Asking survey respondents whether they believe that police generally act in normatively appropriate ways is important, because we can then assess empirically what specific ways the police need to behave if they are to be seen to be generally acting appropriately (see the manuscript for discussion). The second dimension of legitimacy is felt moral duty to obey the police (for discussion, see Posch et al. 2021; Trinkner, 2019). It is important that respondents agree with the statements not because he or she would comply because he or she has little choice, but because of a social, legal, or moral tie, where people are constraining their behavior out of a sense of legal citizenship. As such, asking people whether it is their ‘duty’ to obey the police hopes to capture a positive sense of obligation (something that one is expected or required to do out of moral or legal obligation) and what Raz (1986: 135) calls pre-emptive consent (where these feelings of duty and obligation do not reference to the content of any particular decision or rule).

Willingness to cooperate with the police was measured using a vignette. The items can be seen to reflect an underlying latent continuum. At the lower end we would find a group of individuals who are unlikely and unwilling to call upon and assist the criminal justice system in response to an indicative illegal act. At the upper end we would find a group of individuals who are likely and willing to call upon and assist the criminal justice system in response to an indicative illegal act by all of reporting a specific crime to the police, identifying the culprits in question to the police, and giving evidence in court.

REFERENCES

Jackson, J., Bradford, B., Hough, M., Myhill, A., Quinton, P., & Tyler, T. R. (2012). Why do people comply with the law? Legitimacy and the influence of legal institutions. British Journal of Criminology, 52(6), 1051-1071. https://doi.org/10.1093/bjc/azs032

Jackson, J., Bradford, B., Stanko, E. A., & Hohl, K. (2013). Just Authority? Trust in the Police in England and Wales. Oxon: Routledge.

Pósch, K., Jackson, J., Bradford, B., & MacQueen, S. (2021). “Truly free consent”? Clarifying the nature of police legitimacy using causal mediation analysis. Journal of Experimental Criminology, 17(4), 563-595. https://doi.org/10.1007/s11292-020-09426-x

Raz, J. (1986). The Morality of Freedom. Clarendon Press.

Trinkner, R. (2019). Clarifying the contours of the police legitimacy measurement debate: A response to Cao and Graham. Asian Journal of Criminology, 14(4), 309-335. https://doi.org/10.1007/s11417-019-09300-4

Supplement S2. Statistical modeling: Specification and estimation

In this supplementary section we describe in more detail the specification, motivation and estimation of the statistical models that were used in our analysis.

Let Xji=(X1ji,X2ji,X3ji,X4ji,X5ji,X6ji, X7ji1ji2ji) denote the exogenous variables, i.e. the ones furthest to the left in the path diagram in Figure 1, for respondent i=1,…,nj in country j=1,…,J=30. These include the observed covariates (age and dummy variables for female gender and upper secondary and higher education levels; X1ji – X4ji respectively), plus the observed measures of perceived distributive justice (X5ji), lawfulness of the police (X6ji) and the respondent’s worry about crime (X7ji), together with the latent variables for procedural justice (η1ji) and effectiveness (η2ji). We denote the rest of the latent variables as η3ji for normative alignment with the police, η4ji for duty to obey the police, and η5ji for willingness to cooperate with the police and the courts. Our model of interest is then of the form

η3ji=β30j+β3xjXji+ε3ji                        (1)η4ji=β40j+β4xjXji+ε4ji                        (2)η5ji=β50j+β5xjXji+β53jη3ji+β54jη4ji+ε5ji   (3){\eta_{3ji} = \beta_{30j} + \beta_{3xj}\mathbf{X}_{ji} + \varepsilon_{3ji}\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ (1) }{\eta_{4ji} = \beta_{40j} + \beta_{4xj}\mathbf{X}_{ji} + \varepsilon_{4ji}\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ (2) }{\eta_{5ji} = \beta_{50j} + \beta_{5xj}\mathbf{X}_{ji}^{*} + \beta_{53j}\eta_{3ji} + \beta_{54j}\eta_{4ji} + \varepsilon_{5ji}\ \ \ (3) }

Each of the five latent variables ηkji is measured by three observed items Yklij, for k=1,…,5 and l=1,2,3. The measurement model for each of them is of the form

Yklij= τkl+λklηkji+δklij                   (4)Y_{klij} = \ \tau_{kl} + \lambda_{kl}\eta_{kji} + \delta_{klij}\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ (4)

where the τ-terms (measurement intercepts) and λ-terms (factor loadings) are parameters and the δ-terms are normally distributed residuals with mean 0 and variances var(δklji)=θkl2var(\delta_{klji}) = \theta_{kl}^{2}, and with all δklij independent of each other. This is a conventional factor analysis measurement model, separately for each of the five latent variables. Each parameter in (4) is taken to have the same value in every country j. This is the assumption of measurement equivalence of the items across the countries, as discussed further below.

We estimate the models using a “two-step” method of estimation. In its first step, only the parameters of the measurement models are estimated, separately for the measures of each latent variable and omitting all other variables. This means for each of η1ji,…,η5ji we fit a factor analysis model (4), to the sample of all n= jnjn = \ \sum_{j}^{}n_{j} respondents across all the countries. In each model the latent variable is taken to be normally distributed, with its mean fixed at 0 and variance at 1 for identification of the model. This implies that the scale of each latent variable is defined so that its marginal distribution across all the countries has mean 0 and variance 1. No further identification constraints are required in the rest of the estimation.

In the second step, the structural model (1)-(3) is estimated, combined with the measurement models (4) but with all the measurement parameters fixed at their estimated values from the first step. The likelihood function in this second step is what it would be for a structural equation model where (1)-(3) and (4) were estimated together, except that the step-1 estimates are substituted for τkl\tau_{kl} and λkl\lambda_{kl} as known values. Thus, only the structural parameters (the βs and σs) are estimated in this step.

The basic idea and general properties of two-step estimation of this kind are described for different types of latent variable models by Bakk and Kuha (2018), Rosseel and Loh (2022), and Kuha and Bakk (2023). Two-step estimates of the structural parameters are consistent and have essentially similar large-sample properties as “one-step” estimates that would be obtained by fitting both structural and measurement parts of the model together. The two-step approach is motivated by two broad considerations. Conceptually, by separating the estimation of the measurement and structural models, it also separates the effective definition of the latent variables from other variables in the structural model. This way, the measurement model is not affected even if the structural model is changed or misspecified. In our analysis, the operational definition of each of the five latent variables is thus determined only by their own indicators, without any contribution (even indirect ones) from the other variables in the model. Practically, two-step estimation can be computationally much less demanding than one-step estimation, especially for complex structural models. In our analysis, one-step estimation would involve fitting all of (1)-(4) to the pooled sample of all the countries, requiring the estimation of 1965 distinct parameters at once. The two-step approach is substantially simpler, because its second step for (1)-(3) can be done separately for each country.

We also calculate estimates of the country-level marginal means of the variables. For the latent variables this is done by fitting for each of them separately a model of the form ηkji N(μkj, σkj2)\eta_{kji} \sim \ N(\mu_{kj},\ \sigma_{kj}^{2}) in the second step, with the same measurement model from step 1. This is a structural model ηkji\eta_{kji} where its mean μkj \mu_{kj}\ and variance σkj2\ \sigma_{kj}^{2} depend only on the country k. We also denote by μkj \mu_{kj}\ the population means of the observed measures of distributive justice, lawfulness and worry about crime, which are estimated by their means in the sample. These estimates of μkj\mu_{kj} for both latent and observed variables were calculated using the survey weights that are available for the surveys.

We used the lavaan package (Rosseel 2012) in the R language (R Core Team 2022) to carry out both steps of the estimation (using a likelihood-based approach as described above, rather than the closely related SAM formulation of Rosseel and Loh 2022, which is also implemented in lavaan). The weighted estimates of the country means of the variables were calculated using also the the lavaan.survey (Oberski 2014) and survey packages (Lumley 2010). The R code that was used for the analysis is included in supplementary materials as section S4.

Using the statistical models for comparative analysis

The basic approach of our empirical analysis is to compare estimated regression coefficients of the model (1)-(3) between and within countries, in order to examine the estimated effect sizes in the theoretical model in Figure 1 of the main manuscript across these diverse social, political and legal contexts. We also examine how the estimated means of the variables (μkj\mu_{kj}) vary across the countries, and how the regression coefficients correlate with the estimated country-level means μkj\mu_{kj} of perceived procedural justice.

For these comparisons to be meaningful, we have to first be able to assume that the variables are defined in the same way across all the countries. The first methodological part of this assumption is the careful cross-national harmonization of the survey questions, as discussed in the data section of the paper. Second, for the latent variables we make the assumption of formal measurement equivalence across the countries, meaning that the parameters of the measurement models (4) do not depend on country k. This means that we impose on each latent variable an operational definition which is the same in each country and which is determined by the measurement model for the variable which is estimated from the pooled data for all the countries together. The estimated coefficients of all the variables then refer to the variables thus defined and are comparable across the countries in this sense.

Another type of comparison is between coefficients of different variables within and between countries. For instance, we may want to compare the role of instrumental motives (perceptions of the effectiveness of the police in terms of crime reduction, and personal worries over risk of victimization) to the role of normative motives (procedural justice, normative alignment and duty to obey). Attempting a full answer to such comparisons is likely to be complex, here as in any regression analysis, and to involve substantive considerations outside the statistical analysis. We aim only for a commonly used limited version of them, by considering “standardised” coefficients that refer to similarly defined units of measurement for each variable. Here the standardisation is achieved by construction for the variables of interest (excluding the background covariates age, gender and education), because all of them are defined in such a way that their standard deviations across all the countries are 1. For the latent variables this is imposed when the measurement models are estimated, and the observed measures of perceived lawfulness, distributive justice and fear of crime are directly standardized in this way. Every regression coefficient β of these variables is then in a “fully standardised” form where a difference of 1 unit of standard deviation in this sense in the corresponding explanatory variable is associated with an expected difference of β standard deviation units in the response variable.

For the country means μkj\mu_{kj} of individual latent variables, a useful reference value is provided by the overall mean across the countries, which is fixed at 0. For example, a country for which the mean of perceived procedural justice is positive has a higher mean level of this variable than do these countries on average.

REFERENCES

Bakk, Z., & Kuha, J. (2018). Two-step estimation of models between latent classes and external variables. Psychometrika, 83(4), 871-892. https://doi.org/10.1007/s11336-017-9592-7

Kuha, J., & Bakk, Z. (2023). Two-step estimation of latent trait models. arXiv preprint arXiv: 2303.16101. https://doi.org/10.48550/arXiv.2303.16101

Lumley, T. (2010). Complex Surveys: A Guide to Analysis Using R. Hoboken, NJ: John Wiley and Sons.

Oberski, D. (2014). lavaan.survey: An R package for complex survey analysis of structural equation models. Journal of Statistical Software, 57(1), 1-27. https://doi.org/10.18637/jss.v057.i01

R Core Team. (2022). R: A Language and Environment for Statistical Computing. Vienna: R Foundation for Statistical Computing.

Rosseel, Y. (2012). lavaan: An R package for structural equation modeling. Journal of Statistical Software, 48(1), 1-36. 10.18637/jss.v048.i02

Rosseel, Y., & Loh, W. W. (2022). A Structural After Measurement approach to Structural Equation Modeling. Psychological Methods. Advance online publication. https://doi.org/10.1037/met0000503

Supplement S3. Tables of results from the estimated models

Note: The estimation code and detailed outputs from the models are shown in separate Supplement S4. They include the estimates for those parameters and standard errors that are not included in the tables in this Supplement S3.

Table S3a. Measurement loadings in estimated factor analysis measurement models

Construct

Item 1

Item 2

Item 3

Procedural justice

0.579

0.573

0.533

Effectiveness

1.737

1.821

1.428

Obligation to obey

2.156

2.728

2.495

Normative alignment

0.727

0.789

0.685

Cooperation

0.482

0.813

0.771

Note: See Supplement S1 for the wordings and response options of the measurement items for each construct

Table S3b. Sample sizes and R2 statistics in the structural equation models (SEMs).

Country

n

Normative alignment

Duty to obey

Cooperation

Austria

2,256

0.45

0.32

0.16

Belgium

1,699

0.50

0.14

0.06

Bulgaria

2,430

0.46

0.18

0.11

Croatia

1,608

0.52

0.17

0.06

Cyprus

1,068

0.61

0.30

0.07

Czech Republic

2,384

0.49

0.06

0.06

Denmark

1,573

0.44

0.14

0.09

Estonia

1,792

0.44

0.12

0.10

Finland

1,875

0.40

0.21

0.13

France

1,719

0.57

0.13

0.06

Germany

3,021

0.39

0.11

0.06

Greece

2,714

0.66

0.26

0.03

Hungary

1,560

0.50

0.13

0.07

Ireland

2,537

0.62

0.26

0.14

Israel

2,241

0.39

0.13

0.08

Lithuania

1,621

0.45

0.16

0.03

Netherlands

1,823

0.55

0.12

0.06

Norway

1,540

0.47

0.20

0.09

Poland

1,746

0.45

0.13

0.04

Portugal

2,149

0.33

0.09

0.03

Russia

2,594

0.40

0.23

0.03

Spain

1,879

0.68

0.19

0.06

Slovakia

1,841

0.45

0.05

0.07

Slovenia

1,381

0.48

0.08

0.05

Sweden

1,487

0.59

0.15

0.08

Switzerland

1,501

0.55

0.05

0.05

Ukraine

1,923

0.28

0.14

0.05

United Kingdom

2,333

0.57

0.18

0.10

South Africa

2,388

0.37

0.20

0.07

US

1,603

0.70

0.20

0.14

Table S3c: Estimated regression coefficients in the structural equation models (SEMs) for normative alignment with the police, separately for each country.

Country

Procedural

justice

Distributive justice

Lawfulness

Effectiveness

Fear of

crime

Austria

0.459

0.190

0.045

0.046

-0.060

Belgium

0.636

0.064

0.062

0.191

0.001

Bulgaria

0.481

0.079

0.029

0.185

-0.072

Croatia

0.612

0.040

0.021

0.145

0.010

Cyprus

0.534

0.033

0.122

0.293

0.100

Czech Republic

0.556

0.074

-0.018

0.246

0.005

Denmark

0.566

0.025

0.086

0.132

-0.020

Estonia

0.507

0.001

0.078

0.168

0.064

Finland

0.481

0.033

0.074

0.142

0.027

France

0.502

0.061

0.018

0.401

0.125

Germany

0.433

0.075

0.023

0.198

0.015

Greece

0.543

0.063

0.082

0.247

0.022

Hungary

0.626

0.054

0.038

0.166

0.032

Ireland

0.601

0.047

0.054

0.221

0.018

Israel

0.385

0.036

0.212

0.210

0.062

Lithuania

0.561

0.033

-0.053

0.168

0.046

Netherlands

0.581

0.016

0.018

0.336

0.031

Norway

0.557

-0.019

0.045

0.158

0.041

Poland

0.397

-0.016

0.042

0.246

0.078

Portugal

0.328

0.022

0.039

0.156

-0.034

Russia

0.344

0.057

0.084

0.286

-0.019

Spain

0.764

0.004

0.065

0.137

-0.022

Slovakia

0.269

0.031

0.103

0.385

-0.006

Slovenia

0.621

0.038

-0.026

0.132

-0.024

Sweden

0.635

-0.032

0.041

0.149

-0.018

Switzerland

0.567

0.050

0.090

0.179

-0.007

Ukraine

0.465

0.073

-0.163

0.252

0.054

United Kingdom

0.577

0.078

0.049

0.196

0.016

South Africa

0.236

0.138

-0.029

0.378

0.021

US

0.494

0.044

0.098

0.248

0.024


Table S3d: Estimated regression coefficients in the structural equation models (SEMs) for duty to obey the police, separately for each country.

Country

Procedural

justice

Distributive justice

Lawfulness

Effectiveness

Fear of

crime

Austria

0.317

-0.039

0.033

0.258

0.050

Belgium

0.313

0.012

0.038

0.124

0.026

Bulgaria

0.288

0.015

0.013

0.196

-0.035

Croatia

0.305

-0.103

0.191

0.153

-0.094

Cyprus

0.358

-0.106

0.041

0.157

0.011

Czech Republic

0.218

0.023

-0.030

0.065

0.007

Denmark

0.303

0.042

0.039

0.087

-0.008

Estonia

0.243

-0.063

0.130

0.211

0.007

Finland

0.267

0.020

0.054

0.147

0.030

France

0.202

0.034

-0.003

0.161

0.038

Germany

0.260

0.045

0.038

0.149

0.072

Greece

0.271

0.041

0.032

0.145

0.066

Hungary

0.292

0.010

0.056

0.104

0.007

Ireland

0.368

-0.004

0.112

0.087

-0.039

Israel

0.299

0.120

-0.053

-0.023

0.004

Lithuania

0.234

0.084

0.044

0.189

0.079

Netherlands

0.229

0.010

0.009

0.188

0.026

Norway

0.420

-0.018

0.026

0.090

0.055

Poland

0.175

0.053

0.101

0.141

0.120

Portugal

0.134

0.035

0.049

0.077

-0.029

Russia

0.221

0.059

0.055

0.203

-0.012

Spain

0.288

-0.025

0.068

0.136

0.028

Slovakia

0.112

0.011

0.010

0.152

0.044

Slovenia

0.444

-0.093

-0.039

0.026

-0.107

Sweden

0.393

-0.033

0.052

0.078

-0.001

Switzerland

0.178

-0.012

0.079

0.044

0.027

Ukraine

0.258

0.139

-0.082

0.101

-0.074

United Kingdom

0.216

0.023

0.054

0.187

0.050

South Africa

0.046

0.008

-0.024

0.300

0.006

US

0.274

-0.009

0.037

-0.012

-0.002

Table S3e: Estimated regression coefficients in the structural equation models (SEMs) for cooperation with the police, separately for each country.

Country

Procedural

justice

Effectiveness

Fear of crime

Duty to obey

Normative alignment

Austria

0.295

-0.217

-0.127

0.035

0.107

Belgium

0.091

0.021

-0.006

0.007

0.049

Bulgaria

0.140

-0.010

-0.001

0.031

0.110

Croatia

0.095

0.077

-0.088

-0.021

0.003

Cyprus

0.289

-0.135

-0.050

-0.040

-0.091

Czech Republic

0.121

0.014

-0.033

0.020

0.090

Denmark

0.094

-0.086

-0.076

0.118

0.054

Estonia

0.255

-0.088

-0.027

0.027

0.148

Finland

0.087

-0.071

-0.068

0.150

0.163

France

0.133

-0.097

-0.006

0.037

0.094

Germany

0.102

0.026

-0.016

-0.001

0.074

Greece

0.043

0.105

0.015

-0.031

0.027

Hungary

0.067

-0.024

-0.003

0.112

0.143

Ireland

0.134

-0.011

-0.082

-0.001

0.151

Israel

0.081

-0.001

0.101

0.159

0.009

Lithuania

-0.093

0.162

-0.035

0.037

0.094

Netherlands

0.127

-0.032

0.040

0.092

0.016

Norway

0.088

-0.065

-0.096

0.021

0.166

Poland

0.094

-0.061

0.048

0.003

0.092

Portugal

0.016

-0.048

0.004

0.027

0.040

Russia

0.071

0.073

0.009

0.025

0.030

Spain

0.204

-0.062

0.017

0.067

-0.045

Slovakia

0.289

-0.017

-0.005

0.019

-0.071

Slovenia

-0.016

0.129

-0.010

-0.066

-0.027

Sweden

0.180

0.061

-0.138

0.064

0.057

Switzerland

0.132

-0.062

-0.056

-0.002

0.071

Ukraine

0.107

-0.050

0.052

0.004

0.072

United Kingdom

0.212

-0.075

-0.050

0.115

-0.030

South Africa

0.210

-0.046

0.013

0.013

0.100

US

0.146

-0.186

-0.017

0.010

0.208

Comments
0
comment
No comments here
Why not start the discussion?