LEGITIMACY OF CRIMINAL JUSTICE ORGANIZATIONS AND DEMOCRATIC
STRUCTURES IN COUNTRIES OF AFRICA: A COMPARISON OF 33 NATIONS
By
Nolan Thomas Gillespie

A THESIS
Submitted to Michigan State University
in partial fulfillment of the requirements
for the degree of
Criminal Justice – Master of Science
2017

ABSTRACT
LEGITIMACY OF CRIMINAL JUSTICE ORGANIZATIONS AND DEMOCRATIC
STRUCTURES IN COUNTRIES OF AFRICA: A COMPARISON OF 33 NATIONS
By
Nolan Thomas Gillespie
Prior research on legitimacy in advanced (full) democracies suggests that procedurally
just actions of criminal justice actors relate to higher levels of perceived legitimacy by citizens.
The bulk of extant research on legitimacy usually focuses on policing, yet some researchers
understand the importance of broadening this scope by studying a more thorough representation
of the criminal justice apparatus. Conceptual frameworks using trust to measure legitimacy have
yielded intriguing results. An important direction of legitimacy research is to inquire about how
democratic structures effect legitimacy. In an age of rising globalization, democracy and
governance have become increasingly important to citizens in less advanced countries. For
historical reasons, Africa remains a unique continent to study concerning democracy and
criminal justice organizations. The aim of this research is to examine perceived legitimacy in
criminal justice organizations in less advanced democracies, specifically those that are
considered flawed, hybrid, or authoritarian. More specifically the study examines how
procedural fairness, corruption, social capital, and degrees of democracy effect trust in police and
courts among African countries that differ significantly on various democracy metrics with data
drawn from 33 African nations. The inquiry attempts to reconcile the implications of democratic
society with perceptions of criminal justice organizations. Findings and implications are
discussed.

Copyright by
NOLAN THOMAS GILLESPIE
2017

Dedicated to the people with the courage to follow their hearts.

iv

ACKNOWLEDGEMENTS

It always seems necessary to remain grateful for the help and encouragement one may
receive in consideration of any accomplishment. It is undoubtedly difficult to relegate all of the
individuals that have helped during this process in a page, but some key individuals deserve
recognition. My profound thanks to:
Mahesh Nalla for guidance and allowing me the opportunity to work under his tutelage.
Sheila Maxwell and Charles Corley for guidance and critical suggestions.
Jina Lee and James Carr for offering input and showing kindness throughout this project.
Oluwatobi Ishola for supporting me during a difficult time.
Hal Wilson for offering strong support and conversation in a time of need.
Betsy Ford and Bettie Cervantes for encouraging me to fight through difficult times.
Jordan Brooks for understanding the value of a pun and supporting me to continue.
The Third Tribe for knowing how to find the seriousness of a child at play.
Ronda and Bill Bokram, and Mary Lou Hanish for their love and support.
Greg and Cecille Glatz for always believing in me and treating me like a true son.
My mom, dad, and brothers for reassuring me to follow my heart and never give up.
Coco for being my best friend and offering profound love, our greatest journey awaits.
And to the staff in the School of Criminal Justice for helping me to fulfill all of the necessary
requirements.

v

TABLE OF CONTENTS

LIST OF TABLES

vii

LIST OF FIGURES

viii

Chapter 1 Introduction

1

Chapter 2 Legitimacy and Democracy
2.1 Legitimacy
2.2 Trust as a Measure of Legitimacy

7
8
11

Chapter 3 Predicting Legitimacy
3.1 Social Capital
3.2 Democratic Structure
3.3 Corruption
3.4 Demographic Variables

21
23
25
27
29

Chapter 4 The Present Study
4.1 Context: Africa
4.2 Method and Data
4.3 Dataset
4.4 Research Model
4.5 Measurement
4.5.1 Dependent Variables
4.5.2 Independent Variables

32
32
35
36
37
39
39
41

Chapter 5 Analysis

47

Chapter 6 Discussion and Conclusion

60

REFERENCES

65

vi

LIST OF TABLES

Table 1.

Summary of Studies Relevant to Legitimacy (Trust and Confidence)

14

Table 2.

Economist Democracy Rankings

33

Table 3.

Years of Independence

35

Table 4.

Variables in Conceptual Model

39

Table 5.

Summary Table of Survey Questions for Afrobarometer

44

Table 6.

Descriptive Statistics for Variable Categories

48

Table 7.

Bivariate Logistic Regression Summary for Selected Variables

49

Table 8.

Logistic Regression Summary for Trust in Police (N = 21,640)

50

Table 9.

Logistic Regression Summary for Trust in Courts of Law (N = 21,459)

55

Table 10.

Logistic Regression Summary for Democratic Structure: Police

57

Table 11.

Logistic Regression Summary for Democratic Structure: Courts of law

58

vii

LIST OF FIGURES

Figure 1.

Conceptual model – dependent, independent, and control variables

viii

38

Chapter 1
Introduction
Legitimacy and its role in legal systems remains one of the most important discourses in
studying the foundation of social order (Morris, 2008). Trust, a key construct of legitimacy,
relates to a larger discussion on democracy and remains an important outcome measure for
legitimacy research. Trust in institutions of government remains one of the most important
principles in supporting a democracy (Damme, Pauwels, & Svensson, 2015). Consequently, a
considerable increase in research utilizing trust as a key component of legitimacy has occurred
(Bradford, Jackson, & Stanko, 2009; Cao, 2015; Hohl, Bradford, & Stanko, 2010; Murphy,
Mazerolle, & Bennett, 2014; Tyler, 2011).

Although trust and legitimacy are not synonymous,

the former plays a key role in conceptualizing the latter.
Moreover, these concepts are increasingly important due to globalization. As
globalization continues, democracy has become the means to increase global connectivity
through information technology, trade, and security agreements (Bacevich, 2002).
Governmental structures that enhance economic growth and provide the foundation for increases
in standards of living are desirable when traditionally impoverished citizenry are aware of better
alternatives. The legitimacy of governmental entities, measured in part by citizen trust, remains
an important topic in the context of criminal justice research. Police, courts, and other criminal
justice organizations maintain order, in turn effectively allowing governmental structures to
impose direction to society. Prior research on trust in the larger context of legitimacy, however,
has mainly focused on advanced democracies (Mazerolle, Bennett, Davis, Sargeant, & Manning,
2013; Tyler, 2007a; 2011).

1

Considering the common predictors utilized in relation to legitimacy, researchers
regularly neglect governmental structure. Variables encompassing political structures account
for the theoretical importance context and environment play in explaining criminal justice
operations. Shifts in political power, aimed at specific policy goals, can explain changes in
regular criminal justice operations (Hagan, 1989). From this, a conclusion can be drawn that
political context and structure of government have implications relevant to citizen views of
legitimacy in criminal justice organizations. The inquiry of this connection between democratic
structures and perceptions of legitimacy further stimulates discussions regarding the value of
democracy. This is especially important in political climates that do not historically mirror the
ideals of Western democracy. In other words, by using democracy variables as predictors of
legitimacy one can gather a better understanding of democracy in practice. In addition, research
less often inquires as to how governmental structures effect overall legitimacy within a criminal
justice context (Anderson & Tverdova, 2003).
Democracy, in its ideal form offers the opportunity for citizens to pursue freedoms while
protecting certain political rights (Boutros-Ghali, 1995). Increased global connectivity has
allowed the spread of democratic ideals. Globalization has made democracy more pervasive
throughout the world due to the hegemonic power of the United States (Bacevich, 2002) which
seemingly provokes interest in the practical value of democratic systems. In other words,
globalization increases the probability that citizens in less advanced nations will strive for
improved governmental structures. Aside from increasing global connectivity, it seems
imperative to elucidate the pragmatic value democracy has in bolstering legitimacy. Hence the
importance of trying to identify the value of this form of government by asking whether or not
certain political structures enhance legitimacy in key criminal justice entities permitted to

2

maintain order. Citizen trust in criminal justice organizations is one way to capture how
democratic ideals affect legitimacy.
In general, most extant research on legitimacy focuses on a single criminal justice
organization, like policing. Surprisingly, little research utilizes a more thorough representation
of the criminal justice apparatus, especially beyond police, courts, and corrections (Holtfreter,
2016). Police commonly demand a major focus of such research, but court officials remain
intertwined in the system (Skolnick, 1966). On one hand, police find a unique position in society
due to the authority of utilizing force in administering legal systems (Kääriäinen, 2007). On the
other hand, put simply, the criminal justice apparatus does not function solely due to the efforts
of policing entities. Courts, the executive branch of government, congress or parliament, and
local municipalities, although disjointed and dysfunctional at times, work collectively with
police to promulgate, enforce, and carry out the implications of existing legal systems. For these
reasons, police and courts are the focus of this study instead of using a singular approach and
remain vital in maintaining social order (Tyler, 2007a). The study of additional criminal justice
entities would have occurred if the dataset allowed.
Furthermore, researchers using the procedural justice model have supported the
importance of trust as a measure of legitimacy (Murphy & Cherney, 2011; Tyler, 2011). In
addition, both procedural justice and corruption have strong support for predicting trust in
criminal justice organizations (Hinds & Murphy, 2007; Kääriäinen, 2007; Murphy, 2009;
Tankebe, 2010; Tyler, 2004; Wu & Sun, 2009). Predicting trust, in the larger context of
legitimacy, has commonly focused on variables dealing with demeanor and attributes of criminal
justice practitioners and citizen perceptions of such individuals. More specifically, procedural
justice, performance, professionalism, and demographic variables are common predictors used

3

by researchers studying trust and confidence in criminal justice organizations (Bradford et al.,
2009; Jackson & Bradford, 2010; Hinds & Murphy, 2007; Kääriäinen, 2007; Murphy, 2009;
Murphy & Cherney, 2011; Tyler, 2004). The present study uses procedural justice, corruption,
social capital, and degree of democracy to identify differences in legitimacy of police and courts.
A common tendency found in many studies is the use of citizen trust in police (Bradford et al.,
2009; Jackson & Bradford, 2010; Kääriäinen, 2007; Murphy & Cherney, 2011) related to the
concept of legitimacy. A thorough review of these predictors will be discussed in Chapter 3.
Due to the nature of the study, relevant extant research may fall outside the scope of
traditional criminal justice problems. Therefore, pertinent political science research can
effectively support some of the conceptual foundations necessary in studying such phenomena.
Political structure and legitimacy are inherent in maintaining order and increasing the capacity
for citizens to overcome the state of nature; this line of thought reveals major themes in political
theory (Edmundson, 2013). Research in criminal justice oftentimes demands a multidisciplinary
approach to adequately deal with theoretical assumptions (Bernard & Engel, 2001). This type of
interdisciplinary approach is not new to criminal justice in consideration of its origins as a
separate academic study (Payne, 2016).
Sparingly considered is comparative research on the subject (Kääriäinen, 2007)
especially in less advanced democracies. Extant literature often focuses on a single country or
traditionally western governments resulting in the dearth of studies relevant to less advanced
nations. This line of thought relates directly to the focus or lack thereof, that academia has had
on measuring democracy in Africa in conjunction with criminal justice themes. The discussion
then turns to the history of democracy in Africa that remains pivotal when pursuing such a
research topic. As a result, understanding the meaning of democracy and the role it has played

4

on the continent of Africa will allow a more accurate understanding of research findings. This
demands, at least, a rudimentary historical account of the African nations utilized in approaching
this research question.
The aim of this research is to focus on citizen perceptions of legitimacy in criminal
justice organizations in democracies comparatively different from typical western (advanced)
governmental structures. Research pertinent to legitimacy remains imperative given the position
the criminal justice system has in regulating society. The determinants of legitimacy remain at
the forefront of criminal justice research given the importance of the criminal justice apparatus
and continue to uncover pragmatic methods for improving governance. One should note that
democracy acts as a secondary background to understanding the determinants of perceptions
relevant to legitimacy.
More specifically, this study will attempt to examine the importance procedural justice,
corruption, social capital, and democracy play in shaping views of legitimacy. By studying this
relationship within the context of African nations, less advanced democracies act as a basis for
comparison. That is, if variations exist in perceptions of legitimacy, democracy level stands as a
potential determinant of this effect. Utilizing the Economist Intelligence Unit (EIU) Democracy
Standard, different governmental structures (flawed democracies, hybrid regimes, and
authoritarian regimes) will be the standard for comparing (Economist Intelligence Unit [EIU],
2015) how democratic structures influence legitimacy. These governmental structures allow
insight into the role degrees of democracy play in citizen perceptions of the legitimacy of
criminal justice organizations. The value of this context creates an important foundation for
learning the practical significance political structures have in influencing the legitimacy of
criminal justice entities. One gathers whether or not these perceptions vary across different

5

international borders depending on the level of democracy of African nations. Africa is a
continent that includes extremely diverse cultures and various forms of government
(Uwizeyimana, 2012) that offers an important foundation for fruitful research. Given that trust
remains fundamental in governing citizens, this concept acts as an effective measure of
legitimacy (Tyler, 2011) requisite for the proper functioning of democratic nations (Damme et
al., 2015). Citizens find methods of governance justifiable when trust exists in governmental
constructs (Börzel & Risse, 2016). The criminal justice apparatus acts as a defining feature in
governance, but can only exist as a justifiable establishment with the approval of citizens, hence
the importance of legitimacy.

6

Chapter 2
Legitimacy and Democracy
The study of legitimacy remains one of the most fundamental endeavors relevant to
criminal justice (Tyler, 2007a) and has a long tradition within the discipline (Meško & Tankebe,
2014). The importance of legitimacy stems from fundamental discussions regarding the state
and the populace. Public service provision does not principally rest on the monopoly of force
attributed to the state, but on its legitimacy derived from the people (Tyler, 2007a). This line of
thinking neatly fits into the logic of democracy and the responsibilities such a government has to
its citizens.
Moreover, both the theoretical underpinnings of democratic ideals and the practical
themes in the EIU Democracy Standard stem from the origins of democracy. It is widely known
that democracy arose in the ancient world and continues to spark interest and application
worldwide (Boutros-Ghali, 1995; Fleck & Hanssen, 2006). At the risk of oversimplification,
democracy encapsulates the fact that “political legitimacy derives from the people” (BoutrosGhali, 1995). The word democracy from its Greek origins means rule by the people (Fleck &
Hanssen, 2006). In other words, democracy is fundamentally the extension of political
involvement of the people, or masses, instead of a select few (Beckman, 2014). Even more, the
corollary of such political involvement is to serve the interest of the people instead of only those
in power (Bates, 2010). From this, themes like pluralism, citizen involvement in political affairs,
and civil liberties derive and encompass the metrics used by the EIU Democracy Standard in
describing levels of democracy (EIU, 2015). The parameters of the discourse now center on
defining legitimacy and understanding the importance of the concept. Legitimacy and
democracy, however, take on various forms depending on governmental structures.

7

Nevertheless, many researchers have focused their efforts on legitimacy within the context of
criminal justice (Bottoms & Tankebe, 2012; Meško & Tankebe, 2014; Tyler 2007).
2.1 Legitimacy
To begin, many researchers have defined and measured legitimacy (Beetham, 1991;
Zelditch, 2001; Mazerolle et al., 2013; Tyler 2007). This has made research on the matter
difficult because consensus remains vexed. Consequently, an effort to study legitimacy
necessitates a clear understanding of the concept. Common definitions of legitimacy regularly
stem from Weber’s notion of compliance as a key aspect resulting in deferral to authoritative
institutions (Jackson & Sunshine, 2007). Further, legitimacy has very little to do with force, but
mainly relates to the belief that an authority has the right to carry out certain actions (Zelditch,
2001). A strong foundation exists for researchers to build and adapt certain conceptualizations of
legitimacy. Literature dealing with legitimacy, however, often lacks clear conceptualizations
(Cao, 2015). Despite the convoluted nature of some research, there are common measures and
definitions of legitimacy utilized by researchers and academics.
As a foundation, Weber’s work on legitimacy remains important for research on the
subject (Tyler, 2004). Weber (1978) presented a typology that includes three distinct forms of
legitimacy. The first, legal authority, is established by rules (Weber, 1978). The second
authority stems from tradition, but mainly rests upon individual discretion to believe in the value
of customs (Beetham, 1991). The final form of authority involves the charisma of an individual
with agreeable qualities (Weber, 1978). Beetham (1991) argued how many researchers have
adopted the foundation established by Weber’s typology. Weber’s work, despite its lasting
effect, may include fundamental errors especially in consideration of the legitimacy of modern
states (Beetham, 1991). This point is particularly important regarding democratic structure and

8

legitimacy of political institutions. Despite this argument, it seems fair to defend Weber’s
conception of legal authority that rests on rules and regulatory structure. This type of authority
relates to legal systems and logically connects to establishing law to instill order. Weber was
explaining that belief in authority by those in subordinate positions is also an important condition
when regarding legitimacy (Sunshine & Tyler, 2003). Bottoms and Tankebe (2012) developed
this thought, that legitimacy involves the subordinate, by arguing that legitimacy is dialogic or
involves discourse between pertinent parties.
Contrary to Weber’s approach, Beetham (1991) argued for legitimacy in a manner that
reflected a multifaceted concept, not wholly dependent on belief. Legitimate political power
only exists under certain conditions. Power attained and implemented by rule of law, applicable
to laws accepted by a given community, and consented to by a populace ensures legitimacy
(Beetham, 1991). By approaching legitimacy in this way Beetham (1991) overcomes the major
flaw he portrayed in Weber’s work that centered on belief in different forms of legitimacy. With
this formulation of legitimacy, one is able to make a claim regarding the actual legitimacy of an
entity (Beetham, 1991). Notwithstanding this advancement, belief or consent remains a
necessary condition for political authority.
The relevant point from Beetham’s analysis, for the purpose of this research, is twofold.
Drifting from Weber’s explanation of different forms of legitimacy, Beetham explained that
legitimacy involved various levels or conditions. This advancement developed the concept in a
manner to address illegitimate power (Beetham, 1991). Judgments regarding illegitimate power
play a role in addressing the international community and evaluating political systems in
countries with questionable regimes. Such accounts of legitimate power are increasingly
important in the age of globalization because the international community has a better

9

opportunity to pressure less advanced governments to evolve toward advancements that favor
citizens. Secondly, portraying legitimacy in a multifaceted manner elucidates its complexity
(Beetham, 1991). This multifaceted manner, as discussed above, conceptualizes legitimacy in a
way that includes necessary and sufficient conditions that must be present for a governmental
entity to reach legitimacy. This idea, although anachronistic to Tyler’s (1990) penultimate
legitimacy research, allows insight for the issues that have plagued conceptualization of the term.
Moreover, the concept of legitimacy has developed over time and has led to various
definitions. Despite variations and difficulties with clarity, extant research has supported useful
conceptions of the term (Bottoms & Tankebe, 2012; Tyler, 2007a). In this context, researchers
have come to understand legitimacy as recognition by the public of a right to impose authority
(Bottoms & Tankebe, 2012; Tyler, 2006). Some claim it is a concept that deals with the present,
the right to impose authority now as opposed to the future (Meško & Tankebe, 2014), but this
temporal consideration only relates to legitimacy defined as a justified right. This discourse
adopts the former definition of legitimacy centering on authorization from citizens (Kelman &
Hamilton, 1989) and a collective perception that the activities of an entity are desirable
(Suchman, 1995). This definition ostensibly develops the involvement of measures capturing
expectations as a necessary part of authorizing, consenting, and desiring criminal justice service
provision or governmental structures in general. The culmination of understanding legitimacy
leads to its practical value in research. Legitimacy remains an important factor in attempting to
increase authority pertinent to governing (Tyler, 1990). Given that legitimacy relates to a
“power that is acknowledged as rightful (Beetham, 2013, p. 19),” beliefs from citizens about law
enforcement officials clearly manifest as important measures.

10

Likewise, the measurement of legitimacy has been a major discussion in policing often
leaving the state of academia in confusion. This consternation stems from inadequate and often
imprecise theoretical language (Cao, 2015) and the numerous measures used to capture
legitimacy (Gau, 2013; Jackson & Bradford, 2010; Reisig, Bratton, & Gertz, 2007; Tankebe,
2013). Tyler’s (1990) original conception of legitimacy included institutional trust and the
obligation to obey. Gau (2013), however, argued for the importance of moral authority in
regards to understanding legitimacy. Bottoms and Tankebe (2012), although offering conceptual
developments, focused on the dialogic nature of legitimacy that include different measures.
Some studies have compliance and cooperation as dependent measures of legitimacy (Meško &
Tankebe, 2014). Notwithstanding the issues found in conceptualization, there seems to be
unanimity in the fact that legitimacy is a profound concept composed of fundamental and
interrelated parts.
2.2 Trust as a Measure of Legitimacy
It is important to note that some researchers clearly identify trust and legitimacy as
separate outcome measures. Although trust and legitimacy are separate concepts, the current
research project does not attempt to focus on these separately. The aim of this research is to use
trust as a measure of legitimacy, not as a separate theoretical concept as done in some research
studies (see Hawdon, 2008). Many researchers have considered trust to be a critical part of
legitimacy (Bradford, Murphy, & Jackson, 2014; Jackson & Bradford, 2009; Meško & Tankebe,
2014; Tyler, 1990) which plays a necessary and sufficient role in the research at hand. Trust
also remains one of the most important principles in supporting democracy (Damme, Pauwels, &
Svensson, 2015) and allows profound insight into legitimacy. From this, one can gather why

11

trust remains a major focus of legitimacy research in advanced democracies and the crucial role
legitimacy plays in criminal justice activities.
Legitimacy is commonly understood by way of trust and confidence in police (Cao, 2015;
Hinds & Murphy, 2007; Murphy et al., 2014; Sifrer, Meško, & Bren, 2015; Tyler, 2011). Using
Cao’s (2015) analysis one must keep in mind the importance of clarifying theoretical concepts.
Despite the close connection of trust and confidence, these are undoubtedly separate concepts
(Cao, 2015). It is important to note that in the event researchers conflate these terms,
interchangeable usage leads to questionable results. Depending on the validity of the measures
used by such studies, findings related to trust and confidence should undergo scrutiny to prevent
conclusions stemming from theoretical errors. This type of conflation has inhibited criminal
justice research on legitimacy due to the range and imprecise measurement of terms (Cao, 2015).
Of chief importance, in consideration of criminal justice research, Tyler’s (1990) seminal
study on legitimacy was the impetus for earnest inquiry of the matter (Meško & Tankebe, 2014)
and led to the common adoption of trust as an important dependent measure. Utilizing a
telephone survey, Tyler (1990) gathered insight regarding legitimacy and especially the effect
procedural justice had in predicting the dependent measures used in his study (trust and the
obligation to obey). Trust, or support for the police, was captured using four indicators related to
respect, honesty, pride, and support (i.e. ‘On the whole Chicago police officers are honest)
(Tyler, 1990). This “general affective orientation toward authorities” (Tyler, 1990, p. 47)
represented trust. Researchers like Wolfe and colleagues (2016) explained this by drawing on
Tyler’s (1990; 2004) approach portraying legitimacy as the obligation to obey and trust in the
authority of law enforcement officials.

12

By returning to Tyler’s (1990) original conception one can test measures that have been
widely used. Following Wolfe and colleagues (2016) and others (Hinds & Murphy, 2007; Hohl
et al., 2010; Murphy & Cherney, 2011; Murphy et al., 2014; Tyler, 2004; Tyler, 2011), using
trust as a measure of legitimacy, despite several conceptual advancements, affords this study to
compare results with a bulk of research on the subject. In addition, the arguments of several
researchers, including Tyler (1990), Wolfe and colleagues (2016), and Murphy and Cherney
(2012) will be used in supporting trust as a necessary measure of legitimacy. It should be noted
again, however, that trust and legitimacy are indeed different concepts (Bottoms & Tankebe,
2012). In other words, since this project only uses trust as a measure of legitimacy, one should
not confuse the terms as synonymous, but should understand that the former plays a role in
explaining the latter. Legitimacy is a concept encompassing many components but this project
only captured one component (trust) given certain limitations explained in Chapter 4.
More specifically, Wolfe and colleagues (2016) measured legitimacy by following
Tyler’s original model involving the obligation to obey and trust. Trust was captured with a
single indicator asking respondents if they agree or disagree (1 = strongly disagree to 4 =
strongly agree) with “The police can be trusted to make decisions that are right for my
community (p. 263).” Jackson and colleagues (2012) also adopted this process of trust in police
by asking relevant questions regarding procedural justice and effectiveness. The specific
indicators were not outlined in their study. In studying outcome measures related to legitimacy,
Sargeant and associates (2014) measured trust in police with two indicators asking if respondents
trust police in the community and if they have confidence in the police. Likert scales were used
to capture responses ranging from 1 (strongly disagree) to 5 (strongly agree) (Sargeant, Murphy,
& Cherney, 2014). Murphy and Cherney (2012) also adopted the importance of trust in

13

capturing the concept of legitimacy. An argument consistent throughout the study mentioned the
use of trust and confidence in measuring legitimacy (Murphy & Cherney, 2012). Murphy and
Cherney (2012) used five indicators for legitimacy including “I have confidence in police” and “I
trust police (p. 189).” Likert scales were used again to capture responses ranging from 1
(strongly disagree) to 5 (strongly agree) (Murphy & Cherney, 2012).
The aforementioned researchers, among others, have adopted similar conceptual models
used by Tyler (1990) which confirms the framework of the present study. Several studies based
on Tyler’s contribution focused on trust as a measure of legitimacy (Bradford et al., 2009;
Jackson & Bradford, 2010; Murphy & Cherney, 2011; Sargeant et al., 2014; Murphy et al., 2014;
Tyler, 2004; 2005; 2011). Trust has also become a major focus of researchers as an outcome
measure in and of itself instead of a component of larger concept like legitimacy. The table
below provides a thorough account of research using trust and/or confidence as a dependent
variable. The studies found in the table have informed the conceptual model of studying the
relationship between procedural justice predictors and trust as a dependent measure, some
directly relate to the larger discussion on legitimacy.
Table 1. Summary of Studies Relevant to Legitimacy (Trust and Confidence)
Author(s)

IV

DV (Effects)

Measurement

Tyler, 2004

*procedural justice

Trust and
confidence,
Citizen
cooperation

-Statements of agreement
-“The police are generally honest;
I respect the police; I feel proud
of the police”

Trust and
confidence,
Citizen
cooperation

Institutional trust (8-item scale)
Motive-based trust (3-item scale)

-police performance

Tyler, 2005

-institutional trust, motive-based
trust, distributive justice, performance
measures, and demographics (for
cooperation)
-different models

14

Table 1. (cont’d)
Kääriäinen,
2007

-Individual level; social networks,
experiences of corruption, among
others
-Country level variables

Trust

-score of 0-10 how much you
personally trust police

Hinds &
Murphy,
2007

-predicting legitimacy; procedural
justice, distributive justice, police
performance, education level, age
-predicting satisfaction; legitimacy,
procedural justice, distributive justice

*Confidence
(legitimacy
overall)

Wu & Sun,
2009

Demographic variables, perceived
political power

Trust

Bradford et
al., 2009

-Police contact

Trust and
confidence

- Legitimacy; 4-item scale: ‘I
have confidence in the police,
police do their job well, I have
great respect for the police,
people should always follow the
directions of police officers even
if they go against what they think
is right’
Trust in police (single measure):
1 (don’t trust at all), 2 (don’t
really trust), 3 (trust to a degree),
and 4 (trust a lot)
Trust and confidence
-3 indices to account for specific
aspects of trust and confidence

-Fairness (procedural justice model)

Hohl et al.,
2010

-Experiment; leaflet drop

*Confidence in
local area policing

*police communication/contact

– 5 point scale ranging from 1 =
‘very poor’ to 5 = ‘excellent’; e.g.
how good a job do you think the
police are doing in their local area
- three components of trust and
confidence in the police
Confidence (PSA23; global
confidence or overall confidence)
Single indicators; two measures

- trust in police effectiveness
and fairness
- trust in police effectiveness, trust in
police fairness, and trust in police
engagement and shared values

Trust and
confidence

Murphy &
Cherney,
2011

-procedural justice; various predictors

Trust and
confidence,
Citizen
cooperation

-scale of five items; assess
feelings of respect and
confidence: e.g. ‘I have
confidence in the police’

Tyler, 2011

-Police performance is not found to be
that strong of a predictor

Trust and
confidence

-Several national surveys use the
‘trust and confidence’ index
-asks people to express
confidence in the police as an
institution or in the ability of the
police to protect citizens against
crime
Trust in police: none at all (1),
not very much (2), quite a lot (3),
and a great deal (4), binary
coding for analysis

Jackson &
Bradford,
2010

-PSA23
confidence

*procedural justice

Sun et al.,
2012

-social capital and political
participation

Trust

15

Table 1. (cont’d)
Sargeant et
al., 2013

-procedural justice, police
performance, demographics

Trust and
cooperation

Mazerolle et
al., 2013
(review of
literature)
Murphy et
al., 2014

-forms of legitimacy policing

Trust/confidence
in police

-various elements of procedural
justice; demographics, general
procedural justice, police effectiveness

Trust &
confidence,
cooperation

Cao, 2015

Conceptual study

Confidence, trust,
and satisfaction
with the police

Damme et
al., 2015

-path model on pgs. 21 & 24

-*moral alignment
used to represent
trust
-cooperation is the
main DV in the
study

-police-led legitimacy interventions

-trust in police effectiveness and
procedural fairness leading to moral
alignment leading on to obey the
police and cooperation

Trust: 2 items capturing
institutional trust (‘I trust the
police in my community’ and ‘I
have confidence in the police in
my community’); Likert
Trust police, confidence in police,
satisfied with the way police do
their job; perceived legitimacy
(many measures)
Likert scale, 1-5, higher meant
more trust and confidence; four
items; e.g. ‘I felt that the police
officer was trustworthy’)
Satisfaction is internal; trust and
confidence are external
-trust is viewed as synonymous
with social capital
-confidence part of political trust
and used to mean a generalized
support for police
‘The police generally have the
same sense of right and wrong as
I do’ (D21), ‘The police stand up
for values that are important to
people like me’ (D22) and ‘I
generally support how the police
usually act’ (D23). Respondents
were asked to score the extent to
which they agreed with these
statements on a range from 1,
agree strongly to 5, disagree
strongly

With this in mind, trust commonly means an expectation that relates to how an entity will
operate favorably in the future (Barbalet, 2009). This idea of expectations stems from values and
beliefs about an entity. Trust, however, is not relegated to future upshots. Trust in an entity
concerns expectations of current and future behavior (Jackson & Gau, 2015). Some researchers
have argued that this future orientation creates a separation between trust and legitimacy
(Tankebe, 2013). Therefore, the argument is that trust does not represent an effective measure of
legitimacy. Attempting to disassociate trust from legitimacy based on present and future actions

16

seems faulty. The fallacy here, one can argue, stems from overlooking legitimacy as an attribute
entrusted, in part, by an audience. The concept of legitimacy equally concerns the future because
subordinates, assuming similar agreeable processes, believe in the authority. These conceptual
components stem directly from definitions of legitimacy concerned with general perceptions and
authorization by citizens of allowing authority and order, which is why trust remains an
important component.
In other words, trust is a key aspect of legitimacy and permits authority to exercise
power. A belief in authority deals with the present and the future, unless something occurs to
change that belief. If one holds a belief regarding the existence of legitimacy now, then this also
implies future existence in the context of social order. For example, if one believes in the
electoral process to appoint legitimate (in part, trustworthy) officials, then after an election the
officials voted into office are viewed as legitimate throughout their terms unless proven
otherwise. Due to belief in the electoral process, implications regarding future expectations
related to the legitimacy of the officials arise. The belief in the legitimacy of the officials is not
limited to the present; these beliefs also concern the future and continue to exist unless
something corrupts these views. The belief in overall legitimacy operates in the same manner
and, more importantly, the right to exercise power comes from belief or trust. The distinction
mentioned earlier, between the belief to impose authority and the right to rule is important. The
temporal position of legitimacy may depend on the definition used by researchers. In the
abstract, legitimacy envelops profound concepts related to social order that apply to the present
and the future, unless the status quo shifts. This temporal issue may stem from applications of
trust. The legitimacy of political authority does not fade away because one individual no longer
adheres to this belief.

17

Put elegantly, “One can believe that something is the case, or one can simply conceive of
its being the case without assenting to it” (Gorman, 1993, p. 90). The belief in legitimate
political authority is collective, especially in a democratic society. Such belief rests on a tacit
agreement that can be traced back to Socratic philosophical traditions discussing obligations to
the state (Medina, 1990). This point is fundamental in consideration of social contract theory
and regards some form of consent from the populace. To reiterate, however, trust and legitimacy
are not synonymous terms (Kaina, 2008). The argument, based on previous research, establishes
trust as an important part of legitimacy and exists as a common measure.
Tyler (2005) attempted to explain two distinct forms of trust. The first, motive-based
trust, relates to assumptions about the motives and intentions of an entity (Tyler, 2005). The
second, institutional trust, centers on beliefs regarding the amount of probity and care an entity
possesses (Tyler, 2005). For the purposes of this study, institutional trust stands as the measure
of legitimacy. The reason for this adoption finds clarification in the measurement section. As a
note, the use of institutional trust overcomes the arguments above questioning the utility of trust
in measuring legitimacy. Institutional trust clearly identifies the importance belief plays in the
legitimacy of political institutions.
In like manner, the inquiry regarding the origins of trust remains pivotal. Additionally,
trust also theoretically relates to the antecedent discussion of belief in authority. In other words,
in framing a discussion of the origins of trust one can explain the importance of the term in
conceptualizing legitimacy and in terms of its theoretical utility in explaining the foundation of
society. In democratic society, trust exists as one of the most important principles (Damme,
Pauwels, & Svensson, 2015; Mishler & Rose, 2001). The question then arises concerning the
origins of trust in society.

18

Competing political theories regarding trust explain its origins (Mishler & Rose, 2001).
One can argue that cultural and institutional explanations of the origins of trust are not exactly
competing theories. The former highlights the importance of trust as a learned behavior in early
stages of life and the latter focuses on the consequences of political performance (Mishler &
Rose, 2001). It seems both of these theories can simultaneously explain the origin of trust, albeit
in different forms. To clarify, a cultural explanation of trust concerns an innate feature of
society. This feature concerns the aforementioned collective trust in the fabric of society; this is
the tacit agreement found in the tenets of social contract theory. The basic conditions of society
demand trust insofar as such belief will enhance quality of life (Tyler, 2007a). This collective
trust often influenced by individual occurrences, relates mainly to interactions with political
institutions (Mishler & Rose, 2001). An institutional explanation of trust stems from these
interactions. Beliefs about the care and honesty of public servants can change which either
increase, decrease, or deteriorate to a stage of complete disapproval from the populace.
Both cultural and institutional theorists agree that trust is endogenous, but the latter adds
the critical nature of performance in creating and enhancing trust (Mishler & Rose, 2001). This
agreement is enough to utilize both theories in explaining that the origin of trust stems from the
basic conditions of social life (Tyler, 2007a) which are learned at an early age, but enhanced or
diminished by interactions with political institutions (Mishler & Rose, 2001). Consequently, the
study of trust not only remains theoretically interesting due to its profound nature regarding the
relationship shared between governments and citizens, but it actually leads to practical benefits
(Bradford, Jackson, & Stanko, 2009; Sunshine & Tyler, 2003). Democracy begins to deteriorate
when trust is lacking. This is mostly due to the lack of legitimacy that results from little trust
(Damme et al., 2015).

19

Based on this section dealing with trust as a measure of legitimacy, a note of caution
seems necessary. This project focuses on the predictors relevant to influencing legitimacy. Trust
is the only output measure of legitimacy used by the present study. Therefore, it may at times,
be used interchangeably but does not exist as lexically equivalent to legitimacy. Trust, in the
proper context, only acts as a matter of linguistic convenience given its place in the current
project as a measure of legitimacy. Clearly defined, trust is a component of legitimacy and deals
with expectations of how an entity will operate favorably in the future (Barbalet, 2009).
Legitimacy is a broader concept encompassing trust that deals with the belief in legally
constructed rules and the right of political institutions to instill authority by way of the
established legal constructs (Weber, 1978). Legitimacy deals with the role of certain individuals
in the social order as opposed to trust, which deals with expectations of behaviors (Hawdon,
2008).

20

Chapter 3
Predicting Legitimacy
Procedural justice theory has become a popular approach commonly adopted by
academics focused on studying predictors of legitimacy (Bottoms & Tankebe, 2012). Rarer have
been studies utilizing procedural justice theory to explain the legitimacy of the criminal justice
apparatus, or at least a more thorough representation of the criminal justice system (Tyler,
2007b). That is to say that most studies solely focus on policing (Mazerolle et al., 2013; Murphy
et al., 2014; Murphy & Cherney, 2011; Sargeant et al., 2014; Tyler, 2011). Policing may well be
the most distinct feature of the criminal justice system due to force (Kääriäinen, 2007), but it
does not represent the other critical entities involved in the complex endeavor of maintaining
order (Skolnick, 1966). Criminal justice systems vary throughout the world and, at times,
influence how citizens trust the organizations that compose such systems (Kääriäinen, 2007).
There are, however, similarities found within the different criminal justice systems seen in
various countries, mainly to protect citizens and control crime (Kääriäinen, 2007). Regardless of
the degree of democracy in the nations included in this study, all have some form of policing and
court system. It is possible, and all the more likely in consideration of some nations in Africa,
that instability prevents conditions of normal political order. In other words, hybrid and
authoritarian regimes produce political institutions vastly different from full or flawed
democracies. Therein lies one of the most important parts of the study, which Chapter 4
discusses at a greater length.
Considering the popular approach of procedural justice predictors, a brief outline of
procedural justice theory is important. Procedural justice theory was born out of Tyler’s (1990)
work attempting to discover relevant predictors of legitimacy (Harkin, 2015). The model Tyler

21

(1990) adopted utilized two indicators for legitimacy, trust and the obligation to obey and
focused on independent variables that related to fairness in treatment by criminal justice officials.
His work created the foundation for explaining the role procedural justice plays in influencing
legitimacy. The theory outlines the potential for criminal justice practitioners to increase
legitimacy by operating in a fair manner (Tyler, 2004; 2011). Procedural justice essentially
means “…judgments about the fairness of the processes by which people are treated (Tankebe,
Reisig, & Wang, 2016).”
Put differently, procedural justice involves perceptions about fair procedures and the
treatment citizens receive from authority figures (Murphy et al., 2014). The practical
implications of procedural justice theory have led to many studies attempting to improve
different dimensions of public life (Bradford et al., 2009; Skogan, 2006; Sunshine & Tyler, 2003;
Tyler, 2004). For example, enhancing community relationships with police (Murphy & Cherney,
2011; Sunshine & Tyler, 2003; Tyler 2004, 2005), increasing legitimacy in courts (Canada &
Watson, 2013; Laxminarayan, 2012; Sprott & Greene, 2010), and changing the treatment of
prisoners (Beijersbergen, Dirkzwager, Eichelsheim, Van der Laan, & Nieuwbeerta, 2015;
Beijersbergen, Dirkzwager, & Nieuwbeerta, 2016). Procedural justice theory also implies an
important connection to the discourse regarding legitimacy. In connection to legitimacy, the
theory indicates that beliefs in authority “…are reasonable, honest, deduced autonomously and
generally smart and well informed (Harkin, 2015).”
Researchers have produced numerous models in attempting to capture the importance of
procedural justice in predicting legitimacy. In reviewing research, as established previously,
trust – although not synonymous with legitimacy – was repeatedly used as a dependent variable
(Kääriäinen, 2007; Sun, Hu, & Wu, 2012; Wu & Sun, 2009). The discussion now turns to

22

specific independent variables that researchers have used in testing a relationship with
legitimacy. For example, Tyler and Huo (2002) conducted research that strongly supported
procedural fairness as a main predictor of the reactions people have of police. Bottoms and
Tankebe (2012) also found that perceived procedural fairness played an important role in
determining compliance.
Further, researchers have used numerous predictors that build on the aforementioned
work by Tyler (1990) by implementing the logic of procedural justice. Born out of the multitude
of measures used by researchers outside of the procedural justice model, the endeavor of
attempting to study predictors of legitimacy is often confusing, yet some measures have
successfully supported consistent findings. Measures that have repeatedly shown promise are
analytical components of procedural justice (Bradford et al., 2009; Murphy et al., 2014; Tyler,
2004; 2011). The procedural justice model holds that citizens assess the fairness of criminal
justice organizations when providing public services (Tyler, 2004). For example, Murphy and
Cherney (2011) found that views of fairness affected police legitimacy. One indicator in the
scale for procedural justice stood for equality of treatment (Murphy & Cherney, 2011). These
consistent findings support that the foundation of trust comes from views about how entities
exercise authority (Tyler, 2004). A strong connection for studying the relationship between
procedural justice variables and indicators of legitimacy exists in extant research (Bradford et al.,
2009; Jackson & Bradford, 2010; Mazerolle et al., 2013; Murphy & Cherney, 2011; Murphy et
al., 2011; Sargeant et al., 2013, Tyler, 2004; 2011).
3.1 Social Capital
Similar to the concept of legitimacy, social capital tends to vary by definition (Sun et al.,
2012). Pierre Bourdieu is often cited as a major contributor to research on social capital (Ihlen,

23

2005). Based on Bourdieu’s original definition, social capital emerges as a resource in society
utilized for instrumental goals, often securing power in various forms (Ihlen, 2005). Definitions
that are more common explain social capital as connections involving norms and trust that allow
stakeholders to reach mutual goals (Putnam, 1995). In considering this definition, trust in
community members or facets of private societal structures, builds cohesiveness and allows for
greater success in reaching a highly functional society. In the same way, agreeable relationships
with criminal justice officials play a role in creating the effective means for the pursuit of
individual and collective goals. Social capital variables are relevant in the context of this
research given the effect social capital has on legitimacy. The effect, in general, reveals that
higher social trust in community members (a common measure of social capital; see Hawdon,
2008; Macdonald & Stokes, 2006) leads to favorable views of local police (Macdonald & Stokes,
2006). Social capital is also commonly measured by way of community membership and
activity (Sun et al., 2012) – which likely leads to higher levels of social trust. Social connections
that transcend individual perceptions often lead to favorable views of political structures
(Putnam, 2000) hence the connection of utilizing social capital as a potential influence in the
perceived legitimacy of police and courts.
Several relevant studies focused on community trust that relates to the larger discussion
of legitimacy. Research on forms of social trust supports the current project given the use of
trust as a measure of legitimacy. Trust comes in different forms and is not relegated to belief in
certain entities. That is, in dealing with social capital, one can learn whether trust in community
members leads to higher levels of perceived legitimacy; in the present study, this would mean
higher levels of trust. Social trust or bonding social capital, as Hawdon (2008) outlines,
theoretically relates to community memberships and various connections at the individual level.

24

Social capital has also gained attention in recent years producing favorable results in predicting
trust (Kääriäinen, 2007; Sun et al., 2012). In attempting to explore the variation of trust between
neighborhoods, Hawdon (2008) provided a theoretical argument for the importance of social
capital in determining legitimacy.
Kääriäinen (2007) found support for using social capital as a predictor for trust.
Kääriäinen (2007) noted that trust is “related to a larger debate on social capital” (p. 412). This
conclusion points toward a theoretical connection between the two concepts. In attempting to
explain the variation between racial and community variation in trust of police, Macdonald and
Stokes (2006) discovered that social capital in various community forms leads to higher levels of
trust in police. Finally, Sun and colleagues (2012) also utilized social capital in attempting to
predict trust. The researchers argued for the importance of this variable based on governmental
changes occurring in a “post-Mao era of economic reform…” (Sun et al., 2012, p. 90). The
study supported the strength of using social capital as a predictor for trust in police and
highlighted the importance of studying political context. The present study benefits from
research that has explained the inherent connections between social capital and components of
legitimacy.
3.2 Democratic Structure
Procedural justice research focuses on several indicators relative to fairness in
procedures. This approach, however, does not account for governmental constructs. Most
legitimacy research reflects the context of the American criminal justice system (Tyler, 2007).
This is important because the United States is a democratic nation, often ranking favorably in
democratic metrics (EIU, 2015). Further, procedural justice research mainly focuses on English
speaking nations with specific legal systems (Hough, Jackson, Bradford, Myhill, & Quinton,

25

2010). Prior research indicates that political institutions operating in democratic societies
receive trust (Booth & Richard, 1998; Zmerli & Newton, 2008). This is an important finding
relevant to police and court entities. Consequently, democratic structure remains important
because overall governmental styles might play a role in how citizens view criminal justice
organizations. In an exploratory manner, as a nuance to the procedural justice model, type of
government allows important insight into determinants of legitimacy.
Prior research has also used democratic structure to predict confidence in private
security guards (Nalla, Maxwell, & Mamayek, 2017). Nalla, Maxwell, and Mamayek (2017)
created an index of democracy composed of four indicators representing overall democracy
ranking, state legitimacy, public service indexes, and age of democracy. Future research on the
subject of legitimacy should consider measures like this. Despite this example, democratic
structure has not been a focus of research on the legitimacy of traditional criminal justice
organizations. Zmerli and Newton (2008) studied the effect of social trust on satisfaction with
democracy. In capturing satisfaction with democracy, democratic structure is not the main focus.
To this end, studies relevant to legitimacy, studying democratic structure as a predictor are either
non-existent or unknown which supports one of the major values offered by the present study.
Nalla and Mamayek (2013) offered a gateway to creating a proxy measure for democratic
structure. Using the EIU’s (2015) standard for ranking democracies around the world, Nalla and
Mamayek (2013) studied democratic police, accountability, and citizen oversight in Asia.
Democratic structure played a role in displaying how a country ranked from full democracy to
authoritarian regime (Nalla & Mamayek, 2013). This framework for comparing democratic
structure provides a basis for creating a proxy measure to gauge how political context may
influence perceived legitimacy. Given the amount of research on legitimacy and democracy, it

26

seems relevant to begin studying democratic structure as a predictor. The present study aims to
provide further insight regarding the importance of a macro variable like democratic structure.
3.3 Corruption
Corruption has also become a major focus in attempting to predict the legitimacy of
political institutions (Delhey & Newton, 2005; Kääriäinen 2007; Kubbe, 2013; Madan & Nalla,
2015). In addition, corruption, like other indicators related to procedural justice, exists as an
important concept relevant to government (Kääriäinen, 2007). Kääriäinen (2007) argued, “A
corrupt system of government is not able to fulfill the requirement of equality… (p. 413).”
Corruption has been defined as, “the misuse of public office for private gain” (Sandholtz &
Koetzle, 2000, p. 32). In many ways, corruption exists as procedural justice in the negative.
That is to say, corruption exists as the opposite of what one would expect in dealing with
trustworthy and fair servants of the state. To clarify, corruption is the opposite of fairness and
represents a form of unequal treatment. Corruption not only grasps an important indicator of
legitimacy, but also acts as a strong measure for the quality of a given government (Uslaner,
2008). The importance of this connection remains self-evident and will aid in explaining the
variation between perceptions regarding legitimacy and levels of democracy.
The value of legitimacy research is apparent, but increasing our understanding only
occurs by approaching research in new and interesting ways. Most of the aforementioned
articles attempt to understand what predictors affect indicators of legitimacy in the affirmative
(Hinds & Murphy, 2007). In other words, researchers tend to study how legitimacy is increased.
This is logical considering the practical implications of studies with positive results. In
attempting to increase legitimacy, researchers are usually trying to find ways to enhance the
relationship between policing entities and citizens (Sunshine & Tyler, 2003; Tyler, 2004).

27

Corruption, or a variable representing the potential for decreases in legitimacy, (Kääriäinen,
2007) are not studied as often.
Consequently, this remains a strong avenue for future research to enhance the study of
legitimacy. New ways to add to the discipline in regards to the legitimacy of criminal justice
organizations remains imperative. Different research angles provoke the advancement of
legitimacy research. Some researchers have already begun plunging into this method utilizing
perceived corruption as a pathway to studying relationships to legitimacy. For example,
Sunshine and Tyler (2003) argued that when negative perceptions of police exist it is more
difficult to achieve citizen cooperation. Tankebe et al. (2016) also mentioned the concept of
corruption utilizing the Global Corruption Barometer to compare the United States to Ghana.
Damme et al. (2015) mentioned corruption and historical antecedents that have decreased
legitimacy in Ghana, as well. Wu and Sun (2009) mentioned the issue of corruption in Chinese
policing, arguing that corruption leads to lower trust in police. Madan and Nalla (2015) reported
corruption as a useful predictor in police satisfaction. Kubbe (2013) provides a thorough account
of corruption, trust, and democracy, but the study mainly aims at themes relevant to political
science. In spite of these studies concerning corruption, the majority of extant research has
implemented other variables.
Kääriäinen (2007) provides a strong foundation for considering corruption as a predictor
of trust. This research centered on the quality of government in relation to corruption. The
degree of government corruption is an essential measure of the quality of government
(Kääriäinen, 2007). The main assumption resulting from this is that the general level of
government corruption will explain lack of public trust in police (Kääriäinen, 2007). Corruption
in general government influences the views citizens have of individual political institutions. In

28

addition to other procedural justice variables, corruption allows further understanding of how
trust varies in criminal justice organizations. Filling major gaps in research necessitates utilizing
a comprehensive approach. In other words, attempts to understand what decreases legitimacy in
policing and courts augments explanatory value instead of consistently focusing on traditional
predictors that increase legitimacy. This topic seems to reinvigorate a complicated search for
predictors that increase legitimacy and important components like trust in criminal justice
organizations.
3.4 Demographic Variables
Finally, demographic variables influence perceived legitimacy of criminal justice
organizations and officials. Age, gender, education, and employment have been associated with
legitimacy and outcome measures encompassed by legitimacy like trust and confidence (Damme
et al., 2015; Kääriäinen, 2007; Murphy, 2009; Sargeant et al., 2014; Tankebe et al., 2016; Tyler,
2005; Wolfe et al., 2016). In the context of legitimacy related research, Murphy (2009) found
that demographic variables play a small role in determining outcome measures, however,
included age, educational level, and gender as control variables. Murphy’s (2009) findings relate
to research using perceived legitimacy and procedural justice as predictors of satisfaction, but
remain relevant to the context at hand.
Wolfe and researchers (2016), however, found that gender was significant in studying
the effect of procedural justice on evaluations of police legitimacy. Yet age and education were
not significant (Wolfe et al., 2016). A study by Tyler (2005) focused on factors relevant to
cooperation and institutional support, found a similar finding for education, but age was
significant indicating that older respondents cooperate with police and support institutions at
higher levels compared to younger respondents. Kääriäinen (2007), in comparison to Murphy’s

29

(2009) study, found age, gender, education, and employment to significantly influence public
trust in police. Males tended to have less trust along with unemployed individuals (Kääriäinen,
2007). In addition, older and more educated individuals had more trust in police (Kääriäinen,
2007).
Other studies using demographic variables have confirmed the use of those in the present
study. Sargeant and colleagues (2014) used age, sex, and education in studying trust in police.
Their first model revealed the significance of all three, but after adding variables relevant to
procedural justice and police performance, only sex remained indicating that males have less
trust in police. Tankebe, Reisig, and Wang (2016) indicated that levels of cooperation and
compliance vary with individual characteristics. Their study used police legitimacy as an
independent variable in attempting to predict cooperation with police and compliance with the
law. The research by Tankebe and associates (2016), although different from the present study,
fits into the bulk of research on legitimacy given the arguments therein.
Tankebe, Reisig, and Wang (2016) only used three demographic variables, two of which
the present study adopted. Their study revealed that males tend to have lower levels of
cooperation and compliance and older respondents tended to have higher levels of both outcome
measures. Based on previous research, demographic variables are not a major focus of studies
on legitimacy and outcome measures relevant to themes related to legitimacy. Researchers,
however, often include these as control variables to find the importance sample characteristics
may play in this research. The previous examples studying legitimacy and relevant themes
support the use of the demographic variables used in the present study.
In summary, this review of research has revealed two major apertures. One stems from
the lack of research focused on two dependent variables related to separate entities of the

30

criminal justice apparatus. Police and courts are the focus of this study. The other stems from
research commonly focused on advanced democracies. This study attempts to elucidate findings
in the context of emerging democracies that has potential to produce insightful research on
legitimacy.

31

Chapter 4
The Present Study
4.1 Context: Africa
To begin, the aforementioned stressed the importance of the components of procedural
justice in democracy. To be clear, the notions of legitimacy and equality of treatment are
fundamental in democratic thinking. In addition to the components of procedural justice,
democracy also encompasses other important principles. These principles are important in
identifying what political systems qualify as democracies. The measurement of democracy
includes several different metrics including the amount of citizen participation in government,
the protection of citizen rights, maintenance of social order, quality of public services and ability
to operate devoid of severe political pressure, and equality in the implementation of the rule of
law (Nalla & Mamayek, 2013). The EIU Democracy Standard (2015) adequately captures the
aforementioned categories by highlighting pluralism, political culture, and civil liberties as
important practical themes.
In consideration of the metrics used to gauge democracy, countries are classified into four
categories, but only flawed democracies, hybrid regimes, and authoritarian regimes (EIU, 2015)
are considered. Table 2 shows what countries fall into each category included in the study. An
authoritarian regime, in a certain sense, stands as the antithesis of a full democracy. Researchers
usually focus on full democracies and are not the focus at hand. Flawed democracies are similar
to full democracies regarding civil liberties and electoral process, but have several weaknesses.
These weaknesses include considerations of governance, immature political culture, and lower
levels of political participation (Nalla & Mamayek, 2013). Hybrid regimes have significant

32

shortcomings hampering electoral process and major issues relating to flaws in political culture,
citizen participation, and overall government function (EIU, 2015; Nalla & Mamayek, 2013).

Table 2. Economist Democracy Rankings
Flawed democracy
1. Botswana (28)
2. Cabo Verde (32)
3. South Africa (37)
4. Ghana (53)
5. Tunisia (57)
6. Lesotho (64)
7. Namibia (72)
8. Zambia (73)
9. Senegal (75)

Hybrid regime
10. Benin (87)
11. Mali (88)
12. Tanzania (91)
13. Malawi (92)
14. Kenya (93)
15. Uganda (96)
16. Liberia (100)
17. Madagascar (103)
18. Burkina Faso (106)
19. Morocco (107)
20. Nigeria (108)
21. Mozambique (109)
22. Sierra Leone (111)

Authoritarian regime
23. Algeria (118)
24. Niger (121)
25. Cameroon (126)
26. Togo (130)
27. Côte d’Ivoire (132)
28. Egypt (134)
29. Guinea (136)
30. Swaziland (138)
31. Zimbabwe (141)
32. Burundi (150)
33. Sudan (151)

*EIU rankings in parentheses (out of 167 countries) (EIU, 2015)

Another major issue stems from widespread corruption in political institutions (Nalla &
Mamayek, 2013). Authoritarian regimes rank lowest for reasons stemming from oppressing
political pluralism, controlled elections; abuse of civil liberties; and ominous consequences for
political dissent (Nalla & Mamayek, 2013). Authoritarian regimes create the foundation for
unequal society and typically operate with little resources (Goldsmith, 2005). Yet these
governments remain strong enough to impose a police state on its citizens (Kääriäinen, 2007).
When confronting issues involving political institutions, criminal justice systems, and
government structures, some historical accounts are necessary to consider. Neglecting context is
a grave mistake in attempting to grasp why certain views of government arise or ways in which
particular political structures operate. Africa, in the context of democracy, remains one of the
most interesting continents in the world. Nations in Africa have a long history of colonization
and civil wars (Bates, 2010; Mattes & Bratton, 2007) that provoke intriguing studies. A

33

rudimentary background of democracy in Africa will allow for a better understanding and insight
into some of the perceptions and democracy rankings of the countries included in the study.
Furthermore, countries like the United States are important to mention here due to
globalization and the imposition of democracy on nations with other political systems. World
history shows that colonization, the World Wars, and the Cold War significantly influenced
global adoption of democratic principles (Bacevich. 2002; Rappaport, 1975). Africa is no
exception to outside forces, which remain imperative to consider when studying political
structures. The involvement of the international community in aiding nations with building
democratic systems deserves discussion in further detail, but falls outside the scope of this study.
To emphasize rudimentary concepts of historical accounts relevant to Africa, three points
are critical. The first accounts for the fact that most African countries have relatively young
democratic structures (Bates, 2010). In addition, colonization did not lead to the implementation
of democratic systems, even if the power-holders were European nations with varying degrees of
democracy during the colonization period (Uwizeyimana, 2012). This can account for many of
the issues African countries have in attempting to establish democratic criminal justice systems.
In short, these countries have not had the opportunity to evolve the foundation of democratic
principles into functional practices. Table 3 shows the years of independence. One can see the
majority of these countries are still relatively new regarding sovereignty.
Next, during the age of enhanced globalization, the neglect of Africa resulted from the
inability of foreign markets to establish lucrative enterprises and trade agreements on the
continent (Bacevich, 2002). A corollary of this results in capital flight due to many unstable
governments and corrupt practices (Nega & Schneider, 2012). The inability to control economic
foundations inhibits many African nations to overcome authoritarian regimes that are often

34

corrupt (Nega & Schneider, 2012). It seems paradoxical, but authoritarian regimes with weak
economies temporarily increase political power (Bates, 2010).
Table 3. Years of Independence
Flawed democracy
Botswana
Cabo Verde
Ghana
Lesotho
Namibia
Senegal
South Africa1
Tunisia
Zambia

Indepen.2
1966
1975
1957
1966
1990
1960
1961
1956
1964

Hybrid regime
Benin
Burkina Faso
Kenya
Liberia
Madagascar
Malawi
Mali
Morocco
Mozambique
Nigeria
Sierra Leone
Tanzania
Uganda

Indepen.
1960
1960
1963
1847
1960
1964
1960
1956
1975
1960
1961
1964
1962

Authoritarian regime
Algeria
Burundi
Cameroon
Cote D’ivoire
Egypt
Guinea
Niger
Sudan
Swaziland
Togo
Zimbabwe

Indepen.
1962
1962
1960
1960
1960
1958
1960
1956
1968
1960
1980

*

CIA – The World Factbook (https://www.cia.gov/library/publications/the-world-factbook/)
Majority rule occurred in 1994, but 1961 a republic was declared
2
Indepen. = Independence
1

Lastly, democracy varies on the continent, but one must bear in mind that claims of
democratic structure do not equate to views of liberal democracy seen in many western countries
(Uwizeyimana, 2012). Countries have firmly established cultures and views that invariably
influence the creation of political structure (Uwizeyimana, 2012). The conclusion based on these
premises is that Africa does not have a unique form of democracy, but varying degrees of liberal
democratic structures (Uwizeyimana, 2012). This line of thinking plays a role in addressing the
United States, globalization, the international community, and democracy building, but falls
outside the scope of this research.
4.2 Method and Data
This study will examine how procedural fairness, corruption, social capital, and degrees
of democracy influence citizen perception of legitimacy in police and courts in African nations.

35

Using an additional external democracy standard as a basis for comparison, the research aims at
gaining insight regarding the effect democratic structure has on legitimacy. Nalla and
Mamayek’s (2013) study of democratic policing in Asian countries employed the use of EIU
Democracy Standard. Building on this foundation of studying non-traditional governments
outside of the western world, the EIU standard will aid in making conclusions about less
advanced governments. This external standard will allow for a more accurate depiction of how
government structure affects legitimacy. In addition, the standard acts as the major contextual
basis for predicting perceptions related to legitimacy.
4.3 Dataset
Variables relevant to procedural justice, social capital, and type of government are tested
to see their influences on trust in police and courts using data from the Afrobarometer. Round 5
of the Afrobarometer was conducted between late-2011 and mid-2013 in 34 African countries
and finalized by 2015 (Afrobarometer Network, 2015). This particular study, however, only
uses 33 countries. Due to the high democracy ranking of Mauritius, it was discarded. The data
originally collected by Afrobarometer was used to measure “citizen attitudes on democracy and
governance, the economy, civil society, and other topics” (Afrobarometer Network, 2014). The
data for each country result from nationally representative random, multi-cluster samples (a
minimum of 1,200 respondents) (Afrobarometer Network, 2014). Mattes and Bratton (2007)
argued that the Afrobarometer has several flaws including the lack of a holistic representation of
Africa, under-sampling certain countries and ignoring unreformed authoritarian regimes.
Afrobarometer has arguably confronted the final issue mentioned above including 11
authoritarian regimes in Round 5. Despite some of the qualms, the Afrobarometer collects useful
data that can lead to interesting research on less advanced democracies.

36

The data was collected utilizing face-to-face interviews with a randomly selected sample
(Afrobarometer Network, 2014). Various sampling methods were used to ensure representative
results. The countries in this current project utilized multi-cluster stage sampling with
stratification when necessary (Afrobarometer Network, 2015). All interviewers attend a five-day
Afrobarometer training seminar to gain familiarity and confidence in the survey method
(Afrobarometer Network, 2014). Teams of four interviewers and one field supervisor travel to
the survey sample area within 48 hours of the training (Afrobarometer Network, 2014). On
average interviews usually took one hour and only conducted after consent of the participant
(Afrobarometer Network, 2014).
Funding for Afrobarometer research comes from various sources including Mo Ibrahim
Foundation, Swedish International Development Cooperation Agency, Department for
International Development, United States Agency for International Development, and World
Bank (Afrobarometer Network, 2016).
4.4 Research Model
The conceptual model used for this research is portrayed in Figure 1. The figure
explains the factors that potentially contribute to citizen perceptions of legitimacy. Once again,
the present study captures legitimacy by using a single indicator for each criminal justice
organization (trust in police and trust in courts).
The following hypotheses are tested in this study:
Main Hypothesis: Favorable citizen perceptions of procedurally just officials and increased
social involvement will be significantly associated with higher levels of trust in the police and
courts. Null: Favorable citizen perceptions of procedurally just officials and increased social

37

involvement will not be significantly associated with higher levels of trust in the police and
courts.
Secondary Hypothesis: It is anticipated that higher degrees of democracy (flawed) will lead to
greater trust in police and courts. Null: It is anticipated that higher degrees of democracy
(flawed) will not lead to greater trust in police and courts.
Figure 1. Conceptual Model – dependent, independent, and control variables

38

4.5 Measurement
Each variable included in the conceptual model is outlined in Table 4.

Table 4. Variables in Conceptual Model
Variables

Minimum

Maximum

Mean/SD

0
0

1
1

.78/.42
.86/.35

0

3

1.52/.99

1
0
0
0

3
3
3
3

2.08/.41
1.58/.88
1.25/.84
.40/.85

0
0
0

1
1
1

.45/.50
.35/.48
1.66/.47

1

3

2.02/.74

18
0
1
1

105
1
3
2

1

Dependent Variables
Trust in Police
Trust in Courts
Independent Variables
Procedural justice2
Equality of treatment
Corruption
Perceived government corruption3
Perceived police corruption4
Perceived corruption of courts4
Bribed police5
Social Capital
Member of religious group6
Member of voluntary association6
Activity in the community7
Democratic Structure
Type of government (Authoritarian ref.)8
Demographics
Age - Continuous
Employment (Employed)
Education9 (No formal schooling/Primary school ref.)
Gender (Male)

37.19/14.59
.33/.47
1.55/.63
1.50/.50

1

Value Labels: 0=Not at all, 1=Some/A lot
Value Labels: 0=Never, 1=Rarely, 2=Often, 3=Always
3
Value Labels: 1=None, 2=Some, 3=All (*=Not asked in Madagascar)
4
Value Labels: 0=None, 1=Some of them, 2=Most of them, 3=All of them
5
Value Labels: 0=Never, 1=Once or twice, 2=A few times, 3=Often
6
Value Labels: 0=Not a member, 1=Member/Leader
7
Value Labels: 0=No, 1=Yes
8
Value Labels: 1=Flawed, 2=Hybrid, 3=Authoritarian
9
Value Labels: 1=No formal schooling/Primary school completed, 2=Secondary school/Post-secondary
qualifications, 3=Some university/Post-graduate
2

4.5.1 Dependent Variables
In the past prior research has used several measures to capture legitimacy. Many include
the component of trust (among other indicators). Trust was the only component of legitimacy
available given the limits found in the dataset. Ideally, keeping in mind that this project adopted
39

Tyler’s (1990) original conception of legitimacy, trust and obligation to obey would have offered
a robust operationalization of the concept. Following Wu and Sun’s (2009) and Sun and
colleagues (2012) use of a single measure for trust, this study adopted a similar basic
operationalization. In addition, Wolfe and colleagues (2016) only used a single indicator for
trust, but offered a robust measure of legitimacy by adding indicators for the obligation to obey.
Jackson and Bradford (2010) also use single-item measures for the relationship between
confidence in police and indicators for trust. However, Jackson and Bradford (2010) tested
single-item measures for independent variables; trust was composed of three items. Other
researchers commonly use single measure dependent variables. For example, Kääriäinen (2007)
used a single item measure for citizen trust in police using a scale from 0 to 10 ranking levels of
trust in certain institutions.
For this study, the following question was used for police and courts: How much do you
trust each of the following, or haven’t you heard enough about them to say: The Police? The
question regarding Courts of law utilized the same format. The original questions were
measured on a 4-point Likert scale, ranging from Not at all to A lot. Following Sun and
colleagues (2012) use of a dummy variable, trust was dichotomized into 0 = Not at all and 1 =
Some/A lot. This binary coding allows binary regression analysis. This is necessary because
“the parallel lines assumption required for ordinal regression is violated” (Sun et al., 2012, p.
93). The weakness of using trust as a single indicator (Bradford et al., 2009) is openly
acknowledged; however remains the only method to test the influences of legitimacy using
Tyler’s (1990) original conception. The existing dataset did not offer more complex
representation of legitimacy as found in pertinent studies (see Bradford et al., 2009; Tankebe et
al., 2016; Wolfe et al., 2016).

40

4.5.2 Independent Variables
Equality of treatment, variables involving corruption, social capital, type of government,
and demographics are the main independent variables. Previous research on the subject
informed the use of these variables as predictors for legitimacy.
It has been argued that operationalization of procedural justice captures how fairness
affects decisions and the treatment of citizens (Reisig et al., 2007). Researchers commonly
measure procedural justice as a two-dimensional concept (Reisig et al., 2007). The dataset was
extremely limited in regards to adequate indicators for procedural justice. Murphy and Cherney
(2011) created a six-item scale including ‘Police treat everyone equally.’ This study only
adopted one indicator similar to their measure: In your opinion, how often, in this country: Are
people treated unequally under the law? This indicator is similar to the indicator in Murphy and
Cherney’s (2011) six-item scale and captures one dimension that was mentioned in many
measures of procedural justice. A 4-point Likert scale was used ranging from Never to Always.
Four variables captured corruption including three dealing with perceptions of police,
courts, and general government. Perceptions of corruption involve views related to fairness.
The final corruption variable dealt with bribing police to avoid punishment. Research also
supports the connection between corruption and predicting trust in political institutions
(Anderson & Tverdova, 2003). Variables capturing corruption dealt with respondents that
experienced an act of corruption and their perceptions of the existing corruption in police, courts,
and the government in general. For perceptions of corruption in police and courts (judges and
magistrates), the following question was used: How many of the following people do you think
are involved in corruption, or haven’t you heard enough about them to say: Police? The

41

question regarding judges and magistrates utilized the same format. A 4-point Likert scale was
used ranging from None to All of them.
A 5-item scale was created to capture government corruption including office of the
President/Prime Minister, government officials, local government councilors, tax officials, and
members of parliament. The above question was also used for these offices. A 3-point Likert
scale was used for general government corruption ranging from None to All of them.
A single indicator was used to capture acts of corruption: In the past year, how often, if
ever, have you had to pay a bribe, give a gift, or do a favor to government officials in order to:
Avoid a problem with the police (like passing a checkpoint or avoiding a fine or arrest)? A 4point Likert scale was used ranging from Never to Often. The aforementioned measures
encompass different components of concepts relevant to procedural justice theory.
Sun, Hu, and Wu (2012) tested social capital as a predictor of citizen trust in police.
Kääriäinen (2007) also used social capital, mainly centered on social networks, as a predictor of
trust. Following Sun and colleagues (2012) and Kääriäinen (2007) ideas concerning social
capital, scales were made for two social capital variables, member of a community group and
activity in the community. The 2-item scale for member of a community group included this
question: Let’s turn to your role in the community. Now I am going to read out a list of groups
that people join or attend. For each one, could you tell me whether you are an official leader, an
active member, an inactive member, or not a member: A religious group that meets outside of
regular worship services? The second question used the same format, but asked about ‘Some
other voluntary association or community group.’ The original Likert scale included four
categories, but was dichotomized into Not a member and Member/Leader. This scale returned

42

an unfavorable Cronbach’s alpha of .496. Therefore, each of the variables that were used to
make the scale were utilized in the final model as separate variables.
The 2 item scale for activity in the community included this question: Here is a list of
actions that people sometimes take as citizens. For each of these, please tell me whether you,
personally, have done any of these things during the past year. If not, would you do this if you
had the chance: Attended a community meeting? The second question used the same format, but
asked ‘Got together with others to raise an issue?’ A dichotomized variable was also created
including two categories, No and Yes.
Following Nalla and Mamayek’s (2013) use of the EIU Democracy Standard for
democratic policing, this study grouped the 33 countries into three categories: flawed
democracies, hybrid regimes, or authoritarian regimes. Type of government was a proxy
measure for the EIU Democracy Standard. A variable was created for this proxy measure
combining the countries into a 3-point Likert scale ranging from Flawed to Authoritarian.
Demographic variables included in the model are gender, age, education, and
employment. Age was left as a continuous variable, while employment was dichotomized
between Not employed and Employed. There was substantive reason for creating dummy
variables for education given the information included in the original categories. The categories
were condensed as follows: 1=No formal schooling/Primary school completed, 2=Secondary
school/Post-secondary qualifications, 3=Some university/Post-graduate. This research model is
supported by existing studies concerning procedural justice (Bradford et al., 2009; Murphy et al.,
2014; Tyler, 2004; 2011), social capital (Kääriäinen, 2007; Sun et al., 2012), corruption (Delhey
& Newton, 2005; Kääriäinen 2007; Kubbe, 2013; Madan & Nalla, 2015), and important

43

discussions regarding democratic government and trust (Booth & Richard, 1998; Hough,
Jackson, Bradford, Myhill, & Quinton, 2010; Tyler, 2007; Zmerli & Newton, 2008).
Table 5. Summary Table of Survey Questions for Afrobarometer
Trust (0=Not at all, 1=Some/A lot)
How much do you trust each of the following, or haven’t you heard enough about them to say:
The Police?
How much do you trust each of the following, or haven’t you heard enough about them to say:
Courts of Law?
Equality of Treatment (0=Never, 1=Rarely, 2=Often, 3=Always)
In your opinion, how often, in this country: Are people treated unequally under the law?
Corruption (1=None, 2=Some, 3=All); (0=None, 1=Some of them, 2=Most of them, 3=All of them)
How many of the following people do you think are involved in corruption, or haven’t you heard
enough about them to say: Office of the President/Prime Minister, Members of Parliament,
Government officials, Local government councilors, Tax officials (General government)
How many of the following people do you think are involved in corruption, or haven’t you heard
enough about them to say: The Police?
How many of the following people do you think are involved in corruption, or haven’t you heard
enough about them to say: Judges and Magistrates?
In the past year, how often, if ever, have you had to pay a bribe, give a gift, or do a favor to
government officials in order to: Avoid a problem with the police (like passing a checkpoint or
avoiding a fine or arrest)?
*(0=Never, 1=Once or twice, 2=A few times, 3=Often)
Social Capital (0=Not a member, 1=Member/Leader); (0=No, 1=Yes)
Let’s turn to your role in the community. Now I am going to read out a list of groups that people
join or attend. For each one, could you tell me whether you are an official leader, an active member,
and inactive member or not a member: A religious group that meets outside of regular worship
services?
Let’s turn to your role in the community. Now I am going to read out a list of groups that people
join or attend. For each one, could you tell me whether you are an official leader, an active member,
and inactive member or not a member: Some other voluntary association or community group?
Here is a list of actions that people sometimes take as citizens. For each of these, please tell me
whether you, personally, have done any of these things during the past year. If not, would you do this
if you had the chance: Attended a community meeting?
Here is a list of actions that people sometimes take as citizens. For each of these, please tell me
whether you, personally, have done any of these things during the past year. If not, would you do this
if you had the chance: Got together with others to raise an issue?

Given the limits of the dataset, courts were the only additional criminal justice entity that
could be used to represent a more thorough representation of the criminal justice apparatus.

44

Admittedly, secondary data often limits the ability of researchers to bridge the gap between
conceptualization and measurement. Two issues arise from this division. One is that many
measures used in this study would benefit by additional indicators. Next, this easily complicates
studying a more thorough representation of the criminal justice apparatus when only certain
political institutions were represented by the data. Nevertheless, this study aims at providing an
insight into less advanced democracies in the context of criminal justice service provision.
A critical point also rests on explaining the limitations of the measure for legitimacy. To
reiterate, an ideal model testing legitimacy in criminal justice organizations would include robust
measures. Utilizing Tyler’s (1990) conception, the current project would benefit by having
variables that represent obligation to obey, the other crucial measure used to capture legitimacy.
Single variables related to trust in police and trust in courts were the only available measures that
fit within the context of legitimacy. The present study, however, still offers a valuable
framework for discovering new insights regarding the legitimacy of criminal justice
organizations.
Based on the variables, the statistical approach will be informed by Sun and colleagues
(2012) study of trust. The measurement of trust in the dataset, like Sun and colleagues (2012),
requires binary regression analysis. The dependent variables can be dichotomized into responses
of “Not at all” and “Some/A lot.” This statistical analysis will elucidate whether or not the null
form of the hypotheses can be rejected. It also allows insight into how the independent variables
influence trust in police and courts. It is expected that favorable views of officials based on
procedurally just perceptions, views of lower levels of corruption and less experience with
corruption, increased involvement in social capital, and higher levels of democracy will lead to

45

higher levels of trust in police and courts. More specifically, flawed democracies should have
the highest level of citizen trust as opposed to authoritarian regimes.

46

Chapter 5
Analysis
The purpose of the present study was to elucidate the determinants of legitimacy in
nations with varying democratic structures. Notwithstanding democratic structure as a basis for
comparison, the focus of the study is to understand what contributes to the legitimacy of criminal
justice organizations. Based on the dataset, to discover the relationship between predictors and
legitimacy, binary coding was used for the dependent variables of trust in police and trust in
courts of law. The original scale used by the Afrobarometer only allowed for a Chi square test
since the dependent variables were categorical. Sun and colleagues (2012) had a similar
problem, but dichotomized trust as to use binary regression analysis. Thus, the present study
adopted this approach to find how various predictors influence legitimacy.
Table 6 shows the category distribution for cases and descriptive statistics for
predictors. This distribution of cases allows insight regarding how respondents answered
survey questions. This information also ensures that each category had enough cases to
produce relevant statistics.

47

Table 6. Descriptive Statistics for Variable Categories
Variable
Trust in Police
0 = Not at all
1 = Some/A lot
Trust in Courts
0 = Not at all
1 = Some/A lot
Equality of Treatment
0 = Never
1 = Rarely
2 = Often
3 = Always
Perceived Government Corruption
1 = None
2 = Some
3 = All
Perceived Police Corruption
0 = None
1 = Some of them
2 = Most of them
3 = All of them
Perceived Corruption of Courts
0 = None
1 = Some of them
2 = Most of them
3 = All of them
Bribed Police
0 = Never
1 = Once or Twice
2 = A Few Times
3 = Often
Member of Religious Group
0 = Not a member
1 = Member/Leader
Member of Voluntary Association
0 = Not a member
1 = Member/Leader
Activity in the Community
0 = No
1 = Yes
Type of Government
1 = Flawed
2 = Hybrid
3 = Authoritarian

Mean/SD
.78/.42

N

%

11,257
39,228

22.3
77.7

7,075
42,014

14.4
85.6

9,461
13,062
18,419
8,446

19.2
26.4
37.3
17.1

1,652
28,698
4,542

4.7
82.2
13.0

4,380
19,066
15,562
8,004

9.3
40.6
33.1
17.0

7,207
22,853
10,050
4,098

16.3
51.7
22.7
9.3

26,240
3,245
2,191
1,996

77.9
9.6
6.5
5.9

28,141
23,064

55.0
45.0

33,268
17,802

65.1
34.9

17,252
33,829

33.8
66.2

13,204
22,791
14,392

26.2
45.2
28.6

.86/.35

1.52/.99

2.08/.41

1.58/.88

1.25/.84

.40/.85

.45/.50

.35/.48

1.66/.47

2.02/.74

48

Table 7 portrays information for bivariate logistic analysis for independent variables in
the model.
Table 7. Bivariate Logistic Regression Summary for Selected Variables
Police
B

SE

Courts
ExpB

N. R^2

B

SE

ExpB

N. R^2

Procedural Justice
Equality of treatment

-.357***

.012

.700

.031

-.420***

.014

.657

.035

Perceived govt. corruption

-1.047***

.030

.351

.051

-1.188***

.035

.305

.058

Perceived police corruption

-.881***

.014

.415

.137

-.532***

.016

.587

.046

Perceived corruption of courts

-.558***

.014

.572

.058

-.875***

.016

.417

.117

Bribed police

-.393***

.014

.675

.036

-.365***

.016

.694

.027

-.019

.022

.981

.000

.140***

.026

1.150

.001

-.078***

.022

.925

.000

-.038

.027

.963

.000

.171***

.022

1.186

.002

.217***

.027

1.242

.002

Flawed

.255***

.030

1.105

.004

.255***

.035

1.291

.002

Hybrid

-.195***

.026

.823

.004

.113***

.030

1.120

.002

Corruption

Social capital
Member of religious group
Member of voluntary
association
Activity in the community
Democratic structure

Authoritarian (ref.)
*p < .05. **p < .01. ***p< .001.
N. R^2 = Negelkerke R^2

Demographic variables were not included in the bivariate analysis considering these
were not the main focus of the study. Bivariate logistic regression allows one to gather the
significance of an independent variable without the interaction effects of multivariate analysis.
The results are promising revealing a statistical significance at the .001 for most variables. To
avoid redundancy, a description of how these variables effect perceptions of legitimacy will be
included in the multivariate analysis portrayed in Tables 8 and 9.
Table 8 summarizes the results for respondents from the Afrobarometer dataset. This
statistical analysis reveals many important findings relevant to the perceived legitimacy of
criminal justice organizations. Before interpreting results several important issues emerge
regarding the model used for analysis. Binary logistic regression requires that the dependent

49

variable is dichotomous, hence the changes mentioned in the methods section. Logistic
regression does not allow multicollinearity diagnostics. Therefore, independent variables were
tested with a regular regression ignoring the overall results (coefficients and significance).
Since multicollinearity concerns relationships among the predictors, the procedure was
sufficient in determining that the model did not include any multicollinearity issues aiding in
finalizing the logistic regression model.
Table 8. Logistic Regression Summary for Trust in Police (N = 21,640)
Model 1
B

Model 2

Model 3

SE

ExpB

B

SE

ExpB

B

SE

ExpB

Background information
Age

.005***

.001

1.005

.003*

.001

1.003

.002

.001

1.002

Employed

-.130***

.033

.878

-.070*

.035

.932

-.037

.035

.963

Mid-level Education

-.307***

.034

.736

-.251***

.037

.780

-.276***

.037

.759

Highest Education

-.314***

.059

.730

-.214***

.063

.808

-.268***

.063

.765

.007

.032

1.007

-.069

.034

.933

-.044

.035

.957

-.240***

.018

.787

-.233***

.018

.792

Perceived govt. corruption

-.222***

.046

.801

-.213***

.046

.809

Perceived police corruption

-.710***

.023

.492

-.692***

.023

.501

Perceived corruption of courts

-.113***

.023

.893

-.120***

.024

.887

Bribed police

-.187***

.018

.829

-.191***

.018

.826

Member of religious group

.014

.036

1.014

Member of voluntary association

-.055

.038

.946

Activity in the community

.286***

.040

1.331

Flawed

-.029

.049

.972

Hybrid

-.291***

.041

.747

Lowest Education (ref.)
Male
Procedural Justice
Equality of treatment
Corruption

Social capital

Democratic structure

Authoritarian (ref.)
Nagelkerke R^2

.011

.171

.178

*p < .05. **p < .01. ***p< .001.

In addition, some categorical variables were treated as dummy variables to allow for
simpler interpretation and to reveal a clearer significance of certain coefficients. The scale for

50

general government corruption returned a Cronbach’s alpha score of .846 indicating that the
variables have an acceptable level of internal consistency. The two indicators for member to a
community group revealed a Cronbach’s alpha score of .496 indicating that the variables have
a low level of internal consistency. Consequently, the two items were used independently as
social capital variables. Separate models were tested, utilizing only one of the indicators from
the member to a community group scale to see if the model could be enhanced. The final
model included both indicators because excluding such variables can lead to higher levels of
bias. Excluding variables in these situations often leads to model misspecification (Britt &
Weisburd, 2014). The Nagelkerke R2 was utilized to gain an understanding of the model’s
precision. It has been noted, however, that the use of Nagelkerke R 2 has not reached a
scientific consensus (Britt & Weisburd, 2014).
Beginning with Model 1, testing the effects on trust in police, age, employment status,
and education were statistically significant at the .001 level. Older respondents are more likely
to have higher levels of trust in police. Similarly, employed individuals are likely to have less
trust in police than individuals that are unemployed. Further, the coefficients indicate that
those with higher levels of education are more likely to have lower levels of trust in police
compared to those with less education. Put differently, one unit increase in age increases the
odds of trust in police by a factor of 1.005. Similarly, employment decreases the odds of trust
in police by a factor of .878 compared to unemployed individuals. On the other hand, one unit
increase in education decreases the odds of trust in police by a factor of .736 for individuals
with secondary school through post-secondary qualifications compared to those with less
education.

51

Furthermore, one unit increase in education decreases the odds of trust in police by a
factor of .730 for individuals with some university through post-graduate education compared
to those with no schooling and completion of primary school. More education tends to lead to
higher likelihoods of having lower levels of trust in police. The aforementioned statements
deal with the odds of outcomes, which is the principle attribute of logistic regression, and are
vital when portraying findings. The odds ratios are a measure of association that determine the
odds that an effect will happen due to an independent variable chosen for the model (Britt &
Weisburd, 2014).
In Model 2, age, employment status, and education were statistically significant, at
various levels. Age and employment status decreased in statistical significance from the .001
level to the .05 level, respectively.

Age and employment status disappeared in Model 3

indicating that an intervening explanation exists (Sun & Wu, 2015). Given that these variables
were no longer statistically significant when adding social capital variables and democratic
structure, one may suspect the intervening relationship may exist between these variables. The
dummy variables for education were statistically significant at the .001 level. More
importantly, Model 2 also revealed that stronger perceptions of unequal treatment, higher
levels of perceived corruption in general government, higher levels of perceived corruption in
police and courts, and persons that bribed police are more likely to have lower levels of trust in
police. These findings remain consistent with research regarding procedural justice variables
and legitimacy. For example, an increase of one unit in perception of unequal treatment
decreases the odds of trust in police by a factor of .787. The corruption variables decrease the
odds of trust in police by various factors ranging from .492 to .893.

52

Of greater importance, Model 3, which was the final model, included five new variables
that attempted to capture a more macro perspective of society. With the disappearance of age
and employment status, the dummy variables for education were the only demographic
variables to remain significant at the .001 level. The education variables indicate that
individuals with higher levels of education are likely to have lower levels of trust in police.
Perception of unequal treatment and the corruption variables all remained significant at the
.001 level. Once again, stronger perceptions of unequal treatment, higher levels of perceived
corruption in general government, higher levels of perceived corruption in police and courts,
and persons that bribed police are more likely to have lower levels of trust in police. Activity
in the community returned a favorable Cronbach’s alpha score of .744. Therefore, it remained
in the model as one variable. It was also found to be statistically significant at the .001 level.
This indicates that one unit increase in activity in the community increases the odds of higher
levels of trust in police by a factor of 1.331. The fact that this variable was significant will be
the impetus for discussion in the following chapter.
Finally, type of government, a proxy measure for democratic structure was dummied
allowing a comparison of flawed and hybrid regimes to the reference category of authoritarian
regimes. The variable for countries categorized as flawed democracies was not significant. On
the other hand, countries categorized as hybrid regimes tend to decrease the likelihood of trust
in police. Governments considered hybrid regimes will decrease the odds of trust in police by
a factor of .747 for compared to authoritarian regimes. Model 3 had a Nagelkerke R2 of .178
indicating that about 18% of the variation in the dependent variable is explained by the
independent variables in the model. The increase of the Nagelkerke R 2 from the first to the
second model portrays the importance of variables related to procedural justice in explaining

53

legitimacy as measured by trust. Note that the warning above remains relevant when
considering the use of the Nagelkerke R 2.
For trust in courts of law, Model 1 revealed very different results in comparison to trust
in police. Age was not statistically significant, but gender was at the .05 level different from
the previous logistic regression on trust in police. Men are more likely to have higher levels of
trust in courts of law compared to women. The significance of gender, however, disappears in
Model 2. Employment status was also significant at the .05 level and the dummy for education
recording a significance for both variables at the .001 level. An interesting point of
comparison shows that employed individuals are likely to have more trust in courts of law, but
less trust in police. In Model 1, education, similar to its effect on trust in police, remained an
important factor in decreasing the likelihood of having higher levels of trust in courts of law;
this remained true for the final model. Employment status and education were also significant
in Model 2. Moreover, Model 2 portrayed how stronger perceptions of inequality and
variables related to views and experience of corruption, decease the likelihood of having
higher levels of trust in courts of law.
For example, one unit increase in experience with bribing police deceases the odds of
having higher levels of trust in courts of law by a factor of .861. Perceived corruption of police
was significant at the .01 level, but the other corruption variables along with perception of
unequal treatment were statistically significant at the .001 level. Perceived corruption of police
was the only difference in comparison to the previous model. Like its counterpart, the model for
courts of law reveals the importance of variables related to procedural justice.
Model 3 added the social capital variables along with type of government. This addition
increased the Nagelkerke R2 from .140 to .142. Even lacking scientific consensus on the

54

usefulness of the Nagelkerke R 2, one can gather that other variables not included in the model
account for much of the variability in the dependent variable.
Table 9. Logistic Regression Summary for Trust in Courts of Law (N = 21,459)
Model 1
B

Model 2

SE

ExpB

B

SE

Model 3
ExpB

B

SE

ExpB

Background information
Age

.000

.001

1.000

Employed

-.001

.002

.999

-.002

.002

.998

.097*

.041

1.102

Mid-level Education

-.184***

.042

.832

.108*

.043

1.114

-.160***

.044

.852

.100*

.043

1.105

-.158***

.045

Highest Education

.854

-.309***

.070

.734

-.173*

.074

.841

-.150*

.075

.861

.085*

.039

1.088

.010

.041

1.010

.023

.042

1.024

-.263***

.021

.769

-.261***

.022

.771

Perceived govt. corruption

-.358***

Perceived police corruption

-.065*

.051

.699

-.345***

.052

.708

.028

.937

-.069*

.028

Perceived corruption of courts

.933

Bribed police

-.703***

.028

.495

-.705***

.028

.494

-.149***

.020

.861

-.152***

.021

.859

Member of religious group

.046

.044

1.047

Member of voluntary association

-.026

.046

.975

Activity in the community

.205***

.048

1.228

Flawed

.134*

.058

1.144

Hybrid

.099*

.047

1.104

Lowest Education (ref.)
Male
Procedural Justice
Equality of treatment
Corruption

Social capital

Democratic structure

Authoritarian (ref.)
Nagelkerke R^2

.003

.140

.142

*p < .05. **p < .01. ***p< .001.

That said activity in the community was statistically significant at the .001 level while the
dummy variables for type of government were significant at the .01 level. Given the positive
coefficients, both variables increase the odds of having higher levels of trust in courts of law.
This is very different from the previous model on trust in police. This will be an important
point of contrast that requires attention in the following chapter. Individuals that are more
active in the community, by way of informal interactions, are more likely to have higher levels

55

of trust in courts of law. For trust in courts of law, citizens in countries categorized as flawed
democracies will most likely have higher levels of trust. Countries categorized as flawed
democracies increase the odds of trust in courts of law by a factor of 1.144 compared to
authoritarian regimes. Similarly, hybrid regimes increase the odds of trust in courts of law by
a factor of 1.104 compared to authoritarian regimes. Flawed democracies clearly increase the
log of odds in trust in courts of law by a greater factor than hybrid regimes.
In addition to the previously mentioned logistic regression models, a filtering process
allowed for the selection of cases from each democratic structure. For example, a logistic
regression with the same predictors in the previous models was run for respondents from
countries with flawed democracies. The same was done for respondents from hybrid regimes
and authoritarian regimes. This allowed for a comparative analysis between the predictors
based on respondents from each democratic structure. Comparative models were run for each
dependent variable, trust in police and trust in courts of law. Tables 10 and 11 portray the
logistic regression results using democratic structure as the basis for comparison. One can see
the disappearance of the significance of education within the flawed democracy model for
police compared to the model shown in Table 8. The number of cases were still numerous
enough to reveal any statistically significant variables. It seems reasonable to suspect that
another explanation exists for the disappearance of this variable.
Similar to the previous overall logistic regression, respondents in flawed democracies are
more likely to have lower levels of trust in police when they perceive higher levels of inequality,
corruption of police, and experiences with bribing police. These variables were significant at the
.001 level. General government corruption and corruption of courts were not significant in this
model. Activity in the community was significant at the .01 level. This indicates that persons

56

more active in the community are more likely to have higher levels of trust in police.
Employment status was significant at the .05 level indicating that employed persons are more
likely to have lower levels of trust in police.
Table 10. Logistic Regression Summary for Democratic Structure: Police
Flawed (n= 5,631)

Hybrid (n= 9,789)

Authoritarian (n= 6,220)

B

SE

ExpB

B

SE

ExpB

B

SE

ExpB

Age

.003

.002

1.003

.003

.002

1.003

.001

.003

1.001

Employed

-.153*

.073

.858

-.010

.049

.990

.001

.075

1.001

Mid-level Education

.021

.077

1.021

-.400***

.049

.670

-.257***

.076

.774

Highest Education

.221

.145

1.248

-.460***

.095

.631

-.245*

.112

.783

-.056

.072

.946

-.059

.049

.945

-.028

.068

.973

-.163***

.035

.850

-.205***

.025

.815

-.371***

.039

.690

Perceived govt. corruption

-.093

.108

.912

-.258***

.064

.773

-.179*

.085

.836

Perceived police corruption

-.637***

.048

.529

-.684***

.032

.504

-.732***

.046

.481

Perceived corruption of courts

.015

.048

1.015

-.113***

.033

.893

-.246***

.046

.782

Bribed police

-.318***

.047

.728

-.147***

.025

.863

-.199***

.032

.819

Member of religious group

.085

.075

1.089

-.013

.053

.987

.004

.070

1.004

Member of voluntary association

.090

.088

1.094

-.104

.054

.901

-.073

.073

.930

Activity in the community

.220**

.075

1.247

.341***

.059

1.406

.261***

.080

1.298

Background information

Lowest Education (ref.)
Male
Procedural Justice
Equality of treatment
Corruption

Social capital

Nagelkerke R^2

.11

.165

.226

*p < .05. **p < .01. ***p< .001.

In comparison, the same effect did not exist for authoritarian or hybrid regimes regarding
employment status. Education, in the models for hybrid and authoritarian regimes, was
significant at the .001 level and revealed that higher levels of education most likely lead to lower
levels of trust in police compared to persons with less education. The variables akin to
procedural justice played a stronger role in accounting for the likelihood of higher levels of trust
in police in countries considered hybrid and authoritarian regimes. The Nagelkerke R2 steadily

57

increased from flawed to authoritarian regimes. It increased from .11 to .226 indicating that the
independent variables are more effective at explaining trust in police for authoritarian regimes.
The model for courts of law returned similar results with several nuances. Demographic
variables did not play a role in flawed democracies and authoritarian regimes, but education was
important for hybrid regimes insofar as individuals with secondary school and post-secondary
qualifications are more likely to have lower levels of trust in courts of law compared to those
with less education. One would think that corruption in police would influence trust in courts of
law merely because both entities operate within the same larger system.
Table 11. Logistic Regression Summary for Democratic Structure: Courts of Law
Flawed (n= 5,560)

Hybrid (n= 9,714)

Authoritarian (n= 6,185)

B

SE

ExpB

B

SE

ExpB

B

SE

ExpB

Age

.000

.003

1.000

-.002

.002

.998

-.002

.003

.998

Employed

-.073

.088

.929

.134

.061

1.143

.149

.086

1.161

Mid-level Education

.108

.093

1.114

-.291***

.064

.748

-.139

.086

.870

Highest Education

.066

.167

1.068

-.202

.118

.817

-.104

.126

.901

.039

.087

1.040

-.017

.061

.983

.042

.078

1.042

-.142***

.042

.867

-.239***

.031

.787

-.405***

.092

.667

Perceived govt. corruption

-.214

.125

.807

-.467***

.073

.627

-.237**

.092

.789

Perceived police corruption

-.032

.060

.968

-.024

.040

.976

-.147**

.051

.863

Perceived corruption of courts

-.469***

.063

.626

-.703***

.041

.495

-.886***

.052

.412

Bribed police

-.209***

.055

.811

-.108***

.029

.897

-.183***

.036

.833

Member of religious group

.158

.092

1.172

-.130*

.066

.878

.234**

.079

1.264

Member of voluntary association

.047

.107

1.048

.015

.067

1.016

-.104

.082

.902

Activity in the community

.256**

.090

1.292

.189*

.074

1.208

.203*

.091

1.225

Background information

Lowest Education (ref.)
Male
Procedural Justice
Equality of treatment
Corruption

Social capital

Nagelkerke R^2

.063

.133

.228

*p < .05. **p < .01. ***p< .001.

This, however, was untrue for flawed and hybrid regimes. Respondents in authoritarian
regimes mirrored closely to the overall logistic regression model. The conclusion being that one

58

unit increase in perceived corruption of police decreases the log of odds in trust of courts of law
by a factor of .863. An important nuance was the effect of membership to a religious group.
This variable was not significant in the flawed democracies model, but was significant at the .05
level for hybrid regimes and at the .01 level for authoritarian regimes. Respondents in hybrid
regimes that were members of religious groups are likely to have lower levels of trust in courts
of law. This variable was more significant in authoritarian countries, but had the opposite effect;
membership to a religious group increases the likelihood of having higher levels of trust in courts
of law. The variations also show how democratic structure can influence views of legitimacy
given the differences between each model. These comparative models lead to important
implications.

59

Chapter 6
Discussion and Conclusion
Legitimacy of criminal justice organizations largely depends on how citizens view
personnel. This exists as a major theme in previous research. Views of fairness and corruption
strongly influence how individuals judge the legitimacy of criminal justice actors. Similar to
perceptions, experience with corruption also decreases the likelihood of legitimacy of police and
court officials. Findings involving these predictors remain consistent with previous research. In
consideration of policy, focusing on affable behaviors will give criminal justice organizations a
better opportunity for remaining effective in a given community. Social order depends on the
enforcement and adjudication of laws, but in large part depends on the willingness of community
members to work with criminal justice officials.
The purpose of this study was to reveal the relationship of various predictors with
perceptions of legitimacy in two major criminal justice organizations. Previous research
repeatedly shows the importance of procedural fairness in determining favorable views of
legitimacy in police and courts. The findings in this research suggest the same that citizens view
criminal justice officials more favorably when they are not corrupt and treat people equally.
Certain types of social capital also influenced the odds of having higher levels of perceived
legitimacy in police and courts. The interesting point here rests on how more structured types of
social capital, like member of a religious group and member of a community group, did not
influence legitimacy in either police or courts for the models including all cases. Membership to
a religious group was significant on two occasions when limited to respondents from certain
governmental regimes. The effect was different for hybrid regimes in comparison to
authoritarian regimes. Respondents in authoritarian regimes with membership to religious

60

groups are more likely to have higher perceived legitimacy in courts of law. Theoretically, it
seems possible that religion or religious activity, acts as a way to confirm power. One need not
look any further than the standard of living in countries like Sudan, the seeming religious
position dictators take in such governmental structures, and how citizens remain subdued.
Religion in this situation potentially adds to the structure of control and leads to higher perceived
legitimacy in courts of law.
On the other hand, respondents in hybrid regimes that were members of religious groups
were more likely to have lower levels of perceived legitimacy in courts of law. Philosophically,
religion acts as a basis to believe in something more and adhere to something abstract potentially
leading to a decrease in legitimacy of governmental structures. Conversely, activity in the
community led to the likelihood of higher levels of legitimacy in police and courts for hybrid
regimes. It is possible that attending community meetings and getting together with community
members to raise issues led to higher frequencies of interactions, direct or indirect, with criminal
justice officials that happened to be favorable. These favorable interactions, then led to the
likelihood of higher perceived legitimacy of police and courts. Clearly, perceived corruption of
these agents of the state and experience with corruption counteract the likelihood of higher
degrees of legitimacy. It makes sense that the opposite would be true for the influence of activity
in the community on perceptions of legitimacy. Further research on different forms of social
capital in relation to legitimacy seems beneficial.
Finally, a strong influence existed for democratic structure on the likelihood of levels of
legitimacy. An interesting finding is that democratic structure effects legitimacy in police and
legitimacy in courts of law differently. Hybrid regimes, in comparison to authoritarian regimes,
decreased the likelihood of higher degrees of legitimacy in police. By contrast, flawed and

61

hybrid regimes, in comparison to authoritarian regimes, increase the likelihood of higher degrees
of legitimacy in courts of law. Hybrid regimes have the same effect on levels of legitimacy, but
to a lesser degree than flawed democracies compared to authoritarian regimes. Considering both
models, the analysis suggests that respondents tend to have different perceptions of police and
courts. Further examinations should study if differences between perceptions of police and
courts are universal considering democratic structure and levels of legitimacy.
Studies like Madan and Nalla’s (2015) research relevant to police legitimacy in India are
an important foundation for extending research in this area. Legitimacy research generally
focuses on respondents in advanced democracies and hardly ever account for democratic
structure. Research on legitimacy tends to focus on micro predictors like perceptions and
experience with fairness, moral alignment, and beliefs regarding obedience. Future studies
should begin to develop the importance of macro predictors that might account for some of the
variation in views of legitimacy. Despite the significance of this variable, future research must
pursue ways of developing a more thorough operationalizion of democratic structure. When
using comparative analysis for selected cases, respondents in flawed democracies differed
greatly from those in authoritarian regimes. The explanatory power of the model for
authoritarian regimes was higher for variations in perceptions of legitimacy. This indicates that
different factors are more relevant for legitimacy when using democratic structure as a basis for
comparison. More research seems requisite to clarify this relationship.
There were several limitations in the present study. The first, although not necessarily a
limitation, was the use of secondary data. There are numerous advantages when using existing
data, mainly lower costs and efficiency. At the same time, the variables in the dataset were
rather limited in regards to legitimacy. This project provides a basis for original research

62

designed to capture robust predictors relevant to legitimacy of police, courts, and other criminal
justice entities in less advanced democracies. The Afrobarometer dataset, however, offered
sufficient information to approach the research questions and add to extant research on the
subject.
Furthermore, one of the major limitations was operationalizing legitimacy. The variables
in the dataset only allowed for the use of trust. Many researchers argue against the use of a
single indicator for capturing the depth of variables like trust (see Bradford et al., 2009; Jackson
& Sunshine, 2007; Sunshine & Tyler, 2003). Tyler’s (1990) original conception of legitimacy,
based on two indicators (trust and obligation to obey), was the conceptualization adopted by the
current project. Given the limited variables, only trust could be captured as an indicator of
legitimacy. Tyler’s conception has received many criticisms with continued research on the
legitimacy of criminal justice organizations (see Bottoms & Tankebe, 2012; Meško & Tankebe,
2014; Tankebe, 2013). The state of the research, however, remains vexed on two important
points; the first relating to the conceptualization of legitimacy and the use of procedural justice
predictors. It is safe, at this point, to continue investigating the relationship with Tyler’s original
model, especially in non-traditional countries. At some point in the future, research might
definitively state that this model is obsolete and ineffective, along with Tyler’s conception of
legitimacy.
Although type of government returned favorable results in predicting views of legitimacy
in criminal justice organizations, more research seems necessary to confirm the use of proxy
measures in this area. This approach is questionable considering the proxy measure merely
included cases for respondents in certain countries. In other words, it might behoove researchers
to develop comprehensive measures for democracy in an effort to clarify validity. In the context

63

of private security guards, Nalla and colleagues (2016) created a more comprehensive measure
for democratic structure. A thorough measure similar to the one used in their study would
behoove future researchers studying the effect of democratic structure on legitimacy. This would
also allow for a more complex statistical analysis, like structural equation modeling, leading to a
more profound account of how democracy and perceptions of democratic structure influence
legitimacy of criminal justice organizations. Future research might benefit from studying
continents like Africa and less advanced democracies. Understanding the legitimacy of criminal
justice organizations depends on several factors and improvements remain imperative. One
possible improvement involves elucidating citizen views of legitimacy under different
governmental structures as done in the current study.

64

REFERENCES

65

REFERENCES

Afrobarometer Network. (2014). Round 6 Survey Manual [Survey file]. Retrieved from
http://www.afrobarometer.org/sites/default/files/survey_manuals/ab_r6_survey_manual
en.pdf
Afrobarometer Network. (2015). Data codebook for a Round 5 Afrobarometer survey in 34
African countries [Codebook file]. Retrieved from
http://www.afrobarometer.org/data/merged-round-5-codebook-34-countries-2015
Afrobarometer Network (2016). Funding. Retrieved from
http://www.afrobarometer.org/about/funding
Anderson, C. J. & Tverdova, Y.V. (2003). Corruption, political allegiances, and attitudes toward
government in contemporary democracies. American Journal of Political Science, 47(1),
91-109. doi:10.1111/1540-5907.00007
Bacevich, A. J. (2002). American empire: The realities and consequences of U.S.
diplomacy. Cambridge, Mass: Harvard University Press.
Barbalet, J. (2009). A characterization of trust, and its consequences. Theory and Society, 38(4),
367–382.
Bates, R. H. (2010). Democracy in Africa: A very short history. Social Research, 77(4), 11331148.
Beckman, L. (2014). Democracy and the right to exclusion. Res Publica, 20(4), 395-411.
doi:10.1007/s11158-014-9253-y
Beetham, D. (1991). Max Weber and the legitimacy of the modern state. Analyse & Kritik, 13(1),
34-45. doi:10.1515/auk-1991-0102
Beetham, D. (2013). Revisiting Legitimacy, Twenty Years On. In J. Tankebe & A. Liebling
(eds.), Legitimacy and Criminal Justice: An International Exploration. Oxford: Oxford
University Press.
Beijersbergen, K. A., Dirkzwager, A. J. E., Eichelsheim, V. I., Van der Laan, Peter H, &
Nieuwbeerta, P. (2015). Procedural justice, anger, and prisoners’ misconduct: A
longitudinal study. Criminal Justice and Behavior, 42(2), 196-218.
Beijersbergen, K., Dirkzwager, A., & Nieuwbeerta, P. (2016). Reoffending after release: Does
procedural justice during imprisonment matter? Criminal Justice and Behavior, 43(1),
63-82. doi:10.1177/0093854815609643
Bernard, T. J., & Engel, R. S. (2001). Conceptualizing criminal justice theory. Justice Quarterly,
18(1), 1-30. doi:10.1080/07418820100094801
Booth, J. A., & Richard, P. B. (1998). Civil society and political context in Central America.
American Behavioral Scientist, 42(1), 33-46. doi:10.1177/0002764298042001004

66

Börzel, T. A., & Risse, T. (2016). Dysfunctional state institutions, trust, and governance in areas
of limited statehood: Governance and trust. Regulation & Governance, 10(2), 149-160.
doi:10.1111/rego.12100
Bottoms, A., & Tankebe, J. (2012). Beyond procedural justice: A dialogic approach to
legitimacy in criminal justice. The Journal of Criminal Law and Criminology (1973- ),
102(1), 119-170.
Boutros-Ghali, B. (1995). Democracy: A newly recognized imperative. Global Governance,
1(1), 3-11.
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.
doi:10.1080/10439460802457594
Bradford, B., Murphy, K., & Jackson, J. (2014). Policing, procedural justice and the
(re)production of social identity. British Journal of Criminology, 54(4), 527–550.
Britt, C., & Weisburd, D. (2014). Statistics in criminal justice. Boston, MA: Springer US.
doi:10.1007/978-0-387-34113-2
Canada, K. E., & Watson, A. C. (2013). “Cause everybody likes to be treated good”:
Perceptions of procedural justice among mental health court participants. American
Behavioral Scientist, 57(2), 209-230. doi:10.1177/0002764212465415
Cao, L. (2015). Differentiating confidence in the police, trust in the police, and satisfaction with
the police. Policing: An International Journal of Police Strategies & Management,
38(2), 239-249. doi:10.1108/PIJPSM-12-2014-0127
Damme, A. V., Pauwels, L., & Svensson, R. (2015). Why do swedes cooperate with the police?
A SEM analysis of Tyler's procedural justice model. European Journal on Criminal
Policy and Research, 21(1), 15. doi:10.1007/s10610-013-9224-4
Delhey, J., & Newton, K. (2005). Predicting cross-national levels of social trust: Global pattern
or nordic exceptionalism? European Sociological Review, 21(4), 311-327.
doi:10.1093/esr/jci022
Edmundson, W. A. (2013). Politics in a state of nature. Ratio Juris, 26(2), 149-186.
doi:10.1111/raju.12009
Fleck, R., & Hanssen, F. (2006). The origins of democracy: A model with application to
Ancient Greece. The Journal of Law & Economics, 49(1), 115-146. doi:10.1086/501088
Gau, J. M. (2013). Consent searches as a threat to procedural justice and police legitimacy: An
analysis of consent requests during traffic stops. Criminal Justice Policy Review, 24(6),
759-777.
Goldsmith, A. (2005). Police reform and the problem of trust. Theoretical Criminology, 9(4),
443-470. doi:10.1177/1362480605057727

67

Gorman, M. M. (1993). Hume's theory of belief. Hume Studies, 19(1), 89-101.
doi:10.1353/hms.2011.0448
Hagan, J. (1989). Why is there so little criminal justice theory? Neglected macro- and microlevel links between organization and power. Journal of Research in Crime and
Delinquency, 26(2), 116-135. doi:10.1177/0022427889026002002
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. doi:10.1375/acri.40.1.27
Hohl, K., Bradford, B., & Stanko, E. (2010). Influencing trust and confidence in the London
metropolitan police. British Journal of Criminology, 50(3), 491-513.
doi:10.1093/bjc/azq005
Holtfreter, K. (2016). Procedural justice and legitimacy across the criminal justice system:
Introduction to the special issue. Criminal Justice and Behavior, 43(1), 5-6.
doi:10.1177/0093854815609068
Hough, M., Jackson, J., Bradford, B., Myhill, A., & Quinton, P. (2010). Procedural justice, trust,
and institutional legitimacy. Policing, 4(3), 203-210. doi:10.1093/police/paq027
Ihlen, Ø. (2005). The power of social capital: Adapting Bourdieu to the study of public relations.
Public Relations Review, 31(4), 492-496. doi:10.1016/j.pubrev.2005.08.007
Jackson, J., & Bradford, B. (2009). Crime, policing and social order: On the expressive nature of
public confidence in policing. British Journal of Sociology, 60(3), 493–521.
Jackson, J., & Bradford, B. (2010). What is trust and confidence in the police? Policing, 4(3),
241-248. doi:10.1093/police/paq020
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. doi:10.1093/bjc/azs032
Jackson, J. & Gau, J. M. (2016). Carving Up Concepts? Differentiating Between Trust and
Legitimacy in Public Attitudes Towards Legal Authority. In E. Shockley (ed.),
Interdisciplinary perspectives on trust: Towards theoretical and methodological
integration. Cham: Springer International Publishing. doi:10.1007/978-3-319-22261-5
Jackson, J., & Sunshine, J. (2007). PUBLIC CONFIDENCE IN POLICING: A neodurkheimian perspective. The British Journal of Criminology, 47(2), 214-233.
doi:10.1093/bjc/azl031
Kääriäinen, J. T. (2007). Trust in the police in 16 European countries: A multilevel analysis.
European Journal of Criminology, 4(4), 409-435. doi:10.1177/1477370807080720

68

Kaina, V. (2008). Legitimacy, trust and procedural fairness: Remarks on Marcia Grimes’ study.
European Journal of Political Research, 47(4), 510-521. doi:10.1111/j.14756765.2008.00773.x
Kelman, H.C., Hamilton, V.L. (1989). Crimes of Obedience. New Haven, CT: Yale Univ. Press
Laxminarayan, M. (2012). Procedural justice and psychological effects of criminal proceedings:
The moderating effect of offense type. Social Justice Research, 25(4), 390-405.
doi:10.1007/s11211-012-0167-6
Madan, M., & Nalla, M. (2015). Exploring citizen satisfaction with police in India the role of
procedural justice, police performance, professionalism, and integrity. Policing-an
International Journal of Police Strategies & Management, 38(1), 86-101.
doi:10.1108/PIJPSM-06-2014-0063
Mattes, R., & Bratton, M. (2007). Learning about democracy in Africa: Awareness, performance,
and experience. American Journal of Political Science, 51(1), 192-217.
doi:10.1111/j.1540-5907.2007.00245.x
Mazerolle, L., Bennett, S., Davis, J., Sargeant, E., & Manning, M. (2013). Procedural justice and
police legitimacy: A systematic review of the research evidence. Journal of
Experimental Criminology, 9(3), 245-274. doi:10.1007/s11292-013-9175-2
Medina, V. (1990). Social contract theories: Political obligation or anarchy?: Savage, MD:
Rowman & Littlefield.
Meško, G., & Tankebe, J. (2014). Trust and legitimacy in criminal justice: European
perspectives. Cham: Springer.
Mishler, W., & Rose, R. (2001). What are the origins of political trust?: Testing institutional and
cultural theories in post-communist societies. Comparative Political Studies, 34(1), 3062. doi:10.1177/0010414001034001002
Morris, C. (2008). State Legitimacy and Social Order. In J. Kühnelt (Ed.), Political
Legitimization without Morality? (pp.15-32). Dordrecht: Springer.
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. doi:10.1375/acri.42.2.159
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. doi:10.1177/0004865811405260
Murphy, K., & Cherney, A. (2012). Understanding cooperation with police in a diverse society.
The British Journal of Criminology, 52(1), 181-201. doi:10.1093/bjc/azr065

69

Murphy, K., Mazerolle, L., & Bennett, S. (2014). Promoting trust in police: Findings from a
randomised experimental field trial of procedural justice policing. Policing &
Society,24(4), 405-424. doi:10.1080/10439463.2013.862246
Nalla, M. K., & Mamayek, C. (2013). Democratic policing, police accountability, and citizen
oversight in Asia: An exploratory study. Police Practice and Research, 14(2), 1-13.
doi:10.1080/15614263.2013.767091
Nalla, M.K., Maxwell, S.R. & Mamayek, C.M. (2017). Legitimacy of Private Police in
Developed, Emerging, and Transitional Economies. European Journal of Crime,
Criminal Law, and Criminal Justice, 25(1), 76-100.
Nega, B., & Schneider, G. (2012). Things fall apart: Dictatorships, development, and democracy
in Africa. Journal of Economic Issues, 46(2), 371-382. doi:10.2753/JEI00213624460212
Payne, B. K. (2016). Expanding the boundaries of criminal justice: Emphasizing the "S" in the
criminal justice sciences through interdisciplinary efforts. Justice Quarterly, 33(1), 1-20.
doi:10.1080/07418825.2015.1068837
Putnam, R. D. (1995). Tuning in, tuning out: The strange disappearance of social capital in
America. PS: Political Science and Politics, 28(4), 664-683.
Putnam, R. D. (2000). Bowling alone: The collapse and revival of American community. New
York: Simon & Schuster.
Rappaport, A. (1975). A History of American Diplomacy. New York: Macmillan.
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.
doi:10.1177/0093854807301275
Sandholtz, W., & Koetzle, W. (2000). Accounting for corruption: Economic structure,
democracy, and trade. International Studies Quarterly, 44(1), 31-50. doi:10.1111/00208833.00147
Sargeant, E., Murphy, K., & Cherney, A. (2014). Ethnicity, trust and cooperation with police:
Testing the dominance of the process-based model. European Journal of Criminology,
11(4), 500-524. doi:10.1177/1477370813511386
Skogan, W. G. (2006). Asymmetry in the impact of encounters with police. Policing and Society,
16(2), 99-126. doi:10.1080/10439460600662098
Skolnick, J. H., & University of California. Center for the Study of Law and Society. (1966).
Justice without trial: Law enforcement in democratic society. New York: Wiley.
Suchman, M. C. (1995). Managing legitimacy: Strategic and institutional approaches. The
Academy of Management Review, 20(3), 571-610. doi:10.5465/AMR.1995.9508080331

70

Sun, I., Hu, R., & Wu, Y. (2012). Social capital, political participation, and trust in the police in
urban china. Australian & New Zealand Journal of Criminology, 45(1), 87-105.
doi:10.1177/0004865811431329
Sun, I., & Wu, Y. (2015). Arab Americans' confidence in police. Crime & Delinquency, 61(4),
483-508. doi:10.1177/0011128711420103
Sunshine, J., & Tyler, T. R. (2003). The role of procedural justice and legitimacy in shaping
public support for policing. Law & Society Review, 37(3), 513-548. doi:10.1111/15405893.3703002
Tankebe, J. (2010). PUBLIC CONFIDENCE IN THE POLICE: Testing the effects of public
experiences of police corruption in Ghana. The British Journal of Criminology, 50(2),
296-319. doi:10.1093/bjc/azq001
Tankebe, J. (2013). Viewing things differently: The dimensions of public perceptions of police
legitimacy. Criminology, 51(1), 103-135. doi:10.1111/j.1745-9125.2012.00291.x
Tankebe, J., Reisig, M., & Wang, X. (2016). A multidimensional model of police legitimacy: A
cross-cultural assessment. Law and Human Behavior, 40(1), 11-22.
doi:10.1037/lhb0000153
Tyler, T. R. (2004). Enhancing police legitimacy. The Annals of the American Academy of
Political and Social Science, 593(1), 84-99. doi:10.1177/0002716203262627
Tyler, T. R. (2007a). Legitimacy and criminal justice: An international perspective Russell Sage
Foundation. doi:10.7758/9781610445412
Tyler, T. R. (2007b). Procedural justice and the courts. Court Review, 44(1-2), 26.
Tyler, T. R. (2011). Trust and legitimacy: Policing in the USA and Europe. European Journal of
Criminology, 8(4), 254-266. doi:10.1177/1477370811411462
Uslaner, E. M. (2008). Corruption, inequality, and the rule of law: The bulging pocket makes the
easy life. Cambridge; New York: Cambridge University Press.
Uwizeyimana, D. E. (2012). Democracy and Pretend Democracies in Africa: Myths of African
Democracies. Law, Democracy & Development, 16
Weber, M. (1978). Economy and society: An outline of interpretive sociology. (Roth, G., &
Wittich, C., Trans.). Berkeley: University of California Press. (Original work published
1922)
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.
doi:10.1007/s10940-015-9263-8
Wu, Y., & Sun, I. Y. (2009). Citizen trust in police: The case of china. Police Quarterly, 12(2),
170-191. doi:10.1177/1098611108330228

71

Zelditch, M. (2001). Processes of legitimation: Recent developments and new directions. Social
Psychology Quarterly, 64(1), 4-17.
Zmerli, S., & Newton, K. (2008). Social trust and attitudes toward democracy. The Public
Opinion Quarterly, 72(4), 706-724. doi:10.1093/poq/nfn054

72