The importance of procedural justice and police performance in shaping intentions to cooperate with the police: Does social identity matter?

Abstract

Research consistently reveals that public perceptions of procedural justice and police performance are important for fostering citizens’ willingness to cooperate with police, with procedural justice being more important than police performance. Identifying factors that motivate people’s intentions to cooperate with police is the focus of the present study. Of particular interest will be how people’s affiliations with different groups in society moderate their responses to questions about their willingness to cooperate with police. The study utilizes survey data from 10,148 Australian residents and demonstrates that procedural justice, police performance, and identity each predict people’s intentions to cooperate with police. The findings also reveal that identity can moderate citizens’ concerns about procedural justice and police performance when predicting cooperation.

Keywords

Cooperation ethnicity police police performance procedural justice social identity

Introduction

Encouraging cooperation with police is essential for effective crime control. Without public assistance, police would struggle to maintain order and apprehend offenders. A growing body of research has demonstrated that police will be best equipped to foster cooperation from the public if they are effective and utilize strategies that are seen to be procedurally fair. The Group Value Model underpins much of our understanding about why procedurally fair treatment matters so much to people (Lind and Tyler, 1988). The model explains the connection between procedural justice and cooperation as a function of social identification. When police – as important representatives of the state – use procedural justice, they communicate messages about one’s value and standing in society and reaffirm a sense of ‘group identity’ or societal membership among citizens. When citizens are treated with respect and fairness, their social standing within society is upheld, encouraging allegiance to group norms and cooperation with group authorities (Sunshine and Tyler, 2003). In a law enforcement context, adhering to group norms includes obeying the laws of the land and being a good law-abiding citizen. Group norms relate to the moral norms that would be expected of people in a particular country. Respect for authority is another normative concern that has relevance in a law enforcement context (see Bradford et al., 2014; Manstead, 2000). The Group Value Model further argues that authorities should be more effective in eliciting cooperation from those who identify more strongly with them because people have a greater sense of obligation to obey those they have closer connections to.

In diverse societies, however, eliciting cooperation may not be as straightforward as the Group Value Model suggests. As we move further into the 21st century, Western nations are trending towards increasing ethnic, racial, and cultural diversity. Authorities such as the police are subsequently faced with greater differences between and within groups relating to the expression of cultural norms, values, identities, and attitudes (Rose, 1993; Tyler, 2005). Also, social diversity is not simply linked to markers of ethnicity or culture, but is increasingly evident in the lifestyle, economic, and political choices people adopt. Hence, identity has become fragmented, with individuals less likely to automatically identify with institutional authorities because they are not seen as representing their values or interests. The consequence is that the notion of a unifying identity (that is, thinking of oneself as American, Australian or British) has lost meaning for some groups (Calhoun, 1994; Rutherford, 1998).

One of the core assumptions underlying the Group Value Model is that individuals identify with or want to identify with social groups in society, with some groups being more desirable than others. But what happens when people do not identify with the norms and values of the dominant group in society? For example, an individual may identify strongly with their own ethnic, cultural, or racial subgroup, but little with the superordinate nation-state in which they live. If the authority represents a group to which an individual feels little or no attachment, then cooperation with the authority may not be driven by allegiance to the wider societal group or to procedural justice concerns. This may have implications for the ability of police to foster trust or encourage cooperation using procedural justice. In fact, some scholars go so far as to argue that strong loyalty to subgroups can damage social cohesion within a society, which can lead to an increase in social disorder and conflict between groups (see Glazer, 1997; Schlesinger, 1992).

Within the procedural justice literature there is empirical support for the key assertions put forth by the Group Value Model (for example, Huo, 2003; Huo et al., 1996). To date, however, the importance of social identity has still not been investigated adequately with regard to the consequences of procedural justice for citizens’ intentions to cooperate with legal authorities. In the current study, therefore, we examine the role that social identity plays in shaping the intentions of citizens to cooperate with police. Specifically, we are interested in testing whether identification with an ethnic or racial subordinate group within society – to the exclusion of identifying with the broader superordinate group (that is, Australia) – affects social relations between police and those they govern.

The importance of procedural justice

Since Tom Tyler’s work was published in the early 1990s, the procedural justice literature has focused on how procedural justice can influence people’s law-abiding behaviour. In evaluating whether one has received procedural justice from authorities, research has shown that people focus on four key relational aspects: (1) whether authorities treat them with ‘dignity and respect’; (2) whether the motives of authorities are seen to be ‘fair’ (that is, do they act in the best interest of citizens and do they display concern for their well-being); (3) whether decisions made by authorities are based on facts, not biases or personal opinions (that is, ‘neutrality’); and (4) whether people have the opportunity to ‘voice’ their concerns before an authority makes a decision (Tyler, 1990).

Of course, people’s evaluations of authorities are also shaped by their views of the effectiveness of those authorities. Such an ‘instrumental view suggests that police can increase support from the public when they (a) effectively control crime and criminal behaviour and (b) create a credible risk of detection and sanction for those who break the law’ (Hinds and Murphy, 2007: 28).

Studies conducted in the USA have consistently found that procedural justice matters more to people than instrumental concerns when predicting how they will respond to police (for example, Sunshine and Tyler, 2003; Tyler and Huo, 2002). In other words, research has found that people are generally concerned with whether police treat them with respect, fairness, and neutrality, and offer people a voice in decisions. Instrumental factors such as whether the police effectively perform their job seem to matter less in predicting people’s willingness to cooperate with police. Recent research in Europe and Australia has also demonstrated the dominance of procedural justice concerns over instrumental concerns in determining people’s intentions to cooperate with police (for example, see Jackson et al., 2013; Murphy et al., 2008; Reisig et al., 2012). However, although research suggests procedural justice better explains cooperation with police in general population surveys, recent research conducted in non-Western societies, or research that uses ethnic minority group samples, questions the applicability of this finding to all groups and contexts. Such research tells us that we still have more to learn about the contexts in which procedural justice versus instrumental factors matter most (see Murphy and Cherney, 2012; Sargeant et al., 2014; Tankebe, 2009).

The influence of social identity

In a diverse society, people can hold multiple affiliations (Huo et al., 1996; Meredyth et al., 2010; Rutherford, 1998; Tajfel and Turner, 1986). People can identify primarily with their own ethnic or racial subgroup, they can identify primarily with the country in which they live (that is, a national identity), or they can choose to affiliate themselves with both, or neither.

Whether these different identities can moderate the effect of procedural justice or instrumental concerns on people’s evaluations of authorities has received increasing attention of late. In the context of workplace relations, Huo et al. (1996) used survey data collected from ethnic minority employees to ascertain whether procedural justice or instrumental factors dominated their assessments of their work supervisors. Employees were asked to comment on the fairness of their work supervisor and their willingness to accept their decisions. They were also asked about their level of identification with their workplace and with their own ethnic/racial subgroup. For those who identified strongly with their workplace, procedural justice evaluations of the authority were more important than instrumental evaluations when determining employees’ perceptions of fairness and acceptance of their supervisor’s decisions. In other words, when superordinate identity was high, procedural justice mattered most. In contrast, instrumental factors were the key drivers of decision acceptance and perceptions of fairness for those who weakly identified with the superordinate group but who identified strongly with their own ethnic/racial subgroup.

Although interesting, Huo et al.’s study cannot tell us how these processes may operate in a law enforcement context. Their study also did not examine superordinate allegiance to a nation-state or unifying entity (that is, the USA). Nor did it examine the moderating effects of identity on people’s intentions to cooperate with an authority. Huo et al. simply examined how identity interacted with procedural justice and instrumental factors to predict people’s evaluations of the authority. Whether Huo et al.’s findings are applicable for understanding people’s willingness to cooperate with legal authorities remains untested. As such, the current study aims to extend the application of the Group Value Model to a law enforcement context to better understand how authorities such as the police might achieve greater cooperation from different members of the public.

To date, only six published studies have explored the intersection between identity and justice concerns in the context of law enforcement (see Bradford, 2014; Bradford et al., 2014; Huo, 2003; Murphy, 2013; Murphy and Cherney, 2011; Tyler and Huo, 2002). Each study has its weaknesses and none examines how different types of identity interact with both procedural justice and police performance to influence people’s intentions to cooperate with police. For example, Tyler and Huo’s (2002) study did not assess whether the minority groups in their study identified strongly or weakly with US society or their own ethnic/racial group; identification was based purely on the race of the individual (that is, whether the person self-categorized themselves as African-American, White, or Hispanic). Huo’s (2003) study did not examine people’s willingness to cooperate with police; only acceptance of a law enforcement officer’s decisions was examined. Although studying intentions to cooperate with police, Murphy (2013) examined only people from an ethnic minority background; those from a non-minority background were excluded. Murphy (2013) also did not examine how identity processes interacted with perceptions of police performance to predict cooperation. Murphy and Cherney (2011) failed to test the importance of procedural justice and instrumental factors to cooperation across different levels of subordinate and superordinate identification (they examined only the main effects of procedural justice and identity on cooperation). Finally, Bradford’s recent work has been unable to shed light on whether social identity interacts with procedural justice to influence cooperation with police. In Bradford’s (2014) study, the identity measure was one of citizenship (UK citizen or not). Whether people identified psychologically with their own ethnic subgroup or the UK was not assessed. Finally, Bradford et al.’s (2014) study did not examine how identity interacted with procedural justice to predict cooperation with police; that study was concerned with how procedural justice and identity impacted upon people’s evaluations of the legitimacy of police. Given the issues apparent in each of these studies, we cannot yet be certain how social identification processes interact with both procedural justice or instrumental concerns to predict people’s intentions to cooperate with police.

The current study examines how procedural justice and instrumental factors shape people’s intention to cooperate with police when they identify either weakly or strongly with an ethnic subgroup versus the superordinate nation-state of Australia. As such it extends the application of the Group Value Model to a law enforcement context to understand how identity influences people’s intentions to cooperate with police. If the Group Value Model is correct in its assertions, then we would expect procedurally fair policing to matter more to those who identify more strongly with the superordinate nation-state of Australia (this is, because police are important representatives of Australia). In contrast, for those who feel little or no attachment to Australia, and who instead identify strongly with their own ethnic subordinate group (or who think it important to remain separate from the dominant group), procedural justice may matter less. Here, instrumental factors, such as whether the police deter offenders and effectively deal with crime in their community, may dominate their intentions to cooperate with police.

Method and methodology

Sample and procedures

We draw on data collected in 2010 and 2011 from 10,148 Australian residents who participated in wave three of the Australian Community Capacity Study (ACCS). The sample includes a booster sample of 908 people from an ethnic minority background. The ACCS involved undertaking a large-scale survey of people living in and around Brisbane and Melbourne. Brisbane and Melbourne are the capital cities of the states of Queensland and Victoria in Australia, with populations of approximately 2 million and 4 million people, respectively. The ACCS survey sample includes residents living in 298 randomly selected suburbs¹ in the Brisbane Statistical Division and the Major Statistical Region of Melbourne, as well as a sample of people from Indian, Vietnamese and Arabic-speaking backgrounds living in and around these suburbs.

For the primary ACCS sample, households were randomly selected from within the 298 suburbs using the electronic telephone directory and Random Digit Dialling. Participants were selected from the household if they had most recently celebrated a birthday and were over 18 years of age (many of the Brisbane respondents had already participated in waves one and two of the ACCS). Interviews were conducted using Computer Assisted Telephone Interviewing (see Mazerolle et al., 2012).

The ethnic sample was selected to complement the primary ACCS sample. Indian, Vietnamese and Arabic-speaking groups were chosen because these groups are known to have had problematic relationships with police (Mason, 2009; Meredyth et al., 2010; Poynting, 2006). The ethnic participants were selected using a surname-based method, because Random Digit Dialling was deemed unviable for a small target population. A list of common surnames for Indian, Vietnamese and Arabic-speaking people was collated and the list of names was employed in combination with the electronic telephone directory and post-code matching to select a sampling pool of 10,800 households. Participants in this pool were randomly contacted by phone and were selected if they were over 18 years of age and were next to celebrate a birthday. Interviews were then arranged and conducted face-to-face, using pen-and-paper surveys in the participant’s preferred language (see Murphy et al., 2012).

Response rates for both samples were calculated in line with the American Association for Public Opinion Research (2011) guidelines. For the primary sample the response rate was 50.1 percent. For the ethnic sample the response rate was 43.2 percent.² Participants in the combined sample were aged between 18 and 99 with a mean age of 50. With the inclusion of the ethnic sample, the full sample has a comparable proportion of overseas-born residents to the national population (Australian Bureau of Statistics, 2011). Of the respondents: 40.2 percent were male; 68.4 percent were married or in a de facto relationship; 39.4 percent had a university education; 59.2 percent were in full- or part-time employment; and 30.1 percent were born overseas. The median household income was AUD60,000–79,999. Compared with the national population, the sample is slightly over-representative of women, older people, and people with university qualifications (Australian Bureau of Statistics, 2011). These differences are common in Australian surveys (for example, Murphy et al., 2010) and telephone surveys more broadly (for example, Pickett et al., 2012). We did not weight the sample because we determined the differences were not large enough to warrant this procedure. Instead, we control for demographic characteristics that are associated with attitudes towards police in the analyses (Brown and Benedict, 2002).

Measures

Identical measures of cooperation with police, procedural justice, police performance, superordinate identity, subordinate identity, separatist identity, and demographic/ control characteristics were measured for the primary ACCS sample and the ethnic booster sample (see Table 1 for items comprising key scales). The three policing measures are based on the work of policing scholars in the USA (Sunshine and Tyler, 2003), the UK (Jackson and Bradford, 2010), and Australia (Murphy et al., 2010); these measures have been widely validated in previous studies.

Table 1.

Principal components factor analysis with Varimax rotation for policing constructs.

Policing constructs	Factor
	1	2	3
Cooperation with police
…call police to report a crime?			.79
…help police find someone suspected of committing a crime by providing information?			.86
…report dangerous or suspicious activities to police?			.80
…willingly assist police if asked?			.80
Procedural justice
Police treat people fairly	.76
Police treat people with dignity and respect	.81
Police are always polite when dealing with people	.77
Police listen to people before making decisions	.77
Police make decisions based upon facts, not their personal biases or opinions	.75
Police performance
Dealing with problems that concern you		.75
Preventing crime		.87
Keeping order		.83
Solving crime		.81
Eigenvalues	5.20	2.21	1.50
Explained variance	39.99	16.98	11.57

Note: Values < .30 not shown.

Cooperation with police was measured with four items gauging how likely the participant would be to cooperate with police (alpha = .84). Cooperation was measured on a 1 (unlikely) to 5 (very likely) scale, with higher scores indicating greater intentions to cooperate, rather than actual cooperative behaviour. Procedural justice was measured with five items that captured general perceptions of both the quality of treatment and the quality of decision-making by police (alpha = .86). Each of the four relational aspects of fairness, respect, neutrality, and voice were assessed via this scale. Police performance served as the instrumental variable. It was measured with four items assessing participants’ general beliefs about how good a job police do at preventing and controlling crime and disorder in their community (alpha = .88). The procedural justice and police performance items were scored on Likert scales – 1 (strongly disagree) to 5 (strongly agree) – with higher scores indicating more favourable evaluations.

Because prior research has problematized the measurement of policing variables (Reisig et al., 2007), we conducted a factor analysis of the policing variables to assess the underlying factor structure. Eigenvalues indicate that a three-factor solution provides a good fit for the data: cooperation with police, procedural justice, and police performance load on distinct factors (see Table 1). Bartlett’s test of sphericity (X² = 60381.05(78) p < .001) and the Kaiser-Meyer-Olkin test (.89) indicate that the factor analysis is appropriate for the observed data.

Three single-item measures of identity were examined. Each question was presented one after the other in the survey. Prior to presenting the identity questions, respondents were asked to answer the questions by ‘thinking about how you view yourself within your neighbourhood’. To measure superordinate identity, participants indicated their level of agreement with the following statement: ‘I see myself first and mainly as a member of the Australian community’. Hence, superordinate identification assessed allegiance to a nation-state. To measure subordinate identity, participants responded to: ‘Within Australia, I see myself first and mainly as a member of my racial/ethnic group’. The subordinate identity scale assesses strength of identification with a subgroup nested within the broader superordinate group. A separatist identity was also assessed via the following statement: ‘People from my ethnic/racial group should try to keep a separate cultural identity’. This last item assessed the degree to which respondents felt it important not to assimilate fully into Australian culture (that is, strength of identification with a subgroup external to the superordinate group). Although it may be that this latter question was interpreted by some respondents as measuring the importance of upholding one’s ethnic traditions, while still supporting Australian values and culture, the results presented in Table 3 suggest that this was unlikely; low mean scores on this variable suggested this was not an attractive identity to respondents, even to minority group respondents. For each of the three identity questions, a 1 (strongly disagree) to 5 (strongly agree) Likert scale was used, with higher scores indicating stronger identification.³

Demographic and other control variables were also employed. These were age, gender (0 = female; 1 = male), education (1 = no schooling to 7 = post-graduate qualifications), employment (0 = employed; 1 = unemployed), prior contact with police (0 = no contact; 1 = has had prior contact with police), ethnicity (0 = ethnic minority; 1 = non-minority), and language (0 = speaks a language other than English at home; 1 = speaks English at home).

The ethnicity variable was created by coding respondents’ self-reported ancestry; respondents who indicated they originated from Africa, the Middle East, Asia, or Oceania were categorized as coming from an ethnic minority background (N = 1686; 15.9 percent of the sample), while those from Australia, Europe, and the Americas were categorized as the non-minority group. Although this method of classifying people into an ethnic minority or non-minority group is less than perfect, it does give a rough guide to the ethnic minority status of respondents. The ethnicity variable was also found to be strongly correlated with the language variable (r = .57, p < .001).

Limitations of method

It is important that we highlight a few limitations of our study before presenting the results. They should be taken into account when interpreting the results. First, the data upon which our conclusions are drawn come from a cross-sectional survey. Data of this kind offer evidence that predicted relationships exist, but they leave open the issue of causality between variables. Longitudinal data, or experimental methodologies, would be more effective for teasing apart the causal relationships between our variables of interest.

Second, many of the coefficients reported in Table 4 are small, despite being significant. Although large sample sizes enable greater power to detect small differences, they do run the risk of inflating Type 1 errors. Although there are statistically significant relationships between many of our key variables and intentions to cooperate, we need to interpret the findings with caution until they can be replicated in future research.

Third, we should note that the overall variation explained in the models is small, suggesting that other factors not included in our models may better predict people’s intentions to cooperate. In many of the studies reported in the Introduction, cooperation was not the dependent variable of interest. Perhaps identity plays a more important role in predicting evaluations of authorities or decision acceptance than it does in predicting intentions to cooperate with authorities. Future studies should explore what additional factors may influence people’s intentions to cooperate with police.

Fourth, our study examined people’s intentions to cooperate with police. One can therefore question whether our respondents’ intentions to cooperate with police will result in actual cooperative behaviour in an encounter with a police officer. Although empirical research suggests that intentions can translate into actual behaviour (see Ajzen, 1985), our study is unable to definitely demonstrate whether this will occur. Future research should aim to examine how ethnicity and identity interact with procedural justice and police performance to predict actual cooperation with police.

Finally, some of our other measures can also be criticized. We have already highlighted the limitation of using single-item measures of identity and also the fact that our identity items might be interpreted differently by different respondents (see note 3), but our procedural justice and police performance measures can also be criticized for assessing general perceptions of police rather than personal experiences of fair treatment or poor police performance. These general measures are commonly used in the procedural justice literature, but future research may wish to examine how personal experiences of unfair police treatment might go on to shape individuals’ willingness to cooperate with police, especially if that police officer is a member of a different ethnic group (that is, an outgroup member) from the respondent.

Results

Preliminary analyses

We first calculated descriptive statistics and bivariate correlations for our variables (see Tables 2 and 3). Descriptive statistics in Table 2 indicate a high degree of willingness to cooperate with police (Mean = 4.42). On average, respondents also generally believed the police perform their duties well and are procedurally just. It is interesting to note, however, that there is some variation in the identity measures. Respondents tended to identify strongly with Australia (Mean = 4.05), but they were less likely to hold a strong separatist identity (Mean = 2.36). The subgroup identity measure also fell below the midpoint of the scale (Mean = 2.69). Given that the majority of our full sample were of Anglo-Saxon ancestry, the low mean scores for the subordinate and separatist identity variables could indicate that few people from an Anglo-Saxon background think of themselves in terms of their ethnicity or race.

Table 2.

Descriptive statistics and bivariate correlations for the full sample.

	Mean (SD)	1	2	3	4	5	6	7	8	9	10	11	12	13
1. Cooperation with police	4.42 (0.55)	1
2. Procedural justice	3.76 (0.64)	.25*	1
3. Police performance	3.75 (0.68)	.28*	.50*	1
4. Superordinate id	4.05 (0.72)	.18*	.13*	.13*	1
5. Subordinate id	2.69 (1.12)	−.09*	.01	.01	−.03*	1
6. Separatist id	2.36 (1.01)	−.12*	−.03*	−.04*	−.05*	.32*	1
7. Age	50.07 (15.82)	.09*	.08*	.08*	.07*	−.01	−.06*	1
8. Education	4.89 (1.39)	.04*	−.06*	−.08*	.02*	−.15*	−.00	−.20*	1
9. Contact with police	41.80%	.07*	.00	−.02*	−.00	−.11*	−.08*	−.11	.09*	1
10. Gender	40.20%	−.06*	−.04*	−.09*	.03*	−.01	.06*	.02	.08*	.07*	1
11. Employment	40.80%	−.02*	.05*	.07*	−.03*	.09*	.01	.39*	−.25*	.10*	−.11*	1
12. Language	77.90%	.16*	.05*	.06*	.07*	−.27*	−.35*	.29*	−.13*	.09*	−.07*	.08*	1
13. Ethnicity	84.10%	.11*	.06*	.08*	.09*	−.17*	−.19*	.30*	−.16*	.07*	−.07*	.07*	.57*	1

Notes: Contact with police (0 = no contact; 1 = contact); Gender (0 = female; 1 = male); Employment (0 = employed; 1 = unemployed); Language (0 = speaks a language other than English at home; 1 = speaks English at home); Ethnicity (0 = minority; 1 = non-minority). Percentages shown for dichotomous variables, with percentage referring to the ‘1’ category.

p ⩽ .05

Table 3.

Mean and standard deviation scores for the three identity variables across different ethnic groups.

	Australian	African/Middle Eastern	Asian	European/American
	(N = 1730)	(N = 366)	(N = 1013)	(N = 6866)
	Mean (SD)	Mean (SD)	Mean (SD)	Mean (SD)
Superordinate identity	4.06 (0.67)	4.06 (0.90)	3.79 (0.85)^{a b}	4.08 (0.69)^c
Subordinate identity	2.61 (1.07)	3.36 (1.16)^a	3.03 (1.08)^{a b}	2.61 (1.10)^{b c}
Separatist identity	2.17 (0.87)	2.98 (1.20)^a	2.73 (1.07)^{a b}	2.30 (0.98)^{a b c}

Notes:

Significantly different from Australian at p < .001.

Significantly different from African/Middle Eastern at p < .001.

Significantly different from Asian at p < .001.

Table 3 presents the mean scores for the three identity measures across four different groups: (1) those who reported Australian ancestry; (2) those with African or Middle Eastern ancestry; (3) those with Asian or Oceanic ancestry; and (4) those with European or American ancestry. Those with Anglo-Saxon ancestries (Groups 1 and 4) were indeed less likely to identify with an ethnic/racial subordinate group than the African/Middle Eastern or the Asian group. Similarly, the African/Middle Eastern and Asian group were also significantly more likely to hold a separatist identity. It can also be seen in Table 3 that Asians were less likely to identify strongly with the superordinate group (that is, Australia) when compared with the other three groups.

Returning to Table 2, procedural justice and police performance were both positively correlated with cooperation. Those who felt the police are more likely to use procedural justice and are more effective in dealing with crime were more likely to indicate a general willingness to cooperate with police. Identification was also related to intentions to cooperate. Those who identified more strongly with the superordinate nation-state of Australia were more likely to say they would cooperate with police, and those who identified more strongly with their own ethnic subgroup were less likely to say they would cooperate. Those who identified strongly with a separatist identity were also less likely to say they would cooperate with police.

Further, the superordinate identity variable was positively associated with perceptions of procedural justice, whereas a separatist identity was negatively associated with procedural justice (although only weakly so); those who identified strongly with the superordinate group were more likely to believe police use procedural justice, whereas those with a stronger separatist identity were less likely to view police as procedurally fair. Subordinate identity was not correlated with procedural justice. Similarly, superordinate and separatist identities were positively and negatively correlated with police performance, respectively, while subordinate identity was not related to police performance. In other words, those with a strong superordinate and weak separatist identity were more likely to view police as effective.

Regression analysis

We constructed an Ordinary Least Squares (OLS) regression model predicting people’s intentions to cooperate with police. Variables were entered in groups across three steps to examine the individual effects of control variables, procedural justice and police performance, and identity on intentions to cooperate. Interaction effects between the three identity measures and procedural justice and police performance variables, respectively, were then entered into the model at a fourth step. All variables were centred prior to entry into the model. The results appear in Table 4.

Table 4.

OLS regression of independent variables on intentions to cooperate with police.

	Model 1		Model 2		Model 3		Model 4
	B (SE)	β	B (SE)	β	B (SE)	β	B (SE)	β
Intercept	4.02 (.03)		4.03 (.03)		4.10 (.03)		4.10 (.03)
Age	.00 (.00)	.10***	.00 (.00)	.08***	.00 (.00)	.07***	.00 (.00)	.07***
Gender	−.07 (.01)	−.06***	−.05 (.01)	−.04***	−.05 (.01)	−.05***	−.05 (.01)	−.05***
Education	.03 (.00)	.06***	.03 (.00)	.08***	.03 (.00)	.07***	.03 (.00)	.07***
Employment	−.07 (.01)	−.06***	−.07 (.01)	−.06***	−.06 (.01)	−.05***	−.06 (.01)	−.05***
Language	.16 (.02)	.12***	.16 (.02)	.12***	.11 (.02)	.08***	.11 (.02)	.09***
Ethnicity	.02 (.02)	.02	.00 (.02)	.00	−.01 (.02)	−.00	−.01 (.02)	−.01
Contact with police	.08 (.01)	.07***	.07 (.01)	.07***	.07 (.01)	.06***	.07 (.01)	.06***
Procedural justice			.13 (.01)	.15***	.12 (.01)	.14***	.12 (.01)	.14***
Police performance			.16 (.01)	.20***	.15 (.01)	.19***	.15 (.01)	.19***
Subordinate identity					−.01 (.01)	−.03**	−.01 (.01)	−.03**
Superordinate identity					.09 (.01)	.12***	.09 (.01)	.12***
Separatist identity					−.03 (.01)	−.06***	−.03 (.01)	−.05***
Subordinate identity × Procedural justice							.00 (.01)	.00
Superordinate identity × Procedural justice							−.02 (.01)	−.02
Separatist identity × Procedural justice							−.02 (.01)	−.02*
Subordinate id × Police performance							.01 (.01)	.01
Superordinate identity × Police performance							−.01 (.01)	−.01
Separatist identity × Police performance							.02 (.01)	.03*
R ²	.04		.13		.15		.15
F change	60.66***		496.57***		68.44***		2.76**
df	9821		9819		9816		9810

Notes: Contact with police (no contact = 0; contact = 1); Gender (female = 0; male = 1); Employment (other = 0; unemployed = 1); Language (0 = speaks a language other than English at home; 1 = speaks English at home); Ethnicity (0 = ethnic minority; 1 = non-minority).

p ⩽ .05; **p ⩽ .01; ***p ⩽ .001.

We find in Model 1 that all control variables, apart from the ethnicity variable, predict intentions to cooperate with police. Those who were older, who were female, and who were more educated were more likely to say they would cooperate with police, supporting prior research (see Skogan, 2006). Those who were in work were also more likely to say they would cooperate, so too were those who spoke English at home. We might expect that those who do not speak English would feel more marginalized in society, thereby being less willing to engage with key institutions. Surprisingly, those who have had more personal contact with police indicated a greater willingness to cooperate. Findings from other countries tend to show that police contact results in more negative assessments of police (see Bradford et al., 2009; Skogan, 2006). Perhaps when our respondents had actual contact with police they were pleasantly surprised by the treatment they received; here, vicarious experiences or stories relayed by others about unfair treatment were not borne out during actual encounters. It was also somewhat surprising to find that ethnicity did not predict intentions to cooperate. Studies generally find that ethnic minority groups are less willing to engage with police (for example, Murphy and Cherney, 2011; Skogan, 2006). We did not find this here (for similar findings, see Oliveira and Murphy, 2014; Weitzer, 2010).

Each of the significant control variables continued to predict intentions to cooperate, even after entry of new variables in Models 2 to 4. In Model 2, both procedural justice and police performance measures were added. Here, procedural justice and police performance were both positively and significantly associated with intentions to cooperate with police. Those who believed the police are more procedurally fair and more effective in dealing with crime were more willing to say they would cooperate with police. Specifically, procedural justice and police performance appeared to be of similar importance to Australians in determining their intentions to cooperate, with performance being slightly more important.

Model 3 reveals that identity is also predictive of intentions to cooperate. Those who identified more strongly with Australia were more likely to say they would cooperate with police, whereas those who identified more strongly with a subordinate ethnic group were less likely to cooperate. Those with a stronger separatist ethnic identity were also less likely to say they would cooperate with police. Of the three identity variables, superordinate identity played the biggest role in predicting cooperation.

The Group Value Model predicts that procedural justice concerns should carry more weight if provided by an authority that represents a group one identifies strongly with, whereas instrumental concerns should matter more to those who hold little attachment to the authority. The interaction effects presented in Model 4 partially support these claims. Model 4 shows that the effect of procedural justice on intention to cooperate was moderated only by the separatist identity variable. Specifically, the interaction between procedural justice and separatist identity was negative and significant (β = −.02, p < .05), suggesting that procedural justice had a weaker effect on cooperation for those who held a strong separatist identity; procedural justice was more important to those with a weak separatist identity. The lack of interaction effects between the superordinate identity and procedural justice measure and between the subordinate identity and procedural justice measure indicates that procedural justice had an equivalent positive effect on intentions to cooperate for both low and high superordinate identifiers and for both low and high subordinate identifiers.

Interestingly, the separatist identity variable also moderated the effect of police performance on cooperation (β = .03, p < .05). The positive interaction term suggests that instrumental factors matter more to those with a strong separatist identity than to those with a weak separatist identity. There were no significant interactions between the superordinate identity and police performance variables or the subordinate identity and police performance variables.

Discussion of findings

Prior research suggests that procedural justice is effective for promoting people’s willingness to cooperate with police, more so than instrumental factors. Yet some research indicates that this may be contingent on social or contextual factors; specifically, the group that one identifies with in society may determine whether procedural justice or instrumental factors are more important to people (Sargeant et al., 2014). The present study sought to test this claim, and specifically tested whether police could utilize procedural justice to nurture intentions to cooperate with police in a country where people are free to openly express different group affiliations.

We found a number of interesting findings. First, our results support previous studies by showing that procedural justice directly promotes people’s intentions to cooperate with authorities. Second, we found that police performance was also important for promoting intentions to cooperate with police. In fact, police performance mattered slightly more than procedural justice in our study. Past Australian research has tended to reveal that police performance is given more weight in Australia than in the USA (for example, Murphy et al., 2008). Hinds and Murphy (2007) suggest this may be due to the heavy emphasis given to individuals’ rights in the USA, making procedural justice concerns more salient to people in the USA. Third, our findings reveal that identity also plays some role in shaping intentions to cooperate. Those who identified more strongly with Australia were more inclined to cooperate with police, whereas those who identified more strongly with their own ethnic subgroup, or who believed it important to maintain an ethnic identity separate from an Australian identity, were less inclined to cooperate. Where identity seems to be less reliable is in moderating the effect of procedural justice or police performance on willingness to cooperate with police.

We found that only the separatist identity variable moderated the effect of procedural justice and police performance on intentions to cooperate. The interaction between the procedural justice and separatist identity variables revealed that those with a strong separatist identity placed less importance on procedural justice compared with those with a weak separatist identity when deciding whether they would consider cooperating with police. The significant interaction between the police performance and separatist identity variables also revealed that those with a strong separatist identity focused more strongly on police performance compared with those with a weak separatist identity. Neither the superordinate nor the subordinate identity variables moderated the effect of procedural justice or police performance on intentions to cooperate.

So what do these results tell us? First and foremost, the predictions made by the Group Value Model appear to have some merit in a law enforcement context when predicting intentions to cooperate with police. The Group Value Model predicts that procedural justice considerations should be less relevant, and instrumental factors should be more relevant, in interactions with an authority that represents an ‘outgroup’ one does not care about. The significant interaction effects between the separatist identity and both the procedural justice and the police performance variables supports this prediction. We should note, however, that there are also other non-identity-based explanations for why separatists may have responded this way. It has been previously suggested that police performance should be particularly important to people in contexts where personal security is of central concern (Tankebe, 2009). In other words, concerns about safety and security can lead to an enhanced focus on the ability of police to be effective. Tankebe’s (2009) empirical research in Ghana supports this. He finds that people living in a high-crime context are less concerned about procedural justice and more concerned with the ability of the police to actually prevent and control crime. It has also been noted that resettlement following the ‘strains of war and refugee trauma, as well as resettlement issues’ can contribute to greater hardship experienced by ethnic minority migrants (see Meredyth et al., 2010; Sargeant et al., 2014). These hardships can include a greater likelihood of experiencing poverty, victimization, and crime. We suggest that such experiences might influence a distinct and distrustful attitude towards authorities such as the police, leading to an ‘us’ versus ‘them’ stance towards society (that is, a separatist identity). Further, such experiences and views may have contributed to separatists focusing less on procedural justice and more on whether the police are effective in dealing with their safety concerns.

Whatever the explanation of this particular finding, we should stress that the Group Value Model cannot explain the entire pattern of results observed in our study. The model has difficulty explaining the absence of interaction effects between the superordinate identity and procedural justice variables, and between the subordinate identity and procedural justice variables. Our results revealed that procedural justice mattered equally to those who identified both strongly and weakly with the superordinate nation-state of Australia and to those who identified both strongly and weakly with an ethnic subordinate group. Given that Huo (2003) found that both superordinate and subordinate identity interacted with procedural justice when predicting people’s willingness to accept police decisions, we expected to find similar effects when predicting intentions to cooperate. The sample size in the current study is very large and has the power to detect subtle differences. Yet the fact that there was an absence of interaction effects between these variables challenges some of the Group Value Model’s assertions, at least in an Australian law enforcement context and when predicting people’s intentions to cooperate with police.

Our results in fact suggest that most people in Australian society value fair and respectful treatment from authorities as well as effective policing, regardless of whether the authority is a representative of their own group or not. Why might we not have obtained some of our predicted effects? One possible explanation relates to the way in which people identify with groups in Australia. Perhaps strong identification based on race/ethnicity is not as salient to people in Australian culture as it is in other countries, such as the USA where Huo et al.’s (1996) and Huo’s (2003) studies were undertaken. People living in Australia may also not see identifying with multiple groups as being in conflict. Our results suggest that Australians can identify with an ethnic subgroup but that they value their Australian identity more so. As can be seen in Table 3, compared with the ‘Australian’ and ‘European/Americans’, the ‘Asian’ and ‘African/Middle Easterners’ appeared to place importance on their ethnic subgroup identity, but all respondents also scored very high on the superordinate identity scale, suggesting strong affiliation with Australia (with little variation across the four groups). Perhaps we might have obtained different results had respondents’ ethnic identification been more important to them than their Australian identity. Second, the identity measures used in our study may have influenced our outcomes. Our superordinate identity measure assessed identification with the nation-state of Australia. Huo et al.’s (1996) workplace study assessed respondents’ identification not with the USA but rather with their workplace. Had our superordinate identity variable instead been a measure of identity with a group viewed as an ‘outgroup’ (for example, Anglo-Saxon group rather than Australia), the study might have produced different findings.

So where do our findings leave the Group Value Model and the study of identity processes in the context of procedural justice research? Given the significant main effects of identity on cooperation, and the fact that the separatist identity variable moderated the effect of procedural justice on intentions to cooperate, we suggest the study of identity in relation to procedural justice is important and warrants continued investigation. However, our findings also indicate that procedural justice research may be better served if scholars consider other variables not studied here, and the impact that they may have on cooperative behaviour. The Group Value Model predicts that identity plays a major role, if not the key role, in determining whether people will subsequently respond to procedural justice and cooperate with authorities. We observed quite low variation in our models. This suggests that understanding intentions to cooperate among different groups in society requires consideration of many more factors than those we have included in our models. In fact, it is highly likely that a variety of social, ideological, and contextual issues are at play when determining whether people will cooperate or not with police. Indeed, new research is beginning to consider issues such as the moral alignment citizens share with policing goals, or the important role of neighbourhood context, in explaining perceptions of police (for example, Jackson et al., 2012; 2013; Kochel, 2012; Weitzer, 2010). Future research should weigh up the relative importance of identity for predicting cooperation when considered alongside other additional factors not included in our study.

Conclusion

The challenge for policing agencies around the world is to identify and understand the main drivers that shape people’s willingness to cooperate voluntarily with police. This can be a particular challenge in multicultural nations where different norms, values, and identities may mean police–citizen interactions can be difficult to navigate. One of the arguments raised by opponents of multiculturalism is that identification with ethnic subgroups can threaten social unity (Schlesinger, 1992). Our results contradict this argument. The absence of an interaction effect between our subordinate identity and procedural justice variables suggests that fair and respectful treatment by police can encourage a general intention to cooperate with authorities without requiring that individuals shun their ethnic subgroup. The fact that those who identify strongly with a subordinate ethnic group identity respond equally as well to procedural justice policing as those who do not identify with an ethnic minority group suggests that police are unlikely to encounter difficulties in all contacts with minorities. Our findings instead suggest that it is separatists – those who place a great deal of emphasis on retaining a separate culture distinct from Australian culture – who may pose a greater concern for police, and for whom procedural justice policing approaches may not work. It is not having an ethnic identity per se that appears to be the problem, but instead the degree of marginalization or disengagement one chooses to adopt from mainstream society that may reduce engagement and a willingness to cooperate with police (for similar findings, see Murphy and Cherney, 2012).

Our results suggest that, for those who feel marginalized from mainstream society, emphasizing successes in tackling crime in the community may prove a more effective policing strategy for promoting cooperation. In other words, police may benefit substantially by communicating with disengaged minority groups about the strategies and successes they have had in dealing with crime in local neighbourhoods. If police can communicate that they are doing a good job in dealing with problems that concern citizens, these messages may strike a chord with those who may be more instrumentally focused. This does not mean that police should withdraw the use of procedural justice from those groups they believe may be marginalized. Treating people fairly and with respect, being neutral in decisions, and providing people with voice is an entitlement that all have the right to receive. Such an approach may also assist in strengthening minorities’ identification and affiliation with mainstream society (see Bradford et al., 2014), thereby making procedural justice a more effective strategy in the long term.

Footnotes

Funding

We would like to acknowledge the funding support from the Australian Research Council (Grant Numbers: DP1093960 and RO700002).

Notes

References

Ajzen

(1985) From intention to action: A theory of planned behaviour. In: Kuhl

Beckmann

(eds) Action Control: From Cognition to Behavior. New York: Springer, 11–39.

American Association for Public Opinion Research (2011) Standard Definitions: Final Dispositions of Case Codes and Outcome Rates for Surveys. Ann Arbor, MI: American Association for Public Opinion Research.

Australian Bureau of Statistics (2011) Census QuickStats. URL (accessed 9 May 2015): http://www.abs.gov.au/websitedbs/censushome.nsf/home/Census.

Bradford

(2014) Policing and social identity: Procedural justice, inclusion, and cooperation between police and public. Policing and Society 24(1): 22–43.

Bradford

Jackson

Stanko

(2009) Contact and confidence: Revisiting the impact of public encounters with the police. Policing and Society 19(1): 20–46.

Bradford

Murphy

Jackson

(2014) Officers as mirrors: Policing, procedural justice and the (re)production of social identity. British Journal of Criminology 54(4): 527–550.

Brown

Benedict

(2002) Perceptions of the police: Past findings, methodological issues, conceptual issues and policy implications. Policing: An International Journal of Police Strategies and Management 25(3): 543–580.

Calhoun

(ed.) (1994) Social Theory and the Politics of Identity. Oxford: Blackwell.

Glazer

(1997) We Are All Multiculturalists Now. Cambridge, MA: Harvard University Press.

10.

Haslam

(2004) Psychology in Organizations. London: Sage.

11.

Hinds

Murphy

(2007) Public satisfaction with police: Using procedural justice to improve police legitimacy. Australian and New Zealand Journal of Criminology 40(1): 27–42.

12.

Huo

(2003) Procedural justice and social regulation across group boundaries: Does subgroup identity undermine relationship-based governance? Personality and Social Psychology Bulletin 29(3): 336–348.

13.

Huo

Smith

Tyler

Lind

(1996) Superordinate identification, subgroup identification, and justice concerns: Is separatism the problem; Is assimilation the answer? Psychological Science 7(1): 40–45.

14.

Jackson

Bradford

(2010) What is trust and confidence in the police? Policing 4(3): 241–248.

15.

Jackson

Bradford

Hough

Myhill

Quinton

Tyler

(2012) Why do people comply with the law? Legitimacy and the influence of legal institutions. British Journal of Criminology 52: 1051–1071.

16.

Jackson

Bradford

Stanko

Hohl

(2013) Just Authority: Trust in the Police in England and Wales. London: Routledge.

17.

Kochel

(2012) Can police legitimacy promote collective efficacy? Justice Quarterly 29(3): 384–419.

18.

Lind

Tyler

(1988) The Social Psychology of Procedural Justice. New York: Plenum Press.

19.

Manstead

(2000) The role of moral norms in the attitude–behavior relation. In: Terry

Hogg

(eds) Attitudes, Behavior, and Social Context: The Role of Norms and Group Membership. Mahwah, NJ: Lawrence Erlbaum Associates Publishers, 11–30.

20.

Mazerolle

Wickes

Cherney

Murphy

Sargeant

Zahnow

(2012) Australian Community Capacity Study Wave 3 Brisbane, Wave 1 Melbourne: Technical Report No 3. Institute for Social Science Research, University of Queensland.

21.

Mason

(2009) Violence against Indian students in Australia: A question of dignity. Current Issues in Criminal Justice 21(3): 461–465.

22.

Meredyth

McKernan

Evans

(2010) Police and Vietnamese-Australian communities in multi-ethnic Melbourne. Policing 4(3): 233–240.

23.

Murphy

(2013) Policing at the margins: Fostering trust and cooperation among ethnic minority groups. Journal of Policing, Intelligence and Counter Terrorism 8(2): 184–199.

24.

Murphy

Cherney

(2011) Fostering cooperation with police: How do ethnic minorities in Australia respond to procedural justice-based policing? Australian and New Zealand Journal of Criminology 44(2): 235–257.

25.

Murphy

Cherney

(2012) Understanding cooperation with police in a diverse society. British Journal of Criminology 52(1): 181–201.

26.

Murphy

Hinds

Fleming

(2008) Encouraging public cooperation and support for police. Policing and Society 18(2): 136–155.

27.

Murphy

Mearns

(2010) The 2009 Crime, Safety and Policing in Australia Survey: Survey Methodology and Preliminary Findings. Alfred Deakin Research Institute Working Paper No. 17. Geelong, VIC: Alfred Deakin Research Institute.

28.

Murphy

Cherney

Wickes

Mazerolle

Sargeant

(2012) The Community Capacity Survey – Face-to-face ethnic minority interviews: Methodology and Preliminary Findings. ARC Centre for Excellence in Policing and Security. Brisbane: Griffith University.

29.

Oliveira

Murphy

(2014) Race, social identity, and perceptions of police bias. Race and Justice. Online first doi: 10.1177/2153368714562801.

30.

Pickett

Chiricos

Golden

Gertz

(2012) Reconsidering the relationship between perceived neighborhood racial composition and whites’ perceptions of victimization risk: Do racial stereotypes matter? Criminology 50(1): 145–186.

31.

Poynting

(2006) What caused the Cronulla riot? Race and Class 48(1): 85–92.

32.

Reisig

Bratton

Gertz

(2007) The construct validity and refinement of process-based policing measures. Criminal Justice and Behavior 34(8): 1005–1028.

33.

Reisig

Tankebe

Mesko

(2012) Procedural justice, police legitimacy, and public cooperation with the police among young Slovene adults. Journal of Criminal Justice and Security 14(2): 147–164.

34.

Rose

(1993) Of every hue and caste: Race, immigration, and perceptions of pluralism. Annals of the American Academy of Political and Social Science 530: 187–202.

35.

Rutherford

(ed.) (1998) Identity: Community, Culture, Difference. London: Lawrence & Wishart.

36.

Sargeant

Murphy

Cherney

(2014) Ethnicity, trust and cooperation with police: Testing the dominance of the process-based model. European Journal of Criminology 11(4): 500–524.

37.

Schlesinger

(1992) The Disuniting of America: Reflections on a Multicultural Society. New York: Norton.

38.

Skogan

(2006) Asymmetry in the impact of encounters with police. Policing and Society 16(2): 99–126.

39.

Sunshine

Tyler

(2003) The role of procedural justice and legitimacy in shaping public support for policing. Law and Society Review 37(3): 513–548.

40.

Tajfel

Turner

(1986) The social identity theory of intergroup behavior. Psychology of Intergroup Relations 81: 7–24.

41.

Tankebe

(2009) Public cooperation with the police in Ghana: Does procedural fairness matter? Criminology 47(4): 1265–1293.

42.

Tyler

(1990) Why People Obey the Law: Procedural Justice, Legitimacy, and Compliance. New Haven, CT: Yale University Press.

43.

Tyler

(2005) Policing in black and white: Ethnic group differences in trust and confidence in the police. Police Quarterly 8(3): 322–342.

44.

Tyler

Huo

(2002) Trust in the Law: Encouraging Public Cooperation with the Police and Courts. New York: Russell Sage.

45.

Weitzer

(2010) Race and policing in different ecological contexts. In: Rice

White

(eds) Race, Ethnicity, and Policing: New and Essential Readings. New York: New York University Press, 118–139.