By 
 

Robert P. Peacock 

 
 
 
 
 

 
 

 

A DISSERTATION 

 

 

 

 

APPLYING TYLER’S PROCESS-BASED MODEL OF LEGITIMACY  

IN A DEVELOPING COUNTRY 

Submitted to 

Michigan State University 

in partial fulfillment of the requirements 

for the degree of 

Criminal Justice – Doctor of Philosophy 

2018 

 

 

 

 
 
 

 

 

 

 

 
 
 
 

 

APPLYING TYLER’S PROCESS-BASED MODEL OF LEGITIMACY  

IN A DEVELOPING COUNTRY 

 

ABSTRACT 

By 
 

Robert P. Peacock 

 

 

Few national attributes are likely to predict countries with track records for good 

governance and human rights than their citizens’ judgments on the fairness and legitimacy of 

their law enforcement institutions.  Since Tom Tyler first articulated his theoretical model of 

police legitimacy in 1990, a wave of scholars in North America, Western Europe, and Australia 

continue to demonstrate that the public’s perceptions of procedural justice are the strongest 

predictor of their judgments on the agency’s legitimacy which then predicts the public’s 

willingness to comply with the law or cooperate with the police.  Countries across the globe 

increasingly turn to the Tylerian theory of police legitimacy to model these judgments and guide 

vital procedural justice reform.  Unfortunately, the theory of police legitimacy has failed to 

predict the willingness to cooperate with the police in countries still confronted with serious 

development issues.  Tests of Tyler’s theory of police legitimacy in Ghana, Jamaica, and China 

failed in finding a significant role for Tyler’s construct of legitimacy in the process-based model.  

This study supports past studies testing the full model in developing countries in finding that 

legitimacy did not mediate the relationship between perceived procedural justice and willingness 

to cooperate with the police in Ukraine.   

New variables added to the model test Tankebe’s proposition that the weakness in the 

Tylerian model is likely due to the lack of a normative commitment between the public and law 

enforcement agencies in developing countries.  The finding that postmaterialism, as an indicator 

of normative values, significantly predicted the willingness to cooperate with police supports 

 

Tyler’s argument that shared normative values drive the significant paths in the model.  Despite 

the existence of this normative pathway to associating with cooperation with the police, 

trustworthiness did not mediate the model.  Countries plagued by the twofold scourge of 

systemic corruption and poor governance appear to diverge from developed countries in how 

their citizens evaluate law enforcement institutions.  This may explain why legitimacy in 

developing countries fails to reflect the shared values that Tyler argues favor or rebuff 

cooperation with the police.   

The study’s data analysis found a very high correlation and lack of divergence between 

respondents’ evaluations of the two antecedent variables in the model, procedural justice and 

police effectiveness.  In countries where the police are indifferent to serving the public and are 

closely associated with organized crime and corruption, the public appears to struggle to 

determine if an officer doing their job (or in the opposite case, avoiding their responsibilities) 

should be evaluated in terms of effectiveness or procedural justice.  The overlap between these 

two judgements appears to be much greater than in North America and Western European tests 

of the model.  This may explain why Ukraine continues the findings in developing countries that 

procedural justice and effectiveness have the strongest correlation among all the variables. 

Finally, the adoption of multi-group confirmatory factor analysis to a test of the Tylerian 

model found only a modest level of bias in how respondents interpret test items of procedural 

justice across different demographic groups in Ukraine.  The most important of the findings of 

partial invariance across different demographic groups is in regard to direct experience with the 

police.  The study supports the current de-facto practice in legitimacy research that does not limit 

survey samples to citizens with recent interaction with the police but assumes that judgments of 

procedural justice can also be based on global attitudes.    

 

 

 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

 
 
 
 
 

 

Copyright by 
ROBERT P. PEACOCK 
2018 

 

 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

 
 
 

 

I lovingly dedicate this dissertation to my wife and best friend Tetyana  

and our incredible children Alex and Maria. 

 

 

 

v 

 

 

I would like to thank my graduate advisor and committee chairperson Dr. Sanja Kutnjak 

ACKNOWLEDGEMENTS 

 

Ivković.  She has been a constant source of support and guidance from the first day of my 

doctoral studies to the defense of my dissertation.  I also would like to thank my dissertation 

committee members, Drs. Edmund McGarrell, Meredith Gore, and Joe Hamm for their patience 

and flexibility in working with a dissertation writer located halfway around the world.   

Not long after the public survey for this dissertation was completed in Ukraine, the 

Maidan Protests began the start of a revolution that would dramatically change my own future as 

well.  A carefully laid plan to complete my dissertation while managing a sleepy Department of 

Justice project office of eight employees in Kyiv was turned upside down.  By the start of 2015, I 

would be overseeing an office that at one point exceeded 320 full-time and part-time advisers, 

trainers, and staff tasked with independently recruiting, selecting, training, and standing up a 

national patrol police department that would reach 18,000 new line and command officers by 

2018.  Over a 12-month period, my project team was recruiting, training, and helping stand up a 

police agency the size of the Richmond (Virginia) Police Department every two weeks.  I am 

grateful to the U.S. Department of Justice for giving me this opportunity and then providing me 

with the additional management support and flexible schedule to allow me to return to 

completing my doctorate. 

Finally, this dissertation would not exist without the support of my wife Tetyana and our 

two beautiful children Maria and Alex.  Their sacrifices over the eight years of my doctoral 

studies, include following me to the other side of the globe and back twice, gave me the 

opportunity to complete this dissertation.  I now look forward to fulfilling my end of our pact as 

we make our last move to find our future home and open a new chapter in our lives.  

 

 

vi 

 

TABLE OF CONTENTS 

 
LIST OF TABLES  
 
LIST OF FIGURES 
 
CHAPTER I Introduction 
 
 
 
 
 
CHAPTER II Literature Review  
 
 

Statement of the Problem 
Significance of the Study 
Context of the Study 
Research Questions 

Social Control Roots 
Legitimacy Scholarship Redux 
Birth of Procedural Justice 
Legitimacy as a Measurable Construct 
Classic Process-based Model of Legitimacy 
Normative vs. Instrumental Predictors 

Development of the Tylerian Model 
 
  
  
  
 
 
Criticism of the Process-based Model 
 
 
 
 
 
 
 
Normative Values and Legitimacy 
 
 

Role of Postmaterialism 
Role of Fear of Crime 

Uncertainty on Limiting Data Sample 
Debate on Authoritarianism in the Model 
Concern for the Validity of Model Indicators 
 
 
Additional Model Antecedents 
The Weak Empirical Case in Developing Countries 

  Convergence 
  Divergence 

 
 
 
CHAPTER III Methodology  
 
 

Procedural Justice 
Police Effectiveness 
Postmaterialism 
Fear of Crime 
Police Legitimacy 

Survey Data 
Independent and Mediating Variable Measurement 
 
 
 
 
 
Dependent Variable Measurement 
 
Cooperation with the Police 
Analytical Plan 

 
 
 
 
 
 
 
 
 
 

 

 
 
 
 
 
 

 
 
 
 
 
 
 
 
 

 

 

vii 

ix 

x 

1 
3 
5 
6 
9 

11 
11 
11 
13 
16 
16 
18 
 21 
23 
24 
27 
28 
29 
29 
31 
31 
34 
36 
42 

46 
46 
49 
49 
49 
49 
50 
50 
50 
51 
51 

 

Exploratory Factor Analysis 
Separating the Two Legitimacy Indexes 
Correlation Coefficients 

CHAPTER IV Results and Discussion 
 
 
 
 
 
 
 
 
 
 
 
 
 

54 
54 
54 
58 
59 
62
63 
67 
Does Procedural Justice Predict Cooperation with the Police? 
Does Legitimacy Mediate the PJ – Cooperation Relationship? 
67 
Does Legitimacy Mediate the Effectiveness – Cooperation Relationship?  70 
70 
 
 
73 
79 
Do Postmaterialism Values Predict Cooperation with the Police? 
81 
Does Fear of Crime Predict Cooperation with the Police?   
81 
82 
84 
85 
85 
87 
87 
90 
90 
90 

Preliminary Tests  
 
 
 
Latent Variable Analysis 
Path Analysis 
 
 
 
 
 
 
 
Does Interaction with the Police Bias Perceptions of Procedural Justice? 
 
 
 
 
Control Variables’ Normative-Path Analyses 
 
 
 
 

OLS Comparison 
Instrumental vs. Normative Concerns 

Police Interaction 
Gender 
Age 
Higher Education 

 
 
 
 
 
 
 
 
 
 
CHAPTER V Conclusion 

Police Interaction 
Gender 
Age 
Higher Education 

95 
95 
97 
98 
100 
102
104 
105 
105 
106 
107 
108 

110 
111 
113 

123 

Postmaterialism and A Lack of Normative Commitment to Police 
Supporting the Consensus on Validity Problems 
Blurred Distinction Between Effective and Just Procedures 
Adding Effectiveness to the Legitimacy Construct in Developing Countries 
The Future Role of Obligation to Obey 
Invariance Supports General Survey Sample 
Current Study’s Limitations 
 
 
 
 

Absent Variables 
Corruption as a Control 
Need for Mixed Methods 
Vicarious or Process-based Model 

APPENDIX A: Test Items from Sunshine & Tyler 
APPENDIX B: Ukrainian Survey Instrument 

 

 
 
 
 
 
 
 
 
  
 
 
APPENDICES 
 
 
 
REFERENCES 
 

 

 

viii 

55 

57 

58 

59 

66 

71 

74 

82 

85 

85 

86 

 

LIST OF TABLES 

 

 
 
Table 1:  Descriptive Characteristics of Test Items 

Table 2: EFA Factor Loadings for Predictor Variables 

Table 3: Comparison of Cronbach’s Alphas in Ukraine versus U.S. Studies 

Table 4: Correlations Between Key Indicators  

Table 5: Variance Inflation Factor  

Table 6: OLS Regression Model of Police Legitimacy 

Table 7: Normative versus Instrumental Predictors of Legitimacy 
 
Table 8: Metric Fit Indices for Invariance Tests Across Interaction with Police 

Table 9: Metric Fit Indices for Invariance Tests Across Gender 
 
Table 10: Metric Fit Indices for Invariance Tests Across Age 
 
Table 11: Metric Fit Indices for Invariance Tests Across Education Level 
 

 

 

ix 

 

LIST OF FIGURES 

 
 
 
 

Figure 1: 2013 Map of Ukraine 

Figure 2: Classic Process-based Model of Police Legitimacy 

Figure 3: Normative Versus Instrumental Predictors of Police Outcomes 

Figure 4: Postmaterialist Results over Three Decades 

Figure 5: Relative Postmaterialism by GNP/Capita Across Countries 

Figure 6: Conceptual Model with Post-Materialism and Fear of Crime 

Figure 7: Confirmatory Factor Analysis 
 
Figure 8: Path Analysis 
 
Figure 9: Normative/Instrumental Values Ratio Continuum 
 
Figure 10: Police Interaction as a Control in the Normative-Path Model  
 
Figure 11: Gender as a Control in the Normative-Path Model 
 
Figure 12: Age as a Control in the Normative-Path Model 
 
Figure 13: Higher Education as a Control in the Normative-Path Model 

7 

19  

23 

39 

40 

42 

64 

65 

77 

88 

90 

91 

92 

 

 

 

x 

 

CHAPTER I 

Introduction 

Trust in the police and other public order institutions has fallen dramatically across the 

globe over the last half-century (Catterberg & Moreno, 2005; Inglehart, 1999; OECD, 2013).  

Law enforcement institutions often ignore the decline in trust until a singular case of misconduct 

leads to public scandal or worse.  The 2011 London riot, the 2013 Maidan Protests in Ukraine, 

the 2015 Bharuch riots in India, the violent protests in Charlotte, North Carolina in 2016 or the 

Teheran Riots in 2018 were civil disturbances that ignited in communities or societies that did 

not view the enforcement agencies as fair, just, or legitimate.   

In developed countries, the growing perception of police misconduct has been fueled in 

part by the rapid growth in the availability of officer, vehicle, and third-party video surveillance 

of police-public interactions.  As officer relations with communities come under the microscope, 

police agencies have found that the traditional actions used to address public dissatisfaction, such 

as lowering crime rates or other displays of police effectiveness, have not reversed the decline in 

police legitimacy (Sherman, 2001).  In fact, the nearly two-decade drop in violent crime in the 

developed world1 at the turn of the current century has failed to improve public perceptions of 

law enforcement legitimacy in the countries that benefited from the significant reduction in 

violence (Tseloni, Mailley, Farrell &Tilley, 2010).  As a result, police agencies increasingly have 

been forced to look beyond traditional answers to declining public support. 

The theoretical seed for a paradigm shift in the criminal justice literature was planted in 

the late 1970s when Tom Tyler and Robert Folger (1980) made use of a new dispute settlement 

                                                
1 No consensus exists in the social sciences on the categorization of countries based on political or economic 
development, this study will compare Ukraine with a broad range of countries but define the category of developed 
country status to members of the Organization for Economic Development (OECD). 

 

 

1 

 

construct, procedural justice, to measure the public’s judgments on citizen-police interactions to 

predict their satisfaction with the agency.  Though scholars routinely change or add test items, 

most police legitimacy models tested around the globe are based on the classic, process-based 

model that Tyler (1990) introduced in Why People Obey the Law.  Tests of the Tylerian process-

based model of police legitimacy in North America, Europe, and Australia found that the most 

important factors shaping public judgments were not the traditional measures of police 

effectiveness, but rather the residents’ views on the overall quality of decision-making and the 

perception of the police’s treatment of others in the community (Hinds & Murphy, 2007, Hough, 

Jackson & Bradford, 2013; Tyler, 2002).   

The Tylerian process-based model argued that the public perception on the fairness of the 

procedures used by law enforcement is the most influential predictor of the public’s willingness 

to comply with the law and cooperate with officers through the mediating construct of police 

legitimacy (Tyler, 2003).  The timing could not have been more auspicious for researchers and 

practitioners unable to explain the continued decline in public satisfaction over a period that 

many agencies viewed as their most effective in decades.  By placing the focus on each officer’s 

interaction with the public, the Tylerian process-based model offered law enforcement 

institutions an alternative paradigm for reform that breaks from the traditional police emphasis 

on lowering the crime rate or increasing the number of arrests in order to improve public 

perceptions of the agency.  

As legitimacy theory developed, the Tylerian process-based model linking procedural 

justice with agency outcomes became an informal roadmap for governments and police agencies 

in the US, Western Europe, and Australia.  Law enforcement agencies began to focus on public 

perceptions of fair and impartial police practices, particularly in marginalized communities, so as 

 

 

2 

 

to improve public judgments on legitimacy (Quinlan & Ramirez, 2014; Wagner, 2015).  The 

Chicago Police Department, the UK’s Greater Manchester Police, and the Queensland Police 

Service in Australia are three examples of agencies in developed countries using procedural 

justice initiatives to improve institutional legitimacy in the communities they police (Mazerolle, 

Antrobus, Bennett & Tyler, 2013; Skogan, Craen & Hennessy, 2014; Wheller, Quinton, Fildes & 

Mills, 2013).   

Statement of the Problem  

The classic Tylerian model of legitimacy first tested in metropolitan Chicago has grown 

to become one of the most influential theories in the fields of criminology and criminal justice 

(Tyler, 1990; Tankebe et al., 2016).  Worden and McLean (2017a) argue that the “unprecedented 

attention” of government officials, police executives, and the media to police-community 

relations has led to an “explosion” of empirical research on police legitimacy in developed 

countries (pp. 480-481).  A secondary shock wave of this research explosion can be seen in 

developing countries over the last decade (Bradford, Huq, Jackson & Roberts, 2013; Czapska, 

Radomska, & Wójcik, 2016; Jackson, Asif, Bradford & Zakar, 2014; Karakus, 2017; Reisig, 

Tankebe & Mesko, 2014; Rinehart Kochel, Parks & Mastrofski, 2013; Sun, Wu, Hu, & Farmer, 

2017; Tankebe, 2008; Tsushima & Hamai, 2015).  Unfortunately, while theoretical research 

designed for the United States rarely needs adaptation to function in other advanced 

democracies, the more dramatic changes in the cultural setting found in developing countries can 

significantly alter the relationships and results produced by the same theories (Bennett & Flavin, 

1994). 

The growth in researcher and practitioner use of the Tylerian theory of police legitimacy 

is built on the empirical support for the process-based model in the United States, Western 

 

 

3 

 

Europe, and Australia (Tyler, 2002, Hinds & Murphy, 2007; Murphy, 2009; Hough et al., 2013). 

In contrast, the three tests of the Tylerian process-based model that have been conducted in 

developing countries (Jamaica, Ghana, and China) found that the legitimacy construct did not 

mediate the relationship between procedural justice and cooperation with law enforcement 

(Reisig & Lloyd, 2009; Sun et al., 2017; Tankebe, 2009).  Moreover, scholars that limited their 

research to the strongest relationship in the model (procedural justice predicting legitimacy) or 

developed an alternative construct of legitimacy for the model, report that in Pakistan, South 

Africa, and Turkey the role of procedural justice was weaker than found in the previous studies 

based largely in Western Europe and North America (Bradford et al, 2013; Jackson et al., 2014; 

Karakus, 2017).   

The limited support for the classic Tylerian process-based model of legitimacy in 

developing countries suggests that there could be systemic differences between developed and 

developing countries that significantly alter the model’s key relationships or require an entirely 

different model for developing countries.  This uncertainty has led to ambiguity among 

researchers on how the theory should be adapted outside developed countries.  Tankebe (2009) 

proposed that citizens interacting with the police in developing countries may prioritize other 

normative and instrumental goals over fair treatment.  After finding weak support for a 

legitimacy construct in a developing country, Tankebe, Reisig, and Wang (2016) advocated for 

further research into the socio-political variables that may alter the police legitimacy model in 

developing countries.  This study in Ukraine seeks to test two variables in the model that may 

shed light on the failure of the legitimacy indicator that Tyler (2003) has hypothesized as 

mediating the shared values that are the basis of the prediction that procedural justice will predict 

agency legitimacy and the willingness to cooperate with the police. 

 

 

4 

 

Significance of the Study  

In developed countries, the police legitimacy model holds considerable promise as tool 

researchers can use to identify the changes in police training, policies, or procedures that may 

increase the agency’s legitimacy as well as the community’s willingness to comply and 

cooperate with law enforcement.  This same promise has been undermined in developing 

countries by much weaker empirical support for the legitimacy measure and the process-based 

model.   

Legitimate criminal justice institutions are widely perceived in development studies as a 

key precursor to stable and prosperous democracies.  Unfortunately, practitioners and researchers 

seeking to model police legitimacy in developing countries also face unique challenges to adopt 

constructs and relationships that were established through empirical studies limited to OECD 

member countries.  In preparing an assistance program focused on police reform, the 

Development Bank of Latin America recently stated that “the best place to begin building trust 

and legitimacy in the system is not readily apparent” (Ortega, 2016, p. 2).  High crime rates, low 

police wages, institutional corruption, human rights abuses, and limited trust in the effectiveness 

of the police are a few of the challenges facing police researchers and reformers in the 

developing world. 

Whether a Latin American development specialist or a Cambridge criminologist working 

in Ghana, the key challenges are largely the same in adapting a process-based model of police 

legitimacy to countries with transitioning or emerging economies.  Comparative studies suggest 

that the respondents in developing countries will have different priorities driven in part by their 

greater need to prioritize immediate survival over broader societal goals (Inglehart & Baker, 

2000; Rogowski, 1974).  In addition, researchers and practitioners in developing countries must 

 

 

5 

 

account for a highly skeptical public that does not trust official crime statistics or the uncertain 

and sometimes hostile police response to their calls for service.   

This dissertation will be the first test of the process-based model of police legitimacy in a 

republic of the former Soviet Union, but the significance of the study rests on whether the 

hypothesized variables serve as a modest first step to understanding the failure of the legitimacy 

indicator in developing countries to predict the willingness to cooperate with police in countries 

still transitioning in terms of prosperity and good governance.  Specifically, the study introduces 

two measures (postmaterialism and fear of crime) that are hypothesized as reflecting normative 

values that, similar to the legitimacy construct, could predict the police outcome variable in a 

new model to test the theory of police legitimacy.  If the hypothesized measures of normative 

values have a significant association with the willingness to cooperate with the police, the 

finding would infer that the lack of shared normative values with police explains the weak role of 

police legitimacy indicators in tests of the classic Tylerian model in developing countries. 

Context of the Study 

The research context for this explanatory model of the Tylerian theory of police 

legitimacy is Ukraine (see Figure 1); a developing country noted for weak public trust in its 

systemically corrupt government institutions (Darden, 2001; Beck, 2005).  At the time of this 

study’s 2013 data collection, Ukraine was governed by the country’s fourth president, Viktor 

Yanukovych, who would be ousted by 2014 protests against his increasingly authoritarian and 

corrupt administration.  The Ukrainian police (Militia) in 2013 consisted of approximately 

150,000 officers and support personnel2.  Besides municipal divisions, the militia were divided 

                                                
2 In a continuation of Soviet policies, the size of the Ukrainian militia in 2013 was classified. The Chief of the new 
National Police, Khatia Dekanoidze, in 2015 interviews with the author still struggled to find an accurate number of 
personnel from the former Militia largely because of the uncertain organizational boundaries with the larger 
Ministry of Internal Affairs.    

 

 

6 

 

into functional components, such as investigations, district inspectors, juvenile police, traffic 

police, privately-funded security police, and a non-sworn patrol (MIA, 1996).  

Figure	1:	2013	Map	of	Ukraine	

Source: Nations Online Project (http://www.nationsonline.org/oneworld/map/ukraine-administrative-map.htm). 

 
 

As in the Soviet period, the primary role of the post-Soviet militia was protecting the state 

(Beck, 2005).  This meant officers spent most of their days at a desk or on guard duty while the 

traffic and foot patrol units that most often interacted with the public were primarily composed 

of poorly-trained, non-sworn officers (Beck, Barko, & Tatarenko, 2002; Beck & Robertson, 

2009, Shelley, 1999).  The organization barely differed from the militia in the former Soviet 

 

 

7 

 

Republic of Ukraine and the legal underpinning for the structure remained a 1990 Law on Militia 

that actually preceded the independence of Ukraine in 1991.3   

A defining feature of the militia was systemic corruption that, beyond petty bribery, 

involved collaboration among employees or between agencies to extort money from the public 

(Karklins, 2005; Shelley, 1999).  Officers in this period typically paid their way into the ministry 

as well as each new position, while public bribes were often the largest portion of an officer’s 

income (Beck, 2005; Gerasymenko, 2011; Sahakyan, 2015).  Transparency International (2013) 

reported just prior to the study’s survey that nearly 85 percent of the public considered the militia 

to be corrupt or highly corrupt. 

Locating the study in Ukraine allows for the second test of the full Tylerian model in a 

country located in the bottom half of the rankings on institutional corruption and ineffective 

government institutions (TI, 2013; World Bank, 2016).  The cornerstone empirical cases for the 

validity of the Tylerian model were based on surveys in countries that consistently rank in the 

top 15% of the Corruption Performance Index (CPI) and the World Governance Indicators 

(WGI).  Other empirical tests of the model’s core relationships were in OECD member Turkey 

and non-OECD countries China, Ghana and Jamaica, which consistently score in the top 50% of 

the CPI and WGI rankings (Karakus, 2017; Reisig & Lloyd, 2009; Sun et al., 2017; and 

Tankebe, 2009). Akinlabi’s (2018) tests of the full Tylerian model in Nigeria is the first study in 

a country located in the bottom half of the rankings of institutional corruption and ineffective 

government institutions.    

As a result, this study’s data sample comes from a survey that was administered to a 

public that had every reason to view the existing law enforcement as corrupt, ineffective, and 

                                                
3 The Ukrainian Parliament passed the Law on Police that liquidated the Militia in July, 2015. 

 

 

8 

 

largely untrustworthy.  Officers actively participated in crimes that ranged from selling drugs 

seized in police raids to shaking down businesses for protection money (Peacock & Cordner, 

2016).  Moreover, an officer’s success in the organization was dependent on the use of 

connections or bribes to obtain a promotion to a more lucrative position.  Therefore, it is not a 

surprise that Hough and colleagues (2014) found in a comparison of public surveys over this 

period that Ukraine stood out in Europe for the high levels of bribes and low trust in the police.  

Although researchers have not previously reported a measure of agency legitimacy in Ukraine, 

past surveys have repeatedly demonstrated that the majority of the Ukrainian public viewed the 

militia as ineffective and untrustworthy (Beck & Chistyakova, 2002; KHPG, 2013; Kulik, 1994).  

The level of mistrust appeared to be at a nadir at the time of this study’s data collection, when the 

Ukrainian Academy of Science reported that less than nine percent of the public responded in a 

survey that they either “completely trusted” or “mostly trusted the militia” (OSCE, 2013). 

Research Questions 

The objective of this study was to test the influence of measures of societal insecurity in 

the Tylerian theory of police legitimacy as part of the larger goal of adapting the process-based 

model for use in developing countries.  The study addressed the following six research questions 

that are important steps to the further development of the process-based model in developing 

countries. 

Research Question #1. Do perceptions of procedural justice predict cooperation with the 

police in a Ukraine? 

Research Question #2. Do perceptions of police legitimacy mediate the relationship 

between measures of procedural justice and public cooperation with the police in 

Ukraine? 

 

 

9 

 

 

 

Research Question #3. Do perceptions of police legitimacy mediate the relationship 

between measures of effectiveness and public cooperation with the police in Ukraine? 

Research Question #4. Does a post-materialism index predict cooperation with the 

police in Ukraine?  

Research Question #5. Does fear of crime predict cooperation with the police in 

Ukraine? 

Research Question #6. Does measurement invariance hold for the procedural justice 

indicator regardless of whether the respondent had recent experience with the police? 

 

 

10 

 

CHAPTER II 

Literature Review 

Development of the Tylerian Model 

Tom Tyler (1990; 2003) developed the process-based theory of police legitimacy at the 

turn of this century to model voluntary compliance with the law and assistance to the police.  

Traditionally, the criminal justice system was based on a deterrence model that views 

compliance and cooperation with law enforcement as an instrumental calculus by the public to 

minimize costs (punishment) through obeying the law or maximize a reward (eliminating 

disorder in your neighborhood) through cooperation with the police.  Tyler (2009) viewed his 

new process-based model as a means to test an alternative antecedent to police outcomes based 

on the public’s internal motivation (normative values) for choosing to voluntarily obey the law or 

cooperate with the police.  Tyler (2003) hypothesized that a public that perceives procedural 

justice in officer-citizen relations would then judge the law enforcement agency as more 

legitimate and would be more likely to voluntarily obey the law and cooperate with the police. 

Social Control Roots.  With dependent variables that generally represent the opposite of 

committing a crime (obeying the law and cooperating with police), it is surprising that the 

theoretical ties between the process-based model of police legitimacy and the major theories of 

crime has received little scholarly attention.  As with most of the criminology canon, the 

Tylerian model provided a causal relationship for the dependent measures of crime (or its 

absence).  The process-based theory hypothesized that police officers adopting more just 

procedures (as defined by public perceptions of respect, consideration, and fairness) would then 

influence citizens’ decisions on cooperating with police or complying with the law.   

 

 

11 

 

Among the major theories of crime, the Tylerian model shared the key assumptions and 

hypothesized causal relationship with the social control theory of crime.  The social control 

theory and the Tylerian model placed their conceptual relationship in a society defined by 

Durkheim’s focus on norms.  Tyler (1990) in Why People Obey the Law credits Durkheim with 

pioneering the focus in criminology on the norms that explain why societies conform to rules.  In 

the case of the Tylerian model, the same core assumption was that the public’s judgments of 

police are not solely a function of the agency’s instrumental role in deterring crime.  Instead, the 

Tylerian theory argued that the public views the institution as sharing the same values was no 

less important (Sunshine & Tyler, 2003a). 

Control-based theories of crime may have begun with Durkheim, but their influence on 

criminology waned in the middle of the 20th century as scholars focused instead on the strain and 

social learning theories that were explanations on why people offend.  Travis Hirschi’s (1969) 

Causes of Delinquency sparked a return in criminology to the question of why people comply 

with the law.  Hirschi’s tests of social control theory arguably formed the most important 

empirical groundwork for the core assumptions of Tyler’s (1990) process-based theory of police 

legitimacy.   

As with social control theory, Tyler (2003) argued that compliance with the law and 

institutions is based primarily on perceptions of shared values and beliefs.  Social control 

theories place their central focus on positive attachments to institutions and the expectation that 

citizens obey the law if they are constrained by perceptions of common values with their social 

peers and institutions in which they regularly interact (Hirschi, 1969).  In terms of the legitimacy 

model, the more the public viewed the police as sharing the same values, the more likely that 

they perceived the agency as legitimate which in turn influences their decisions to obey the law 

 

 

12 

 

and cooperate with the police.  As a result, Tyler’s (2003) model argued that a society’s 

willingness to comply with the police can be altered by changing the public’s judgment of shared 

values with the law enforcement officers in their communities.  Hirschi’s (1969, 205) description 

of the roots of social control theory also further linked social control theory to legitimacy 

because he credits Weber (1968) as first arguing that the decision to obey the law was not a 

moral decision but a judgment based on the public perception of whether the law or institutions 

are legitimate. 

Legitimacy Scholarship Redux.  During the same socially-turbulent 1960s in which 

Hirschi (1969) revived social control theories, several political scientists returned to Weber’s 

(1968) posthumous work on legitimacy to develop a modern version of the classic legitimacy 

construct.  A concept first chronicled in Greek debates (Beetham, 2002), Weber introduced the 

modern concept of institutional legitimacy to explain the need for government agencies to gain 

voluntary compliance with law or authority given the inherent limits to maintain a standing army 

or constantly turning to armed coercion for compliance and cooperation.  Weber (1968) 

pioneered the idea that Durkheim’s social norms were critical for institutions to gain the 

voluntary compliance of individuals.   

Weber (1968) profiled three legitimating ideologies for institutions: traditional authority, 

charismatic authority, and rational (rule-of-law) bureaucratic authority.  Weber saw these as 

overlapping descriptions that essentially explained how charismatic rulers come to power or 

maintain their authority through tradition, while also allowing for the more modern legitimating 

role of written law.  By the end of the 1960s, a new wave of scholars revived Weber’s nearly 

century-old legitimacy construct, in particular his myth of a rational bureaucratic authority, to 

 

 

13 

 

hypothesize that the adoption of legalistic or just procedures will raise police institutional 

legitimacy. 

The empirical development of the legitimacy construct was closely interwoven with the 

rise of deadly riots in U.S. and European cities over the late 1960s.  The expression “legitimacy 

crisis” was coined in this period to describe the faltering public support for government 

institutions in developed countries that faced mounting crime and societal conflict (Gilley, 2009).  

The first scholars to empirically test Weber’s concept of organizational legitimacy (Aberbach 

and Walker, 1970; Paige, 1971; Wright, 1976) were specifically looking at the role legitimacy 

and public trust had on the public’s choice to riot or comply with the law and its enforcement.  

They used Easton’s (1965) systems model to pioneer the application of quantitative methods to 

support the Weberian assumption that coercive methods (deterrence) are not sufficient for an 

institution to gain legitimacy which, instead, depends on shared norms between the institution 

and the public they regulate (McEwen & Maiman, 1986).   

Tyler (1990) credited these studies for expanding the classic Weberian focus on 

legitimacy as support for a specific authority to a new definition that would be based on diffuse 

support for rules and institutions.  This scholarship demonstrated the impact institutional 

legitimacy, often measured in these studies with an index of political trust, had on compliance 

with the law (see Aberbach and Walker, 1970; Easton, 1975; Paige, 1971; Wright, 1976).  

Besides adapting quantitative methods to institutional legitimacy studies, this wave of post-riot 

scholars also differed from Weber on viewing legitimacy as a product of citizen consent (Dogan, 

1992).  As Weber did not include a role for justice or fairness in his description of legitimacy, his 

theoretical model of legitimacy was often criticized for failing to differentiate between a 

democratic or dictatorial authority.  Blau (1963) and Grafstein (1981) specifically argued that 

 

 

14 

 

Weber’s model of legitimacy does not differentiate between forced compliance (ie. a slave that 

no longer needed to be whipped to comply) and those freely choosing to comply with law or 

authority. 

Starting in the late 1960s, social scientists increasingly defined institutional legitimacy as 

a public belief that the institution is the most appropriate or morally proper for society (Dogan, 

1992).  Easton (1965) provided a foundation for this scholarship by viewing stable society as 

dependent on the values in favor of legal compliance.  Easton (1965, p. 273) described this 

diffuse support as a “reservoir of favorable attitudes or good will” that helps an institution gain 

consent or tolerance for even unfavorable outcomes.  Gamson (1968) argued that this was diffuse 

support.  He differentiated these attitudes from the specific support for institutions that is a 

product of instrumental calculations on whether the institution’s concrete outcomes benefited the 

citizen or not (Gamson, 1968).  Moreover, Gamson argued that institutional legitimacy was a 

significant control in a society to the degree that citizens recognized an obligation to obey and 

developed trust in the institution.  Tyler (1990) later combined these two conditions (obligation 

to obey and trustworthiness) to form his new construct of legitimacy.  Tyler (1990) cited 

Gamson’s (1968) observation that when the reservoirs of diffuse support were high, then New 

England local governments could readily gain support for new initiatives, but when they were 

depleted it grew difficult to govern effectively.   

The early empirical tests of these legitimacy models treated trust and legitimacy as 

interchangeable concepts.  Public policy scholars (e.g., Aberbach, 1969; Aberbach & Walker, 

1970; Paige, 1971; Seligson, 1980) used the concept of trust in institutions to operationalize 

legitimacy in their tests of the impact legitimacy has on compliance with the law.  A second 

wave of scholars (e.g., Brown, 1974; McEwen & Maiman, 1986; Rodgers & Lewis, 1974; Sarat, 

 

 

15 

 

1975) followed with studies that focused on the legitimacy of the law or trust in criminal justice 

institutions as a predictor of compliance with the law.  The late 20th-century studies of 

legitimacy primarily sought to determine the role the construct had on public behavior, but 

scholars rarely examined the precursors of this legitimacy (Hyde, 1983).   

Birth of Procedural Justice.  The second major theoretical breakthrough in the 

development of the Tylerian theory of police legitimacy followed the 1975 publication 

Procedural Justice: A Psychological Analysis by social psychologist John Thibaut and legal 

scholar Laurens Walker.  Their book introduced a process-based explanation of citizen 

perceptions of their interactions with public institutions.  Specifically, Thibaut and Walker’s 

(1975) process-control model was based on the proposition that procedural justice (how justice 

takes place) is just as important in shaping satisfactory resolutions of disputes as the actual 

outcome of the disputes. 

Tyler and other social psychologists initially adopted the same variables used in the 

control model of Thibaut and Walker (Tyler & Folger, 1980).  As a result, the first decade 

following the publication of Procedural Justice produced a wave of empirical studies focused 

solely on the potential impact that perceptions of just procedure had on public satisfaction.  For 

example, in Tyler and Folger’s (1980) pioneering application of the concept of procedural justice 

to police-citizen encounters, they framed their model in terms of whether the respondents’ 

perception of police procedures in the resolution of a dispute influenced their evaluation of 

satisfaction with the law enforcement institution.   

Legitimacy as a Measurable Construct.  In Why People Obey the Law, Tyler (1990) 

formed a two-dimensional definition of legitimacy that would mediate the precursors to 

compliance with the law.  The wave of legitimacy studies in the 1970s had either operationalized 

 

 

16 

 

legitimacy through a support for the police measure (often a survey question on trust in the 

police) or a perceived obligation to obey the law (e.g., Brown, 1974; Rodgers & Lewis, 1974; 

Sarat, 1975).  Brown (1974) was the exception as he used independent, attitudinal variables 

(support for the police as an institution and support for the law in general) to predict compliance 

with the law.  Tyler and colleagues (Sunshine & Tyler, 2003a; Tyler, 1990) chose to combine a 

support for police index (in later studies referred to as trust in the police or trustworthiness of 

police) and the obligation to obey to form his new construct of legitimacy that would predict 

compliance with the law as well as cooperation with police. 

Tyler and colleagues (Sunshine & Tyler, 2003a; Tyler, 1990; Tyler & Huo, 2002) refined 

what would become their trust sub-construct for legitimacy by operationalizing the concept using 

test items that represented different forms of support for the police agency.  Initially, Tyler 

(1990) drew on past studies to devise a four-item index that asked whether the public had 

“respect,” “pride,” “support” for the police and whether they viewed the police as “honest.”  

More test items were added in subsequent studies (Tyler and Huo, 2002; Sunshine & Tyler, 

2003a) that resulted in six, ten, and later a thirteen-item measure for institutional trust.  The new 

test items asked the respondent whether they felt “proud,” “encouraged,” “confident,” 

“protected,” or “not embarrassed” by the police as well as if they “respected,” had the “same 

values,” or “trusted [the police] to make decisions.”  The indexes demonstrated that Tyler and 

colleagues viewed trust in the police as representing a composite of the many different 

judgments of independent facets of an institution’s capacity to be viewed as trustworthy by the 

public.  Sunshine and Tyler (2003a) and an abundance of other studies (e.g., Hawdon, Ryan & 

Griffin, 2003; Murphy, 2013; Paternoster, Brame, Bachman & Sherman, 1997; Wells, 2007; 

 

 

17 

 

Wolfe, 2011) supported the trustworthiness component of the legitimacy construct in criminal 

justice research.  

The second of the two components in Tyler’s (1990) legitimacy construct was the 

obligation to obey.  As with Tyler’s additive index for trustworthiness, the test items for the 

obligation to obey sub-construct have differed between Tyler’s key studies from as low as three 

test items to as high as nine test items.  Tyler (1990) sought to develop a measure of how 

strongly the respondent supported the law regardless of personal feelings.  He drew from Sarat’s 

(1975) study on support for the legal system that ranked respondents based on their willingness 

to support the law even if they personally opposed the required legal action.  As a result, the test 

items that formed this additive index generally ask the respondent if they would obey police 

requests even if they believed the police or the action was wrong.   

The obligation to obey sub-component to legitimacy may be the most disputed in the 

process-based model.  Tyler and Huo (2002) and Tyler (2006) first reported that the test items 

used for the sub-construct resulted in low Cronbach’s alphas while more recent scholarship (Gau, 

2011; Reisig, Bratton & Gertz, 2007) raised questions whether obligation to obey loads on the 

same factor as trustworthiness.  Reisig and colleagues (2007) removed three of the five items 

used to measure obligation to obey because their factor analysis found that those items did not 

load on the construct.  Despite concerns over the validity of some test items used for this 

indicator, researchers continued to find empirical support that obligation to obey predicts 

legitimacy and police outcomes (Gau, 2011; Sunshine & Tyler, 2003a; Tyler, 2006). 

Classic Process-based Model of Legitimacy.  Tyler (1990) used his new 

measure of legitimacy to forge together two disparate, empirically tested relationships: 

Thibaut and Walker’s (1975) control theory expectation that public satisfaction with a 

 

 

18 

 

settlement is predicted by process-based precursors; and the legal and policy studies 

based on Easton’s (1965) systems model in which institutional legitimacy predicts 

cooperation with the authorities or compliance with the law.   

Tyler’s assimilation of the two distinct models was a textbook example of end-to-

end theoretical integration (Hirschi, 1979).  The two theories were merged by equating 

two distinct, though related, variables from the two theories (public satisfaction and 

institutional legitimacy).  Legitimacy would be operationalized by combining 

institutional trust, a concept Tyler (1990) initially termed support for the police, with the 

obligation to obey the authorities.  The new model was process-based because 

perceptions of procedural justice and other encounter-based citizen evaluations, such as 

officer effectiveness, would serve as the precursors of the new police legitimacy 

construct. 

Figure 2: Classic Process-based Model of Police Legitimacy

In Why People Obey the Law, Tyler (1990) first presented his proposed relationship 

between the three primary nodes of his process-based model: procedural justice; the legitimacy 

of the criminal justice institution; and compliance with the law.  At the center of his model 

(Figure 2), Tyler placed his new definition of institutional legitimacy as the mediator between 

 

 

 

19 

 

just procedures and police outcomes.  In the classic process-based model, police legitimacy is 

predicted by a procedural justice construct that also consists of two components.  Quality of 

treatment represents judgments on the degree to which procedures were respectful; quality of 

decision-making represents the degree to which police use reason and law rather than processes 

viewed as biased (Sunshine and Tyler, 2003a).  No element of the police legitimacy model has 

had stronger empirical support than the foundational hypothesis that procedural justice predicts 

police legitimacy (Sunshine & Tyler, 2003a; Tyler, 1990; Tyler & Fagan, 2008; Tyler & 

Wakslak, 2004).  

Tyler and colleagues continue to experiment with all three components of this model 

(procedural justice, police effectiveness, and the outcome variables that grew to include 

cooperation with the police and compliance with the law).  This process-based model of police 

legitimacy has become a leading theoretical model used throughout the social sciences to explain 
the complex relationship between local authorities and the public they serve.  A groundswell of 

studies over the last two decades in developed countries has revealed that perceptions of police 

legitimacy are primarily influenced by judgments about the fairness of procedures and 

assessments of police performance (Hinds & Murphy, 2007; Sunshine & Tyler, 2003; Tyler & 

Huo, 2002).  Legitimacy, in turn, mediates the relationship between procedural justice and the 

dependent variables of cooperation with the police or compliance with the law (Hinds & 

Murphy, 2007; Reisig et al. 2007; Sunshine & Tyler, 2003; Tyler & Huo, 2002).   

The empirical case for the influence of procedural justice and legitimacy on cooperation 

and compliance with legal authorities has been demonstrated across a variety of settings, 

including the police, courts, and corrections (Kaiser, 2016).  Though initially the model was 

supported largely by US-based research, the key relationships between procedural justice, 

 

 

20 

 

legitimacy, and the police outcome variables have been supported by stringent tests in a half 

dozen OECD member countries (Hough et al., 2013; Jonathan-Zamir & Weisburd, 2013; 

Mazerolle et al., 2013; Murphy et al., 2008; Reisig et al., 2014; Tankebe, 2013a; Van Damme, 

Pauwels & Svensson, 2013).  Moreover, the influence of procedural justice on police legitimacy 

and police outcome variables is fairly stable across different ethnic groups and communities 

(Cao, 2011).  Murphy and Cherney (2011) reported that Australia may be an exception as tests 

across different ethnic groups found that procedural justice differed as a predictor of cooperation 

with police.   

Normative vs. Instrumental Predictors.  The process-based model of police legitimacy 

would have drawn far less interest from scholars and practitioners if the chief finding in tests of 

the theory was limited to strong support for the relationship between perceived just procedures 

and institutional legitimacy.  Rather, the most influential outcome of the adoption of the Tylerian 

model in the developed world has been the finding that normative judgments on the fairness of 

police procedures are more influential than instrumental views of effective policing as predictors 

of the public’s compliance and cooperation with the police (Tyler, 1990; Tankebe, 2009). 

Prior to the empirical case made by Tyler and colleagues, most researchers and law 

enforcement practitioners focused on the instrumental motivations for the public’s judgment of 

police legitimacy (Sherman, 1993; Manning, 2001).  This was contradicted by a wave of studies 

using the Tylerian model (e.g., Tyler & Blader, 2000; Tyler & Huo, 2002; Sunshine & Tyler 

2003a) suggesting that instrumental or deterrence-based policies would have limited influence on 

either the agency’s legitimacy or public cooperation and compliance.  After reviewing the 

dominant role of normative factors in a test of the process-based model in several U.S. cities, 

Tyler and Huo (2002) argued that the model of policing based on instrumental factors of 

 

 

21 

 

reducing crime was unlikely to significantly influence public opinion or garner the public's long-

term support for the police.   

Across the criminal justice literature, scholars have drawn from the Tylerian model in 

demonstrating the role procedural justice has on police outcomes ranging from domestic violence 

(Pasternoster et al., 1997) to public disorder and misdemeanor crimes (Sunshine and Tyler, 

2003a).  The Tylerian model paved the way for a research approach in developed countries that 

now assumes that the legitimacy of a law enforcement institution is a function not only of the 

agency’s instrumental role in deterring crime but also the degree that the public views the 

institution as sharing the same values (Murphy, Hinds & Fleming, 2009; Sunshine and Tyler, 

2003b).  As a result, a scholar that planned to test the Tylerian model in a country for the first 

time would start by comparing the influence of the normative values forming perceptions of 

procedural justice with the instrumental judgments of the institutions effectiveness (Figure 3). 

In developed countries, the strong empirical support for the Tylerian process-based model 

has attracted a new wave of research that, besides demonstrating the significance of the key 

relationships in the classic model, focused on establishing the validity of measurements used for 

the key concepts.  Most notably, several scholars (e.g., Gau, 2011, 2014; Reisig et al., 2007; 

Tankebe, 2009) have experimented with the indexes used to operationalize police legitimacy in 

the model.  Tyler and his colleagues’ findings increasingly shape new policies and trainings on 

implementing procedural justice with the hope of raising agency legitimacy in applied settings 

(e.g., Mazerolle, Antrobus, Bennett & Tyler, 2013; Skogan, Craen & Hennessy, 2014; Wheller, 

Quinton, Fildes & Mills, 2013).  This research was a direct result of the strength of evidence 

supporting the model’s core findings in developed countries.  

 

 

22 

 

 

The Tylerian process-based model of police legitimacy, as developed in Tyler (1990) and 

 

Sunshine and Tyler (2003a), became the “most favored [theoretical] instrument in the current 

police literature” (Tankebe, 2009, p. 1273).  As the authors of a leading alternative model of 

police legitimacy, Bottoms and Tankebe (2012, p.121) concede, the research supporting the 

process-based model of police legitimacy is “rightly regarded as the most important 

criminological scholarship on legitimacy currently available.”  Nevertheless, the abundance of 

police legitimacy scholarship includes several criticisms of the process-based model of police 

legitimacy.   

Criticism of the Process-based Model 

A review of the legitimacy literature suggested that the key criticisms of the Tylerian 

theory of police legitimacy could be categorized into five broad groups of concerns: 1) 

uncertainty on limiting data sample; 2) debate on authoritarianism in the model; 3) concern for 

 

 

23 

 

the validity of model indicators; 4) discussions on additional model antecedents; and 5) the weak 

empirical case in developing countries. 

Uncertainty on Limiting Data Sample.  The Tylerian model was referred to as the 

process-based model of police legitimacy because the theory starts from the assumption that 

individual perceptions of officer behavior is critical for the formation of public judgments of 

police legitimacy and willingness to comply with the law and cooperate with that agency.  The 

legitimacy theory’s focus on police-citizen interactions led to confusion over whether the 

process-based model should be tested only on populations that report some type of contact with 

the police.   

The exceptional nature of contact with the police across the globe has required 

researchers to debate as part of their studies whether they should exclude survey respondents 

without the police interactions that shapes their judgments on the fairness of police procedures.  

Two of the largest surveys on police-citizen interaction were the European Social Survey (ESS) 

of 50,000 respondents from 28 European countries and the Bureau of Justice Statistics (BJS) of 

250,000 US residents.  The ESS (2010) survey found that 32.6% of the respondents had contact 

with the police in the last two years (ESS, 2010).  The BJS survey found that more than 26% had 

contact with the police in 2011 (Durose & Langton, 2013).  This suggests that limiting a data 

sample to those with contact in the last one or two years would significantly raise the scale of the 

sample required for a legitimacy study (Hohl, Bradford & Stanko, 2010).   

The decision of researchers on whether to exclude those lacking recent experience with 

police for use in studies of the process-based model changed over time.  Tyler’s first procedural 

justice study of the police focused on making the case that those with personal contact with the 

police significantly differed from the rest of the public in their evaluation of the agency (Tyler 

 

 

24 

 

and Folger, 1980).  Moreover, the two books that arguably did the most to establish the nascent 

theory of police legitimacy--Tyler’s (1990) Why People Obey the Law and Tyler and Huo’s 

(2002) Trust in the Law--tested procedural justice on legitimacy and police outcomes using 

population samples limited to those who reported recent experience with the police or courts.  

Over time, this expectation has been relaxed as some of the most influential research on police 

legitimacy theory, including Sunshine and Tyler (2003a), Tyler and Wakslak (2004), Reisig and 

colleagues (2007), as well as Gau (2011) did not limit their tests of the legitimacy model to just 

respondents that had recent experience with the police.  With a few exceptions (e.g., Tyler and 

Fagan, 2008), scholars over the last decade generally have taken a position on this theoretical 

issue by not requiring recent experience with the police for the data samples used in their tests of 

the process-based model.   

 

Tyler has not directly addressed the debate on the correct survey sample, though he has 

argued that perceptions of procedural justice do not have to result from direct experiences but 

can be vicarious experiences gained through friends, colleagues or the media (Tyler and Darley, 

2000).  Gau (2011) argued that there is limited evidence that police legitimacy or outcomes differ 

based on police contact.  In the U.S., Gau (2011) and Wolfe (2011) reported that police contact 

did not independently predict a significantly different set of judgments on police legitimacy or 

outcomes, while Reisig (2007) found that police contact had only a modest effect on evaluations 

of procedural justice.  Similar findings have been reported in Australia, where Murphy and 

colleagues (2008) found that contact with the police did not significantly influence the judgments 

on legitimacy or willingness to cooperate with the police; and in South Africa, where Bradford 

and colleagues (2013) reported that personal contact with the police had limited influence on the 

same judgments.  

 

 

25 

 

In presenting a case for not limiting samples to citizens with recent police contact, Gau 

(2011) highlighted the findings of Rosenbaum and colleagues (2005, p. 361) that “prior attitudes, 

whether positive or negative, seem to be more important than any type of recent experience with 

the police for determining current attitudes [toward the police].”  This was supported by the 

randomized experiment in which Braga, Winship, Tyler, Fagan, and Meares (2014) found that 

prior interactions with the police did not significantly influence judgments while citizens’ 

attitudes prior to the experience were the most consistent and strongest predictor of their 

evaluations of the police officer’s behavior.  These tests are in line with past criminal justice 

studies that first raised the question whether general perceptions of the criminal justice system 

can be more impactful than personal experiences (Brandl, Frank, Worden, & Bynum 1994; 

Skogan, 2006). 

Besides dividing respondents based on their recent interaction with the police, legitimacy 

researchers generally have tested whether demographic control variables significantly influenced 

the model’s key judgments.  After all, Tyler (1990) stated that his legitimacy theory’s key 

relationships may not be invariant in some circumstances.  In a few cases, researchers 

demonstrated that demographic variables affect respondents’ judgments in the model.  For 

example, Hinds and Murphy (2007) found in Australia that age and education were significant 

predictors of police legitimacy, while Braga and colleagues (2014) in the U.S. found that the 

level of education was a significant predictor of the procedural justice evaluations on police 

behavior.  Wolfe, Nix, Kaminski and Rojek (2016) sought to systematically test the invariance in 

the relationship between procedural justice and police legitimacy by comparing the impact that 

different demographic, experiential, and situational factors had on predictors of police 

legitimacy.  Their test, based on a comparison of OLS regressions, generally supported the 

 

 

26 

 

process-based model’s assumption that group differences, whether by race, gender, or police 

interaction, do not significantly influence the antecedents of police legitimacy (Wolfe et al., 

2016). 

Despite the recent wave of more stringent testing of legitimacy theory, scholars have yet 

to test whether a common understanding exists among respondents of the model’s key concepts 

in order to produce stable psychometric properties (measurement invariance) across different 

sub-groups.  In other words, tests of measurement invariance can determine whether items in the 

survey instrument mean the same things across groups, such as the group with recent experience 

with the police and the group lacking direct interaction with the police.   

In a previous debate on the invariance of a criminological theory across different groups, 

Piquero and Rosay (1998) pioneered the use of multi-group confirmatory factor analysis in a 

criminology study to establish measurement invariance for the self-control index across gender.  

As the use of the process-based model of police legitimacy continued to expand across different 

countries and demographic groups, the comparison of differences in latent indicators is valid 

only to the degree that measurement invariance can be established across the groups being 

compared. 

Debate on Authoritarianism in the Model.  The philosophical debate over the role of 

authoritarianism in institutional legitimacy can be traced to Weber’s (1968) categorization of the 

three means through which an authority can gain legitimacy.  Beetham (1974) and other scholars 

in the post-civil unrest wave of scholarship on legitimacy criticized Weber’s construct for a lack 

of concern whether the legitimacy gained by an institution was compelled through threat of 

violence or through voluntary submission to a governing authority viewed as sharing similar 

morals or beliefs.   

 

 

27 

 

The debate on the role of authoritarianism has continued in the contemporary dispute on 

whether legitimacy should include or exclude coercive submission to authorities.  Since the 

1960s, researchers have generally treated legitimacy as reflecting voluntary support to 

authorities.  As legitimacy models increasingly were used to model police-citizen relationships in 

less democratic countries, some scholars argued that legitimacy constructs also must account for 

the involuntary precursors to compliance or cooperation with law enforcement.  Harkin (2015) 

argued that because an authority produces legitimacy, as well as being a product of it, that 

Tylerian theory should incorporate the role of coercive measures in the public’s perception of 

agency legitimacy.  Likewise, Kochel (2012) warned that across different societies citizens may 

be compelled to obey police out of fear of the consequences of disobedience, rather than because 

they trust police or feel a moral obligation to their authority.  

Concern for the Validity of Model Indicators.  Over the last decade, concerns over the 

measurement of the latent constructs that compose the Tylerian model have become a common 

criticism of police legitimacy theory.  Most notably scholars using factor analysis have found 

that scales used for key indicators often do not load in line with theoretical expectations, while 

test items used for key indicators of independent constructs have problems with their 

convergence or divergence. (Gau, 2011, 2013; Murphy, 2013; Reisig & Lloyd, 2009; Tankebe, 

2008, 2009).   

Scholars establish the validity of a model through the use of stringent testing and 

experimentation with the indicators.  Campbell and Fiske (1959) developed the standard for 

establishing construct validity by examining the relationship within and between the measures 

developed in a theoretical model.  Specifically, a validation process should start with scholars 

establishing that the indicators provide for (a) convergent validity; and (b) discriminant validity. 

 

 

28 

 

Convergence.  Convergent validity is a judgment on whether the theory’s scales, 

operationalized as related in the model, are unidimensional and actually interrelated (Gerbing 

and Anderson, 1988).  Since Tyler and Huo (2002) first raised a concern over the dimensionality 

issue for police legitimacy, three empirical studies (Reisig et al., 2007; Tankebe, 2008; Gau, 

2011) specifically employed different forms of factor analysis to examine the common variance 

between the sub-scales composing legitimacy (trust in police and obligation to obey).  Those 

studies shared a consensus that the two sub-scales forming the process-based model’s indicator 

for legitimacy were not unidimensional and should not be combined into one theoretical 

construct.   

Divergence.  In the simplest terms, divergent validity is the extent to which “a measure is 

novel in the sense of measuring something different from that provided by other measures” 

(Holton, Bates, Bookter, & Yamkovenko, 2007, p.387).  Working in tandem with evaluations of 

the model’s convergent validity, the tests for discriminant validity demonstrate that the indicator 

of a construct is not highly correlated with other scales measuring theoretically disparate 

concepts.  Unfortunately, the classic studies testing the police legitimacy model (notably, Tyler, 

1990; Tyler & Huo, 2002; Sunshine & Tyler, 2003) did not report bivariate correlations for the 

indicators used in their tests.  Reisig, Bratton, and Gertz (2007) and Hinds and Murphy (2007) 

were the first empirical tests of the process-based model that called into question the discriminate 

validity of the indicators.  The studies based on U.S. and Australian data showed strong bivariate 

correlations between the trust sub-scale in the police legitimacy indicator and the antecedent 

measure of procedural justice.  The high bivariate correlations may be a warning to examine the 

indicators further for discriminant validity.   

 

 

29 

 

Researchers have expressed concern that the lack of divergence between the constructs is 

influenced by the respondents’ problems in differentiating their views on very similar normative 

judgments (Reisig, et al., 2007; Maguire & Johnson, 2010; Gau, 2011).  A review of the survey 

items in APPENDIX A, which Sunshine and Tyler (2003a) used in arguably the most cited test 

of the process-based model, suggests that similarities between test items could confound the 

disparate composite measurement scales.  In one example, a test item in the procedural justice 

measure asked the respondent whether the police “make their decision based on the facts, not 

their personal biases or opinions;” while a test item in the police legitimacy scale asks whether 

the respondents “trust the leaders of the NYPD to make decisions that NYPD officers make.” 

Certainly, the process-based model provides a compelling case that the quality of 

decision-making by officers is a different construct from a judgment on trusting those decisions.  

Discriminant validity depends on the survey respondent discriminating between test items so as 

to produce measures of unrelated constructs.  The similarity in the two questions’ emphasis on 

decisions could lead to difficulty for respondents to form responses that are restricted to the two 

separate constructs whose relationship is a cornerstone of the theoretical model.  Reisig and 

colleagues (2007) pointed out that this threat to construct validity may inflate the statistical 

relationship between the independent and dependent variables in the model.   

The factor analysis by Reisig and colleagues (2007) and Gau (2011) further attested to the 

lack of discriminant validity in the classic legitimacy model’s key indicators.  Gau (2011) found 

that trust in police shared considerable variance with the two procedural justice sub-scales of 

quality of treatment (Pearson’s r = .86) and quality of decision-making (Pearson’s r = .79) 

despite these sub-scales representing independent elements in the theoretical model.  Gau’s 

 

 

30 

 

(2011) confirmatory factor analyses found that all three sub-scales loaded on a single latent 

factor.  

Leading legitimacy scholars (e.g., Gau, 2011; Reisig et al., 2007; Tankebe, 2008) have 

advocated for a standardization of the key constructs found in the theory and the adoption of 

stringent tests of the model that would allow for a cumulative evaluation process across a broad 

and diverse set of methods and indicators.  Worden and McLean (2017a), as well as Jonathan-

Zamir and Weisburd (2013), argue that the diverse evaluation methodologies should include 

qualitative methods as well as systematic social observation of police-citizen interactions.   

 

Additional Model Antecedents.  A fourth common theme in the critical literature on 

legitimacy is the need for new antecedents that could add to the model’s explanatory power on 

police legitimacy and the prospects that citizens will comply and cooperate with the police.  

Tyler and Sunshine (2003a) introduced risk and distributive fairness to the classic Tylerian 

model, while Sunshine (2007) made the case for moral identification as a mediator in the model.  

Other scholars continue to identify potential new antecedents to legitimacy and police outcomes 

including lawfulness (Tankebe, 2013a), social identity (Bradford, 2014), and legal cynicism 

(Gau, 2015). 

The Weak Empirical Case in Developing Countries.  The last general theme in the 

critical literature on the Tylerian model focuses on the limited empirical support for the theory in 

the vast majority of the world’s countries which are still emerging in terms of prosperity and 

democratic governance.  Tyler’s process-based model of police legitimacy was tested outside of 

OECD member countries in 2008, when Tankebe reported that the classic process-based model 

of police legitimacy lacked empirical validity as a model of public cooperation with the police in 

Ghana.  Through the end of 2017, the only other published full tests of the process-based model 

 

 

31 

 

in non-OECD countries were in Jamaica and China4, where legitimacy failed as a mediating 

construct in the model (Reisig & Lloyd, 2009; Sun, et al., 2017).  On the other hand, procedural 

justice did modestly predict police outcomes without the legitimacy mediator.  The results in the 

three non-OECD countries did not provide strong support for the process-based model and leave 

unanswered the question of whether the theory serves as a valid model in developing countries.   

The tests of the legitimacy theory in non-OECD countries call into question the Tylerian 

model’s central conclusion that a society’s normative concerns outweigh instrumental factors in 

influencing police legitimacy or police outcomes of compliance and cooperation with the police.  

Tankebe (2009) found that police effectiveness--an instrumental factor--was the only significant 

predictor of police outcomes in Ghana.  The strength of the instrumental antecedent to police 

outcomes in Ghana was supported by Pryce and colleagues (2016) who studied Ghanaian 

immigrants to the United States.  Pryce and colleagues (2016) found that, even after emigrating 

from Ghana, police effectiveness for the immigrants was the primary factor influencing 

perceived legitimacy and the willingness to cooperate with the police.   

Scholars in South Africa (Bradford et al., 2013), and in Pakistan (Jackson et al., 2014) 

found that perceptions of just procedures were a weak predictor of legitimacy in comparison to 

police effectiveness in those two countries.  In China, Sun and colleagues (2017) found that 

procedural justice was a modest predictor of cooperation with the police, but similar to the other 

developing countries, police effectiveness was the strongest predictor of police legitimacy.  

Given the lack of a consensus in social research on the divide between developing and 

developed countries, tests of the Tylerian theory in borderline countries could also shape our 

                                                
4 Two partial tests of the Tylerian model also were conducted in non-OECD countries, but the South African study 
(Bradford et al., 2013) did not include police outcome variables while the study in Trinidad & Tobago (Rinehart 
Kochel et al., 2013) used a limited proxy for the procedural justice construct (the absence of police misconduct)  In 
addition, Karakus (2017) tested the full Tylerian model but used police satisfaction for his mediator in the model. 

 

 

32 

 

understanding of the model’s findings in developing countries.  Hough and colleagues (2013) 

found that the strength with which procedural justice predicted the Tylerian legitimacy construct 

differed across Europe.  Most notably, scholars have tested the model in former Soviet bloc 

countries of Slovenia and Poland that are now members of the OECD (Czapska et al., 2016; 

Reisig et al., 2014).  Although the two countries are now ranked among the world’s wealthiest 

countries, most of the population in Slovenia and Poland were raised in conditions that lacked 

their contemporary levels of prosperity and good governance.   

Reisig and colleagues (2014) found in Slovenia that procedural justice (as well as police 

effectiveness) predicted legitimacy while legitimacy was a significant predictor of cooperation 

with the police.  The study was the first test of the model in a former socialist country, but the 

students surveyed were born after the transition to a market economy and live in one of the 

world’s wealthiest countries with a GNP/person equivalent to the Riverside/San Bernardino 

metropolitan area of California.  The study suggests that despite being socialized by parents that 

were raised in far different circumstances that Slovenian youth respond similarly to respondents 

in developed countries.  In contrast, Czapska and colleagues’ (2016) found in their sample of 

Polish students that neither procedural justice nor police legitimacy were associated with the 

willingness to cooperate with police, instead, police effectiveness alone was the significant 

predictor of cooperation with the police.  

Turkey is another OECD country in which the stringent test of the Tylerian model served 

as a comparison for the role of legitimacy in Ukraine.  Turkey is a rapidly developing country 

with a high human-development ranking at 71 out of 187 countries by the United Nations 

Development Program (UNDP, 2016).  This suggests that Turkey could occupy the middle 

ground between developed and developing nations in their normative expectations toward the 

 

 

33 

 

police.  Karakus (2017) altered the Tylerian legitimacy indicator in his study by replacing 

obligation to obey with an index on satisfaction with the police.  This hybrid measure of the 

Tylerian legitimacy construct proved empirically valid in Turkey.  Nevertheless, the instrumental 

factor of police effectiveness was stronger than procedural justice as a predictor of police 

legitimacy and cooperation with the police (Karakus, 2017).  The model’s test in Turkey differed 

from previous studies in developing countries because the revised legitimacy indicator (adding 

satisfaction with the police sub-index) mediated the relationship between procedural justice and 

the willingness to cooperate with the police.  

Normative Values and Legitimacy 

Past legitimacy scholarship suggests that the problems transitioning the Tylerian theory 

of legitimacy to developing countries could partly be addressed through the addition of new 

measures that would shed light on the degree that Tyler’s expectation of relations shaped by 

shared normative values across judgments on interactions with officers, the legitimacy of the 

agency, and the willingness to cooperate with the police.  After finding weak support for the 

predictive role of legitimacy on predicting willingness to cooperate with the police in a 

developing country, Tankebe, Reisig, and Wang (2016) proposed that across societies different 

variables likely have a significant influence on the model’s key relationships.  Tankebe and 

colleagues (2016) advocated for future studies that would identify variables that could explain 

the failure of legitimacy theory in social contexts different from those in developed countries.   

The pursuit of measures that represent societal differences in criminal justice models is 

not unique to the theory of police legitimacy.  In the early years of police legitimacy research, 

Sherman (1993) argued that different social settings would bring divergent public reactions to 

identical acts of law enforcement.  Sherman’s observation primarily concerned different 

 

 

34 

 

communities, but his argument is no less true in comparing societies found in developed versus 

developing countries.  

In discussing institutional legitimacy, Weber (1968) also accounted for the possibility 

that disparate societies judge their institutions differently.  Weber (1968) did not categorize 

societies in terms of development but rather introduced three distinct paths by which authorities 

claim legitimacy.  Nevertheless, he saw some governments increasingly were developing rational 

bureaucratic authority.  This provides an early argument that rule-of-law policing and rational 

procedures could serve as a pathway to the legitimacy of an institution in a society.  Nonetheless, 

Weber (1968) also argued that the normative values that determine institutional legitimacy could 

be shaped by the ideologies of charismatic and traditional authorities.  In terms of debating 

differences across societies, Weber’s (1968) classification of legitimate authority suggests that 

rational and just procedures may not be the path for an institution to gain legitimacy in a society 

still dominated by charismatic or traditional authorities. 

Contemporary policing literature has rarely integrated different measures of societal 

values in evaluating theoretical models, but one of the few exceptions was, by coincidence, the 

first study of police procedural justice (Tyler & Folger, 1980).  In an attempt to transition the 

legal dispute resolution theory to the topic of policing, the Tyler and Folger (1980) included in 

their study two indexes of values drawn from the research literature on political ideology.  Tyler 

and Folger (1980) reported that the two indexes—developed in the 1950s for comparing political 

leanings--modestly influenced their respondents’ evaluation of the Evanston Police Department.  

The authors did not address the two societal measures in their conclusions but opened the door to 

a new attempt 38 years later to evaluate the impact of social values on the theory of police 

legitimacy.  This study will add to the Tylerian model two variables, postmaterialism and fear of 

 

 

35 

 

crime, that past research has supported as differing considerably across societies and represent 

the degree that normative values are prioritized by the public.  The variables would serve to test 

whether a measure of normative values would predict the cooperation with police which Tyler 

(2004) viewed as a process in which shared normative values were critical for the model’s 

mediating legitimacy indicator to predict cooperation with the police. 

Role of Postmaterialism.  This study will use a socio-cultural variable (Self-Expression 

vs. Survival Factor) that sociologists Ron Inglehart and Wayne Baker (2000) created to reveal 

the social values that prioritize daily survival in relation to normative priorities.  Inglehart and 

Baker developed this index to track how changes over time alter a society’s internal values.  

Scholars developed such postmaterialist indexes in support of Inglehart’s (1990) theoretical 

argument that over the course of their development industrialized democracies also undergo a 

cultural shift as individual values move from an emphasis on the material/survival (instrumental) 

values to prioritizing quality of life/self-expression (normative) values. 

The postmaterialism variable is expected to fit in to a Tylerian theory of police legitimacy 

that is a function of the interplay of instrumental and normative values in the public’s judgments 

of the police.  Tyler (2004) argued that social and cultural factors shape a society’s views on 

whether the normative aspects of police behavior, such as just procedures, outweigh the public’s 

concern for police effectiveness and other instrumental values in deciding whether to comply or 

cooperate with the police.  Specifically, the theory assumes that “internalized values” either 

favor or rebuff compliance with the law and cooperation with the police (Tyler, 2004, p. 86).  In 

adopting this approach, Tyler (1990) credited Easton (1965) for the assumption that the 

socialization processes forming these internal values differ across societies and lead to different 

priorities.  Generally, legitimacy scholars (e.g., Gau et al., 2012; Tyler, 2004) have found the 

 

 

36 

 

most significant cross-cultural differences between countries rather than between different 

societies in a singular country. 

Within the social sciences, postmaterialism measures were developed as part of 

Inglehart’s (1990) theory of value shift and are used to track changes in the internal values of 

different societies (Pavlović, 2011).  Inglehart (1977) labeled those favoring normative values as 

postmaterialists5.  Past studies of the theory of values shift suggested that postmaterialists in a 

society are more likely to value just procedures and less likely to simply accept orders to obey if 

they do not consider their views are reflected in those decisions (Inglehart, 1977; Inglehart, 1996; 

Inglehart, 1999; Clark et al., 1999).  Postmaterialist indicators mirrored the Tylerian model’s 

expectation that the public’s normative values demand fairness and openness to dialogue on the 

part of officers and would determine their judgments on obeying orders and trusting the police.  

Inglehart (1990) argued that societies slowly adopt postmaterialist values as younger generations 

are raised in periods of physical security and economic stability in which normative values 

increasingly compete for the more traditional concerns over personal survival (Curry & 

O’Connell, 2000).   

Inglehart (2012) developed his values model by combining two antagonistic schools of 

thought: Weber’s view that values shape economic development and Marx’s contradictory view 

that the first step in a transition is economic change which produces new values.  Inglehart’s 

theory of values shift took a middle road in viewing developments in the economy and 

transformations in societal values as reinforcing each other over time.  Moreover, the theory of 

values shift argued that the newly acquired values are found across different economic strata in a 

developed country.  This fell in line with findings of Gau and colleagues (2012) that the 

                                                
5 This study uses the term postmaterialism for the index used in the study (self-expression vs. survival factor) as 
well as other indexes intended to measure internal normative values in a society.   

 

 

37 

 

normative value of procedural justice in influencing perceptions of legitimacy held up in tests of 

the Tylerian model across different communities regardless of whether the locality was classified 

as high or low in terms of economic disadvantage (Gau, Corsaro, Stewart & Brunson, 2012).  

The foundational assumption of the theory of values shift was that for every country and 

their system of governance there is a unique set of norms, values, and other orientations 

consistent with that authority (Rogowski, 1974).  Moreover, if these values reflect norms learned 

while citizens are young, changes in values across a society would require the passage of 

considerable time (Parsons, 1935; Inglehart, 1990).  In the criminal justice canon, both Tyler 

(2004) and Sherman (1993) argued that internalized values are developed over time through 

judgments on the police’s use of fair procedures.  This suggested that differences in cultural 

values shape the balance between the public’s instrumental and normative judgments of the 

police across countries.  Outside policing, Brockner and colleagues (2001) also argued that 

societal values likely shaped perceptions of procedural justice and other attributes of the 

organizations they studied across Europe, Asia, and Latin America. 

Studies applying postmaterialist indexes in developed countries demonstrated that the 

postmaterialist scores in those countries rose dramatically over eras of relative prosperity (Figure 

4).  Moreover, a scatter plot of global data (see Figure 5) suggested that a proportional 

relationship existed in which citizens in countries farther along in political and economic 

development demonstrated greater concern for postmaterialist values (e.g., fair and just 

procedures, intolerance of corruption) while prioritizing far less the instrumental concerns for 

personal materialism and survival (Inglehart, 1996).    

 

 

38 

 

 

The scale and unique nature of Inglehart’s research on values likely accounts for his rank 

as the second most cited Political Science researcher of his generation (Abramson, 2011).  As 

part of this research, Inglehart and colleagues developed several different indexes to measure 

changes in post-materialism values over time.  Inglehart and Baker’s (2000) self-expression vs. 

survival index offered scholars a concise 4-item index to measure the degree that a society 

expresses post-materialist values over the materialist values of survival. 

Scholars have continued to debate whether the self-expression vs. survival index and 

other postmaterialist measures demonstrate the central tenets of Inglehart’s Theory of Values 

Shift, particularly in regard to data from Eastern Europe and developing countries (e.g., Kotze & 

Lombard, 2002; Pavlović, 2015).  Scholars no longer view postmaterialism measures (such as 

the self-expression vs. survival index) as antecedents to democracy, but rather just one indicator 

of the democratic values that underpin the representative institutions in a country (Hiskey & 

 

 

39 

 

Bowler, 2005).  Nevertheless, this study’s focus on postmaterialist measures serving as an 

indicator of the degree that normative values drive judgments continues to find support in the 

literature.   

 

Criticism of the postmaterialism indexes most often is from scholars who continue to use 

the postmodern measures in their own research but disagree with text item development or the 

theoretical claims made based on the values reported by the indexes (Davis, Dowley & Silver, 

1999).  Davis and Davenport (1999) criticized postmaterialist scholars for not continuing to test 

and validate the indexes that have been unchanged decades after the test items were first 

 

 

40 

 

developed through factor analyses in the 1970s.  Scholars also criticized the degree that 

Inglehart’s indexes are overly sensitive to changes in respondent priorities based on short-term 

economic fluctuations in unemployment (Clarke, Dutt & Rapkin, 1991) or the inflation rate 

(Duch & Taylor, 1993).  

The criticisms of Inglehart’s theory of values shift as an explanation of change in 

societies did not diminish the value of the postmaterialism measures to social science as an 

instrument that weighs the concern for material security against normative concerns for 

tolerance, self-expression, and personal rights (Abramson, 2011, Pavlović, 2015).  For example, 

the past use of postmaterialism indexes to predict public perceptions of government performance 

(e.g., Fernandez and Kuenzi, 2010), suggested that they could serve as antecedents to police 

legitimacy and police outcome variables.  

Citizen expectations did not remain flat over this post-industrial period, but rather Gilley 

(2009) argued they rose steadily which presented an even greater challenge to maintain 

legitimacy.  The postmaterialist indexes offer a potential measure to capture the shifting values 

on which public perceptions of institutional (police) legitimacy are formed.  This study 

hypothesizes that a measure that captures the degree that the public prioritizes normative values 

(rather than survival values) would predict the inner normative values that Tyler argues 

determines the public’s willingness to cooperate with the police.  Specifically, this study will test 

whether a measure of postmaterialism predicts the outcome variable of cooperation in the 

Tylerian model (see Figure 6).  

The postmaterialism measure may shed light on whether normative values shape public 

judgments on officers, the agency, and the public’s willingness to cooperate with the police.  In 

developed countries, the relationship has been repeatedly demonstrated through the strong role 

 

 

41 

 

that legitimacy, in comparison to the instrumental concern for police effectiveness, has in 

mediating just or unjust officer behavior and predicting cooperation with the police (Sunshine & 

Tyler, 2003a).  By contrast, past studies in developing countries suggested that legitimacy will be 

a weak mediator in the process-based model (e.g., Czapska et al., 2016; Karakus, 2017; Reisig & 

Lloyd, 2007).  Therefore, postmaterialism, which is a predictor of normative over instrumental 

values, can demonstrate whether normative judgments are associated with police outcomes in a 

developing country.  A positive test would suggest that the weakness of the legitimacy indicator 

in developing countries may not reflect normative values but another factor that prevents 

normative attachments to law enforcement agencies and limits the mediating role of the 

trustworthiness indicator in the model.  

Role of Fear of Crime.  The Fear of crime measure in this study is intended to represent 

the degree that instrumental fears for personal security supersede a citizen’s normative priorities 

in evaluating officer behavior and cooperation with the police.  A subjective measure shaped by 

 

 

 

42 

 

a broad range of individual and neighborhood characteristics, the fear of crime in past studies 

appeared to behave independent of the actual victimization rate (e.g., DuBow, McCable & 

Kaplan, 1978; Hale, 1996; Leiker, 2003).  Most notably, scholars have demonstrated that crime 

is weakly correlated with a neighborhood’s level of fear of crime (Lewis & Salem, 2016) and 

that, while rising crime can lift the level of fear of crime, the fall in crime had less of an impact 

on public fears (Taylor & Hall, 1986).   

More than 350 empirical studies of fear of crime have been conducted across a wide 

range of countries (Fernandez and Kuenza, 2010; Leiker, 2003).  US-based research 

demonstrated that neighborhood disorder and the lack of informal neighborhood social ties 

facilitated or inhibited the fear of crime and produced significant differences across and between 

different communities (Hunter, 1979; LaGrange et al., 1992; McGarrell et al., 1997).  In a 23-

country study across Europe, Hummelsheim and colleagues (2011) found that, because of factors 

that included neighborhood social incivilities and civil ties between neighbors, the fear of crime 

indicator acts as a social insecurity that differed across countries. 

The behavior of the fear of crime variable suggested that it can act as a catch-all measure.  

The fear of crime did not act as a targeted instrument for the public’s concern that they could be 

victimized in the future, but rather the fear of crime had “its roots in something more diffuse than 

the perceived threat of some specific danger in the immediate environment” (Garfalo & Laub, 

1978).  Scholars came to accept that the fear of crime represented a wide variety of other 

community perceptions including fear of strangers, fear of other ethnicities, as well as anxiety 

over neighborhood incivilities and physical disorder (Garfalo & Laub, 1978; Taylor, 1999).  

Studies in developing countries also found that fear of crime was an insecurity that acted as a 

sponge absorbing over time a wide variety of citizens’ concerns over social and physical 

 

 

43 

 

instability in their daily life (Bennett & Flavin, 1994; Dammert & Malone, 2003).  This 

description suggested that the fear of crime index could represent the impact that extended 

periods of instability in citizens’ physical environment have on the weight they place on 

instrumental over normative values in their judgments on police behavior and outcomes.  

As a result, this study posits that to the degree that fear of crime represents normative 

judgments it can serve as a predictor of the willingness to cooperate with the police.  Fernandez 

and Kuenza (2010), for example, found across 32 countries in Latin America and Africa that 

citizens’ perception of public safety was as important a factor as any socio-economic variable in 

predicting support for and satisfaction with the state of democracy in the country.  In a Latin 

American survey, Perez (2003) drew a similar conclusion that increased fear of crime in a 

country reduces the level of public support for criminal justice institutions.  As a result, this 

study proposes that the level of fear of crime in a society will be associated with the role of inner 

normative values that Tyler identifies as shaping the decision to cooperate with the police.  

Police legitimacy scholars used fear of crime indicators in a few previous studies.  

Sunshine and Tyler (2003a) developed a latent variable for police effectiveness that combined 

test items on police performance, neighborhood conditions, and the fear of crime.  However, the 

latent variable for police effectiveness did not significantly impact legitimacy or cooperation 

with the police.  Hence, Sunshine and Tyler (2003a) argued that U.S. tests of the legitimacy 

theory demonstrate that instrumental concerns are less influential than the procedural justice 

construct in predicting agency legitimacy and police outcomes.  

In developing countries, limited research on the role of physical insecurity in the Tylerian 

theory of legitimacy left considerable uncertainty on the role for the latent variable as a mediator 

in the process-based model.  Karakus (2017) found that fear of crime, along with other 

 

 

44 

 

neighborhood safety measures, did not prove a significant predictor of either police legitimacy or 

cooperation with police in Turkey.  In Ghana, Tankebe (2013b) proposed that anxiety over crime 

likely was a major reason the Tylerian model failed to significantly predict legitimacy or police 

outcomes, but he did not include the variable in his regression models.  The influence of a 

physical insecurity measure on citizen judgments in developing countries was supported by 

Fernandez and Kuenza’s (2010) finding in Latin America and Africa that citizens gave greater 

weight to personal safety than economic growth and development in evaluating their government 

institutions.  

Sunshine and Tyler (2003a) defined the fear of crime as an instrumental construct 

because citizens make a decision on the desirability of a condition that directly affects them. 

Hummelsheim and colleagues (2011) argued that fear of crime also served as a reflection of 

social insecurity in societies.  This study posits that this insecurity influences the balance of 

normative and instrumental values in citizens which allows fear of crime to associate with the 

internalized normative values that Tyler argues determine the willingness to cooperate with the 

police.  The study will contrast fear of crime and legitimacy in the strength of their association 

with the model’s outcome variable, cooperation with the police.   

 

 

 

45 

 

 

CHAPTER III 

Methodology 

The chief objective of this study was to test the Tylerian process-based model of police 

legitimacy in a developing country and explore whether two new variables, hypothesized to 

reflect normative values, predicted the public’s willingness to cooperate with the police.  This 

quantitative study used a path analysis on an existing community survey to test the key 

relationships of the classic process-based model of police legitimacy including the influence of 

the new variables on the police outcome cooperation with the police. 

Prior to performing the path analysis, the study reported validity test results in support of 

the effort to construct more valid indicators that started with Reisig and colleagues (2007) and 

has expanded across a nascent wave of stringent tests of the process-based model of police 

legitimacy (Gau, 2011, 2013; Murphy, 2013; Reisig and Lloyd, 2009; Tankebe, 2008, 2009).  

Finally, this study introduced the use of multi-group confirmatory factor analyses in validating 

the Tylerian model by testing the procedural justice construct for measurement invariance across 

groups separated by their experience with police and demographic characteristics.   

Survey Data 

The study’s data were collected through a face-to-face community survey of Ukrainian 

residents in 2013.  The survey was administered by the U.S. Department of Justice’s 

International Criminal Investigative Training Assistance Program (ICITAP) in Ukraine.  The 

ICITAP Project Office in Kyiv financed the administration of the survey to serve as the pre-

intervention data for a community policing initiative.  While serving as an adviser to the ICITAP 

office, the study’s author developed the questionnaire.   

 

 

46 

 

The survey sought to collect pre-intervention neighborhood views on the district officers 

operating in the experimental and control sections of three Ukrainian cities located in different 

corners of the country (Kamianets-Podilskyi, Lutsk, and Nova Kakhovka).  The U.S. Department 

of Justice contracted the Ukrainian office of Europe’s leading consumer survey firm GfK to 

collect the data using door-to-door interviews in the participating police districts.  The 

questionnaire included a wide variety of test items and indexes used by researchers to measure 

perceptions of police legitimacy and performance.   

The sample used for the study consisted of 1,964 completed questionnaires.  GfK 

collected more than twice this number of questionnaires but those conducting the interviews 

were allowed to mark an option 99 for questions that the respondent gave the response “hard to 

answer.”  Although researchers generally seek to avoid dropping observations from a sample, 

this study did not use 2,041 questionnaires in which the choice “hard to answer” was given for 

one of the 21 test items that compose this test of the Tylerian model of police legitimacy6.   

The 1,964 questionnaires were completed by adult Ukrainian citizens between 18 and 91 

years of age with a median age of 44.1, SD=16.7.  In line with the general population (State 

Statistics Service of Ukraine, 2001) the survey respondents were 54.9% female with more than 

one-third (36.0%) possessing a two-year or more university degree.  This truncated list did not 

differ from the original sample except in the case of higher education.  The list deletion increased 

the portion of respondents with a higher education degree but was closer to the national average 

in Ukraine of 35% (State Statistics Service of Ukraine, 2001).   

                                                
6 The questionnaires that used the “hard to answer” responses were dropped because: the missing 
data was negatively associated with education-level which ruled out a multiple imputation 
solution for the missing items; and the path analyses did not find a difference in the core findings 
of the process-based model whether run on the original data sample or the smaller set of 
completed questionnaires.  

 

 

47 

 

The pared list contributed to the unequal sample distribution between the three surveyed 

cities as the smaller Kamianets-Podilskyi sample accounted for 20% more of the “hard to 

answer” responses than the other two cities.  Kamianets-Podilskyi residents viewed the 

performance of the militia more negatively than the other two cities which may have impacted 

their willingness to answer some of the questions in the survey.  A path analysis of the Tylerian 

model on separate data sets for each city found that, though the results differed in terms of the 

strength of some parameters, the study’s key findings in terms of significant relationships and 

directions were uniform across each of the cities. 

The survey was intended to provide a representative sample of the experimental and 

control districts in each of the three cities and was not intended to generalize to either national or 

city-wide populations.  The interviews in each city were distributed based on addresses listed for 

city districts assigned to the experimental or control group for the community policing project.  

The list of addresses to use within a district were selected at random with the only limit that no 

more than five completed questionnaires could come from addresses in the same apartment 

building.  If the interviewer discovered more than one eligible adult at an assigned address, they 

would interview the adult whose birth date was the most recent.  The survey teams did not record 

a door-to-door response rate in the three cities.  The surveyors would no longer pursue an 

address after a refusal to participate or three failures to find someone at that address.   

Shortly after the survey was completed, the DOJ community policing program was put on 

hold because of the start of social turmoil that later became known as the Maidan Revolution.  

Subsequently, the ICITAP project office refocused on supporting the new Ukrainian 

government’s priorities to liquidate the former militia structure and stand up a new police 

service.  The US Department of Justice ICITAP agreed to permit the use of the orphaned data for 

 

 

48 

 

research purposes.  The MSU Institutional Review Board ruled that the study protocol and 

ICITAP survey data met the criteria of exempt status according to Federal regulations. 

Independent and Mediating Variable Measurement 

 

Procedural Justice.  The procedural justice scale is conceptualized in the classic 

Tylerian process-based model as formed by two dimensions: the sub-construct quality of 

treatment and the sub-construct quality of decision-making (Tyler and Huo, 2002).  The survey 

operationalized procedural justice using the eight test items that Reisig and colleagues (2007) 

evaluated from Sunshine and Tyler’s (2003a) indexes for procedural justice.  The five test items 

forming the quality of treatment (APPENDIX B: Questions 9.1 – 9.5) and the three test items for 

quality of decision-making (APPENDIX B: Questions 9.6 – 9.8) were measured on a 4-point 

Likert scale ranging from 1 (strongly disagree) to 4 (strongly agree).  

 

Police Effectiveness.  The five-item index used in this study to measure service quality 

or effectiveness of the police was first developed by Rinehart Kochel, Parks and Mastrofski 

(2013) for a public survey in Trinidad & Tobago.  The test items forming this latent variable 

(APPENDIX B: Questions 8.1 – 8.5) generally asked if the police were prompt or effective in 

preventing crime.  The responses were measured on a Likert scale ranging from 1 (strongly 

disagree) to 4 (strongly agree).   

 

Postmaterialism.  This study is based on Inglehart and Baker’s (2000) self-expression 

versus survival factor that consists of four test items measured on a Likert scale similar to that 

used for the procedural justice test items (APPENDIX B: Questions 21.1 – 21.4).  The variable 

has been used over several decades in modernization and democratization studies and is usually 

calculated as an additive scale that represents the degree that a society expresses normative, self-

expression values over the materialist values of personal survival. 

 

 

49 

 

 

Fear of Crime.  This latent variable was operationalized using four survey items drawn 

from Ferrero and Grange’s (1987) study on alternative measures of the fear of crime 

(APPENDIX B: Questions 8.6 – 8.9).  As with all the indexes, the test item terminology was 

adjusted for use in Ukraine.  For example, there is no direct parallel in either Russian or 

Ukrainian for the word neighborhood.  Instead, the survey’s questions on respondent views 

regarding crime and disorder in their neighborhood were altered to refer to an area one half 

kilometer around their home.   

 

Police Legitimacy.  Since Tyler (1990) first proposed this construct, the process-based 

model has used a two-dimension measure of police legitimacy that combines an obligation to 

obey index with an index that operationalizes police trustworthiness.  The survey data used the 

two test items for obligation to obey and the three test items for police trustworthiness that Reisig 

and colleagues (2007) validated after testing the factor loadings, inter-item correlations, and 

multicollinearity of the larger set of test items used in Sunshine and Tyler’s (2003a) study of the 

process-based model of police legitimacy.  This latent variable is based on five test items 

(APPENDIX B: Questions 9.13 – 9.17) measured on a 4-point Likert scale ranging from 1 

(strongly disagree) to 4 (strongly agree).   

Dependent Variable Measurement 

This survey initially intended to test the dual outcome variables of compliance with the 

law and cooperation with the police that were part of Sunshine and Tyler’s (2003a) influential 

test of the process-based model of police legitimacy.  The GfK survey team conducted an initial 

pre-survey with two dozen respondents to test the draft questionnaire in the first half of October, 

2013.  The pre-survey suggested that a high portion of questionnaires would not be fully 

completed if the survey included the test items on compliance with the law.  Most of the pre-

 

 

50 

 

survey respondents refused to choose a response to the compliance questions.  The refusals 

spanned the full range of test items including queries on littering, speeding, or drunk driving that 

were included in Tyler’s (1990) first police legitimacy study that used compliance as an outcome 

variable.   

As a result, the project team agreed with the survey team to just use the willingness to 

cooperate with law enforcement as the sole dependent variable on the questionnaire.  The 

decision to limit the outcome variable to only cooperation with the police is in line with previous 

tests of the police legitimacy theory in developing countries (Czapska et al., 2016; Karakus, 

2017; Reisig et al., 2014; Sun et al., Tankebe, 2009).  Other researchers may have encountered 

similar difficulties in collecting responses on legal compliance questions which would explain 

the lack of published tests of the Tylerian process-based model using compliance with the law as 

an outcome variable. 

Cooperation with the Police.  The cooperation with the police index in this study 

recorded the respondent’s willingness to take the initiative over four different scenarios ranging 

from calling to report a crime to providing information on a potential suspect (APPENDIX B: 

Questions 11.1 – 11.4).  The latent variable of cooperation with the police is based on four test 

items that Reisig and colleagues (2007) and Gau (2011) supported in their construct validity 

studies of the process-based model of police legitimacy.   

Analytical Plan 

This study examined whether material and physical insecurity mediated the relationship 

between the public’s judgments on police use of just procedures and the public’s willingness to 

cooperate with the police.  The research required first testing the key indicators and relationships 

of the process-based model using Ukrainian data and then determining whether the hypothesized 

 

 

51 

 

mediators offered an alternative path for explaining the relationship between procedural justice 

and decision to cooperate with the Ukrainian militia. 

As a first step, the study sought to determine if there were validity problems in the 

operationalization of the Tylerian model in Ukraine.  Following the lead of a number of scholars 

that have raised concerns with the process used to develop the model’s measurements (e.g., Gau, 

2011, 2013; Maguire & Johnson, 2010; Reisig et al., 2007; Tankebe, 2009), the analytical plan 

made use of robust validity testing prior to examining the model’s key predictions.  An 

Exploratory Factor Analysis (EFA), was used to determine factor loadings.  Moreover, the 

analysis made use of traditional tests of correlation coefficients and Cronbach’s alphas.     

The structural equation-modeling (SEM) tools of Stata 15 were used to conduct path 

analyses of the key relationships in the Tylerian model.  SEM analysis is ideally suited to 

examine a well-tested and theorized model, such as the Tylerian theory of police legitimacy.  

SEM models were used to test the five research questions applying the Tylerian theory of police 

legitimacy in developing countries and whether the two hypothesized variables reflecting 

normative values predict the willingness to cooperate with the police in Ukraine.   

The last component of the data analysis combines measurement model and structural tests 

to address the sixth research question on whether possessing recent experience with the police 

bias the procedural justice scores in the process-based model of police legitimacy.  Tests for 

measurement invariance in the indicators were not previously part of published studies on the 

Tylerian model.  This study used a sequence of multi-group confirmatory factor analyses (MG-

CFA) to test the procedural justice construct for measurement invariance across groups.  

Procedural justice is the construct in the Tylerian theory that is hypothesized to reflect a 

community’s experience with police officers.  The MG-CFA tools also determined whether 

 

 

52 

 

gender, age, and education level, as well the research question of direct versus indirect 

experience with the police, bias the model’s psychometric measure of procedural justice.   

 

 

 

 

53 

 

Preliminary Tests 

CHAPTER IV 

Results and Discussion 

Prior to conducting the tests to assess the study’s key research questions, Stata 15 was 

used for an exploratory analysis prior to examining the validity of the model’s indicators.  This 

study will make use of a “cumulative evaluation process” for improving a model’s construct 

validity (Campbell and Fiske, 1959).  Most notably, the study followed the lead of legitimacy 

scholars (e.g., Gau, 2011; Reisig et al., 2007) in pursuing a factor analysis of the study’s 

indicators, an analysis of the indicator’s Cronbach’s alpha coefficients, and the bivariate 

correlation coefficients.   

Exploratory Factor Analysis.  The study used SEM Builder in Stata 15 to perform an 

exploratory factor analysis (EFA) of the model’s key measures.  An EFA is most often used to 

uncover the key factors in collected data, but in this case, an EFA was chosen as an initial test of 

the validity of the measurements used for the theoretically-supported process-based model of 

police legitimacy.  As a first step in preparing for an EFA, the test items were examined for 

sampling adequacy.  Specifically, the Bartlett test of sphericity, χ2(210) = 27,943.7, p < .001, 

supported the lack of an identity matrix and, consequently, the appropriateness of using factor 

analysis on this data set.  A Kaiser-Meyer-Olkin test, KMO = .940, similarly suggested that there 

is adequate common variance among the 23 test items operationalized as predictor variables in 

this study.   

Next, the distribution of the test items used in the scales for the model were examined for 

normality.  As demonstrated in Table 1, the skewness and kurtosis results across the study’s 27 

test measures generally were within the range viewed as adequate for univariate normality.  Test  

 

 

54 

 

Table 1: Descriptive Characteristics of Test Items 

Test Item 

Mean 

Standard 
Deviation 

Skewness 

Kurtosis 

 

 

2.99 
2.90 
2.86 
2.91 
3.07 
3.01 
3.00 
2.77 

2.82 
2.69 
3.09 
2.09 
2.07 

2.94 
2.92 
2.88 
2.87 
2.86 

3.43 
3.30 
3.48 
3.24 

 

 

 

 

Procedural Justice 
  9.1 Respect 
  9.2 Listen 
  9.3 Treat Fairly 
  9.4 Respect Rights 
  9.5 Courtesy 
  9.6 Fact-based 
  9.7 Explain 
  9.8 Fairly 
Legitimacy 
  9.13 Basic Rights 
  9.14 Trust 
  9.15 Job Well Done 
  9.16 Accept 
  9.17 Do as Told 
Police Effectiveness 
  8.1 Duties 
  8.2 Order 
  8.3 Try to Help 
  8.4 Prompt 
  8.5 Prevent 
Cooperation 
  11.1 Call 
  11.2 Report 
  11.3 Accident 
  11.4 Info 
Fear of Crime 
  8.6 Fear Alone 
  8.7 Worry Property 
  8.8 Reason to Fear 
  8.9 Love Ones 
Post Materialism Index 

 

0.72 
0.79 
0.82 
0.80 
0.74 
0.76 
0.75 
0.83 

0.70 
0.80 
0.79 
0.88 
0.92 

0.82 
0.83 
0.87 
0.88 
0.86 

0.69 
0.80 
0.71 
0.87 

 

 

 

 

 

-0.61 
-0.45 
-0.41 
-0.58 
-0.68 
-0.59 
-0.60 
-0.30 

-0.63 
-0.42 
-0.74 
0.35 
0.31 

-0.56 
-0.51 
-0.55 
-0.50 
-0.41 

-0.93 
-0.98 
-1.29 
-0.97 

-0.16 
-0.07 
-0.03 
-0.69 
 0.87 

 

 

 

 

 

2.41 
2.79 
2.70 
2.92 
2.57 
2.69 
2.98 
2.55 

2.82 
2.72 
2.96 
2.29 
2.04 

2.93 
2.81 
2.71 
2.59 
2.55 

2.27 
2.49 
4.35 
2.64 

 

 

 

 

1.83 
1.92 
1.96 
2.43 
2.94 

2.59 
2.54 
2.53 
2.99 
    12.01 

1.10 
1.01 
0.99 
0.99 
       2.27 

item 11.3 (Reporting of Accidents) in the cooperation with the police index was an exception 

because the kurtosis coefficient was outside the range (±3) generally viewed as adequate 

(Onwuegbuzie & Daniel; 2002).  A high and positive kurtosis coefficient can lead to 

 

 

55 

 

underestimating the variance of the variables, which increases the chance of a “false positive” or 

Type I error rate (Tabachnick & Fidel,1996).   

After reviewing the theoretical arguments for keeping the accident variable in the 

cooperation with the police index, the test item was dropped from the data set leaving three test 

items for the latent measure of cooperation with the police.  Although common in past studies 

using a cooperation index, the test item (Reporting of Accidents) may be problematic.  The 

reporting of accidents in Ukraine is not strictly a question of judgments about the police but may 

manifest other concerns in a country where only a minority of drivers comply with the law 

requiring car insurance.  Therefore, the removal of the accident test item to improve the 

normality of the data may refine the dependent variable to better reflect the willingness of the 

respondent to cooperate with the police in the context of Ukraine.   

The next step in the analysis was an examination of the Eigen values and scree plots of 

the factors in the EFA test.  The model extracted five factors from the 23 test items (Table 2).  

Oblique rotation (Promax) was used because past literature suggests the test items are correlated.  

The test items for most factors loaded together on separate factors but a single factor 

incorporated the test items forming the procedural justice and the trustworthiness scales. 

Although supportive of the hypothesized factor structure, the EFA reaffirmed the most 

commonly discussed construct validity concerns for the Tylerian model, the lack of convergence 

for the obligation to obey and trustworthiness subscales in the EFA (Tyler & Huo, 2002; Gau, 

2011, Reisig et al., 2007).  The lack of convergence between the two subscales forming the 

legitimacy measure is in line with previous studies that reported the two sub-scales were not 

unidimensional and should not be combined in a construct theorized as unidimensional.  The 

factor analysis’ results regarding legitimacy’s unidimensionality serve as a warning that 

 

 

56 

 

researchers applying the process-based model of legitimacy should not rely solely on Cronbach’s 

Alpha to test the validity of the theory’s indicators.  As Hirschi and Selvin (1967) argued a half 

century ago, Cronbach’s alpha may suggest a reliable and internally consistent indicator for an 

index whose test items actually load on two or more independent constructs.   

Table 2: EFA Factor Loadings for Predictor Variables 

Item 

 

Loadings: Five Factors 
Comp. 
Comp. 

Comp. 

Comp. 

Comp. 

1 

2 

3 

 

 

 

 

4 

 

 

5 

 
1. The local police treat citizens with respect. 
2. … take the time to listen to people. 
3. … treat people fairly. 
4. … respect citizens rights. 
5. …. are courteous to people with whom they come in 

  Quality of Decision-Making 1. The local police make decisions based upon the facts. 
 
 

2. … explain their decisions to those involved. 
3. …  make decisions that lead to the fair resolution of 

to contact. 

problems. 

Construct 

(Cronbach’s Alphas) 

 

Procedural Justice (α = .93) 
  Quality of Treatment 
 
 
 
 

Police Legitimacy (α = .74) 
  Trustworthiness 
 

 
Police Legitimacy (α = .74) 
  Obligation to Obey 

 

Police Effectiveness (α =.91)  
 

 
 
 
 
Cooperation (α = .81) 
 
 

 

Fear of Crime (α =.84) 
 

 

 

 

 

.672 
.702 
.720 
.759 
.820 
.827 
.823 
.704 

 
 

 
 
 
 
 
 
 
 
 
 

 
 
 

 

 

 

 
 
 
 
 
 
 
 
 
 
 
 
 
 

 
 

.784 
.758 
.750 
.756 
.700 

 
 
 

 
 
 

 

 

 

 
 
 
 
 
 
 
 
 
 
 
 
 
 

 
 
 
 
 
 
 
 
 
 

 
 

.815 

.866 

.843 

.747 

 
 
 
 
 
 
 
 
 
 
 
 
 
 

 
 
 
 
 
 
 
 
 
 

 
 
 

 

 

 

 
 
 
 
 

 
 
 

 
 
 

 
 

.943 

.958 

 
 

 
 
 
 
 

.852 
.893 

.814 

 
 

 

 

 

are right for your community. 

 
1. People’s basic rights are protected by the local police.  .690 
2. The local police can be trusted to make decisions that 
.691 
.673 

3. Most local police officers can do their job well. 
 
1. You should accept police decisions even if you think 

 

2. You should do what the police tell you even if you 

they are wrong. 

disagree. 

duties properly. 

1. The local police know how to carry out their official 

2. … are able to maintain order on the streets.  
3. … try to help citizens. 
4. … respond promptly when needed. 
5. … prevent crime. 
 
1. I would you call the police to report a crime?  
2. ….. report a suspicious activity near your home to the 

3.  …. provide information to the police to help catch a 

police.  

suspected criminal. 

 
1. In the half-kilometer area round your home, do you 

feel afraid to walk alone at night. 

2. When I am away from home, I worry about the safety 

of my property. 

3. There is a reason to be afraid of becoming a victim of 

crime in my community. 

4. I worry about others in my home when they head 

outside at night. 

 

57 

 

 

The Tylerian model’s key indicators also were examined for internal consistency using 

Cronbach’s Alpha (Table 2).  This coefficient served to measure the reliability of the predictor 

scales.  A comparison of the coefficients from the Ukrainian sample and the U.S. data samples 

(Gau, 2011; Reisig et al., 2011) demonstrated considerable similarity between the alpha 

coefficients of the key indicators in the two countries (see Table 3).  As in the U.S. tests, the 

alpha coefficients suggested internal consistency (a > 0.70) across the test items used to form 

procedural justice and legitimacy scales.   

Table 3: Comparison of Cronbach’s Alphas in Ukraine versus U.S. Studies 

 

Procedural Justice 
Legitimacy 

Ukrainian 
Sample 

.94 
.74 

Southern 
California 

.92 
.81 

National  

U.S. 
.90 
.72 

     Sources: Gau, 2011 and Reisig, Bratton & Gertz, 2007 
 
 
Separating the Two Legitimacy Indexes.  No aspect of the Tylerian model’s 

 

measurement validity has received as much criticism as the lack of a sufficient relationship 

between the two variables (trustworthiness and obligation to obey) to defend combining them 

into a single legitimacy construct (Gau, 2011; Reisig et al., 2007; Tankebe, 2008; Tyler and Huo, 

2002).  The validity tests in Ukraine demonstrated the lack of divergence of the two legitimacy 

sub-indexes as noted in the weak correlation between trustworthiness and obligation to obey and 

the lack of a common factor in their EFA.  In the first published exploratory analysis on the 

factors in the Tylerian model, Gau (2011) found in a PCA of U.S. data that her measurement 

model proved a poor fit to the data and chose to split Tyler’s hypothesized indicator for 

legitimacy to produce a model with an adequate fit. 

 

 

58 

 

Gau (2011) in her U.S. study and Reisig and colleagues (2014) in a European study 

established the precedent to separate trustworthiness and obligation to obey to form independent 

mediating indicators of legitimacy in the Tylerian model.  The use of the trustworthiness 

indicator as a mediator in a legitimacy model is not unprecedented and can be found in Tyler’s 

organizational behavior research.  Specifically, De Cremer and Tyler (2007) reported that trust 

was a significant mediator between procedural justice and employee cooperation with authority 

in a business setting.  In order to best match Tyler’s intended obligation concept, a single-item, 

observed variable was selected as the indicator for obligation to obey in the model (APPENDIX 

B: Question 9.17 - You should do what the police tell you even if you disagree). The second 

option (Question 9.16 - You should accept police decisions even if you think they are wrong) has 

been criticized in past studies for the suggestion that obligation to obey is the same thing as 

amorally doing something you believe is wrong (e.g., Jackson, Bradford, Hough, Myhill, 

Quinton & Tyler, 2012).  Test item 9.17 simply requires obeying something that involves a 

disagreement without necessarily involving a question of right and wrong.   

Correlation Coefficients.  The bivariate relationships between the key indicators in the 

model were evaluated using correlation coefficients (Pearson’s r).  Following the suggestion of 

the EFA and the lead of others in the legitimacy literature, the tests included the separate 

indicators for legitimacy in the model (trustworthiness and obligation to obey; Table 4).   

Table 4: Correlations Between Key Indicators 

 
1.  Procedural Justice 
2.  Trustworthiness 
3.  Effectiveness 
4.  Obligation to Obey 
5.  Cooperation 
6.  Postmaterialism 
7.  Fear of Crime 
 **p < .001, *p < .01 

1 
- 

 .76** 
 .79** 
 .18** 
 .28** 
 .01 
-.30** 

2 
 
- 
 .66** 
 .25** 
 .23** 
-.01 
-.30** 

3 
 

 

- 
 .13** 
 .24** 
-.01 
-.32** 

 

 

59 

4 
 

 
 

- 
 .03 
 .05 
-.15 

5 
 

 
 
 

- 

 .11** 
-.15** 

 
 
 

6 
 

 
- 
.01 

 
 
 

7 
 

 
 
- 

 

The correlation coefficients demonstrated the lack of convergence between the two 

subscales of legitimacy in the Ukrainian path model, r = 0.25, p < .001.  The strongest 

correlations were between procedural justice and police effectiveness, Pearson’s r = .79, p < 

.001, and between procedural justice and trustworthiness, Pearson’s r = .76, p < .001.   

The lack of divergence between the procedural justice and legitimacy indicators in the 

model is likely influenced by the respondents’ problems in differentiating their views on very 

similar normative judgments (Gau, 2011; Maguire & Johnson, 2010; Reisig et al., 2007).  A 

review of the survey test items used for the classic process-based model (see Sunshine and Tyler; 

2003a) suggests potential problems due to the similarity of the questions in the disparate 

composite measurement scales.  This study used only the test items that were supported by past 

validity tests (Gau, 2011; Reisig et al., 2007), but questions over the ability of respondents to 

discriminate between test items measuring separate constructs remain.  For example, a test item 

forming the procedural justice indicator asks the respondent whether “the local police respect 

citizens’ rights;” while a test item in the trustworthiness scale that forms the legitimacy indicator 

asks whether the “people’s basic rights are protected by the local police.   

Certainly, Tyler’s model is built on a theory that argues that the quality of decision-

making of officers in past interactions is a different construct from a judgment on trusting the 

local police agency to protect citizen rights.  Nonetheless, the respondents are most likely to 

focus on the common mention of police and citizens’ rights for test items that then measure two 

independent constructs in the model.  Moreover, test items in the indexes for the procedural 

justice indicator do not always follow the theoretical case that they are intended to evaluate past 

experiences with the agency’s officers in order to provide contrast to the test items intended to 

measure judgments on the agency as a whole that form the legitimacy construct.  As a result, test 

 

 

60 

 

items found in different indicators (notably, procedural justice and trustworthiness) may 

compromise the constructs’ distinctiveness if respondents treat them as too similar. 

The high correlation coefficients between procedural justice and trustworthiness is 

corroborated by the common loadings in the EFA for procedural justice and trustworthiness.  

This is in line with past studies making use of factor analysis and correlation analysis.  Notably, 

Reisig and colleagues (2007) and Gau (2011; 2013) raised questions on the lack of discriminant 

validity between the indicators of procedural justice and trustworthiness of the agency in tests 

with U.S. data.   

The Ukrainian results not only replicate past validity problems but suggest that using the 

model’s key indicators in developing countries may also result in an additional problem. The 

coefficient between the two indexes forming the procedural justice indicator, Pearson’s r = .84, p 

< .001, is nearly equal to the coefficient between procedural justice and police effectiveness, 

Pearson’s r = .79, p < .001, or the coefficient found between procedural justice and 

trustworthiness, Pearson’s r = .76, p < .001.  Bollen and Lennox (1991, p.305) argue that 

“within-construct correlations must be greater than between-construct correlations” to have 

sufficient discriminant validity.  This is the case in the current Ukrainian results, but by a margin 

small enough to suggest problems with discriminant validity exists across these three indicators 

in the model. 

The only previous factor analysis of a Tylerian model using the police effectiveness 

indicator in another developing country reported similarly mixed results.  Although Karakus 

(2017) found that the test items for police effectiveness and procedural justice load on separate 

factors in his PCA of Turkish data; the correlation coefficient, r = .62, between police 

effectiveness and procedural justice is the single strongest bivariate relationship in his study.  

 

 

61 

 

This supported the findings in Ghana, where Tankebe (2009) reported that the bivariate 

relationship between procedural justice and effectiveness is the largest correlation, Pearson’s r = 

.56, p < .001.   

Past tests of the Tylerian model suggest that citizens in developing countries may view 

elements of police effectiveness (e.g., an officer arriving at your door in response to a call for 

assistance) as a part of their judgments on procedural justice.  In North America or Western 

Europe, police officers performing their required duties would likely be judged in terms of their 

effectiveness, but in societies policed by corrupt and indifferent law enforcement, citizens may 

evaluate the experience (e.g., an officer arriving at your door in response to a call for assistance) 

as combining effectiveness and procedural justice.  Of course, the opposite scenarios also explain 

this relationship if citizens evaluate police behavior as equally unjust and ineffective.  This 

would lead to considerable overlap in developing countries between the judgments of procedural 

justice and effectiveness and may account for the high correlation between the two indicators.   

Latent Variable Analysis 

Although the EFA addresses questions on the unidimensionality of indicators, Brown 

(1986) argues that a CFA is a more stringent test of construct validity.  A CFA allows for a 

“better measure of the latent variable,” because the instrument isolates the shared variance 

between measures from their unique variances (Acock, 2013).   

After specifying the model, overall test of fit was tested for the models.  Because of their 

susceptibility to sample size and model complexity, χ2 tests nearly always reject models with 

large (n > 400) sample sizes (Hooper et al., 2008).  Therefore, large sample-size studies 

generally must utilize other test statistics to assess model fit.  Hu and Bentler (1999) found that 

the use of the root mean square error of approximation (RMSEA) reduced errors in model fit 

 

 

62 

 

judgments in large-sample studies.  Sivo, Xitao, Witta and Willse (2006) reported that RMSEA 

and standardized root mean square residual (SRMR) were among the only fit tests that did not 

require different cut off values as sample size increased.  The baseline comparison index that will 

test model fit is the Tucker-Lewis index (TLI) that stands out for penalizing for simply adding 

parameters to gain model fit.  This study’s latent variable analyses fit was assessed via the 

RMSEA, SRMR, and the TLI as a baseline comparison.  In assessing model fit, this study used 

the following cutoff criteria: RMSEA < .08; TLI > .95; SRMR < .08.  These criteria generally 

indicate acceptable model fit (Hooper et al., 2008; Hu & Bentler, 1999; Sivo et al., 2006). 

A confirmatory factor analysis examined the relationships between the manifest variables 

and the hypothesized latent factors that will be part of the study’s research questions (see Figure 

7).  The CFA provided support for the latent variables representing the constructs first identified  

in the EFA.  The CFA had an adequate fit of the data across the goodness of fit tests: χ2(231) = 

1,648.2, p < .001; TLI = .93; SRMR = .06; RMSEA = .05.  The TLI fit test was slightly below 

the rule of thumb (.95), but is known to penalize a model that includes experimental variables 

that have low correlations with the other variables in the factor analysis (Kenny, 2015).  This was 

the case in the CFA as the correlation coefficients for the hypothesized normative variables 

(postmaterialism and fear of crime) were trivial (R2 = .01) in half their bivariate relationships. 

Path Analysis 

After using the preliminary data tests, EFA, and CFA to establish the test items and latent 

variables that would model Ukrainian judgments on police behavior and outcomes, SEM was 

used to test the key relationships in the Tylerian model of police legitimacy (Figure 8).  The path  

  

 

 

63 

ε8 .32

ε9 .36

ε10.63

q9_13_basicrights

q9_14_trust

q9_15_jobwell

.82

.8

.61

-.022

 

Figure 7: Confirmatory Factor Analysis 

 

ε4 .34

q9_1_respect

ε5 .3

q9_2_listen

ε6 .29

q9_3_treatfair

ε7 .27

q9_4_respect

q9_5_courteous
ε14.46

q9_6_factbased
ε15.39

ε16.42

q9_7_explain

ε17.33

q9_8_fairly

.81

.84
.85

.78
.76

.82

.88

-.016
.78

.86
Procedural_Justice
.74

Trustworthiness

2.3
Obligation
1

.85

Effectiveness

.81

.83

q8_2_order

-.043
.81

.81

.86
q8_4_prompt

PostMaterialism

5.3

1

.12

Cooperation

.24

.26
.29

-.027

Fear_of_Crime

-.018

.7

.84

.69

q8_6_fear

.79
q8_8_reasontofear

q11_1_call

ε11.37

q11_2_report ε12.32

.8

.82
.71

-.078

q11_4_info ε13.49

q8_5_prevent

q8_1_duties

ε1 .34

ε2 .31q8_3_try

ε18.35

ε19.34

ε3 .26

ε20.51q8_7_worry

ε22.29 q8_9_lovedones

ε21.38

ε23.53

 

All estimates p < .001 
Goodness of Fit Measures: χ2(263) = 2,019.64, p < .001; TLI = .934; SRMR = .061; RMSEA = .05

 

 

64 

 

Figure 8:  Path Analysis 

ε4 .35

1_respect

ε5 .3

2_listen

ε6 .29

3_treatfair

ε7 .27

4_respect

ε17.46

5_courteous

ε18.39

6_factbased

ε19.42

7_explain

ε20.33

8_fairly

ε10.32

ε11.36

ε12.63

13_basicrights

14_trust

15_jobwell

.82

.61

.8

ε9 .22

PostMaterialism

5.3

1

.81

.84
.84

.85

.74

.78
.76

Trustworthiness

**
.8

*
.096

-.051

**
.3

**
.25

*
-.11

Obligation ε8 .97

.042

Proc_Justice

**
.85

Effectiveness

.82

.81

.83

.81

.81

.86

2_order

4_prompt

5_prevent

fear

ε23.51

ε2 .313_try

ε21.35

ε22.34

1_duties

ε1 .35

ε3 .26

**
.12

1_call

ε14.37

.8
Cooperation ε13.9
.82
.71

*
-.079

2_report

ε15.32

4_info

ε16.49

-.032

.7

Fear_of_Crime
.69

.84

.79

lovedones

worry

reasontofear

ε26.53

ε24.38

ε25.29

 

   ** p < .001;* p < .01; non-significant estimates represented by .25 lines. 
   Goodness of Fit Measures: χ2 (267) = 1,988.02, p < .001; TLI = .936; SRMR = .050; RMSEA = .057; R2 = 0.99. 

 

 

65 

 

analysis estimated the strength of the parameters and relationships among Tyler’s antecedents, 

the legitimacy indicator, and the willingness to cooperate with police as well as the role of the 

two variables (postmaterialism and fear of crime) hypothesized in the study to predict 

cooperation with the police.  The initial run of the path model had a poor fit of the data across the 

goodness of fit tests: χ2(267) = 3,905.0, p < .001; TLI = .87; SRMR = .08; RMSEA = .08. 

Given the high correlation between the model’s two antecedents, a covariate was placed between 

procedural justice and effectiveness.  After this covariance was included the SEM had an 

adequate fit to the data: χ2(267) = 1,988.02, p < .001; TLI = .94; SRMR = .05; RMSEA = .06.   

The high correlations among some of the scales used in the SEM led to testing the 

indicators for possible multi-collinearity.  The presence of multi-collinearity could limit the 

significance of the relationships between the collinear indicators.  Therefore, a variance inflation 

factor (VIF) test checked whether the four variables regressed on cooperation demonstrated 

multicollinearity (Table 5).  The VIF mean (2.47) was well below the general rule-of-thumb 

(4.0) for potential multicollinearity.  Moreover, the study’s large sample size further reduced 

concerns of multicollinearity.  Grewal and colleagues (2004) found large sample size 

significantly reduced the concern for Type II errors in SEM analyses 

Table 5: Variance Inflation Factor 

Variable 

Procedural Justice 
Effectiveness 
Trustworthiness 
Obligation to Obey 
              Mean VIF 

VIF 
3.60 
2.68 
2.52 
1.07 
2.47 

1/VIF 
.28 
.37 
.40 
.94 
 

 

The strength of the key relationships in the path analysis address the first five of the 

study’s research questions on whether: procedural justice predicts cooperation with the police; 

 

 

66 

 

legitimacy mediates the relationship between procedural justice and cooperation; legitimacy 

mediates the relationship between effectiveness and cooperation; post-materialism values predict 

cooperation with the police; and the fear of crime predicts cooperation with the police.  

Does Procedural Justice Predict Cooperation with the Police?  The study used the 

path model to assess the dependent relationship between the key indicators in Research Question 

#1 (Do perceptions of procedural justice predict cooperation with the police in Ukraine?).  In the 

path model, the regression coefficient between procedural justice and cooperation was 

significant, β = 0.30, p < .001, and in the direction hypothesized (the greater the perceptions of 

procedural justice, the higher the willingness to cooperate with the police). Therefore, the null 

hypothesis can be rejected for the research question on whether procedural justice has a direct 

association with the outcome variable cooperation with the police. 

Does Legitimacy Mediate the PJ – Cooperation Relationship? The SEM also served 

to test Research Question #2 (Do perceptions of police legitimacy mediate the relationship 

between measures of procedural justice and public cooperation with the police in Ukraine?).  In 

this study, legitimacy was split into the two separate components of trustworthiness and 

obligation to obey.  The SEM found that in each case procedural justice was associated with the 

variables of trustworthiness and obligation to obey.  Nevertheless, the path analysis also 

demonstrated that trustworthiness and obligation to obey did not have a significant relationship 

with the outcome variable cooperation with the police.  As a result, the null hypothesis cannot be 

rejected because obligation to obey and trustworthiness cannot mediate the relationship if they 

are not associated with the outcome variable of cooperation. 

In the SEM, the two antecedent variables in the model were limited to associations with 

trustworthiness and obligation to obey.  Specifically, procedural justice explained 78% of the 

 

 

67 

 

variance in the latent variable of trustworthiness, R2 = .78, while accounting for just 3% of the 

variance in the obligation to obey variable, R2 = .03.  The SEM also provided standardized beta 

coefficients for the key relationships.  In this path analysis, the regression coefficient between 

procedural justice and trustworthiness was far stronger, β = 0.89, p < .001, than the coefficient 

representing the strength of the relationship between procedural justice and obligation to obey, β 

= 0.16, p < .001.  These results are in accordance with past U.S. studies that found a weaker 

relationship between procedural justice and obligation to obey than between procedural justice 

and the trustworthiness component of legitimacy (Reisig et al., 2007 and Gau, 2011).  

In past U.S research, the obligation to obey has frequently not had a relationship with the 

outcome variable (Reisig et al., 2007).  In the tests in two countries near Ukraine, Reisig and 

colleagues (2014) found in a Slovenian student sample that trustworthiness alone predicted 

cooperation, while Karakus (2017) found a legitimacy construct that combined trustworthiness 

and public satisfaction predicted cooperation with the police.  Tankebe (2008) in Ghana and Sun 

and colleagues (2017) in China also found that neither trustworthiness or obligation to obey 

mediated the model. 

In comparing the two indicators of legitimacy in the Ukrainian path model, procedural 

justice had the weakest relationship with the legitimacy indicator obligation to obey, β = 0.25, p 

< .001.  Obligation to obey did not have a significant association with cooperation, which 

prevents the indicator from serving as a mediator in the process-based model.  As a result, the 

Ukrainian study supports the decision of some researchers (e.g., Hinds & Murphy, 2007; 

Karakus, 2017) that have chosen not to include obligation to obey in their legitimacy construct.   

The path analysis with the split legitimacy indicators demonstrated that Ukraine stands 

out among developing countries in the strength of the association between procedural justice and 

 

 

68 

 

the trustworthiness indicator of legitimacy, β = 0.80, p < .001.  A lack of discriminant validity 

between procedural justice and trustworthiness may serve to exaggerate the degree of 

association, but a wealth of studies primarily with U.S. data provide a precedent for a significant 

relationship between procedural justice and legitimacy.   

The strong association between procedural justice and trustworthiness in the Ukrainian 

path analysis differs from the findings in other developing countries, such as Pakistan, Nigeria, 

and Ghana, where tests found a weaker association between procedural justice and police 

legitimacy (see Akinlabi, 2018; Jackson et al., 2014; Tankebe, 2009).  Sun and colleagues (2017) 

found that in China procedural justice predicted legitimacy but not as strongly as police 

effectiveness predicted legitimacy.  This difference across countries may mirror the findings of 

Hough and colleagues (2013) that European countries differ considerably across Europe from the 

economically prosperous West Europe to the still developing East Europe in the strength with 

which procedural justice predicts indicators of police legitimacy.  The Ukrainian path analysis 

supported the case made by past legitimacy researchers (e.g., Gau, 2011; Reisig et al., 2007; 

2014; and Tankebe, 2009) who advocate for a process-based model of police legitimacy that uses 

the trustworthiness measure as an independent mediator between procedural justice and the 

cooperation with the police.   

The lack of a mediating role for the two indicators of legitimacy in the path analysis is in 

line with the previous tests of the classic Tylerian model in developing countries.  In the two 

tests from Central Europe, Czapska and colleagues (2016) report that neither procedural justice 

nor police legitimacy predicted cooperation in Poland, while Reisig and colleagues (2014) 

indicate that legitimacy explained a modest amount of variation in cooperation with the police in 

Slovenia.  The Ukrainian result adds one more study to the growing list of research suggesting 

 

 

69 

 

that the Tylerian construct of a legitimacy is not a valid mediator in the process by which 

procedural justice influences police outcomes in even prosperous developing countries. 

Does Legitimacy Mediate the Effectiveness – Cooperation Relationship?  The path 

model failed to support the hypothesis proposed in Research Question #3 (Do perceptions of 

police legitimacy mediate the relationship between measures of effectiveness and cooperation 

with the police in Ukraine?).  As established in Research Question #2, a mediating role for the 

two legitimacy indicators was undermined by the lack of an association between the two 

legitimacy indicators and cooperation with police.   

The path analysis found that effectiveness was modestly associated with the two 

mediating indicators.  Specifically, the regression coefficient between effectiveness and 

trustworthiness was, β = 0.10, p < .001, in the direction hypothesized.  The greater the perception 

of police effectiveness the higher the judgments of legitimacy.  The regression coefficient 

representing the strength of the relationship between effectiveness and obligation to obey also 

was quite modest, β = -0.11, p < .001, and in a direction different than hypothesized.  The greater 

the perceptions of police effectiveness, the less support for an obligation to obey the police.   

OLS Comparison.  The weak role of effectiveness compared to procedural justice in 

predicting the indicators of legitimacy had not been found in previous tests of the Tylerian model 

in a developing country.  Therefore, an Ordinary Least Squares (OLS) regression test was run to 

determine the standardized regression coefficients of procedural justice and effectiveness in 

predicting Tyler’s classic, two-dimensional indicator of legitimacy and for demonstrative 

purposes compare them to the large set of past OLS regression tests of the predictors of 

legitimacy. 

 

 

70 

 

As a first step, a first-generation multivariate method, OLS regression, tested the strength 

of police effectiveness and other predictors of the two-dimensional measure of police legitimacy.  

The standardized regression coefficients allowed for a comparison with the results of previous 

research on the antecedents of legitimacy in other countries that primarily used OLS regression 

for their results.  The past OLS tests of the antecedents to legitimacy used the classic construct of 

trustworthiness and obligation to obey to form the dependent variable.  This study’s OLS 

regression used the classic indicator for legitimacy in the case of this one test to allow for a 

comparison of the relative strength of the predictors in relation to previous research results.   

The Tylerian legitimacy scale was regressed on the respondents’ judgments of procedural 

justice, police effectiveness, and the demographic variables for gender, age, and education level 

(see Table 6).  The significance of the police effectiveness indicator, β = 0.16, p < .001, in the 

Ukrainian data was not a surprise given the strength of effectiveness in past tests of the model in 

developing countries (Akinlabi, 2018; Jackson et al., 2014; Karakus, 2017; Sun et al., 2017; 

Tankebe, 2009).  The OLS regression reaffirmed the finding from the earlier path analysis on the 

Ukrainian data that the normative values of procedural justice, β = 0.64, p < .001, dominated the 

instrumental variable police effectiveness in predicting police legitimacy.  The demographic 

controls did not have a significant role in the police legitimacy model.   

Table 6: OLS Regression Model of Police Legitimacy 
SE 
0.009 
0.121 
0.016 
0.056 
0.048 

Independent Variables 
Procedural justice 
Police effectiveness 
Age 
Gender 
Education 

         *p < .05, ** p <.01 (two-tailed test). 
 

Coefficient 

0.23 
0.08 
-0.01 
0.09 
0.01 

t-ratio 
0.000** 
0.000** 
0.902 
0.083 
0.014 

β 
0.62 
0.16 
-0.01 
0.03 
0.02 

The dominant role for procedural justice in predicting legitimacy was not the case in 

previous developing country research.  Tankebe (2008) found relative parity between the 

 

 

71 

 

influence of procedural justice and effectiveness in predicting legitimacy in Ghana.  Among the 

countries straddling the status of developed and developing country, studies in Poland and 

Slovenia supported a dominant role for procedural justice in predicting agency legitimacy 

(Czapska et al., 2016; Reisig et al., 2014), while Karakus (2017) found relative parity between 

procedural justice and effectiveness as predictors of legitimacy in Turkey.  

The SEM analysis supported the research question’s null hypothesis in finding that 

perceptions of police performance (effectiveness) did not mediate the impact of perceptions of 

officer procedural justice on public cooperation with the police.  This contradicts the two 

previous studies that examined the relationship between effectiveness and cooperation with the 

police in developing countries.  Tankebe (2009) reported that in Ghana procedural justice 

predicted cooperation only in the model that did not add police effectiveness, while Sun and 

colleagues (2017) reported that police effectiveness in China was one of several significant 

predictors of cooperation with the police and equal in strength to legitimacy and neighborhood 

cohesion.  Though Ukraine is a developing country that scores lower than Ghana and China in 

rankings on institutional corruption and ineffective government institutions (TI, 2013, WGI, 

2016), citizen judgments on police effectiveness appear to have a weaker influence on the 

willingness to cooperate with the police in Ukraine.  The direct relationship between procedural 

justice and cooperation with the police was the sole significant path in the full model.   

The mixed results in the tests of the relationship between procedural justice and 

cooperation in developing countries suggests that the state of effective and democratic 

governance or prosperity in a country does not solely dictate the degree that just police 

procedures can influence police outcomes in developing countries.  In past tests in three 

developing countries and two borderline developed countries (Poland and Turkey), procedural 

 

 

72 

 

justice was found to be weaker than police effectiveness as a predictor of either police legitimacy 

or the willingness to cooperate with the police (Akinlabi, 2018; Czapska et al., 2016; Jackson et 

al., 2014; Karakus, 2017; Tankebe, 2009).  At the same time, this study’s findings in Ukraine and 

past research in Slovenia and China suggest that procedural justice can still serve as a predictor 

of police outcomes in borderline developed or developing countries (Reisig et al., 2014; Sun et 

al., 2017).  The mixed results in these studies leave open the question on whether developing 

countries themselves differ in a key attribute or set of qualities that predict the influence of 

procedural justice in their judgments on police outcomes. 

Instrumental vs. Normative Concerns.  The remarkable growth in research based on the 

Tylerian theory of legitimacy is arguably a direct result of the early findings in the United States 

and then other developed countries that normative values (procedural justice) have a larger role 

predicting police outcomes (e.g., cooperation with the police) than do instrumental values (e.g., 

police effectiveness).  This finding is the key to the theory’s most important policy implication 

that the optimal strategy for building legitimacy and trust in the police is to prioritize the pursuit 

of fair and respectful procedures and not necessarily effective outcomes.  The preeminence of 

normative over instrumental values has been established across a wide spectrum of communities 

in developed countries (Hinds & Murphy, 2007; Hough et al., 2013; Murphy, 2009; Tyler, 2002).  

Nevertheless, the recent wave of research in developing countries suggests far weaker support 

for the role of normative values in predicting police outcomes in countries challenged in terms of 

economic prosperity and good governance (Reisig & Lloyd 2007; Sun et al., 2017; Tankebe, 

2009).  

The SEM analysis in this study found that Ukrainians strongly associate procedural 

justice (β = 0.82) and less strongly police effectiveness (β = 0.22) with the model’s indicator of 

 

 

73 

 

legitimacy, trustworthiness.  To compare this result with the findings from other countries, this 

study proposes a values ratio (v) consisting of the procedural justice coefficient divided by the 

police effectiveness coefficient.  The Ukrainian values ratio, v = 3.72, reveals a strong role for 

the normative values in predicting judgments of police legitimacy.  The values ratio may offer a 

rough means to compare the balance of normative and instrumental values across different 

countries.  For descriptive purposes, Table 7 plotted estimates for the values ratios based solely 

on the coefficients reported between the antecedent variables (procedural justice and police 

effectiveness) and police legitimacy in past studies.  

Table 7: Normative versus Instrumental Predictors of Legitimacy 

Country 

 

Effectiveness 

Proc. 
Justice 
0.74 
0.71 
062 
0.47 
0.33 
0.46 
0.46 
0.35 
0.35 
0.28 
0.34 

0.15 
United States 
0.16 
Sweden 
0.16 
Ukraine 
0.15  
Poland 
0.13 
Sloveniaa 
Japan 
0.27 
Australiab 
0.28 
Ghana 
0.34 
0.40 
Turkey 
0.38  
Chinaa 
0.63 
Pakistan 
aEntries are unstandardized regression coefficients. 
bAlternative studies in Australia report coefficients suggesting a slightly lower ratio. 
Sources of Data: Czapska et al., 2016; Jackson et al., 2014; Karakus, 2017; Murphy & Cherney, 2011; Reisig 
et al., 2014; Sun et al., 2017; Sunshine & Tyler, 2003a; Tankebe, 2008; Tsushima & Hamai, 2015; Van 
Damme et al., 2013. 

Ratio 
4.93 
4.73 
3.94 
3.13 
2.54 
1.70 
1.64 
1.03 
0.88 
0.74 
0.54 

Value ratios compare the relative role of normative versus instrumental predictors of 

 

judgments on police legitimacy.  The studies supplying the coefficients differ in terms of how 

they operationalize the model’s key measures and two of the studies used non-standardized 

coefficients.  Moreover, the studies sometimes differed in the control variables added to the 

regression of legitimacy on procedural justice and effectiveness.  Nevertheless, a ratio of 

 

 

74 

 

coefficients from the same regression equation provides a rough representation of the relative 

balance between the two values.  The quotient of the normative versus instrumental factors in 

Ukrainian judgments on police legitimacy produces a high values ratio for a developing country, 

v = 3.94.  The ratio is more in line with EU-member countries Poland and Slovenia that also 

were once part of the former Soviet bloc.   

With the possible exception of the developing countries in the former Soviet bloc, the 

table of values ratios suggests that countries challenged by systemic corruption and poor 

governance (e.g. Pakistan) have much lower quotients probably because of the strong role that 

police effectiveness has as a predictor in the country.  As a result, those countries with the 

greatest social and economic challenges appear to produce values ratios that are near or below 

1.0 as the instrumental evaluations exceed procedural justice as a predictor of the public’s 

judgments on police legitimacy.   

The overall order of the values table is probably partly the result of the public’s low 

expectations in regard to police performance in poorly governed and corrupt countries.  Tankebe 

(2009) theorizes that in an environment where policing is “characterized by abuse, violence, 

intimidation, and widespread corruption,” (p. 1271) the judgment to obey the police may be 

driven by a fear of the consequences, rather than sentiments of trust in the police or respect for 

their authority.  Specifically, the public in these countries regularly weighs the risk of calling the 

police to a crime scene which can increase the likelihood of further victimization.  In societies 

where the basic functions of the police (e.g., crime prevention, investigation of a reported crime) 

are not taken for granted, procedural justice and legitimacy judgments may be driven to a large 

degree by respondents’ views on the effectiveness of the police.  

 

 

75 

 

As more tests of police legitimacy theory in developing countries fail to demonstrate the 

same strong normative role in judgments on agency legitimacy that were found in developed 

countries, policing scholars must begin to test different explanations for differences in results 

across countries.  This study’s values ratios help to depict the diverse weights that individuals in 

different societies place on the normative and instrumental factors in reaching their judgments 

about the police.  The cross-country comparison of the differences in the values ratio calculations 

can be further highlighted through the use of a continuum that ranks the ratios from lowest to 

highest (see Figure 9).  

The continuum merely serves a demonstrative purpose because there are significant 

differences in the models in which the two regression coefficients were taken to form the ratio 

for each country. The graphic highlights the presence of a cluster of developed and developing 

countries.  Developing countries are primarily located on the lower end, where the instrumental 

values in the denominator of the ratio are greater.  The OECD member countries are spread out 

above the ratio value of 1.5, which likely represents a significantly greater role for normative 

over instrumental factors.  Therefore, the continuum suggests that the relative status of a 

country’s economic and political development predicts the degree that normative values 

influence society’s judgments on the legitimacy of the police.  In other words, the values ratio 

continuum supports the key conclusion of Inglehart’s Theory of Value Shift.  Specifically, 

Inglehart (1990) argued that a society’s relative concern for normative values, such as human 

rights, over more immediate instrumental concerns is based on the degree that a generation 

raised in the country could take survival for granted and increasingly prioritize normative 

concerns beyond subsistence. 

 

 

 

76 

 

Figure 9: Normative/Instrumental Values Ratio Continuum 

	
	
	
	

	
	
	
	

	
	
	
	
	
	
	

	
	
	
	

n
a
t
s
i
k
a
P

y
e
k
r
u
T

a
n
a
h
G

a
n
i
h
C

a
i
l
a
r
t
s
u
A

n
a
p
a
J

	
	
	

	
	

a
i
n
e
v
o
l
S

d
n
a
l
o
P

	
	

e
n
i
a
r
k
U

	
	
	
	
	
	
s
e
t
a
t
S
d
e
t
i
n
U

n
e
d
e
w
S

	

00.250.50.7511.251.51.7522.252.52.7533.253.53.7544.254.54.755

 

Values Ratio 

 
 

Ukraine appears an exception on the continuum because the country is ranked with the 

less developed countries in many developmental criteria, such as institutional corruption and 

ineffective governance, (TI, 2013, WGI, 2016), but is located higher on the values scale 

continuum than developing country peers.  Nevertheless, the Ukrainians share many of the same 

experiences as Central European neighbors in Poland and Slovenia which are generally classified 

as rapidly developing countries and also demonstrated a stronger influence of normative over 

instrumental values in the scores on their values ratios.  Japan’s position near the border between 

developing and developed countries is in line with Inglehart and Welzel’s (2005) findings for 

Japan’s comparative position on the self-expression versus survival index.  Inglehart and Welzel 

(2005) surveys found that the normative versus instrumental concern in Australia were similar to 

those in the United States, which suggests that Australia’s low position in the continuum is rather 

 

 

77 

 

unexpected considering the expectation that procedural justice concerns would be a strong 

predictor of legitimacy in a society that consistently prioritizes self-expression over survival 

values. 

Ukraine’s apparent inflated position for a developing country in the values ratio 

continuum could also be a result of the difficulty Ukrainian respondents have in discriminating 

between normative and instrumental concerns.  The study already noted the unusually high 

correlation between procedural justice and effectiveness in the Ukrainian data.  Moreover, 

procedural justice accounted for nearly three-quarters of the variance in the latent variable of 

effectiveness (R2 = .74).  The strong association suggests that Ukrainian respondents may view a 

fair procedure as one that also is effective to the degree that the officer is simply doing his or her 

job rather than taking part in the systemic system of organized criminal activity generally 

associated with the post-Soviet militia.   

The phenomenon of expectancy disconfirmation developed in consumer research may 

have a role in the Ukraine respondents’ surprising prioritization of normative concerns.  The 

extant literature suggests that communities marked by low expectations on a service are more 

likely to give higher marks for the same behavior (e.g., Cardoza, 1965; Hero & Durand, 1985).  

The phenomenon has often been likened to a person’s evaluation of the temperature outside 

differing based on whether they have spent the last few months in artic or tropical conditions 

which will condition them differently in making their judgments (Van Ryzin, 2013).  Reisig and 

Chandek (2001) argued that expectancy disconfirmation explained positive judgments in a U.S. 

survey on interactions with the police.  Certainly, the Ukrainian public have consistently 

experienced poor law enforcement for several decades which suggests that expectancy 

 

 

78 

 

disconfirmation could help explain the strength of the regression coefficient between values of 

procedural justice and legitimacy. 

The research question on the impact of police effectiveness also asks whether the 

instrumental values of police performance influence perceptions of public cooperation with the 

police.  In Ukraine, the effectiveness variable in the path analysis was not significantly 

associated with cooperation with the police.  This runs counter to the tests in developing 

countries that examined the relationship between effectiveness and cooperation.  Most notably, 

Tankebe (2009) in Ghana and Czapska and colleagues (2016) in Poland found that police 

effectiveness was the sole Tylerian construct that predicted cooperation with the police.  The 

Ukrainian finding that effectiveness did not predict the police outcome indicator appears more in 

line with U.S. tests that found modest or no influence of police effectiveness on residents’ 

willingness to cooperate with the police (Davis & Henderson, 2003; Fagan & Tyler, 2004; 

Sunshine & Tyler, 2003; Warner, 2007).  

Do Postmaterialism Values Predict Cooperation with the Police?  The path analysis 

also served to test Research Question #4 (Do post-materialism values predict cooperation with 

the police in Ukraine?).  The postmaterialism indicator formed as an additive variable from the 

test items of Inglehart and Baker’s self-expression vs. survival index (labeled PMaterialism in 

the path analyses).  Postmaterialism was added as a predictor of the outcome variable in the 

Tylerian model.  The variable is hypothesized as reflecting whether the sampled population 

prioritizes normative versus instrumental values in their judgments on social issues.  The path 

analysis tested whether postmaterialism is associated with the willingness to cooperate with 

police.  In figure 8, the postmaterialism indicator significantly predicted cooperation, β = 0.12, p 

< .001.  

 

 

79 

 

Although a modest association, the Postmaterialism indicator predicted cooperation in the 

direction hypothesized.  Those citizens that gave greater weight to normative values in their 

survey judgments were more likely to cooperate with the police.  The Postmaterialism index is 

the second strongest predictor of cooperation with the police in the Ukrainian model and the 

third largest regression coefficient in the previous tests of independent variables predicting 

cooperation with police in the only tests of the full Tylerian model in developing countries 

(Reisig & Lloyd, 2009; Sun, 2017; Tankebe, 2009).  The addition of the self-expression vs. 

survival index to the model raised (22%) the model’s explanation of variance for the dependent 

variable cooperation with the police (R2 = .11).7 

The goal of the test of Postmaterialism predictor of Cooperation of the Police was to 

place in the model a second variable reflecting normative values.  Procedural justice differs from 

the Postmaterialism index by representing normative judgements on past interactions with police, 

but they support Tyler’s (2009) argument that his model is based on normative values.  Tyler 

(2009) viewed the shared normative values forming legitimacy judgments would also associate 

strongly with the internal motivation (normative values) for choosing to cooperate with the 

police.  The Postmaterialism test supports Tyler’s proposition that normative values drive the 

associations between the key variables.  The significant role of Postmaterialism also serves to 

further isolate the indicator of trustworthiness in Tyler’s models that fails to support Tyler’s 

proposition in developing country tests of the model.  While Ukrainian citizens reach favorable 

normative judgements in their evaluation of procedural justice and in the postmaterialism survey, 

their judgments on trust in the police institution fail to associate with the normative values 

favorable to cooperating with the police.  As a result, the heart of the Tylerian process-based 

                                                
7 Across all the tests of the full Tylerian model in developing countries, the model’s explanation of variance for the 
dependent variable rarely exceeds R2 = .11. 

 

 

80 

 

model (legitimacy) fails to predict the willingness to cooperate with the police in the fourth of 

four tests of the full, classic Tylerian model of police legitimacy in a developing country.   

Does Fear of Crime Predict Cooperation with the Police?  The path analysis also 

served to evaluate Research Question #5 (Does fear of crime predict cooperation with the police 

in Ukraine?).  The path analysis found a significant though modest path in the hypothesized 

direction between fear of crime and cooperation with the police, β = -.08, p < .001, and in the 

direction hypothesized.  Specifically, lower levels of fear of crime were associated with greater 

cooperation with the police.   

The weak and uncertain role of the fear of crime in predicting cooperation with the police 

is not surprising.  The role of the fear of crime indicator was hypothesized in the study as a 

sponge reflecting citizens’ concerns over social and physical instability.  The role of the indicator 

was dependent on whether Ukrainian judgments on the fears of crime and disorder in their 

community reflected the relative strength of normative goals over values of personal survival.  If 

this were the case, the impact on the internal normative values that Tyler proposes shape the 

decision to cooperate with the police was negligible. 

Does Interaction with the Police Bias Perceptions of Procedural Justice? 

This study’s last set of factor analyses tested Research Question #6 (Does measurement 

invariance hold for the procedural justice construct regardless of whether the respondent had 

recent experience with the police?).  In the legitimacy theory literature, the procedural justice 

construct is explained as judgments on the fairness of interactions with the police.  This 

antecedent variable in the model is theorized in the legitimacy literature as reflecting judgments 

on past experience with the police.  Contemporary studies using the Tylerian model are based on 

the expectation that survey participants that lack direct experience with the police will not bias 

 

 

81 

 

the model’s findings.  This validity property is usually referred to as measurement invariance 

which indicates that a common construct (e.g., procedural justice) is being measured across 

different groups.  For this reason, the study adopted a series of CFAs to test the procedural 

justice index in Ukraine on whether measurement invariance holds across respondents whether 

they had an interaction with the police or lacked experience with the police.  

Police Interaction.  Following the lead of the psychometric studies profiled in 

Vandenberg and Lance (2000), this study uses a sequence of multi-group confirmatory factor 

analyses (MG-CFA) to test the procedural justice construct for measurement invariance.  The 

MG-CFA of ordered categorical data is a test based on a series of increasingly restrictive models.  

First, a free parameter CFA across the two groups was performed to determine whether 

configural invariance existed.  The factor-indicator relationships for procedural justice were 

identical in the direction hypothesized (the results on the sequence of four CFAs are presented in 

Table 8).   

Table 8: Metric Fit Indices for Invariance Tests Across Interaction with Police  
 

Δdf 

p value 

ΔCFI  RMSEA 

 
 
Configural Invariance 
4 
Metric Invariance 
4 
Scalar Invariance 
Variance Invariance 
2 
Note: N = 1,964 (No police interaction/n = 1,495, Police interaction/n = 469) 
 

χ2 
7.85 
8.92 
11.57 
243.28 

χ2/df 
1.96 
1.12 
0.96 
17.37 

Δχ2 

 
1.07 
2.65 
231.71 

df 
4 
8 
12 
14 

 

0.34 
0.48 
0.00 

CFI 
0.99 
1.00 
1.00 
0.92 

 

0.00 
0.00 
0.04 

0.03 
0.01 
0.00 
0.13 

After establishing configural invariance, a second CFA that constrained the measurement 

coefficients to be equal (mean = 0; variance = 1) was used to measure the metric invariance that 

is arguably the most important test for determining whether a data set is measurement invariant 

across different groups.  The goodness of fit test for metric invariance was not significant 

between the two groups, χ2(8) = 8.92, p = .39, and the null hypothesis of measurement invariance 

could not be rejected.  The third CFA tested scalar invariance between the two groups by 

constraining not only the coefficients to be equal but doing the same to the intercepts.  The scalar 

 

 

82 

 

invariance CFA also was not significant, χ2(12) = 11.57, p = .48, and the null hypothesis of 

measurement invariance could not be rejected.  

The last MF-CFA test for determining pure measurement invariance (variance 

invariance) imposes constraints on the coefficients, intercepts, and error variances while 

allowing the mean and variance of procedural justice to differ across the two police interaction 

groups.  In this case, the lack of experience with the police grouping was constrained while those 

with recent police interaction could be estimated freely.  As reported in Table 7, this variance 

invariance test had a significant result, χ2(14) = 243.3, p < .001, therefore, the null hypothesis of 

measurement invariance was rejected by the most stringent test of invariance.   

Though the χ2-test was significant, the last confirmatory factor analysis resulted in an 

insignificant latent mean difference (-0.06, p = .22).  In an MG-CFA analysis, a latent mean is 

not reported but rather the difference in means across the two groups.  Therefore, the CFA 

estimated that the difference in latent means between the group with direct experience with the 

police and the group with no recent experience with the police was not significantly different 

from 0.  Measurement invariance has to be rejected because of the instrument bias found in the 

variance invariance test, but the insignificant coefficient that results suggests that experience 

with the police did not meaningfully alter the mean of the resulting latent variable (procedural 

justice).  In other words, across two groups that differed in having experience with police, there 

was no meaningful difference in their common interpretation of the items forming procedural 

justice. 

In a perfect world, measurement invariance would indicate that the respondents across the 

two separate groups (direct experience with the police; no recent police interaction) interpret the 

test items in the exact same manner.  In practice, studies based on large sample sizes rarely 

 

 

83 

 

support the null hypothesis across all four MG-CFA tests.  Instead of demonstrating full 

measurement invariance, psychometric researchers seek measurement indicators that 

demonstrate strong though partial levels of invariance (Steenkamp & Baumgartner, 1998; 

Milfont & Fischer, 2010).   

In making use of MG-CFA tests, a researcher should consider the type of instrument that 

is being tested.  Few question the necessity of demonstrating full measurement invariance for a 

scholastic aptitude test or other standardized instrument in which a small measurement 

invariance would mean a bias favors some respondents over others.  On the other hand, partial 

levels of invariance are common for questionnaires used to assess aggregate judgments on 

institutions.  Most social science studies can accept some bias across groups that does not 

significantly impact the estimates for the indicators used in the model. 

A key question in examining the results of MG-CFA tests is whether this lack of pure 

measurement invariance is caused by instrument bias or a true difference between the two 

groups.  This question is difficult to answer in part because of the lack of past studies 

establishing measurement invariance for the process-based model of police legitimacy.  To help 

address that question, the same MG-CFA tests were performed on the sample’s three 

demographic characteristics (gender, age, and education level) to provide a baseline of 

instrument invariance across other groups. 

Gender.  The four sets of MG-CFAs across gender (see Table 9) were not significant in 

the first three invariance tests under different degrees of constraint.  As with police interaction, 

the last CFA rejected the null hypothesis of measurement invariance after constraining error 

variance.  Nevertheless, the coefficient of procedural justice, β = -0.06; p = .14, that represents 

the difference in means across gender groups, was not significantly different from 0.  This 

 

 

84 

 

suggests that the limited bias that exists across the two genders that did not significantly impact 

judgments on procedural justice. 

Table 9: Metric Fit Indices for Invariance Tests Across Gender  
 

Δdf 

χ2 
9.89 
10.69 
15.66 
206.51 

 
Configural Invariance 
Metric Invariance 
Scalar Invariance 
Variance Invariance 
Note: N = 1,964 (Female/n = 1,078, Male/n = 886)  
Table 10: Metric Fit Indices for Invariance Tests Across Age  
 

Δχ2 
χ2/df 
 
2.47 
0.80 
1.34 
1.30 
4.97 
14.75  190.85 

df 
4 
8 
12 
14 

 
4 
4 
2 

p value 

 

0.22 
0.21 
0.00 

 
Configural Invariance 
Metric Invariance 
Scalar Invariance 
Variance Invariance 
Note: N = 1,964 (< 43 years/n = 983, = or > 43 years/n = 981)  

χ2 
7.85 
8.92 
11.57 
243.28 

χ2/df 
 
1.96 
0.84 
1.12 
0.96 
2.65 
14.75  231.71 

df 
4 
8 
12 
14 

Δ χ2 

 
4 
4 
2 

 

0.35 
0.48 
0.00 

CFI 
0.99 
1.00 
0.99 
0.97 

CFI 
0.99 
1.00 
1.00 
0.96 

ΔCFI  RMSEA 

 

0.00 
0.00 
0.03 

0.04 
0.02 
0.02 
0.12 

 

0.00 
0.00 
0.04 

0.03 
0.01 
0.00 
0.13 

Δdf 

p value 

ΔCFI  RMSEA 

Age.  To perform a test for the impact of age on measurement invariance, a binomial 

variable was created by splitting respondents into a younger group (below 43 years of age) and 

an older group (43 years and older).  The split was formed based on the median age of just over 

43 years for the data sample.  The four sets of CFAs (see Table 10) were not significant across 

the first three invariance tests under different degrees of constraint.  As with previous two sets of 

invariance testing, the last test (variance invariance) produced a significant result, χ2(14) = 243.3, 

p < .001.  Therefore, the null hypothesis of measurement invariance again was rejected.  The 

variance invariance CFA estimated an insignificant coefficient of Procedural Justice, β = 0.01, p 

= .28, that suggests that the difference in means across the two categories of age was not 

significantly different from 0.  This infers that the limited bias that exists across the younger half 

and older half of the data set did not significantly impact judgments on procedural justice. 

Higher Education.  The MG-CFA tests compared the procedural justice model fit across 

a binomial variable measuring the respondents’ educational level.  The first CFA across the two 

groups (college degree or no degree) determined that configural invariance existed because the 

 

 

85 

 

factor-indicator relationships for procedural justice were identical in the direction hypothesized 

(see Table 11).  The second CFA that constrained the measurement coefficients to be equal 

(mean = 0; variance = 1) found that metric invariance was significant between the two groups, 

χ2(8) = 15.77, p = .047, and the null hypothesis of measurement invariance was rejected.  The 

third CFA test of scalar invariance between the two groups that added constraints to the 

intercepts also led to a significant result, χ2(12) = 22.3, p = .03.   

Table 11: Metric Fit Indices for Invariance Tests Across Education Level  
 

Δdf 

p value 

ΔCFI  RMSEA 

 
 
Configural Invariance 
4 
Metric Invariance 
4 
Scalar Invariance 
Variance Invariance 
2 
Note: N = 1,957 No Higher Education/n = 1,438, Higher Education/n = 519)  

χ2 
7.61 
15.77 
22.34 
185.20 

χ2/df 
1.90 
1.97 
1.86 
13.23 

Δχ2 

 
8.16 
11.43  
162.86 

df 
4 
8 
12 
14 

 

0.05 
0.03 
0.00 

CFI 
0.99 
0.99 
0.99 
0.97 

 
0.0 
0.0 
0.026 

0.03 
0.03 
0.03 
0.11 

Though the failure to support metric and scalar invariance across the higher education 

groups already ruled out measurement invariance, the fourth CFA also was estimated to 

determine that model fit was significant, χ2(14) = 185.2, p < .001.  Moreover, this CFA estimated 

a small but significant coefficient for procedural justice, β = 0.01, p = .04.  This suggests the 

process-based model’s latent values for procedural justice would be slightly altered by the higher 

education level of the respondent.   

This study found that differences in characteristics of the participants in the Ukrainian 

sample did not significantly bias the perceptions of procedural justice with the lone exception of 

higher education status.  Nevertheless, the MG-CFA findings suggest that possession of a college 

degree introduces only a small bias in the mean judgments of procedural justice. Therefore, this 

study’s MG-CFA tests across demographic groupings suggest that partial invariance exists 

across the Tylerian instrument in Ukraine.  This general finding strengthens the support for the 

earlier MG-CFA estimate of strong, though partial, invariance between groups based on 

experience with the police and the common practice of using data samples that include those that 

 

 

86 

 

report no recent contact with the police in samples intended for tests of citizens’ judgments on 

procedural justice.  

Control Variables’ Normative-Path Analyses 

This study’s findings of partial measurement invariance and weak or not significant 

coefficients for the procedural justice indicator across the different sub-groups in the Ukrainian 

data sample leaves open the question whether the lack of pure invariance in the MG-CFAs was 

caused by instrument bias or a true difference between the groups.  Therefore, the bivariate 

variables representing experience with the police and different demographic groups were added 

to the Normative-path of the Tylerian process-based model to assess their influence on the key 

indicators that Karakus (2017) describes as the normative pathway to legitimacy and cooperation 

with the police.   

Police Interaction.  A path analysis of the key indicators supported the MG-CFA finding 

that separating the sample based on experience with the police did not have a significant impact 

on procedural justice (see Figure 10).  The regression coefficient between the police interaction 

dummy variable and procedural justice was not significant.  The path analysis also tested 

whether a bivariate variable representing experience with the police in the process-based model 

would have a strong association with the independent indicators for police legitimacy or 

cooperation with the police.  The path model demonstrated that police interaction was a modest 

but significant predictor of the two components of legitimacy (trustworthiness in police, β = 

0.036, p < .01; obligation to obey, β = 0.06, p < .01).  In the two cases, experience with the 

police modestly increased positive judgments of trustworthiness and obligation to obey.  The 

introduction of the police interaction indicator to the model did negligibly influence the only 

predictor of the outcome variable, as the regression coefficient increased (3%) on the path  

 

 

87 

 

Figure 10: Police Interaction as a Control in the Normative-Path Model  

ε10

ε11

ε12

ε2

1_respect

ε3

2_listen

ε4

3_treatfair

ε5

4_respect

ε6

5_courteous

ε7

6_factbased

ε8

7_explain

ε14

8_fairly

.81

.83

.85

.85
ε1
.74

.79

.77

.82

13_basicrights

14_trust

15_jobwell

.82

.8

.61

ε9
Trustworthiness

**

.89

-.046

*
.038
.56

Proc_Just

pol_interaction
**
.35
**

1
*
.061

.16

-.032

Obligation

ε13

-.042

.043

Cooperation

ε15

.8

.83

.71

1_call

ε16

2_report

ε17

4_info

ε18

 

     ** p < .001; * p < .01; non-signficant estimates represented by .25 lines. 
     Goodness of Fit Measures: Satorra-Bentler Scaled χ2(97) = 915.7, p < .001; TLI = .94; SRMR = .03;  
     RMSEA = .06, R2 = .94. 

 

 

88 

 

procedural justice and cooperation with the police (β = 0.35, p < .001).  The police interaction 

indicator did not change the overall correlation coefficient for the Tylerian model, R2 = 0.94.  

The finding that the police interaction variable did not significantly predict cooperation 

with the police in Ukraine was supported by Tankebe’s (2009) finding that police contact did not 

predict cooperation with the police in Ghana.  Nevertheless, the positive impact that experience 

with the police had on the judgments of agency legitimacy is a new finding in a developing 

country.  Tankebe (2008) and Akinlabi (2015) found that contact with the police had a negative 

impact on the judgments of legitimacy in Ghana and Nigeria.  The other test of the role of 

experience with the police was in a developed country.  Murphy (2013) found that personal 

contact with the police had no impact on the predictors of legitimacy in Australia.  Therefore, the 

limited set of tests of police contacts in the Tylerian model suggest that recent experience with 

the police does not impact judgments of procedural justice or the police outcome variable of  

cooperation with police but has a less predictable relationship with the police legitimacy 

construct. 

Gender. Given the modest bias detected in the MG-CFA test, a bivariate gender variable 

was added to a path analysis of the Tylerian model.  As demonstrated in Figure 11, the gender 

indicator in the path model was a small but significant predictor for trustworthiness, β = 0.03, p < 

.01.  The gender variables had insignificant associations with procedural justice, obligation to 

obey, and cooperation.  The introduction of the gender indicator to the model did not change the 

strength of the only predictor of the outcome variable, the procedural justice association with 

cooperation with police, β = 0.34, p < .001; nor did the introduction of the new indicator have an 

impact on the overall correlation coefficient for the Tylerian model, R2 = 0.95.   

 

 

89 

 

Figure 11: Gender as a Control in the Normative-Path Model  
ε11

ε10

ε12

13_basicrights

14_trust

15_jobwell

.82

.8

.61

ε9
Trustworthiness

**
.89
-.044

Proc_Just

**
.34
**

.16

-.033

-.014

*
.036
.91
gender

1

.038

Obligation

-.029

ε13

Cooperation

ε15

.79

.83

.71

1_call

ε16

2_report

ε17

4_info

ε18

ε2

1_respect

ε3

2_listen

ε4

3_treatfair

ε5

4_respect

ε6

5_courteous

ε7

6_factbased

ε8

7_explain

ε14

8_fairly

.81

.83

.85

.85
ε1
.74

.79

.77

.82

 

     ** p < .001; * p < .01, non-signficant estimates represented by .25 lines. 
     Goodness of Fit Measures: Satorra-Bentler Scaled χ2(97) = 896.5, p < .001; TLI = .95; SRMR = .03;  
     RMSEA = .06, R2 = .95     

 

 

90 

 

Figure 12: Age as a Control in the Normative-Path Model  

ε10

ε11

ε12

ε2

1_respect

ε3

2_listen

ε4

3_treatfair

ε5

4_respect

ε6

5_courteous

ε7

6_factbased

ε8

7_explain

ε14

8_fairly

.8

.83

.84

.85
ε1
.74

.79

.77

.82

13_basicrights

14_trust

15_jobwell

.82

.8

.61

ε9
Trustworthiness

.88

Proc_Just

**
.0046
**
.34
**

.16

-.0088

-.035

age

2.6

1

-.026

.032

Obligation

-.029

ε13

Cooperation

ε15

.8

.83

.71

1_call

ε16

2_report

ε17

4_info

ε18

 

     ** p < .001; * p < .01, non-signficant estimates represented by .25 lines. 
     Goodness of Fit Measures: Satorra-Bentler Scaled χ2(97) = 884.1, p < .001; TLI = .95; SRMR = .03;  
     RMSEA = .06, R2 = .95

 

 

91 

 

Figure 13: Higher Education as a Control in the Normative-Path Model  

ε10

ε11

ε12

13_basicrights

14_trust

15_jobwell

.82

.8

.61

ε9
Trustworthiness

**

.88

Proc_Just

.035
**
.34
**

.15

.0091

.6
higher_ed
1

.037

Obligation

-.037

.038

-.031

ε13

Cooperation

ε15

.8

.83

.71

1_call

ε16

2_report

ε17

4_info

ε18

ε2

1_respect

ε3

2_listen

ε4

3_treatfair

ε5

4_respect

ε6

5_courteous

ε7

6_factbased

ε8

7_explain

ε14

8_fairly

.8

.83
.84

.85
ε1
.74

.79
.77

.82

 

     ** p < .001; * p < .01, non-signficant estimates represented by .25 lines. 
     Goodness of Fit Measures: Satorra-Bentler Scaled χ2(97) = 890.9, p < .001; TLI = .95; SRMR = .03;  
     RMSEA = .06, R2 = .95  

 

 

92 

 

Age.  The continues variable for the respondent’s age also was added to the path analysis 

of the Tylerian model (See Figure 12).  The MG-CFA test had used a dummy variable dividing 

the respondents in two by age.  This test used the actual age of the respondents.  The path model  

found that respondent’s age did not share a significant association with procedural justice or the 

legitimacy measures of trustworthiness or obligation to obey.  In addition, age was not a 

significant predictor of cooperation with police.   

Higher Education.  Given the MG-CFA tests finding that procedural justice was 

significantly altered by the respondent’s higher education status, a path model also was run to 

determine if the higher education variable served as a predictor variable in the larger process-

based legitimacy model.  In this case, the path analysis did not find a significant regression 

coefficient between levels of higher education and the model’s latent variables (see Figure 13).  

This suggests that the model’s indicators are not altered based on the respondents’ level of higher 

education.  The lack of a significant role for higher education in the model differs from prior  

research.  Murphy (2013) reported higher education status as a modest predictor of police 

legitimacy in Australia.   

The SEM analyses appear to largely support the findings of the MG-CFAs that infer that 

experience with the police and the study’s three demographic variables did not introduce a 

significant bias in the Tylerian model’s key indicators on his theorized normative pathway.  The 

MG-CFAs established that higher education status was the lone demographic variable affecting 

judgments of procedural justice, but the path models suggest that gender may have a weak 

association with the trustworthiness component of police legitimacy.  Otherwise, the 

demographic variables had no impact on the indicators of legitimacy or the police outcome 

measure of cooperation with the police.  As a result, this study’s analyses provided general 

 

93 

 

support for the process-based model of police legitimacy, notably the normative values pathway, 

as an unbiased instrument for measuring society’s judgments on the behavior and outcomes of 

police in a developing country. 

 

 

 

94 

 

 

CHAPTER V 

 

Conclusion 

The world has changed dramatically since Weber’s pioneering work on legitimacy was 

published posthumously in 1922.  Scholars continue to debate the indicators of legitimacy, but 

increasingly seek to model how the interaction between the institution’s officers and the public 

shape an institution’s legitimacy and the public’s willingness to cooperate with that agency and 

comply with the law.  The explosion of empirical research on the classic process-based model of 

police legitimacy attests to the potential for a theoretical paradigm that relates police officer 

behavior to desired police outcomes (Worden & McLean, 2017a).  Given the even greater 

challenges to gaining public trust and cooperation in developing countries, scholars working 

outside OECD countries must continue to develop and test new indicators that can model the 

antecedents to legitimacy and police outcomes in developing countries. 

Postmaterialism and A Lack of Normative Commitment to Police 

Postmaterialism joined procedural justice as the only significant predictors of cooperation 

with the police in the Ukrainian model.  The two variables reflecting normative values support 

Tyler’s argument that the paths in his model are based on values.  Tyler (2009) argued that the 

shared normative values forming legitimacy judgments would also associate strongly with the 

internal motivation (normative values) for choosing to cooperate with the police.  A wave of past 

studies in North America and Western Europe supported this expectation.  The significant role of 

postmaterialism in the model suggests that the normative preferences of the Ukrainian public this 

variable represents is associated with the willingness to cooperate with the police.  Therefore, 

this study supports Tankebe’s proposition that in developing countries the public’s judgments on 

the trustworthiness and legitimacy of police does not reflect normative commitments.   

 

95 

 

The SEM analysis certainly demonstrated that the Ukrainian public’s judgments on trust 

in the police failed to associate with the normative values favorable to cooperating with the 

police.  As a result, the heart of the Tylerian process-based model (legitimacy) failed to predict 

the model’s outcome variable.  This supports the growing body of research in developing 

countries that through the end of 2017 lacks a test of the full Tylerian model in which Tyler’s 

construct of legitimacy has mediated the relationship between procedural justice and cooperation 

or even had a significant association with the outcome variable (Reisig & Lloyd, 2009; Sun et 

al., 2017; Tankebe, 2009).  

In examining the failure of the classic legitimacy construct in Ghana, Tankebe (2009) 

argues that developing countries do not yet fully identify with the normative goals of supporting 

the civic or public good, but are focused instead on attachments to family, tribe or ethnic group.  

These informal relationships are likely part of the explanation for the weak role of the normative 

legitimacy construct not only in Ghana but in other developing countries as well (Stack, 1986).  

Tankebe (2009) emphasizes the constant challenge to maintain professional policing standards 

when the norms of informal bonds can lead a society to view corrupt acts of civil institutions as 

legitimate if they are in the interests of the oligarchic or tribal hierarchies.   

In Ukraine’s case, informal networks are often seen as more powerful than the formal 

government structures.  Karklins (2005) provided a typology of the rise of the patronage 

networks and oligarchs in former communist countries that have become far more powerful than 

government officials or their civil administrations.  This is certainly the case in Ukraine where 

clans oversee an economy that is still largely patronage-based with weak government regulation 

(Johnston, 2005; Leitner, et al., 2015; Shelley, 1998).  The victim of a crime in Ukraine and 

other developing countries may not necessarily view the police as their best option for seeking 

 

96 

 

assistance.  Instead, the victims may be just as likely to turn to informal or clan-based ties to seek 

justice.  As a result, the conflict between democratic governance and informal ties may explain 

the failure of the path analyses in Ukraine to find an association between the key legitimacy 

measure of trust in the police and the indicator of the will to cooperate with the police.   

The conflict between civic and what Tankebe calls primeval values also may explain how 

tests of the legitimacy model in developing countries indicate a strong role for procedural justice 

in predicting a legitimacy measure that then fails to mediate the relationship with the police 

outcome variables.  In Ukraine, the normative value of procedural justice was a much stronger 

predictor of legitimacy than the instrumental value of effectiveness, but a robust association 

between procedural justice and trustworthiness did not lead to an association between 

trustworthiness and the internalized values that Tyler views as driving the willingness to 

cooperate with the police.  If citizens view a hidden patronage system as no less legitimate than 

state institutions, they may not prioritize police outcomes--complying with the law or assisting 

the police--regardless of their observing just and respectful procedures on the street.  Across the 

four tests of the classic Tylerian model in developing countries, a common failure is the link 

between judgments of trustworthiness and willingness to cooperate with the police. 

Supporting the Consensus on Validity Problems 

The factor and correlation analyses of the Ukrainian data supported past studies 

demonstrating a lack of discriminant validity between key test items (notably, the indexes 

forming procedural justice, effectiveness, and legitimacy) that are intended to measure entirely 

different constructs in the Tylerian theory of legitimacy.  The results suggest further validity 

testing and test item development is needed to ensure that survey respondents can discern 

between constructs so as to produce separate measures for unrelated constructs.  The similarity in 

 

97 

 

questions across different indicators could lead to difficulty for respondents to form responses 

that are restricted to the two separate constructs whose relationship is a cornerstone of the 

theoretical model.  Moreover, this lack of discriminant validity likely inflates the statistical 

relationship between the predictor and dependent variables in the model.  

The analyses of the Ukrainian data also suggest that the discriminant validity problem 

with the Tylerian model may be even greater in developing countries.  Ukraine became the third 

study in developing countries that reported the highest correlation in the model was between 

effectiveness and procedural justice.  Future research that applies the process-based model of 

police legitimacy in a developing country should not only make use of stringent tests for 

discriminant validity between procedural justice and trustworthiness, but also between those 

variables and police effectiveness. 

The study’s test results infer that a fundamental threat to the continued development of 

the legitimacy theory remains the poorly specified indicators. Without continued diverse and 

methodical development of the indicators, the model will continue to suffer convergent and 

discriminant validity problems.  The stringent tests of the model started by Reisig and colleagues 

in 2007 and continuing through this and other studies are necessary to further refine the choice of 

test items and operationalization of the model’s key variables. 

Blurred Distinction Between Effective and Just Procedures   

The path models of the legitimacy theory in Ukraine provide an opportunity to examine 

the role police effectiveness has in the process-based model in a developing country 

characterized by systemic corruption.  The SEM demonstrated that effectiveness was strongly 

associated with procedural justice and trustworthiness, which was not the case in the previous 

tests of the model in developed countries (e.g., Tyler & Huo, 2002; Sunshine & Tyler, 2003a).  

 

98 

 

In Ukraine, the procedural justice indicator accounts for nearly the same amount of variance in 

the latent variable of police effectiveness (R2 = .74), as it explained in the trustworthiness 

indicator (R2 = .79).   

Other tests of legitimacy theory in developing countries did not isolate the trustworthiness 

index in their report on the bivariate correlations.  Nevertheless, studies in Poland, Pakistan, 

China, Turkey, and Nigeria reported that police effectiveness, not procedural justice, was the 

strongest predictor of legitimacy which suggests that in those countries a relatively large 

correlation also exists between effectiveness and legitimacy (Czapska et al., 2016; Jackson et al., 

2014; Karakus, 2017; Sun et al., 2017; Akinlabi, 2018).  In a similar study, Nalla, Hamm, and 

Paek (2017) report that a latent measure for police professionalism in India also strongly 

predicted the indicator for procedural fairness, β = 0.82, p < .001.   

Compared to the respondents in the Chicago courthouse waiting room where Tyler first 

tested his process-based model of legitimacy, the public in developing countries appear to have 

far more difficulty differentiating between just versus effective procedures.  The close 

association between the indicators of procedural justice and police effectiveness in developing 

countries suggests that when the public cannot take for granted that police will either arrive or 

perform their duties when called, then a “just procedure” may simply be the police arriving after 

a call for assistance.  In other words, survey respondents in developing countries may perceive 

test items measuring effective policing as also representing procedural justice in a country where 

a common injustice is ignoring citizens’ requests for assistance or action.  As a result, this issue 

may partly be a measurement problem and future studies should consider experimenting with 

new test items that may reduce the overlap in interpretation of a just versus effective procedure.   

 

99 

 

Tests of the legitimacy theory in developing countries may require variables in the 

procedural justice index that measure perceived injustices while providing greater discriminant 

validity in relation to the test items that measure ineffectiveness.  For example, the Ukrainian 

model included an effectiveness index that asks if the local police know how to carry out their 

official duties properly (APPENDIX B: Question 8.1). This certainly fits the definition of police 

effectiveness, but does it provide the appropriate degree of discriminant validity from the test 

items in procedural justice or trustworthiness that asks if the local police make decisions based 

upon the facts or do their job well (APPENDIX B: Question 9.6; Question 9.15).  In past 

developing country tests of the model, the public appear to view these and other test items in the 

indexes to be asking very similar questions.   

Future research in developing and developed countries should seek to further polarize the 

test items used for different measures in the model.  A good start would be replacing the vague 

description of carrying out duties properly in the effectiveness measure with a test item that asks 

if officers are effective at more concrete functions in the view of the public (e.g., preventing 

crime, solving property crimes, resolving disputes in the neighborhood).  This could help to 

further differentiate the measures of effectiveness and procedural justice.  Testing and retesting 

new indicators could help address the convergent and discriminant validity problems across the 

model’s key antecedents to police legitimacy.  

Adding Effectiveness to the Legitimacy Construct in Developing Countries 

The lack of discriminant validity in the model may also suggest a poorly theorized role 

for police effectiveness in the Tylerian model.  The SEM path models of the Ukrainian data 

support the core expectation of the theory of police legitimacy that just behavior on the part of 

police officers on the street predicts the public’s willingness to cooperate with the police.  

 

100 

 

Nevertheless, the failure of legitimacy to mediate the relationship provides further evidence that 

the Tylerian model of police legitimacy fails to model the process through which procedural 

justice influences police outcomes in developing countries.  Tyler’s legitimacy construct also 

failed as a mediator in China and Poland, while weakly mediating the relationship between 

procedural justice and the police outcome variable in Slovenia (Czapska et al., 2016; Reisig et 

al., 2014; Sun et al., 2017).   

Bottoms and Tankebe (2012) propose adding the trustworthiness and effectiveness 

indicators to a new police legitimacy construct.  The merger of these indexes in a single 
construct would eliminate the concern that effectiveness and trustworthiness lack discriminant 
validity.  In a subsequent study, Tankebe, Reisig and Wang (2016) added police effectiveness to 

a definition of legitimacy and found that, unlike Tankebe’s (2009) previous test of a Tylerian 

construct of legitimacy, the new construct modestly predicted cooperation with the police among 

students in Ghana.   

The Ukrainian findings support the possibility of combining police effectiveness and 

trustworthiness in a common indicator because the two measures are highly correlated and the 

path analysis shows a significant path between the two indicators.  Given the repeated failure of 
the Tylerian legitimacy construct in developing countries (e.g., Reisig & Lloyd, 2007; Tankebe, 
2009), scholars are likely to continue to experiment with new indicators for the police legitimacy 
construct in search of a valid and reliable mediator for a model of police legitimacy.  Future 

research also should seek to clarify what characteristics of a society lead to a stronger role for 

different hypothesized indicators of legitimacy to better understand how the construct operates 

across societies that differ in terms of economic prosperity or levels of corruption and good 

governance. 

 

101 

 

The Future Role of Obligation to Obey 

No construct operationalized in the Tylerian process-based model has received greater 

scrutiny than the question of whether obligation to obey should continue as part of the legitimacy 

construct.  In the case of Ukraine, the model fit the data only splitting the construct and treating 

obligation to obey as an independent mediator in the model.  This decision certainly eliminated 

the most important source of the Tylerian model’s convergent validity issues by separating the 

two indexes that across every published validity test of the model has yet to form a univariate 

indicator.  The more important issue for legitimacy researchers is whether obligation to obey 

should continue to serve as a component of the legitimacy construct.  

Scholars choosing to build their model solely around the trustworthiness indicator would 

effectively move away from the original, Weberian perspective on legitimacy.  Beetham (1974) 

and other legitimacy scholars in the Gamson-era criticized the Weberian concept for treating 

legitimacy gained by an institution through the threat of violence as equal to the legitimacy that 

is a product of voluntary submission to a governing authority viewed as sharing similar morals or 

beliefs with its citizens.   

Today, Tankebe (2009) has been out front among police scholars in arguing that 

obligation to obey should be removed as an indicator of legitimacy in a democracy because 

institutional legitimacy requires citizen consent that is not measured as part of the existing 

obligation to obey measure.  Tankebe specifically cites Evans (2007:80) in stating that “consent 

can only be measured in situations in which individuals can choose between real alternatives.”  

Instead, the sub-construct of obligation to obey seeks to measure the degree citizens blindly 

follow the orders of authorities using test items asking whether “they should obey the law even 

when they think it goes against what they think is right” (Tyler, 1990).   

 

102 

 

In developing his two-part construct of legitimacy, Tyler (1990) credits Gamson (1968) 

who argued that political legitimacy requires an obligation to obey which limits the need for 

coercive measures and used the example of the selective service system that regulates U.S. 

military conscription.  The example of the US selective service system suggests that an otherwise 

coercive action (forced conscription) may still have popular consent demonstrated through the 

election of the authorities.  As research on modeling police legitimacy expands to more and more 

countries, some scholars (e.g., Harkin, 2015; Kochel, 2012) argue that legitimacy models should 

include involuntary precursors to compliance or cooperation with law enforcement.  Kochel 

(2012) reasons that if fear can compel citizens to obey then fear should be a precursor in the 

model predicting police legitimacy.  Certainly, one of criminology’s oldest theoretical models, 

deterrence theory, has long argued that the fear of punishment or retribution can explain a 

considerable portion of human behavior. 

The debate on the legitimacy construct is likely to continue because there does not appear 

to be a ready compromise that pleases the scholars advocating for legitimacy defined by 

voluntary consent and those arguing that coercive governments can still gain legitimacy through 

fear.  Nevertheless, the theoretical debate on the appropriate role for the construct eventually 

must establish empirically that obligation to obey is part of a unidimensional or 

multidimensional measure of police legitimacy that mediates the process-based model.  This 

Ukrainian study adds to the growing literature that finds obligation to obey is weakly associated 

with procedural justice and not a predictor of the public’s willingness to cooperate with the 

police.  Future studies should consider treating the obligation to obey indicator as a separate 

potential mediator in the model.  Moreover, scholars should consider new measures that do not 

reflect an obligation to obey, but rather measures the public’s fear in not complying or 

 

103 

 

cooperating with law enforcement that may be more closely associated with the trustworthiness 

scale in the legitimacy indicator. 

Invariance Supports General Survey Sample 

This study experimented with using multi-group confirmatory factor analysis to test the 

level of group bias in evaluating the procedural justice indicator because of uncertainty over 

whether procedural justice reflects direct or indirect judgments on the publics interactions with 

the police.  The procedural justice indicator was selected for the test because the theoretical 

description of the model refers only to procedural justice as an evaluation of past experiences 

interacting with officers.  The tests on the procedural justice indicator concluded that the 

Ukrainian data produced little bias across groups whether based on demographics or experience 

with the police.  The wide-scale adoption of MG-CFA tests for measurement invariance has 

largely been limited to the world of aptitude tests and other psychometric instruments applied to 

evaluate individual performance, but this study demonstrates the potential for this analysis in 

institutional legitimacy research.   

The MG-CFA tests in Ukraine support the current practice among most legitimacy 

scholars to combine respondents with no recent interaction with the police and those reporting 

recent experience with the police into a single sample for modeling police legitimacy.  Recent 

experience with the police in Ukraine did not impact citizens’ judgments on procedural justice or 

willingness to cooperate with the police while having a trivial influence on the public’s 

evaluation of police legitimacy.  The findings are in line with Worden and McLean’s (2017b) 

mixed methods study (survey and observation of the same police-citizen interactions) that found 

officer interaction with the public is only a modest predictor of judgments on procedural justice.  

 

104 

 

Future scholarship on police legitimacy must begin to treat prior attitudes as an important 

exogenous factor in a model measuring procedural justice and police legitimacy.  

Current Study’s Limitations 

This test of the Tylerian model is based on existing cross-sectional data from three cities 

in Ukraine.  Future tests of the process-based model would benefit from the addition of a few 

variables that were not part of the 2013 survey, such as a measure of perceived corruption.  

Finally, the survey-based approach to examining the public’s judgments on the police limits the 

researcher’s ability to assess for biases across survey responses.   A mixed methods approach is 

needed to seriously address potential predispositions in the surveyed judgments on the police.   

Absent Variables.  After the 2013 pre-test survey, the U.S. Department of Justice project 

team approved a final questionnaire that removed two indexes, distributive fairness and 

compliance with the law, that were viewed as non-essential to the Tylerian model and whose 

removal was expected to improve the rate of completed interview questionnaires.  The 

distributive fairness test items on treatment of immigrant or minority populations led to 

confusion in the pre-test interviews in the three largely homogenous cities, and just a couple of 

the pre-survey respondents were willing to provide answers to even misdemeanor violations of 

the law for the compliance index.   

The distributive fairness indicator has had mixed results as an antecedent of legitimacy in 

tests of the Tylerian theory.  Sunshine and Tyler (2003a) found distributional justice was less 

important as a predictor of legitimacy but the construct is often a significant antecedent to 

legitimacy in tests in developed countries (e.g., Reisig, 2007; Murphy et al., 2008).  The two 

tests in developing countries provide weak results as distributive fairness was not a significant 

predictor of cooperation in Ghana and Jamaica (Reisig & Lloyd, 2009; Tankebe, 2009), while 

 

105 

 

Sun and colleagues (2016) reported that distributive justice was a significant antecedent of 

legitimacy but not the willingness to cooperate with police in China (Sun et al., 2016).  For the 

most part, scholars in developing countries have chosen not to include distributive justice in their 

tests of the Tylerian construct of legitimacy or the full process-based model (Bradford et al., 

2013; Czapska, et al., 2016; Jackson et al., 2014; Karakus, 2017; Reisig et al., 2014).  

As with distributive justice, the compliance with the law index is not considered an 

essential element the Tylerian model and is largely absent from tests of the process-based model 

in developing countries (e.g., Czapska, et al., 2016; Jackson et al., 2014; Karakus, 2017; Reisig 

et al., 2014; Sun et al., 2016, Tankebe, 2009).  The compliance indicator’s test items asking if 

respondents had smoked marijuana or committed petty crimes produced few completed 

responses in the Ukrainian pre-test.  The difficulty in getting respondents to answer compliance 

questions in developing countries likely limits tests of the Tylerian model using compliance as a 

second police outcome variable.  

Corruption as a Control.  Beyond the variables commonly associated with tests of 

Tylerian legitimacy theory, the study would have benefited from the addition of a perceptions of 

police corruption measure in the 2013 survey.  Ukraine is a country plagued by systemic 

corruption but there is uncertainty whether the public’s experience with corruption influences 

their judgments on procedural justice and police outcomes.  Past studies that tested a corruption 

indicator with the antecedents to police legitimacy provide mixed results.  In a survey of 

Nigerian students, Akinlabi (2015) reported that corruption negatively predicted police 

legitimacy, and was nearly equal in strength as the procedural justice measure.  On the other 

hand, Jackson and colleagues (2014) found in a test of an adult sample in Nigeria that police 

corruption was not a significant predictor of the Tylerian components of legitimacy 

 

106 

 

(trustworthiness and obligation to obey).  The mixed findings suggest that future research is 

needed on the role of police corruption in the legitimacy theory in developing countries. 

Need for Mixed Methods.  This study relied on existing quantitative data to examine the 

process-based model of police legitimacy in a developing country. As with all tests of a theory, it 

is important not only to test the quantitative validity of the theoretical model but also to enhance 

the findings by integrating other methods in the research.  The development of a legitimacy 

model that would be valid across a range of countries requires adding a qualitative dimension by 

using focus groups to gather community perspectives on what respondents consider when 

weighing the normative and instrumental norms in reaching judgments on officer procedures, 

institutional legitimacy, and police outcomes.   

A qualitative component to the Ukrainian study would have helped in addressing the 

continued uncertainties over the respondents’ potential confusion in separating the meaning of 

key test items used to measure different constructs in the model.  Moreover, qualitative studies 

also will be critical to determining the degree that pre-existing biases influence public judgments 

on police behavior and police outcomes. 

In a pioneering mixed-methods approach to the Tylerian model, Worden and McLean’s 

(2017) Mirage of Police Reform suggests legitimacy researchers cannot rely solely on survey 

information to evaluate public judgments on officer behavior and agency legitimacy.  In a study 

of two cities in upstate New York, Worden and McLean (2017) reported that a plurality of 

respondents that experienced unfair treatment in the view of an outside observer gave 

consistently favorable evaluations of procedural justice, while a smaller group consistently gave 

poor evaluations regardless of the objective state of procedural justice in the officer-citizen 

interaction.  The adoption of further mixed methods approaches to the study of the process-based 

 

107 

 

model will be critical to determine the degree and direction of the potential bias that Worden and 

McLean found in respondent evaluations to the key Tylerian constructs in the model.  

Vicarious or Process-based Model?  Worden and McLean’s (2017) findings on the 

disconnect between the objective evaluations of police-citizen interactions and those citizens’ 

evaluations of their experiences provides additional support for the argument that vicarious 

experiences greatly outweigh direct experience on the judgments of the police.  The MG-CFA 

tests and an OLS regression used in this study to isolate the effect of experience with the police, 

revealed that recent experience with the police produces little influence on the respondent’s 

judgments on the procedural justice test items.  This finding supports the decade or more trend 

for police legitimacy researchers not to remove those without recent police interaction from their 

survey sample (e.g., Sunshine & Tyler, 2003a; Tyler & Wakslak, 2004; Reisig et al., 2007; Gau, 

2011)  

The expectation that experience with the police has a limited influence on the Tylerian 

model raises a new question.  Does a theoretical model, that is often referred to as the process-

based model because of the antecedent role of citizens interacting with the police, actually 

represent a model of a vicarious and not direct experience with the police?  This study’s 

literature review highlighted the mixed results found in the policing studies that sought to 

compare the influence of vicarious and direct experiences on public perceptions of the police 

(e.g., Brandl et al., 1994; Skogan, 2006).   

In a longitudinal study of juveniles going through the justice system, Kaiser (2016) 

reported that there were significant and distinct pathways for the vicarious and direct experiences 

of procedural justice on legitimacy over time.  Nevertheless, by using a sample of youth highly 

involved in the criminal justice system, Kaiser’s (2016) support for a balance in the influence of 

 

108 

 

direct and vicarious experience may not generalize to the broader population.  More research is 

needed to provide greater understanding of the relative roles of these two sets of perceptions on 

judgments over procedural justice and police outcomes.  Nevertheless, this study’s findings 

support the current de-facto treatment of procedural justice as a global judgment and not just a 

result of direct experience with the police.  Researchers that survey the entire population for their 

studies are making a conscious decision that the model measures vicarious as well as specific 

experiences of police behavior. 

The Tylerian model of police legitimacy has become arguably the most influential 

contemporary theory of criminal justice among scholars and practitioners alike.  This is no less 

true in police studies in developing countries that increasingly have been turning to the Tylerian 

theory to model police legitimacy.  Yet the surge of studies making use of the classic police 

legitimacy model may be missing an opportunity to further refine the model’s key measures and 

our understanding of its core relationships.   

This study is part of the latest wave in Tylerian studies that over the last decade has 
sought to adopt the process-based model to developing countries.  These tests generally have 
found that in developing countries “normative commitments and identification between citizens 
and police are weak or nonexistent” (Tankebe, 2009, p.1283).  This study’s data analysis offers 

the first of hopefully many more tests of variables that could shed light on the how the process-

based model may differ across societies.    

 

 

109 

 

 

 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

APPENDICES 

110 

 

 
 

 
 
 

 
 
 
 
 
 
Procedural Justice 
    Quality of Treatment 
 
 
 
 

 
 
 
 
 
 
 
 
Legitimacy 
    Obligation to Obey 

 

 

 

 

 

APPENDIX A: Test Items from Sunshine and Tyler 
 

Construct 
Procedural Justice* 
    Quality of Decision-Making 

Question Item 

 
    1) Make decisions about how to handle problems in fair 
        ways. 
Asked if the police… 
    3) Treat everyone in your neighborhood with dignity  
         and respect. 
    4) Treat everyone in your community equally. 
    5) Accurately understand and apply the law. 
    6) Make their decisions based on the facts, not their 
        personal biases or opinions. 
Asked how fairly the police decide… 
    7) Who to stop and question on the street. 
    8) Who to stop for traffic violations. 
    9) Who to arrest and take to jail. 
  10) How much they will help people with problems. 
 
    
    2) Do the police treat people fairly? 
Do the police… 
  11) Clearly explain the reasons for their actions. 
  12) Give honest explanations for their actions. 
  13) Give people a chance to express their views before  
         making decisions. 
  14) Consider people’s opinions when deciding what to do. 
  15) Take account of people’s needs and concerns. 
  16) Treat people with dignity and respect. 
  17) Respect people’s rights. 
  18) Sincerely try to help people with their problems. 
  19) Try to find the best solutions for people’s problems. 
  20) The NYPD treats citizens with courtesy and respect. 
   
 
    1) You should accept the decisions made by the police,  
         even if you think they are wrong. 
    2) You should do what the police tell you to do even 
         when you do not understand the reasons for their 
         decision. 
    3) You should do what the police tell you to do, even 
         when you disagree with their decisions. 
    4) You should do what the police tell you to do even  
         when you do not like the way they treat you. 
    5) There are times when it is ok for you to ignore what 
         the police tell you (reversed). 

111 

 

 

 

 

 
 
Legitimacy 
     Trustworthiness 

 
 

 
 

 
 

 
 

 

    6) Sometimes you have to bend the law for things to     
         come out right (reversed). 
    7) The law represents the values of the people in power,  
          rather than the values of people like you (reversed). 
    8) People in power use the law to try to control people  
          like you (reversed). 
    9) The law does not protect your interests (reversed). 
 
   
  10) Overall, the NYPD is a legitimate authority and people  
          should obey the decisions that NYPD officers make. 
  11) I have confidence that the NYPD can do its job well. 
  12) I trust the leaders of the NYPD to make decisions that  
          NYPD officers make. 
  13) People’s basic rights are well protected by the police. 
  14) The police care about the well-being of everyone they  
          deal with. 
  15) I am proud of the work of the NYPD.  
  16) I agree with many of the values that define what the  
          NYPD stands for. 
  17) The police are often dishonest (reversed). 
  18) Some of the things the police do embarrass our city  
         (reversed). 
  19) There are many things about the NYPD and its  
          policies that need to be changed (reversed). 
 

*Sunshine and Tyler (2003a) refer to this scale as “Procedural Fairness,” but over preceding and subsequent studies 
Tyler and colleagues refer to the same set of indicators as procedural justice.  
 

 

 

112 

 

APPENDIX B: Ukrainian Survey Instrument  
(English-Language Version8) 
 
Please provide yes or no answers to the following questions? 
 
A1. Are you a resident at this address? 
 

Yes……………………………………………….. 1 
 No …………………………………...………….. 2 (End of Interview) 

 
A2.  What is your gender?   

Yes……………………………….…………..….. 1 
 No ………………………………………..…….. 2 

 
A3.  How old are you in years?  _______ 
 
 
 
 
 
 
 
 
 

Less than 18 è (End of Interview) 
18 – 25 years old 
26 – 40 years old 
41 – 60 years old 
Older than 61 years 

 

                                                
8 The questionnaire versions used by the survey team were in the Russian and Ukrainian languages with the 
respondent selecting the language for their questionnaire. 

 

113 

 
Q1. Have you heard of the Ministry of Interior Initiative to Improve Public Relations? 

Questions on District Offices of Police 

Yes…………………………………….……..….. 1 
 No ………………………………………..…….. 2 

 
Q2. Have you or someone you know recently participated in a public event planned by the police  
       at an area school or cultural center? 

Yes…………………………………….……..….. 1 
 No ………………………………………..…….. 2 

 
Q3. In the last 12 months, have you visited your local police district office? 

Yes ………...………………………….……..….. 1 
 No ………………………………………..…….. 2 

 
Q4. Do you know where your local district office is located? 

Yes………………………………….………..….. 1 
 No ………………………………………..…….. 2 

 
Q5. In the last 12 months, have you had any contact specifically with your district officer?  

Yes ……………………...……………….…..….. 1 
 No ………………………………………..…….. 2 

 
Q6. Do you know the name or recognize the face of your district officer? 

Yes …………………………………………..….. 1 
 No ………………………………………..…….. 2 

 

 
 
 
 

 

 

114 

 

Questions on the Police and Your Neighborhood 

 
7. Please rate your building and surrounding neighborhood as a place to live compared to others 
     in your city? 
 

Much Worse ………….…………...…..…….…… 1   
Somewhat Worse …….………………..….…..….. 2 
Not worse or better ……………….……..……….. 3 
Somewhat Better ……………….……….….……. 4  
Much Better …………………….……….……….. 5  

 

 

                  
 
Q8. Please state the degree to which you disagree or agree with the following statements on 
your neighborhood.  Please choose an answer between 1 “strongly disagree” to 4 “strongly 
agree.”  Pass the respondent card 1 to select one response from the following for each question.     
 
 

Strongly Disagree ……….………….…….…… 1 
Tend to Agree …………...………….…….…… 2 
Tend to disagree .…………...……...……..…… 3 
Strongly Agree …………...……………….…… 4 
Difficult to Answer ………………………….. 99              

 

 
Police Effectiveness Index 
8.1 The local police know how to carry out their official duties properly. 
   
8.2 The local police are able to maintain order on the streets. 
 
8.3 The local police try to help citizens. 
 
8.4 The local police respond promptly when needed. 
 
8.5 The local police prevent crime. 
 
Fear of Crime Index 
8.6 In the area a half kilometer around my home, there is a place where I am afraid to walk alone 
at night. 
  
8.7 When I am away from home, I worry about the safety of my property. 
 
8.8 There is a reason to be afraid of becoming a victim of crime in my community. 
                                     
8.9 I worry about others in my home when they head outside at night. 
   
 
 

 

115 

 

Q9. Please state the degree to which you disagree or agree with the following statements on 
the relationship between the public and the police.  Please choose an answer between 1 
“strongly disagree” to 4 “strongly agree.”  Pass the respondent the following card 1 to select 
one response from the following for each question.     
 

Strongly Disagree ……….………….…….…… 1 
Tend to Agree …………...………….…….…… 2 
Tend to Disagree .…………...……...……..…… 3 
Strongly Agree …………...……………….…… 4 
Difficult to Answer ………………………….. 99              

 
Quality of Treatment Index 
9.1   The local police treat citizens with respect. 
 
9.2.  The local police take the time to listen to people. 
 
9.3   The local police treat people fairly. 
 
9.4   The local police respect citizens’ rights. 
 
9.5   The local police are courteous to people with whom they come into contact. 
 
Quality of Decision-Making Index 
9.6   The local police make decisions based upon the facts. 
 
9.7   The local police explain their decisions to those involved. 
 
9.8  The local police make decisions that lead to the fair resolution of problems. 
 
Distributive Fairness Index (dropped after pre-test problems) 
9.9   The local police provide the same quality of service to everyone in the neighborhood. 
 
9.10 The local police enforce the law consistently when dealing with all people. 
 
9.11 The local police give immigrants less help.   
 
9.12 The local police provide better service to wealthier citizens in my neighborhood. 
 
Trust in the Police Index 
9.13 People’s basic rights are protected by the local police. 
 
9.14 The local police can be trusted to make decisions that are right for your community. 
 
9.15 Most local police officers can do their job well.  
 
 

 

 

 

 

116 

 

Obligation to Obey Index 
9.16 You should accept police decisions even if you think they are wrong. 
 
9.17 You should do what the police tell you even if you disagree 
 
Q10.  Please state how often the following has occurred.  Please choose an answer between 
1 “never” to 4 “very often.”  Pass the respondent card 2 to select one response from the 
following for each question.     
 

Never ……….………….…….…… 
Not Often …………...………….…….…… 
Often .…………...……...……..…… 
Very Often …………...……………….…… 
Difficult to Answer ………………………….. 99              

1 
2 
3 
4 

 
Officer Mistreatment Index 
10.1 The police in my neighborhood address citizens in a respectful manner and appropriate 

tone?  [Reverse code] 

 
10.2 Officers in my district often use offensive language in talking to the public.  
          
 
10.3  Officers use more physical force than necessary for an incident in my district. 
   
 
11.  Please indicate how likely you would undertake the following actions. Please choose an 
answer between 1 “very unlikely” to 4 “very likely.”  Pass the respondent card 3 to select one 
response from the following for each question.     
 

Very Unlikely ……….………….………..……  1 
Unlikely …………...……………….…….……  2 
Likely .…………...………………..……..……  3 
Very Likely …………...………...……….……  4 
Difficult to Answer ………………………….. 99              

 
Cooperation with the Police Index 
11.1 …call the police to report a crime? 
 
11.2. …report suspicious activity near your home to the police. 
 
11.3. …call the police to report an accident. 
 
11.4. …provide information to the police to help catch a suspected criminal? 
 
 

 

117 

 

Q12. Please state the degree to which you disagree or agree with the following statements 
on the work of the police with the public.  Please choose an answer between 1 “strongly 
disagree” to 4 “strongly agree.”  Pass the respondent card 1 to select one response from the 
following for each question.     
 

Strongly Disagree ……….………….…….…… 1 
Tend to Agree …………...………….…….…… 2 
Tend to Disagree .…………...……...……..…… 3 
Strongly Agree …………...……………….…… 4 
Difficult to Answer ………………………….. 99              

 
Satisfaction with the Police Index 
12.1.  I am generally satisfied with my local police 
 
12.2   I am generally satisfied with the national police. 
 
Informal Social Control Index 
12.3   My neighbors could be counted on to intervene if they found children skipping school and 
          hanging out in our courtyard. 
 
12.4   My neighbors could be counted on to intervene if they found children were spray-painting 
          graffiti on a local building. 
 
12.5   My neighbors could be counted on to intervene if the children were showing disrespect to    
          an adult on a local street. 
 
12.6   My neighbors could be counted on to intervene if a fight broke out in front of their  
          building. 
 
Social Cohesion Component 
12.7   The people around here are willing to help their neighbors. 
 
12.8   The people in this neighborhood have close ties. 
 
12.9   People in this neighborhood can be trusted. 
 
12.10 People in this neighborhood generally don’t get along with each other.  
[Reverse code] 
 
12.11 People in this neighborhood do not share the same values.  
[Reverse code] 
 
Crime and Disorder Index 
12.12 Near your home drunks on the street is a problem 
 
12.13 Near your home narcotic users is a problem 

 

118 

 

 

 

 
Q15. Have you had any interaction with the local police (not the GAI/traffic enforcement) in last  
         12 months?   
 

Yes ……………………...……………….…..….. 1 
 No ………………………………………..…….. 2 

 
Q16. Have you or someone in your household been a victim of crime in the last 12 months? 
 

Yes ……………………...……………….…..….. 1 
 No ………………………………………..…….. 2 

 
Q17. Have you or someone in your household been a victim of crime that you did not report to  
         the police in the last 12 months? 
 

Yes ……………………...……………….…..….. 1 
 No ………………………………………..…….. 2 

 

 
12.14 People loitering near your home late at night is a problem. 
 
12.15 Near your home juveniles vandalizing property is a problem. 
 
Physical Disorder Index 
12.16 Near your home excessively loud neighbors is a problem. 
 
12.17 People dumping trash outside is a problem near your home. 
 
12.18 Abandoned buildings attracting juveniles or strangers is a problem near your home. 
 
12.19 Juveniles painting graffiti on buildings is a problem near your home. 
 

Questions Regarding the Police Over Last 12 Months 

Victimization Test Items 
Q13. Have you reported a crime to militia in the last 12 months—by crime we are specifically  
         referring to an action that violates Ukraine’s criminal code?  
 

Yes ……………………...……………….…..….. 1 
No …………………………………….…..…….. 2 

 
Q14. Have you reported a non-criminal issue to police in last 12 months (noisy  
         neighbor/dangerous dogs/etc.)?   
 

Yes ……………………...……………….…..….. 1 
 No ………………………………………..…….. 2 

119 

 

 

Q18. Have you or someone in your household had an interaction with the GAI (traffic  
         enforcement) in the last 12 months? 

Yes ……………………...……………….…..….. 1 
 No ………………………………………..…….. 2 

Questions on the Work of the Government and the Situation in Ukraine 

 
Q19. Please state the degree to which you disagree or agree with the following statements 
about other government agencies besides the police.  Please choose an answer between 1 
“strongly disagree” to 4 “strongly agree.”  Pass the respondent card 1 to select one response 
from the following for each question.     
 

Strongly Disagree ……….………….…….…… 1 
Tend to Agree …………...………….…….…… 2 
Tend to Disagree .…………...……...……..…… 3 
Strongly Agree …………...……………….…… 4 
Difficult to Answer ………………………….. 99              

 
19.1 Other government agencies do their job well.  
 
19.2 Other government agencies look out for the interests of the local community. 
 
19.3 Other government institutions protect your basic rights. 
 
Postmaterialism Index 
20. Please choose what you personally would consider the top priority from the following 
list of national priorities.  Mark just one choice. 
 
Maintaining order in the nation ………………………………..  1 
Giving the people more say in important political decisions ….  2 
Fighting rising prices ………………………………………….   3 
Protecting freedom of speech …………………………………   4 
 
21. Please state the degree of your disagreement or agreement with the following 
statements Please choose an answer between 1 “strongly disagree” to 4 “strongly agree.”  
Pass the respondent card 1 to select one response from the following for each question.     
 

Strongly Disagree ……….………….…….…… 1 
Tend to Agree …………...………….…….…… 2 
Tend to Disagree .…………...……...……..…… 3 
Strongly Agree …………...……………….…… 4 
Difficult to Answer ………………………….. 99              

 
Self-Expression vs. Survival Index 
21.1   I am very happy. 
 

 

120 

21.2   I would sign a non-governmental petition.                              
 
21.3  Homosexuality is sometimes justifiable.   
 
21.4  One has to be careful trusting people.   
[Reverse code] 
 

Social-Demographic Questions 

 
A4. What is the highest level of school or post-secondary studies you completed? 

Incomplete secondary education ………………………………………………………..  1 
(did not graduate from high school) 
Basic secondary ……………………………………………………………………….... 2 
(Graduated from basic secondary school (8-9 grade) did not complete vocational school) 
Completed secondary ………………..………………………………………………...... 3 
(Completed full secondary school (grades 10-11) or graduated from vocational school) 
Completed Associates Degree …………………….....………………………………….. 4 
Started but not yet completed Higher Education …….............………………………….. 5 
Full or basic higher education ……………………….……………………...……………6 
(Finished University, received the degree of Bachelor or Master) 
Hard to say / refuse to answer …………………………………………………………. 99 

 
A5. What is your main occupation that you spend most of your time? 

Employee (private company, institution, or military) …………………...........……...…. 1 
Self-Employed (contractor, entrepreneur) ………………………..…………………...... 2 
Temporarily Employed (unemployed, maternity leave) ……………………………....... 3 
Retired …………………………………………………………………………....……... 4 
Does not work / housekeeping ………………………………………………………..… 5 
Student ………………………………………………………………………………..…. 6 
Hard to say / refusal to answer …………………………………….…………………..  99 

 
A6. What can you say about the financial state of your family?   

 

 

 

Forced to economize on food ………………………………..……………….…………..  1 
Enough for food. To purchase clothes, shoes, need to save or borrow ……………..……  2 
Enough for food, clothing, footwear. Need to borrow for good suit, mobile phone,  
vacuum cleaner you need to save or borrow …………………….…………………....….. 3 
Enough for basics but the acquisition of expensive items such as washing machine,  
fridge), you must save or borrow …………………………………………………………. 4 
Enough for food, clothing, shoes, expensive purchase. For such purchases, like a  
car or apartment, need to save or borrow …………………………………………………. 5 
Any necessary purchases can be done in any time ………………………………………... 6 
It is hard to say (not on card) ……………………………………………………………….. 99 

 
A7. Do you have children under 18 living with you? 

Read through, there are several options. 
Yes, children less than 6 years ……………………. ……………………………..….…… 1 
Yes, children from 6 to 18 years ……………………………………….………….……… 2 
No children under 18 years / not living with you ………………………………….……... 3 

121 

A8. Do you own or rent your accommodation?  Just one answer. 

Own …………………………………………………………………………………….…. 1 
Rent ……………………………………………………………………………………..… 2 

 
A9. Have You lived in your house less than 2 years?  Just one answer. 

Yes …………………………………………………………………………………….….. 1 
No …………………………………………………………………………………………. 

 

 
 

 

 

122 

 

 
 

 

 
 

 

 
 
 
 
 
 
 
 
 
 
 
 
 

 
 

 
 

REFERENCES 

 

123 

 

 

REFERENCES 

 

 
Aberbach, J. D. (1969). Alienation and political behavior. American Political Science Review, 

63(1), 86-99. 

 
Aberbach, J. D. & Walker, J. L. (1970). Political trust and racial ideology. The American 

Political Science Review, 64(4), 1199-1219. 

 
Abramson, P. R. (2011, April 20). Critiques and Counter-Critiques of the Postmaterialism 
Thesis: Thirty-Four Years of Debate. UC Irvine Center for the Study of Democracy 
Working Paper.  Retrieved at: https://escholarship.org/uc/item/3f72v9q4. 

 
Acock, A. C. (2013). Discovering Structural Equation Modeling Using Stata. College Station, 

TX: Stata Press. 

 
Adler, F (1983) Nations Not Obsessed with Crime. Littleton: Fred B. Rothman & Co. 
 
Akinlabi, O. M. (2015). Young people, procedural justice, and police legitimacy in Nigeria. 

Policing and Society, 27(4), 419-438. 

 
Akinlabi, O. M. (2018). Why do Nigerians cooperate with the police? Legitimacy, procedural 

justice, and other contextual factors in Nigeria. In D. Oberwittler & S. Roche (Eds.), 
Police-Citizen Relations Across the World. London: Routledge. 

 
Bagozzi, R. P. & Yi, Y. (2012). Specification, evaluation, and interpretation of structural 

equation models. Journal of the Academy of Marketing Science, 40(8), 8-34. 

 
Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in social 

psychological research: Conceptual, strategic, and statistical considerations. Journal of 
Personality and Social Psychology, 51, 1173-1182.  

 
Beck, A. (2005). Reflections on policing in post-Soviet Ukraine: A case-study of continuity. The 

Journal of Power Institutions in Post-Soviet Societies, 2. Retrieved at: 
http://pipss.revues.org/294. 

 
Beck, A. Barko, V. Tatarenko, A. (2002). Developing police management training in post-soviet 
societies: Understanding the context and setting the agenda. Police Quarterly, 5(4), 447-
469. 

 
Beck, A. & Chistyakova, Y. (2002). Crime and Policing in Post-Soviet Societies: Bridging the 

Police/Public Divide. Policing and Society, 12(2), 123-137. 

 
Beck, A. Robertson, A. (2009). The challenges to Developing Democratic Policing in Post-

Soviet Societies: the Russian Experience. Police Practice and Research, 10(4), 285-293. 

 

124 

 

 
Beetham, D. (2002). Political Legitimacy. In K. Nash, A. Scott (Eds.), The Blackwell Companion 

to Political Sociology. Malden, MA: John Wiley & Sons. 

 
Beetham, D. (1974). Max Weber and the Theory of Modern Politics. Cambridge, UK: Polity 

Press. 

 
Bennett, R. R. & Flavin, J. M. (1994). Determinants of fear of crime: The effect of cultural 

setting. Justice Quarterly, 11, 345-364.  

 
Besanez, M. E. (2016). A World of Three Cultures. New York: Oxford University Press. 
 
Blau, P. M. (1963). Critical remarks on Weber’s theory of authority. The American Political 

Science Review, 57(2), 305-316. 

 
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. 

 
Bradford, B. (2014). Policing and social identity: procedural justice, inclusion, and cooperation 

between police and public. Policing and Society, 24(1), 22-43. 

 
Bradford, B., Huq, A., Jackson, J., and Roberts, B. (2013). What price fairness when security is 
at stake? Police legitimacy in South Africa. Regulation and Governance, 8(2), 246-268. 

 
Bradford, B., Jackson, J. & Stanko, E. A. (2009). Contact and confidence: revisiting the impact 

of public encounters with the police. Policing and Society, 19(1), 20-46. 

 
Braga, A. A., Winship, C. Tyler, T. R., Fagan, J. & Meares, T. L. (2014). The salience of social 

contextual factors in appraisals of police interactions with citizens: A randomized 
factorial experiment. Journal of Quantitative Criminology, 30, 599-627. 

 
Brandl, S. G., Frank, J., Worden, R. E., & Bynum, T. S. (1994). Global and specific attitudes 
toward the police: Disentangling the relationship. Justice Quarterly, 11(1), 119–134. 

 
Brockner, J., Ackerman, G., Greenberg, J., Gelfand, M. J., Francesco, A. M., Chen, Z. X., 

Leung, K., Bierbrauer, G., Gomez, C., Kirkman, B. L. & Shapiro, D. (2001). Culture and 
procedural justice: The influence of power distance on reactions to voice. Journal of 
Experimental Social Psychology, 37, 300-315. 

 
Brown, G. W. (1974). Adolescent attitudes and lawful behavior. Public Opinion Quarterly, 38, 

98-106. 

 
Brown, R. L. (1986). Efficacy of the indirect approach for estimating structure equation models 
with missing data: A comparison of five methods. Structural Equation Modeling, 1, 287-
316. 

 

125 

 

Cao, L. (2011). Visible minorities and confidence in the police. Canadian Journal of 

Criminology and Criminal Justice, 53, 1-26. 

Cao, L., Yung-Lien, L., Zhao, R. (2011). Shades of blue: Confidence in the police in the world. 

Journal of Criminal Justice, 40, 40-49. 

Campbell, D. T. & Fiske, D. W. (1959) Convergent and discriminant validation by the multitrait-

multimethod matrix. Psychological Bulletin, 56(2), 81-105. 

 
Carmines, E. G. & Zeller, R. A. (1979). Reliability and Validity Assessment. Beverly Hills, CA: 

Sage Publications. 

 
Cardoza, R. N. (1965). An experimental study of consumer effort, expectation, and satisfaction. 

Journal of Marketing Research, 2, 244-249. 

 
Catterberg, G. & Moreno, A. (2005). The individual bases of political trust: Trends in new and 

established democracies. International Journal of Public Opinion Research, 18(1), 31-48. 

Chen, F.F. (2007). Sensitivity of goodness of fit indexes to lack of measurement invariance. 

Structural Equation Modeling: A Multidisciplinary Journal, 14(3), 464-504.  

Chin, W., (1998). The Partial least squares approach to structural equation modeling. In G. A. 

Marcoulides (Ed.)., Modern Methods for Business Research. London: Psychology Press.  

 
Clarke, H. D., Dutt, N., & Rapkin, J. (1991). Measuring values change in western industrialized 
societies: The impact of unemployment. American Political Science Review, 85, 905-920. 

 
Clark, H. D., Kornberg, A., McIntyre, C., Bauer-Kaase, P. and Kaase, M. (1999). The effect of 
economic priorities on the measurement of value change: New experimental evidence. 
American Political Science Review, 93(September), 637-647. 

 
Cohen, L. E., Felson, M. & Land, K. C. (1980).  Property crime rates in the United States: A 

macrodynamic analysis, 1947-1977; with ex ante forecasts for the mid-1980s.  American 
Journal of Sociology, 86(1), 90-118. 

 
Cortina, J. M. (1993). What is coefficient alpha? An examination of theory and applications. 

Journal of Applied Psychology, 78, 98-104.  

 
Covington, J. and Taylor, R. B. (1991). “Fear of Crime in Urban Residential Neighbourhoods: 
Implications of Between- and Within- Neighborhood Sources for Current Models. The  
Sociological Quarterly, 32, 2, 231-249. 

 
Curry, P. & O’Connell, M. F. (2000). Post-materialist values and political preference: Some 

unlikely findings from Northern Ireland. European Journal of Political Research, 37, 19-
30. 

 

 

126 

 

Czapska, J., Radomska, E. & Wójcik, D. (2016).  Police legitimacy, procedural justice, and 
cooperation with the police: A Polish perspective. Journal of Criminal Justice and 
Security, 16(4), 453-470. 

 
Dammert, L. & Malone, M. F. (2003). Fear of crime or fear of life? Public insecurities in Chile. 

Bulletin of Latin American Research, 22(1), 79-101. 

 
Darden, K. A. (2001). Blackmail as a tool of state domination: Ukraine under Kuchma. East 

European Constitutional Review, Spring-Summer 2001, 67-71. 

 
Davis, A. W. (1977). Asymptotic theory for principal component analysis: Non-normal case. 

Australian & New Zealand Journal of Statistics, 19(3), 206-212. 

 
Davis, D. W. & Davenport, C. (1999). Assessing the validity of the postmaterialist index. The 

American Political Science Review, 93(3), 649-677. 

 
Davis, D. W., Dowley, K. M., Silver, B. D. (1999). Postmaterialism in world societies: Is it 

really a value dimension? American Journal of Political Science, 43(3), 935-962. 

 
Davis, R. C. & Henderson, N. J. (2003). Willingness to report crime: The role of ethnic group 

members and community efficacy. Crime & Delinquency, 49 (4), 564-580. 

 
Dogan, M. (1992). Conceptions of legitimacy. In M. E. Hawkesworth, M. Kogan (Eds.), 

Encyclopedia of Government and Politics. New York: Routledge. 

 
DuBow, F., McCable, E., Kaplan, G. (1979). Reactions to Crime: A Critical Review of the 
Literature. Washington, DC:  National Institute of Law Enforcement and Criminal 
Justice: U.S. Government Printing Office. 

 
Duch, R. & Taylor, M. (1993). Postmaterialism and the economic condition. American Journal 

of Political Science, 37, 747-779. 

 
Durose, M. R., Langton, L. (2013). Study Finds Some Racial Differences in Perceptions of 

Police Behavior During Contact with the Public. Bureau of Justice Statistics. Washington 
DC. Retrieved at: https://www.bjs.gov/index.cfm?ty=pbdetail&iid=4915. 

 
Easton, D. (1965). A Systems Analysis of Political Life. Chicago: University of Chicago Press. 
 
Easton, D. (1975). A reassessment of the concept of political support. British Journal of Political 

Science, 5, 435-457. 

 
ESS (2010). European Social Survey Round 5 Data.  Norwegian Centre for Research Data. 

Retrieved at: http://www.europeansocialsurvey.org/data. 

 
Evans, J. D. (1996). Straightforward statistics for the behavioral sciences. Pacific Grove, CA: 

Brooks/Cole Publishing.  

 

127 

 
Fernandez, K. E. & Kuenzi, M. (2010). Crime and Support for Democracy in Africa and Latin 

America. Political Studies, 58(3), 450-471. 

 
Ferrero, K. F. & Grange, R. L. (1987). The measurement of fear of crime. Sociological Inquiry, 

57(1), 70-97. 

 
Flander, B. & Ručman, A. B. (2015). Lost in transition: Criminal justice reforms and the crises 

of legitimacy in Central and Eastern Europe. In G. Mesko & J. Tankebe (Eds.), Trust and 
Legitimacy in Criminal Justice: European Perspectives. New York: Springer. 

 
Gamson, W. A. (1968). Power and discontent. Homewood, IL: Dorsey. 
 
Garfalo, J. & Laub, J. (1978). The fear of crime: Broadening our perspective. Victimology, 3, 

 
Gau, J. M. (2010). Basic principles and practices of structural equation modeling in criminal 

justice and criminology research. Journal of Criminal Justice Education, 21(2), 136-151. 

 
Gau, J. M. (2011). The convergent and discriminant validity and procedural justice and police 

legitimacy: An empirical test of core theoretical propositions.  Journal of Criminal 
Justice, 39, 489-498. 

 
Gau, J. M. (2015). Procedural justice, police legitimacy, and legal cynicism: A test for mediation 

efforts. Police Practice and Research: An International Journal, 16(5), 402-415. 

 
Gau, J. M., Corsaro, N., Stewart, E. A., Brunson, R. K. (2012). Examining macro-level impacts 

on procedural justice and police legitimacy. Journal of Criminal Justice, 40, 333-343. 

 
Gerasymenko, G. (2011). Trafficking in human beings: The roots and realities in Ukraine. Kyiv, 

Ukraine: National Academy of Sciences. Retrieved at: http://www.wiscnetwork.org/ 
porto2011/papers/WISC_2011-600.doc.  

 
Gerbing, D. W. & Anderson, J. C. (1988). An updated paradigm for scale development 

incorporating unidimensionality and its assessment. Journal of Marketing Research, 
XXV, 188-192. 

 
Gilley, B. (2009). The Right to Rule: How States Win and Lose Legitimacy. New York: 

Columbia University Press. 

 
Goudriaan, Lynch, and Nieuwbeerta (2004). Reporting to the police in western nations: A 

theoretical analysis of the effects of social context. Justice Quarterly, 21(4), 933-969. 

 
Grafstein, R. (1981). The Failure of Weber’s Conception of Legitimacy: Its Causes and 

Implications. The Journal of Politics, 43, 456-472.  

242-253. 

 

 

 

128 

 

Green, S. B., Lissitz, R. W. & Mulaik, S. A. (1977). Limitations of coefficient alphas as an index 

of test unidimensionality. Educational and Psychological Measurement, 37, 827-838.  

 
Grewal, R., Cote, J. A. & Baumgartner, H. (2004). Multicollinearity and Measurement Error in 

Structural Equation Models: Implications for Theory Testing. Marketing Science, 
23(4),519-529. 

 
Hair, J. F., Black, W. C., Babin, B. J. & Anderson, R. E. (2010). Multivariate Data Analysis, 7th 

Edition. Upper Saddle, New Jersey: Pearson. 

 
Hale, C. (1996). Fear of crime: A review of the literature. International Review of Victimology, 

 
Haller, M. (2002). Theory and Method in Comparative Study of Values. European Sociological 

12, 221-239. 

Review, 18(2), 139-158. 

 
Harkin, D. (2015). Police legitimacy, ideology and qualitative methods: A critique of procedural 

justice theory. Criminology & Criminal Justice: An International Journal, 15(5), 594- 
612.  

 
Harrington, A. (2009). Confirmatory Factor Analysis. New York, NY: Oxford University Press.  
 
Hawdon, J. E., Ryan, J. Griffin, S. P. (2003). Policing tactics and perceptions of police 

legitimacy. Police Quarterly, 6(4), 469-491. 

 
Hero, R. E. & Durand, R. (1985). Explaining citizen evaluations of urban services: A comparison 

of some alternative models. Urban Affairs Quarterly, 20, 344-354. 

 
Hinds, L. & Murphy, K. (2007). Public satisfaction with police: Using procedural justice to 
improve police legitimacy. The Australian and New Zealand Journal of Criminology, 
40(1), 27-42. 

 
Hirschi, T. (1969). Causes of Delinquency. Berkeley, CA: University of California Press. 
 
Hirschi, T. (1979).  Separate and unequal is better. Journal of Research in Crime and 

 
Hirschi, T. & Selvin, H. C. (1966). False criteria of casualty in delinquency research. Social 

Delinquency, 16(1), 34-38. 

Problems, 13(3), 254-268. 

 
Hiskey, J. T. & Bowler, S. (2005). Local context and democratization in Mexico. American 

Journal of Political Science, 49(1), 57-71. 

 
Hohl, K., Bradford, B., Stanko, E. A. (2010). Influencing trust and confidence in the London 
Metropolitan Police: Results from an experiment testing the effect of leaflet drops on 
public opinion. British Journal of Criminology, 50(3), 491-513. 

 

129 

 

Holton, E. F., Bates, R. A., Bookter, A. I., Yamkovenko, V. B. (2007). Convergent and divergent 

validity of the learning transfer system inventory. Human Resource Development 
Quarterly, 18(3), 385-419. 

Hooper, D., Coughlan, J., Mullen, M. (2008). Structural Equation Modelling: Guidelines for 

Determining Model Fit. Electronic Journal of Business Research Methods, 6(1), 53-60.  

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

 
Hough, M. Jackson, J. Bradford, B. (2014). Trust in justice and the legitimacy of legal 

authorities: Topline findings from a European comparative study. In Body-Gendrot, S. 
Hough, M. Kerezsi, K. Levy, R. Snacken, S. (Eds.) The Routledge Handbook of 
European Criminology. New York: Routledge. 

 
Hu, L. & Bentler, P. M. (1999). Cutoff criteria for fix indexes in covariance structure analysis” 
Conventional criteria versus new alternatives.  Structural Equation Modeling, 6(1), 1-55. 

 
Hummelsheim, D., Hirtenlehner, H., Jackson, J. and Oberwittler, D. (2011). Social insecurities 

and fear of crime: A cross-national study on the impact of welfare state policies on crime-
related anxieties. European Sociological Review, 27(3), 327-345. 

 
Huntington, S. P. (2000). In L. Harrison & S. Huntington (Eds.), Culture Matters: How Values 

Shape Human Progress. New York: Basic Books. 

 
Hyde, A. (1983). The concept of legitimation in the sociology of law. Wisconsin Law Review, 

1983, 379-426. 

 
Icduygo, A., Romano, D. & Sirkeci, I. (1999). The ethnic question in an environment of 

insecurity: the Kurds in Turkey. Ethnic and Racial Studies, 22(6), 991-1010. 

 
Inglehart, R. (1971). The silent revolution in Europe: Intergenerational change in post-industrial 

societies. American Political Science Review, 65(4), 991-1017. 

 
Inglehart, R. (1990). Culture Shift in Advance Industrial Society. Princeton, NJ: Princeton 

University Press.  

 
Inglehart, R. (1999). Postmodernization erodes respect for authority, but increases support for 

democracy. In P. Norris (Ed.), Critical Citizens: Global Support for Democratic 
Governance. Oxford: Oxford University Press. 

 
Inglehart, R. (2008). Changing values among Western publics from 1970 to 2006. West 

European Politics, 31(1), 130-146. 

 

 

130 

 

Inglehart, R. (2012) Modernization and democracy. In V. Inozemtsev and P. Dutkiewicz (Eds.), 
Democracy versus Modernization: A Dilemma for Russia and for the World. New York: 
Routledge.  

 
Inglehart, R. & Baker, W. E. (2000). Modernization, cultural change, and the persistence of 

traditional values. American Sociological Review, 65(1), 19-51. 

 
Inglehart, R. & Welzel, C. (2005). Cultural Change and Democracy: The Human Development 

Sequence. Cambridge, UK: Cambridge University Press. 

 
Innes, M. (2004). Reinventing Tradition? Reassurance, neighborhood security and policing. 

Criminology & Criminal Justice, 4(2), 151-171.  

 
Jackson, J., Asif, M., Bradford, B., and Zakar, M. Z. (2014). Corruption and police legitimacy in 

Lahore, Pakistan. British Journal of Criminology, 54(6), 1067-1088. 

 
Jackson, J., Bradford, B., Hohl, K. & Farrall, S. (2009). Does the fear of crime erode public 

confidence in policing?  Policing, 3(1), 100-111. 

 
Jackson, J., Bradford, B., Hough, M., Myhill, A., Quinton, P. & Tyler, T. (2012). Why do people 

comply with the law? Legitimacy and the influence of legal institutions. British Journal 
of Criminology, 52(6), 1051-1071. 

 
Johnston, M. (2005). Syndromes of Corruption: Wealth, Power, and Democracy. New York: 

Cambridge University Press.   

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

 
Kaiser, K. A. (2016). Procedural Justice and Legal Socialization Among Serious Adolescent 
Offenders. (Doctoral dissertation). Retrieved from ASU Repository: https://repository. 
asu.edu/attachments/170384/content/Kaiser_asu_0010E_15804.pdf. 

 
Karklins, R. (2005). The System Made Me Do It: Corruption in Post-Communist Societies. New 

York: M.E. Sharpe.  

 
Karakus, O. (2017). Instrumental and normative pathways to legitimacy and public cooperation 
with the police in Turkey: Considering perceived neighborhood characteristics and local 
government performance. Justice Quarterly, 34, 1, 25-54. 

 
Kenny, D. (2015, November 24). Measuring Model Fit. Retrieved at: http://davidakenny.net/ 

cm/fit.htm. 

 
KHPG (2013). Kharkiv Human Rights Group Report - Analysis of Human Rights Situation in 

Ukraine.  Retrieved at: http://khpg.org/en/index.php?id=1398077898. 

 

131 

 
Kochel, T. R. (2012). Can police legitimacy promote collective efficacy? Justice Quarterly, 29, 

384-419. 

 
Kulik, A.G. 1994. Otnoshenie naseleniia k militsii. Prestupnost v Ukraine, no. 2. 
 
Kutnjak Ivković, S. (2008). A comparative study of public support for the police. International 

Criminal Justice Review, 18, 406-434. 

 
LaGrange, R. L., Ferraro, K., Supancic, M. (1992). Perceived risk and fear of crime: Role of 

social and physical incivilities. Journal of Research in Crime and Delinquency, 29, 311-
334. 

 
Leiker, J. J. (2003). Fear of Crime in the City: A Multivariate Analysis of Perceptions of 
Community Safety in Twelve Cities. (Doctoral dissertation). Retrieved at ProQuest 
Dissertations and Theses Full Text Database. (UMI No.3120549). 

 
Leitner, J. Meissner, H., Maryna-David, E. (2015). The debate about political risk: How 

corruption, favoritism, and institutional ambiguity shape business strategies in Ukraine. 
In Karasaavvoglou, A. Ongan, S., Polychrodinou, P. (Eds.), EU Crisis and the Role of the 
Periphery. Basel, Switzerland: Springer International. 

 
Licht, A. N., Goldschmidt, C., Schwartz, S. H. (2007). Culture rules: The foundation of the rule 
of law and other norms of governance. Journal of Comparative Economics, 35, 659-688. 

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

Plenum Press. 

 
Maguire, E. R. & Johnson, D. (2010). Measuring public perceptions of the police. Policing: An 

International Journal of Police Strategies & Management, 33(4), 703-730.  

 
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-
64. 

 
MacCallum, R. C., Roznowski, M. & Necowitz, L. B. (1992). Model modifications in covariance 

structure analysis: the problem of capitalization on chance. Psychological Bulletin, 11, 
490-504. 

 
Manning, P. K. (2001). Theorizing policing: The drama and myth of crime control in the NYPD. 

Theoretical Criminology, 5(3), 315-344. 

 
Matsunaga, M. (2008). Item parceling in structural equation modeling: A primer. 

Communication Methods and Measures, 2(8), 260-293. 

 

 

 

132 

 

McEwen, C. A. & Maiman, R. J. (1986). In search of legitimacy: Toward an empirical analysis. 

Law & Policy, 8, 257-272. 

 
McGarrell, E; Giacomazzi, A., Thurman, Q. (1997). Neighborhood disorder, integration, and the 

fear of crime. Justice Quarterly, 14(3), 479-500. 

 
McNamara, R.A., Norenzayan, A. & Henrich, J. (2016). Supernatural punishment, in-group 

biases, and material insecurity: experiments and ethnography from Yasawa, Fiji. 
Religion, Brain & Behavior, 6(1), 34-55. 

 
Milfont, T. L. & Fischer, R. (2010). Testing measurement invariance across groups: Applications 

in cross-cultural research. International Journal of Psychological Research, 3(1), 111-
121. 

 
Millsap, R. E. (2011). Statistical Approaches to Measurement Invariance. New York: Taylor & 

Francis Group. 

 
Morata-Ramirez, M. & Holgado-Tello, F. P. (2013). Construct validity of Likert scales through 

confirmatory factor analysis: A simulation study comparing different methods of 
estimation based on Pearson and polychoric correlations. International Journal of Social 
Science Studies, 1(1), 54-65.  

 
Murphy, K. (2009). Public satisfaction with police: The importance of procedural justice and 

police performance in police-citizen encounters. The Australian and New Zealand 
Journal of Criminology, 42(2), 159-178. 

 
Murphy, K. (2013). Does procedural justice matter to youth? Comparing adults’ and youths’ 

willingness to collaborate with police. Policing and Society, 23(1), 1-24. 

 
Murphy, K. & Cherney, A. (2011). Fostering cooperation with the police: How do ethnic 

minorities in Australia respond to procedural justice-based policing. Australian & New 
Zealand Journal of Criminology, 44(2), 235-257. 

 
Murphy, K., Hinds, L., & Fleming, J. (2008). Encouraging public cooperation and support for 

police. Policing & Society, 18(2), 136-155. 

 
Myhill, A. & Bradford, B. (2012). Can police enhance public confidence by improving quality of 
service? Results from two surveys in England and Wales. Policing & Society, 22(4), 397-
425. 

 
Nalla, M. K., Hamm, J. A., Paek, S. Y. (2017). Is police integrity an important predictor of 

citizen satisfaction in police in post-colonial emerging democracies? The case of India. 
Asian Criminology. Retrieved at: https://link.springer.com/article/10.1007/s11417-017-
9254-9. 

 
Nunnally, J. C. (1978). Psychometric Theory. New York: McGraw-Hill. 

 

133 

 
OECD, (2013). Trust in Government. OECD Publishing. Retrieved at: http://www.oecd-

ilibrary.org/docserver/download/4213201ec007.pdf?expires=1460912086&id=id&accna
me=guest&checksum=53D96C0C72354577FC7D233DA76E0C7E. 

 
Oliveira, A. & Murphy, K. (2014). Race, social identity, and perceptions of police bias. Race and 

Justice, 5(3), 259-277. 

 
Onwuegbuzie, A. J. & Daniel, L. G. (2002). Uses and misuses of the correlation coefficient. 

Research in the Schools, 9, 73-90. 

 
Ortega, D. E. (2016, January 5). Effectiveness versus legitimacy: Use of force and police training 

in Latin America. Washington DC: Brookings Institution. Retrieved at: 
http://www.brookings.edu/blogs/up-front/posts/2016/01/06-police-training-latin-america-
ortega 

 
OSCE (2013, September 30). ODIHR Report: Militia vs. Ukrainian People. Retrieved at: 

https://www.osce.org/odihr/106136?download=true.  

 
Paige, J. M. (1971). Political orientation and riot participation. American Sociological Review, 

36(5), 810-820. 

Ethics, 45, 282-316. 

 
Parsons, T. (1935). The place of ultimate values in sociological theory. International Journal of 

 
Paternoster, R., Brame, R., Bachman, R., & Sherman, L. W. (1997). Do fair procedures matter? 

The effect of procedural justice on spouse assault. Law & Society Review, 31 (1), 163-
204. 

 
Pavlović, Z. (2015). Individual and Country Level Determinants of (Post)Materialist Values in 

Eastern Europe. European Quarterly of Political Attitudes and Mentalities, 4(2). 

 
Peacock, R. & Cordner, G. (2016). “Shock therapy” in Ukraine: A radical approach to post-

soviet reform. Public Administration and Development, 36(2), 80-92. 

 
Peng C. Y. J., Harwell M., Liou S. M., Ehman L. H. (2006). Advances in missing data methods 

and implications for educational research. In S. S. Sawilowsky (Ed.), Real Data 
Analysis. Charlotte, NC: Information Age Publishing. pp. 31–78. 

 
Perez, O. (2003). Democratic legitimacy and public insecurity: Crime and democracy in El 

Salvador and Guatemala. Political Science Quarterly, 118(4), 627-644. 

 
Piquero, A. R. & Rosay, A. B. (1998). The reliability and validity of Gramick et al.’s Self-

Control Scale: A Comment on Longshore et al. Criminology, 36(1), 157-173. 

 

 

 

134 

 

Pryce, D. K., Johnson, D. J., & Maguire, E. R. (2016). Procedural justice, obligation to obey, and 

cooperation with police in a sample of Ghanaian immigrants. Criminal Justice and 
Behavior. Retrieved at: http://journals.sagepub.com/doi/abs/10.1177/00938548166802 
25?journalCode=cjbb. 

 
 
Quinlan, T. L. & Ramirez, D. (2014, December 24). A blueprint for local police reform to 
improve legitimacy. Huffington Post. Retrieved at: http://www.huffingtonpost.com/ 
tara-lai-quinlan/a-blueprint-for-local-pol_b_6369706.html. 

 
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. 

 
Reisig, M. D., & Chandek, M. S. (2001). The effects of expectancy disconfirmation on outcome 

satisfaction in police-citizen encounters. Policing, 24(1), 88-99. 

 
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. 

 
Reisig, M. D., Tankebe, J. Mesko, G. (2014). Procedural justice, police legitimacy, and public 
cooperation with the police among young Slovene adults. Journal of Criminal Justice 
and Security, 2, 147-164. 

 
Rinehart Kochel, T., Parks, R. & Mastrofski, S. D. (2013). Examining police effectiveness as a 
precursor to legitimacy and cooperation with police. Justice Quarterly, 30(5), 895-925. 

 
Rodgers, H. & Lewis, E. (1974). Political support and compliance attitudes. American Politics 

Quarterly, 2, 61-77. 

 
Rogowski, R. (1974). Rational Legitimacy: A Theory of Political Support. Princeton, NJ: 

Princeton University Press. 

 
Rosenbaum, D. P., Schuck, A. M., Costello, S. K., Hawkins, D. E. & Ring, M. K. (2005). 
Attitudes toward the police: The effects of direct and vicarious experience. Police 
Quarterly, 8(3), 343-365. 

 
Rousseau, D. M., Sitkin, S. B., Burt, R. S., & Camerer, C. (1998). Not so different after all: A 
cross-discipline view of trust. The Academy of Management Review, 23 (3), 393-4040. 

 
Sahakyan, A. (2015, June 20). The hydra-headed monster of police corruption in the former 

Soviet Union. Kyiv Post. Retrieved at: http://www.kyivpost.com/opinion/op-ed/armine-
sahakyan-the-hydra-headed-monster-of-police-corruption-in-the-former-soviet-union-
391563.html. 

 
Sarat, A. (1975). Support for the legal system: An analysis of knowledge, attitudes, and behavior. 

American Politics Quarterly, 3, 3-24. 

 

135 

 

 
Seligman, A. B. (1997). The Problem of Trust. Princeton, NJ: Princeton University Press. 
 
Seligson, M. A. (1980). Trust, efficacy, and modes of political participation: A study of Costa 

Rican peasants. British Journal of Political Science, 10(1), p. 75-98. 

 
Shapiro, D. L.; Kirkman, B. L. (1999). Employees’ reaction to the change to work teams: The 
influence of “anticipatory” injustice. Journal of Organizational Change Management, 
12(1), 51-66. 

 
Shapiro, D. L.; Kirkman, B. L. (2001). Anticipatory injustice: The consequences of expecting 

injustice in the workplace.  In J. Greenberg & Cropanzano R. (Eds.), Advances in 
Organizational Justice. Stanford, CA: Stanford University Press. 

 
Shelley, L. I. 1998.  Organized crime and corruption in Ukraine: Impediments to the 

development of a free market economy. Demokratizatsiya: The Journal of Post-Soviet 
Democratization, 6(4), 648-663. 

 
Shelley, L. I. 1999. Post-Socialist Policing: Limitations on Institutional Change.  In Mawby, R. I. 

(Ed.), Policing Across the World: Issues for the Twenty-first Century. New York: 
Routledge. 

 
Sherman, L. W. (1993). Defiance, deterrence, irrelevance: A theory of the criminal sanction. 

Journal of Research in Crime and Deliquency, 30, 445-473. 

 
Sherman, L. W. (2001). Trust and Confidence in Criminal Justice. National Institute of Justice. 

Retrieved at: https://www.ncjrs.gov/pdffiles1/nij/189106-1.pdf. 

 
Sivo, S. A., Xitao, F. E., Witta, E. L. Y Willse, J. T. (2006). The search for “optimal” cutoff 

properties: Fit index criteria in structural equation modeling. The Journal of Experimental 
Education, 74(3), 267-288. 

 
Skogan, W. G. (2006). Asymmetry in the impact of encounters with the police. Policing and  

Society, 16(2), 99-126. 

 
Skogan, W. G., Craen, M. V., Hennessy, C. (2014). Training police for procedural justice. 

Journal of Experimental Criminology. Retrieved at http://link.springer.com/article/10. 
1007/s11292-014-9223-6. 

 
Stack, J. (1986). The Primordial Challenge. Westport, CT: Greenwood Press.  
 
State Statistics Service of Ukraine (2001). Ukrainian Census for 2001. Retrieved from 

http://ukrcensus.gov.ua/eng. 

 
Steenkamp, J. E. M. & Baumgartner, H. (1998). Assessing measurement invariance in cross-

national consumer research. The Journal of Consumer Research, 25(1), 78-90. 

 

136 

 

 

 

 
Tankebe, J. (2008). Police effectiveness and police trustworthiness in Ghana: An empirical 

appraisal. Criminology and Criminal Justice, 8, 185-202. 

 
Tankebe, J. (2009). Public cooperation with the police in Ghana: Does procedural fairness 

matter. Criminology, 47(4), 1265-1293. 

 
Tankebe, J. (2013a). Viewing things differently: The dimensions of public perceptions of police 

legitimacy. Criminology, 51(1), 103-135. 

 
Tankebe, J. (2013b). In search of moral recognition? Policing and eudaemonic legitimacy in 

Ghana. Law & Social Inquiry, 38(3), 576-597. 

 
Tankebe, J., Reisig, M. D., Wang, X. (2016). A multidimensional model of police legitimacy: A 

cross-cultural assessment. Law and Human Behavior, 40(1), 11-22. 

 
Taylor, R. B. & Hale, M. M. (1986). Testing alternative models of fear of crime. Journal of 

Criminal Law and Criminology, 26, 151-189. 

 
Sterne, J. A. C., White, I. R., Carlin, J. B., Spratt, M., Royston, P., Kenward, M. G., Wood, A. 

M. & Carpenter, J. R., 2009. Multiple imputation for missing data in epidemiological and 
clinical research: potential and pitfalls. British Medical Journal, 339, 157-160.  

 
Stolle, D., Soroka, S. & Johnston, R. (2008). When does diversity erode trust? Neighborhood 
diversity, interpersonal trust, and the mediating effect of social interactions. Political 
Studies, 56, 57-75. 

 
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. 

 
Sunshine, J. W. (2007). New York and its Police: The Impact of Perceptions of Justice on Officer 
Credibility. (Doctoral dissertation). Retrieved at ProQuest Dissertations and Theses Full 
Text Database. (UMI No.3222009). 

 
Sunshine, J. W. & Tyler, T. R. (2003a). The role of procedural justice and legitimacy in shaping 

public support for policing. Law & Society Review, 37, 513-548. 

 
Sunshine, J. W. & Tyler, T. R. (2003b). Moral solidarity, identification with the community, and 

the importance of procedural justice: The police as prototypical representatives of a 
group’s moral values. Social Psychology Quarterly, 66(2), 153-165. 

 
Tabachnick, B. G. & Fidel, L.S. (1996). Using Multivariate Statistics (3rd ed.). NewYork: Harper 

Collins College Publishers.    

137 

 

Taylor, R. B. (1999). Crime, Grime, Fear and Decline: A longitudinal look. Washington, D.C.: 
U.S. Department of Justice Office of Justice Programs. Retrieved at: https://www.ncjrs. 
gov/pdffiles1/177603.pdf. 

 
Thibaut, J. W. & Walker, W. L. (1975). Procedural Justice: A Psychological Analysis. New 

York: L. Erlbaum Associates. 

 
TI (2013). Transparency International Corruption Perceptions Index 2013. Retrieved at: 

https://www.transparency.org/cpi2013/results. 

 
Tseloni, A., Mailley, J., Farrell, G., Tilley, N. (2010). Exploring the international decline in 

crime rates. European Journal of Criminology, 7(5), 375-394. 

 
Tsushima, M. & Hamai, K. (2015). Public cooperation with the police in Japan: Testing the 

legitimacy model. Journal of Contemporary Criminal Justice, 3(2), 212-228. 

 
Tyler, T. R. (1988). What is procedural justice? Law & Society Review, 22, 301-355. 
 
Tyler, T. R. (1990). Why People Obey the Law. New Haven, CT: Yale University Press.  
 
Tyler, T. R. (1996). Trust in Organizations. Thousand Oaks, CA: Sage 
 
Tyler, T. R. (1997). The psychology of legitimacy: A relational perspective on voluntary 
deference to authorities. Personality and Social Psychology Review, 1, 323-345. 

 
Tyler, T. R. (1998). Trust and democratic government. In V. Braithwaite & M. Levi (Eds.), Trust 

and Governance. New York: Russell Sage Foundation. 

 
Tyler, T. R. (2001a). Social Justice. In R. Brown & S. Gaertner (Ed.), Blackwell Handbook of 

Social Psychology. London: Blackwell. 

 
Tyler, T. R. (2001b). Trust and law abiding behavior: Building better relationship between 

police, the courts, and the minority community. Boston University Law Review, 81, 361-
406. 

 
Tyler, T. R. (2003). Procedural justice, legitimacy, and the effective rule of law. In M. Tonry 

(Ed.), Crime and Justice. Chicago: University of Chicago Press. 

 
Tyler, T. R . (2004) Enhancing police legitimacy. Annals of the American Academy of Political 

and Social Science, 593, 84-89. 

 
Tyler, T. R. (2005a). Policing in black and white: Ethnic group differences in trust and 

confidence in the police. Police Quarterly, 8, 322-342. 

 
Tyler, T. R. (2005b). Introduction: Legitimating Ideologies. Social Justice Research 18(3), p. 

211-215. 

 

138 

 
Tyler, T. R. (2006). Psychological perspectives on legitimacy and legitimation. Annual Review of 

Psychology, 57, 375-400. 

 
Tyler, T. R. (2009). Legitimacy and criminal justice: The benefits of self-regulation. Ohio State 

Journal of Criminal Law, 7(1), 307-359. 

 
Tyler, T. R. (2011). Trust and legitimacy: Policing in the USA and Europe. European Journal of 

Criminology, 8(4), 254-266. 

 
Tyler, T. R. & Blader, S. (2000). Cooperation in Groups: Procedural Justice, Social Identity, 

and Behavioral Engagement. London: Psychology Press. 

 
Tyler, T. R. & Darley, J. M. (2000). Building a law-abiding society: Taking public views about 

morality and the legitimacy of legal authorities into account when formulating 
substantive law. Hofstra Law Review 28(3): 707-39.  

 
Tyler, T. R. & Degoey, P. (1996). Trust in organizational authorities: The influence of motive 
attributions on willingness to accept decisions.  In R. M. Kramer & T. R. Tyler (Eds.), 
Trust in Organizations: Frontiers of Theory and Research. Thousand Oaks, CA: Sage 
Publications. 

 
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. & Folger, R. (1980). Distributional and procedural aspects of satisfaction with 

citizen-police encounters. Basic and Applied Social Psychology, 1(4), 281-292. 

 
Tyler, T. R. & Huo, Y. J. (2002). Trust in the Law. New York: Russell Sage Foundation. 
 
Tyler, T. R. & Wakslak, C. J. (2004). Profiling and police legitimacy: Procedural justice, 

attributions of motive, and acceptance of police authority. Criminology, 42(2), 253-282. 

 
UNDP (2016).  Human Development Index.  Retrieved at: http://hdr.undp.org/en/countries/ 

 

 

 

profiles/TUR. 

 
Urbach, N. & Ahlemann, F. (2010). Structural Equation Modeling in Information Systems 

Research Using Partial Least Squares. Journal of Information Technology Theory and 
Application, 11(2), 5-40.  

 
Van Damme, A., Pauwels, L., Svensson, R. (2013). Why do Swedes cooperate with the police? 

A SEM analysis of Tyler’s procedural justice model. European Journal on Criminal 
Policy and Research, 21,15-33. 

139 

Van de Schoot, R., De Beuckelaer, A., Lek, K., Zondervan-Zwijnenbur, M. (2015). Editorial: 

Measurement Invariance. Frontiers in Psychology: Quantitative Psychology and 
Measurement.  Retrieved at: http://journal.frontiersin.org/journal/psychology/section/ 
quantitative-psychology-and-measurement#research-topics. 

 
Vandenberg, R. J. & Lance, C. E. (2000). A review and synthesis of the measurement invariance 

literature: Suggestions, practices, and recommendations for organizational research. 
Organizational research methods, 3, 4-69. 

 
Van Dijk, J. (2008). The World of Crime. Thousand Oaks, CA: Sage Publications. 
 
Van Ryzin, G. C. (2013). An experimental test of the expectancy-disconfirmation theory of 

citizen satisfaction. Journal of Policy Analysis and Management, 32(3), 597-614. 

 
Wagner, L. (2015, December 30). Chicago Mayor announces more tasers, training for police. 

NPR Hourly Newscast. Retrieved at: http://www.npr.org/sections/thetwo-
way/2015/12/30/461518814/chicago-mayor-rahm-emanuel-announces-further-police-
reforms. 

 
Warner, B. D. (2007). Directly intervene or call the authorities? A study of forms of 

neighborhood social control within a social disorganization framework. Criminology, 
45(1), 99-129. 

 
Weber, M. (1947). The Theory of Social and Economic Organization. New York: Oxford 

University Press. 

 
Weber, M. (1968). Economy and Society. Berkley, CA: University of California Press. 
 
Wells, W. (2007). Type of contact and evaluations of police officers: The effect of procedural 
justice across three types of police—citizen contacts. Journal of Criminal Justice, 35, 
612-621. 

 
Wheller, L., Quinton, P., Fildes, A. & Mills, A. (2013). The Greater Manchester Police 

Procedural Justice Training Experiment: Technical Report. London: The College of 
Policing. 

 
Wolfe, S. E. (2011). The effect of low self-control on perceived police legitimacy. Journal of 

Criminal Justice, 39 (1), 67-74. 

 

 

 

 
Wolfe, S. E., Nix, J., Kaminski, R. & Rojek, J. (2016). Is the effect of procedural justice on 

police legitimacy invariant? Testing the generality of procedural justice and competing 
antecedents of legitimacy. Journal of Quantitative Criminology, 32(2), 253-282. 

 
Worden, R. E. & McLean, S. J. (2017a). Research on police legitimacy: the state of the art. 

Policing, 40(3), 480-513. 

140 

 

Worden, R. E. & McLean, S. J. (2017b) Mirage of Police Reform: Procedural Justice and Police 

Legitimacy. Oakland, CA: University of California Press. 

 
World Bank (2016). Worldwide Governance Indicators Reports. Retrieved at: 

http://info.worldbank.org/governance/wgi/#reports. 

 
Wright, J. D. (1976). The Dissent of the Governed. New York: Academic Press. 
 

 

141