Public trust in the police: Investigating the influence of police performance, procedural fairness, and police-community relations in Addis Ababa, Ethiopia

Abstract In Western countries, there is conclusive research on public trust in the police and the factors that affect it; however, the situation in developing countries like Ethiopia is largely unknown and unexplored. The current study utilized cross-sectional data from representative 616 households from Addis Ababa, Ethiopia. The study results found lower trust in the local police, less effective in reducing crime, and residents’ unfavorable perceptions of police procedural fairness, accompanied by poor relationships with the community. Moreover, perceptions of police effectiveness, procedural justice, and police-community relations were significant predictors of citizens’ trust in the police. The results of the study provide evidence for the applicability of the underlying assumptions of the performance model, procedural justice and community policing in the context of Ethiopia. As a result, to enhance trust in the police and improve relations with citizens, police should effectively control crime and communicate it, exert authority procedurally, exercise fair and respectful policing, and undertake activities that positively strengthen their relationship with the community. This study has theoretical and practical implications and provides insights for enhancing public trust in the police.


Introduction
In Ethiopia, police practices are often criticized for human rights violations and lack of democratic policing principles (DiNunzio, 2014;Toggia, 2008).Notwithstanding criticisms over police behavior, a dearth of empirical literature systematically reveals concerns and calls for change in policy, strategy, and practice.However, police scholars and researchers emphasized that by influencing policies and procedures, studies on policing and police conduct can alter organizational and operational practices (Alemika, 2009;Tankebe, 2008).There has yet to be much research, particularly on police-citizen trust and the variables that explain and strengthen trust relationships.Nevertheless, research evidence recognized the relevance of public trust in criminal justice institutions in reducing crime, building and maintaining a safer society, improving public relations, and enhancing law enforcement institutions' legitimacy (Boateng, 2018;Sargeant, 2015;Tankebe, 2008).Enhancing trust is essential, but doing so necessitates having a firm understanding of what it means in various socioeconomic contexts and identifying pertinent elements relevant to improve police-citizen relations.
The World Bank (2011) study suggests that effective law enforcement should foster informal social control, improve social cohesion and enhance trust with citizens.Abundant research works had similar findings that public trust in criminal justice institutions improves collaboration and cooperation and empowers law enforcement institutions, all important in reducing crime and improving community safety (Groff, 2015;Mazerolle et al., 2012;Sampson, 2012;Tyler & Huo, 2003).Further, citizens' trust in police influences their willingness to report criminal activity and use informal social control to regulate human behavior (Kirk & Papachristos, 2011;Roberts & Plesnicˇar, 2015;Sampson & Bartusch, 1998).The value of trust relationships between the public and criminal justice institutions, particularly the police institution, is well supported by data.In contrast, there is conflicting literature regarding why people trust criminal justice institutions.For instance, a group of scholars argues that police performance or police effectiveness in controlling crime and providing services better explains and contextualizes the overall trust in the police (Boateng, 2017;Wang & Sun, 2020).This body of research argues that effective crime control builds trust, which in turn helps the police in their crime control efforts.In contrast, another set of academics argued that police procedural fairness continues to have a more significant impact in determining the public's trust in the police (Jackson et al., 2012;Nix et al., 2015;Tyler, 2000).In addition to policing features, prior victimization, perceived crime level, and fear of crime can potentially erode trust in criminal justice institutions across democratic contexts (Singer et al., 2019).
The research described above is based predominantly on the experiences of western countries, leaving the issue in developing nations relatively unstudied.Few studies conducted in Africa have separately supported the crucial role of police performance (Boateng, 2017;Tankebe, 2008) and procedural fairness (Prinsloo, 2019) in determining trust in the police.However, these studies could be more extensive in their scope and provide adequate evidence to generalize to socioeconomically distinct societies in the global south, creating a gap in understanding the topic.In this case, investigating public trust in the police and revealing elements that might affect public confidence in the police in non-Western nations like Ethiopia.
Despite the critical importance that public perception of institutions and the elements that influence trust relationships play in Ethiopia, more needs to be done on these topics.The recent World Values Survey revealed that 40.8 percent of Ethiopians have little faith in the police, which is lower than most public institutions, including courts and the central government (Haerpfer et al., 2022).Other empirical studies have examined factors influencing citizens' perceptions of government institutions' service provisions, including criminal justice institutions (Gudeta, 2019;Wandall, 2015).Gudeta (2019), using the 2007 public opinion survey, found average public trust in criminal justice institutions and stated that institutional performance is critical to understanding people's perceptions of trust and confidence.Wandall (2015) revealed the negative impacts of plural legal governance and its inconsistent implementation on public confidence and weak trust relationships.Further, DiNunzio (2014) noted weak police-citizen relationships accompanied by a lack of confidence and legitimacy.Despite relevant implications, the above studies are broader in their scope, feature theoretical and methodological limitations, and have little to contribute knowledge about citizens' trust in the police-and factors that influence public trust in the police.
The current study examines the impact of procedural justice, police effectiveness, and policecitizen relations on citizens' trust in the police using data from Ethiopia.This study recognizes the importance of citizens' trust in the police as well as the critical roles that these concepts play in policing and the need to understand how views of these concepts are formed.

Conceptualization of public trust in the police
The concept of trust entails the "willingness of a party to be vulnerable to the actions of another party based on the expectation that the other will perform a particular action important to the trustor, irrespective of the ability to monitor or control that other party" (Mayer et al., 1995, p. 712).Whereas trust in criminal justice involves the public's expectation and confidence in the integrity, effectiveness, and justice of the criminal justice system and institutions in their jurisdictions (Sherman, 2002).The concept of police trust involves the confidence that the public has in the police to keep them safe and deliver necessary services (Boateng, 2017).For Sherman (2002), trust in the criminal justice system has three dimensions: belief in the system's integrity, efficacy, and justice.All aspects of trust are valuable for criminal justice in that it makes them more efficient and authoritative, ensures public cooperation, and enhances the willingness of the public to report crimes and provide information (Mazerolle et al., 2013;Singer et al., 2019;Wandall, 2015).
The term "public trust in the police" refers to the public's confidence in police officers' ability to carry out their tasks professionally and ethically (Hardin, 2002, p. 31).Trust in police also refers to the notion that police have people's best interests in mind and can act appropriately under specific conditions (Hardin, 2002).Trustworthiness is essential for police agencies as citizens generally have limited knowledge of police practices and lack expertise in evaluating police performance (Jackson et al., 2012).That is, a person's trust in police tends to be based on limited personal experience that conveys little information about police intentions and characteristics.Researchers have been trying to conceive and find the roots of public trust in the police to enrich police interaction with the public and increase trust relationships' benefits (Nix et al., 2015;Tyler & Huo, 2003;Van Craen, 2013).
The effectiveness and quality of police services, according to proponents of the performance model, are more critical for building public trust in the police than fairness in methods and procedures (Sargeant, 2015;Tankebe, 2008).Procedural fairness proponents, on the contrary, argue that procedural fairness is an essential component of police activity and that the public prefers to evaluate police officers based on the fairness of the tactics they use during interaction and the decisions they make (Nix et al., 2015;Sunshine & Tyler, 2003;Tankebe, 2008).To these researchers, to gain residents' trust, the police must treat everyone equally, regardless of race, gender, or socioeconomic class.
Overall, public trust in police work improves police effectiveness and legitimizes police actions (Goldsmith, 2005), leading to voluntary acceptance, which supports positive behavioral outcomes, including cooperation, compliance, and information sharing-all of which improve police effectiveness (Sunshine & Tyler, 2003).However, not every police institution is regarded as reliable and trustworthy, and some police agencies have severe trust issues, resulting in ineffectiveness.When a police department lacks public trust, gaining public cooperation and compliance is difficult.
Several studies have found that trust in police is essential in reducing crime and violence (Kochel & Weisburd, 2019;Kochel, 2013;Renauer, 2007;Sargeant, 2015).Citizens who trust the police can help by reporting crimes they see and supplying crucial information that leads to the arrest of criminals.In different sociocultural contexts, the above studies contended that police could contribute to collective efficacy and lower crime through effective services, minimizing misconduct, and improving procedural fairness.Thus, a study that interrogates trust in police and factors that shape it can contribute to crime control efforts of the police.

Trust in the police: theoretical overview
Several theoretical perspectives have been proposed to explain why the public trust the criminal justice system, its relevance for practical operations, and how it can be improved or enhanced (Bouckaert et al., 2002;Tyler & Huo, 2003;Van Craen, 2013).These theories were created with the knowledge that a lack of public participation and trust hinders the police's ability to enforce laws and handle crimes that harm the safety and security of societies.Furthermore, it was thought that people who trust the police were more likely to assist them by reporting crimes and covertly enforcing social rules (Nix et al., 2015).
Moreover, in contrast to utilizing instrumental tactics like threatening punishment for criminal activity or providing rewards like crime prevention, approaches to authorities that emphasize establishing trust demonstrated higher prosocial outcomes(O 'Brien & Tyler, 2020;T. C. O. O'brien & Tyler, 2019).The effectiveness of utilizing trust tactics makes sense from both a moral and practical standpoint, diminishing fear among communities struggling with crime and increasing their willingness to become active stewards in providing solutions to social issues (T.C. O. O'brien & Tyler, 2019).Furthermore, trust-based approaches can positively transform cultural attitudes toward authority, creating solid bonds between citizens and those charged with keeping them safe, building goodwill between all parties involved, and creating a healthier society for all.
Performance (outcome-based or instrumental), procedural justice (process-based or expressive), and community policing models are frequently referenced in investigating the value of trust and are widely cited to guide research in the field (Boateng, 2017;Renauer, 2007;Van Craen, 2013).In the present study, performance, procedural justice, and community policing models and assumptions are poised to guide the study to see if they explain the variation of public trust in the police and are applicable in the Ethiopian context.

Performance theory and its explanation of institutional trust
The performance model is one of the most widely utilized perspectives to explain the sources of public trust in institutions (Boateng, 2018;Van Craen, 2013;Sargeant, 2015;Tankebe, 2008;Wang & Sun, 2020).This theory has macro and micro dimensions where the macro performance examines disparities in trust between countries over time due to changes in unemployment rates, economic growth, inflation, political stability, and other factors (Bouckaert et al., 2002;Maarten, 2013).Following this aspect of the theory, the police can be evaluated, as a critical component of criminal justice, by looking at the overall crime rates, insecurity, levels of fear of crime, and neighborhood disorder (Boateng, 2017).
Micro-performance theory, on the other hand, emphasizes variations in trust to changes in (the quality of, or perception of) service delivery of a specific institution, where actual service delivery determines trust reactions (Bouckaert et al., 2002).Specifically, Bouckaert et al. (2002) identified the quality of service, the existence of service, and the level of service as indicators for judging the micro-performance of any institution.This aspect of the theory contends that public expectations of the police in controlling crime, reducing fear of crime, and offering essential services determine the degree to which the public trusts the police (Boateng, 2017;Sargeant, 2015).Accordingly, the public reacts negatively if the police fail to control crime, are incompetent in dealing with crime problems, and inadequately provide required services (Sargeant, 2015;Tyler, 2005).In contrast, if the police successfully combat crime, they will gain public trust and confidence and the public's help and support.Thus, the police's ability or inability to engage in effective crime control and service provision is seen as driving the valence of public evaluations of trust in the police.However, micro explanations are more reliable if individuals or residents are well-informed about the institution under investigation.As a result, scholars preferred micro performance theory to explain public trust in local police, as the local police maintain intimate interactions with people (Boateng, 2017;Sargeant, 2015;Tankebe, 2008).
Studies primarily confined to developed societies established the importance of the effectiveness of the police in controlling crime in understanding citizens' trust in the police (Berthelot et al., 2018;Sargeant, 2015;Telep & Weisburd, 2012).However, few studies have produced results that contradict the assumption of performance theory and instead support the fairness of police practices as a significant determinant of the public's trust in the police (Lim & Kwak, 2022;Roberts & Plesnicˇar, 2015;Sun et al., 2014;Tankebe, 2008).Again, other groups of scholars argue that experience of victimization, fear of crime, and actual crime problems in their environment influence citizen's trust and confidence in the police (Berthelot et al., 2018;Jackson & Bradford, 2009;Renauer, 2007;Sargeant, 2015;Singer et al., 2019).As a result, when individuals feel protected and not at risk of victimization, they show higher trust in the police, and the opposite is also true.
Concerning the developing countries' context, few studies revealed that police effectiveness in controlling crime determines citizens' trust in police and officers' treatment of citizens with respect and dignity, neighborhood disorder, and police corruption or misconduct (Boateng, 2017;Tankebe, 2008).However, the generalizability of the above evidence to other societies is limited in different ways: first, the assumptions of performance theory are primarily tested in western societies, which hinders the ability of a theory to achieve global acceptance in explaining citizens' trust in the police; second, because the evidence from these societies itself shows conflicting results, showing that either performance model alone is not a sufficient factor in determining trust in the police; and third, these studies have only taken into account the perspectives of citizens, not those of the police when describing what influences public-citizen trust relationships.Thus, the current study examined public trust in the police and investigated the applicability of the performance model in explaining trust in the police.

Procedural justice theory and its explanation of institutional trust
Police officers are considered representatives of mainstream society and the rule of law (Gau et al., 2012), and their actions considerably impact the public attitude.As a result, how police officers carry out their duties is a serious concern, as they possess immense powers that can be readily exploited.Procedural fairness refers to people's perceptions of their treatment during decisionmaking procedures (T.C. O. O'brien & Tyler, 2019).Procedural justice also entails treating people with respect and dignity, as well as fairness and appropriateness of methods for exerting authority (Stoutland, 2001).A growing body of research shows that fair policing fosters trust in the police, police legitimacy, and other forms of cooperative and supportive conduct among members of the public (Jackson et al., 2012;Sunshine & Tyler, 2003;Tankebe, 2008;Van Craen & Skogan, 2017).
The procedural justice model, in contrast to the performance model, focuses on the process that leads to trust and where officers' acts are viewed as correct when they listen to and examine people's opinions when they are neutral when presenting excellent explanations for their choice encounters and demonstrating a good knowledge of the laws and practices following the laws (Mazerolle et al., 2012(Mazerolle et al., , 2013;;Tyler & Huo, 2003;Tyler, 2005).In addition, to studies highlighting the importance of procedural justice in policing, the degree to which citizens are treated fairly and justly by the police is a crucial determinant of whether or not those persons will trust, voluntarily collaborate and obey orders.Tyler (1990) and Tyler and Huo (2003) identified four components of procedural justice: citizen participation in the decision-making process; perceived neutrality in the officer's decision-making; treatment with dignity and respect; and belief in the officer's motives.Accordingly, when police incorporate these procedural justice components into their interactions with citizens and suspects, citizens are more likely to trust them, follow their directives, and consider them legitimate.In contrast to Tyler's (2000) characterizations, later studies have treated procedural justice as having two linked components, namely quality of treatment which involves issues of neutrality and public perception that decisions are made based on objective measures and fair and just decision-making where officers must treat citizens with respect and decency (Renauer, 2007).
Regardless of the systematization, several scholars have utilized procedural justice and empirically established that people are more likely to trust the police when they believe police follow fair and unbiased procedures, treat people with respect and dignity, involve citizens in decision making (Maarten, 2013;Mazerolle et al., 2012Mazerolle et al., , 2013;;Sargeant, 2015).Further, a substantial body of research demonstrated the superior importance of procedural fairness over police crime control effectiveness in predicting citizen trust in the police (Jackson et al., 2012;Nix et al., 2015;Sunshine & Tyler, 2003).Following the theoretical and empirical literature, it is hypothesized that favorable perceptions of police procedural fairness will positively influence citizens' trust in the police.Thus, it is expected that people who believe the police to be procedurally fair will indicate higher trust than those who believe.
In contrast, a longitudinal study conducted in the United Kingdom discovered that trust in the police is mainly connected to perceptions of crime and the property crime rate, not to aggregate fear about crime and perceptions of social cohesion or informal social control, or other factors (Sindall, Sturgis, and Jennings, 2012).The findings of the above study support statements by eminent scholars who acknowledged that in some nations, outcome considerations might have a more significant influence on how the public views legal authorities than procedural treatment (Tyler, 1990).
One drawback of the procedural justice approach is that its lack of attention to the past means that any existing distrust can slow the process of trust building, with community members interpreting their new experiences through the lens of the past (T.C. O. O'brien & Tyler, 2019).Thus, authorities can compensate for this problem by complementing procedural justice with diligent work toward reconciliation, emphasizing gestures intended to earn communities' trust; these gestures address the community rather than individuals and acknowledge historical and other reasons for distrust (T.C. O. O'brien & Tyler, 2019).Evidence from experimental research demonstrated that reconciliatory gestures emphasizing trust-building approaches to authority could increase community members' voluntary cooperation with authorities by increasing beliefs about their sincerity and, most consistently, in terms of reporting a crime (O 'Brien & Tyler, 2020).
Regarding the relevance of theoretical viewpoints in developing countries, Tankebe (2008) discovered that in the Ghanaian setting, practical issues like perceptions of police efficacy in preventing crime impacted public trust in the police.However, Tankabe noted that effectiveness in controlling crime does not guarantee trustworthiness that can only be legitimized if it is achieved in a way that respects citizens' dignity.Other researchers also claim that to understand police trust, both performance and procedural justice perspectives are critical (Prinsloo, 2019;Sun et al., 2014).Thus, it is relevant to investigate how police effectiveness and procedural justice influence residents' trust in the police.The present study adds to the existing literature by examining procedural-based and outcome-based sources of trust in the police.

Why was a study conducted in Addis Ababa, Ethiopia?
The study is conducted in Addis Ababa, the capital city of Ethiopia.Addis Ababa is the primate city, economic, cultural and political center, and the most populous city in Ethiopia, home to people from many backgrounds and walks of life.Addis Ababa accounts for about 20% of the country's total urban population and 29% of the GDP, and with a diverse range of people with different socioeconomic backgrounds.Addis Ababa is the world's fourth diplomatic hub and Africa's political capital.The police in Addis Ababa has been criticized for treating the wealthy, those with connections to ethnic political groups, those who support the government favorably while treating the poor, those with less political clout, or those who oppose the government harshly.Furthermore, empirical studies contend that the police in Addis Ababa discriminate against lower socioeconomic classes through arbitrary arrest and imprisonment, as well as by not working in disadvantaged neighborhoods (DiNunzio, 2014).
Further, the police were criticized for failing to prevent or reduce crime, preventing everincreasing and complicated crime on the one hand, and for improper and excessive use of force on citizens.These behaviors undermine the rule of law in Ethiopia and raise serious questions about police behavior and professional and autonomous pursuits and public trust in the police.However, the abovementioned observations lack systematization and an empirical base.Therefore, assessing how citizens form their perceptions of trust in police and interrogating pertinent factors that explain it is needed and can be an exciting contribution.

Data collection and sampling
The data was collected in systematically selected 11 neighborhoods (Katana in Amharic) in Addis Ababa, Ethiopia, using face-to-face structured and unstructured interviews.Addis Ababa is organized along 11 sub-cities, 118 districts and 850 counties.There are significant differences among sub-cities, districts, and neighborhoods in Addis Ababa-by income level, prevailing socioeconomic activities, and population density, and therefore it is expected that perceptions towards the police may vary based on the neighborhood in which a person resides.The selection of the sub-cities and districts is guided by earlier reconnaissance visits of the researcher to the various parts of the city and some consultations made with relevant officers at the Addis Ababa Police Commission.Accordingly, three sub-cities, each having a homogeneous characteristic that purposefully represents Addis Ababa's socioeconomics, were selected, namely Bole, Addis Ketema, and Kirkos sub-cities.Bole is frequently referred to as hosting wealthy and middle-income residents.Addis Ketema sub-city hosts the bus station and the nation's largest market, which also has the highest business activity.Kirkos is home to both the city's commercial center and lower-class neighborhoods.However, we have gathered the data in Kirkos sub-city's lower-income neighborhoods.
A sampling of respondents in the selected neighborhoods was undertaken using a multistage cluster sampling using a list of enumeration areas-the smallest well-defined unit from which household data can be obtained.District-level police stations have designated geographic locations for community policing purposes.The police know these smallest areas of jurisdiction as Katana (synonymous with the county level).However, Katana is not only a geographical area; it may be viewed as a neighborhood with residents who share similar traits and are linked together in social and economic life.Therefore, the actual enumeration areas are randomly selected Katana from each district in this study.Respondents for the household surveys are interviewed face-to-face using the Kobo toolkit, an android-supported mobile application, to collect data.Table 1 shows sampled sub-cities, districts, the number of Katana, counties, or EAs, and the sample size for the respective enumeration areas.
With regards to the sampling, first, the required number of EAs for each of the 11 neighborhoods, including two residential condominium compounds, were obtained from the community policing department of each sub-city police division.The number of EAs obtained was proportional to the population size of the various neighborhoods.The next stage involved a systematic sampling of households within each EA.The last stage involved sampling a household head from each house.As a result, 616 households in total participated in the household survey, with a response rate of more than 99 percent.

Dependent variable
The dependent variable of this study is trust in the police.Sunshine and Tyler's (2003) indicators were adjusted to better suit the purposes of the current study to measure trust in the police.Thus, to gauge trust in the police, residents were asked whether they believed their local police would protect people and property.Response categories were (1) Strongly disagree, (2) Disagree, (3) Undecided, (4) Agree, and (5) Strongly agree.

Independent variables
To measure police effectiveness in controlling crime and providing services, we adapted 6 Likert-type items from the literature with contextualization (Boateng, 2017;Sargeant, 2015).The lead-in question for each item asks the respondent to express whether they agree or disagree with the statement.The items include: "The police are effective in controlling violent crime in your neighborhood," "The police are effective at arresting crime suspects in your neighborhood," "The police are effective in controlling burglary in your neighborhood," "The police are effective in controlling theft in your neighborhood," "The police promptly respond to emergency calls for assistance," "Police do a good job in responding to people in the neighborhood after being victims of crime."we have used the first four categories to assess how effective the police are at controlling crime.In contrast, the latter two items assess police effectiveness in providing services.Response categories were (1) Strongly disagree, (2) Disagree, (3) Undecided, (4) Agree, and (5) Strongly agree.Police effectiveness to control crime and provide services subscale has a mean of 17.33 (SD = 4.39) and an alpha value of 0.749, suggesting good internal reliability.
Six contextually appropriate Likert-type items that ask respondents to evaluate how the local police interact procedurally with people were adapted from earlier studies to measure procedural fairness (Mazerolle et al., 2012;Nix et al., 2015;Tyler, 2005).Every item has a lead-in introduction that asks the respondent to indicate if they agree or disagree with the statement.The statements include, "When the police stop people, they usually handle the situation well," "The police in your neighborhood or district accurately understand and apply the law," and "The police make decisions based upon facts, not their personal biases or opinions," "The police give honest explanations for their actions to the people they deal with, "Most police officers use only the amount of force necessary to accomplish their tasks," and, "The police treat everyone in your neighborhood equally ."Responsecategories that fit the above indicators are (1) Strongly disagree, (2) Disagree, (3) Undecided, (4) Agree, and (5) Strongly agree.Therefore, the responses were summed to form a procedural fairness scale.The procedural justice subscale has a mean of 15.77 (SD = 4.56) and an alpha value of 0.811, suggesting good internal reliability.
How to gauge police-community relations has been informed by the literature on community policing and its pertinent indicators (Crowl, 2017;Gill et al., 2014).Police community relations are evaluated in the current study based on how well local police work with the community to prevent crime, whether the police-community partnership already in place is successful in resolving local issues, and how well the community policing program works to lower crime and the fear of crime.The lead-in question for each item asks the respondent to express whether they agree or disagree with the statement.The indicators include, "The police involve the community to discuss crime problems, to identify causes and provide solutions", "How good a job is a police doing in working together with residents in your neighborhood to solve local problems", and, "Community policing program is effective in reducing crime and fear of crime".Police community relations items have response categories of (1) Strongly disagree, (2) Disagree, (3) Undecided, (4) Agree, and (5) Strongly agree, which are analogous to the previous predictor factors.The police community relations subscale has a mean of 9.11 (SD = 3.32) and an alpha value of 0.711, suggesting good internal reliability.

Control variables.
Control variables for this study include crime victimization, neighborhood crime problem, fear of crime, social cohesion, and informal social control.Crime victimization is measured by asking people if they have experienced robbery, burglary, assault, violent threats, or any other type of crime in the last 12 months (Yes = 1, No = 0).The level of neighborhood crime problem is measured by asking residents about the seriousness of robbery, theft, and burglary in their neighborhood (Not a problem = 1, somewhat a problem = 2, a problem = 3 and a big problem = 4).The neighborhood crime problem subscale has a mean of 6.36 (SD = 2.412) and an alpha value of 0.831, suggesting good internal reliability.
We have measured fear of crime by three items that asked respondents whether they agree or disagree with statements under various scenarios: (1) "I am afraid to walk in my neighborhood at night time by myself," (2) "I worried that someone breaks into my house and stealing my property," (3) "I worried that being attacked or robbed or mugged on the street ."Responsecategories for fear of crime are (1) Strongly disagree, (2) Disagree, (3) Undecided, (4) Agree, and (5) Strongly agree.The fear of crime subscale has a mean of 3.78 (SD = 1.997) and an alpha value of 0.753, suggesting good internal reliability.
The scale items utilized to measure social cohesion and informal social control are derived from the literature yet with contextualization (Kochel, 2013;Sampson et al., 1997).Social cohesion was also measured using seven items using a five-point Likert scale to rate respondents' agreement with the following statements: (1)People around here are willing to help their neighbors."(2)"This is a close-knit neighborhood."(3) "People in this neighborhood can be trusted."(4) "People in this neighborhood share confidential information."(5) "Adults in this neighborhood know who the local children are." (6) People in this neighborhood get together to deal with problems".(7) People in this neighborhood visit each other's homes or talk on the street" The social cohesion subscale has a mean of 27.34 (SD = 4.958) and an alpha value of 0.785, suggesting good internal reliability.
Informal social control was measured using six items asking for the likelihood of their neighbors intervening under various circumstances, (1) children were skipping school and hanging out in undesirable places, (2) children were disrespecting an adult; (3) Underage youth smoking cigarettes or chewing "Khat," (4) a disrespectful husband attacking his wife, (5) a fight broke out in front of their houses, and (6) form a neighborhood watch or undertake patrol to respond to surging crime and disorder.A five-point Likert scale ranging from "very unlikely" to "very likely" was used to rate responses.The informal social control subscale has a mean of 25.67 (SD = 4.246) and an alpha value of 0.853, suggesting good internal reliability.

Analytical plan
The analysis of the quantitative data was conducted using SPSS version 26.In particular, SPSS 26 was used to create descriptive statistics for every variable, analyze bivariate correlations between every variable, and do regression analyses on both the outcome and predictor variables.First, the dependent and independent variables and the respondent's profile are described using descriptive statistics, including frequencies, means, and standard deviation.Second, normal distribution and multicollinearity tests were undertaken before undertaking the bivariate analysis.Using both means and log transform methods, the normal distribution test demonstrated that the data do not qualify for normal distribution or that the relationship between the variables is not linear.Thus, a non-parametric correlation analysis, namely the Spearman rho correlation, was undertaken to assess the strength and significance of the relationship between the variables and whether the independent variables were not highly correlated (i.e., there was no multicollinearity).Indeed, as shown in the correlation matrix Table 4, there is no sign of multicollinearity since all correlation coefficients for independent variables were less than 0.7.Third, since the only outcome variable of the study-trust in the police-is with five ordinal level categories and the data did not follow the normal distribution, ordinal and logistic regression is conducted to determine the effect of the predictor variables on the dependent variable (trust in police).Fourth, the qualitative data were analyzed thematically and embedded in the quantitative data.

Profiles of the respondents
Table 2. displays the research respondents' demographic background.Most of the respondents (67%) were males, with an average age of 35 years (SD = 9.60) and an average of 4.11 household size (SD = 2.02) with an average of 2 (SD = 1.29) household members generating an income.The average duration of residency in the current neighborhood is 20 years (SD = 14.362), with minimum and maximum periods of one year and 56 years, respectively.The youngest person who participated in the survey was 19 years old, while the oldest person was 86 years old.Furthermore, 52 % of the respondents were married, about 54% were run own businesses, 31.82 % were employed, and only 5.52 % were unemployed.Regarding their financial position, the respondents reported that 44.48 %, 41.23 %, and 9.42% said they were coping, having a hard time, and living comfortably.In addition, 62.3 % of respondents said they had completed senior high school or less, with 33 % reporting having completed above high school.

Trust in the police
To understand residents' trust in the police, we have asked if residents trust the police in their community to protect lives and property.The numeric data reveals that the study area's population has lower than average levels of trust in the police (Mean: 2.94 SD = 1.131).Approximately 47% of respondents (n = 288) do not think the police can protect people and their property.In contrast, only 29% of respondents reported that the police in their neighborhoods are trusted to protect people and their property.The remaining 17 percent of respondents have yet to decide whether they could rely on the local police to protect people and property.

Police effectiveness in controlling crime and providing services
In most measures of police effectiveness, respondents do not believe that the local police are successful in controlling any crime in their communities.For instance, 40 percent of respondents (n = 252, mean = 3.02 SD = 1.073), 45 percent of respondents (n = 276, mean = 2.90 SD = 1.091), 43 percent of respondents (n = 270, mean = 2.82, SD = 1.072), and roughly 50 percent of respondents (n = 301, mean = 2.65 SD = 1.058) reported that the police in their neighborhoods were ineffective at controlling violent crime, arresting suspects, controlling burglary and theft, respectively.Moreover, a sizable portion of respondents, between 21 and 26 percent, were undecided regarding the dependability of the neighborhood police in policing crime and upholding order.In contrast, local police are rated better on service provision indicators (Mean = 2.96, SD = 1.025) over crime controlling indicators (Mean = 2.84, SD=.71).For instance, 33% and 37% of respondents said, respectively, that the police in their neighborhood immediately respond to emergency calls and are doing an excellent job in responding to locals who have been the victims of crime.

Police procedural justice
Among the determinants of public trust in the police poised in this study are the fair and just treatment of citizens, police legal knowledge, and the proper exercise of authority.The survey results revealed that local police are even rated lower with procedural justice indicators (mean = 2.628, SD=.759) than other police performance indicators.For instance, 60% of respondents (n = 370) said that when the local police stop people and go about their normal business on the streets, they do not handle the encounters well.Regarding how well local police conduct crucial policing encounters with citizens, only 21% of respondents gave local police a positive evaluation.Approximately 50% of respondents (n = 302) disagreed that the police in their neighborhood or district correctly understood and applied the law regarding knowledge and application.Once more, about half of the respondents said that the local police make decisions based on their own biases or opinions rather than on the facts.
Further, 52 % of the respondents reported that the police do not explain their actions honestly to the people they deal with them.Furthermore, 57 % of the respondents (n = 346) reported that most police officers do not use only the amount of force necessary to accomplish their tasks.Finally, above 60 percent of the respondents do not believe the police treat everyone in their neighborhood equally.

Police community relations.
Compared to the previous factors, local police units are favorably rated on police-community relation indicators.For instance, around 60 percent of the respondents (n = 342) replied that the local police in their neighborhood involve the community to discuss crime problems, identify causes and provide solutions.In contrast, 42% of respondents said that the local police collaborate well with the locals to solve problems in their community.Furthermore, when asked how effective community policing programs are in their neighborhood, 46 percent of the study responded that the initiatives successfully lower crime and the fear of crime.In contrast, a third of the respondents claimed that community policing initiatives are ineffective in lowering crime and the fear of crime.

The relationship between trust in police and the independent variables
The Spearman rho enabled us to determine the relationship between trust in the police, and the independent variables, primarily police effectiveness, procedural fairness, and police-community relation.Table 3. shows the descriptive statistics and key correlation results.As shown in the correlation matrix, trust in police has a statistically significant relationship with police effectiveness in controlling crime and providing services (r = .648**),procedural justice(r = .518**),and policecommunity relations(r = .515**).Moreover, police effectiveness in controlling crime (r = 0.579**) and police effectiveness in providing service (r = 0.589**) each had a positive and moderately strong statistically significant relationship with trust in the police.Residents who indicated that the police effectively control crime and provide services have also indicated they trust the local police to protect lives and property.Similarly, residents who reported that local police are effective in crime control and service provisions equally reported that they trust the local police to protect lives and property.
Concerning procedural fairness, local police that discharges their operations in a police routines are the ones whom the residents trust to protect life and property.The police that does not follow the law, are cruel, disdain civilians, and frequently use force to resolve conflicts are the ones who can least be trusted to safeguard people and property.Police are trusted to protect lives and property when they build solid relationships with the community and use those relationships to address problems in the community.
Overall, findings from the correlation analysis demonstrate that increased police effectiveness in controlling crime and providing services, procedural fairness, and enhanced police-community relations leads to higher levels of public trust in the police and vice versa.Thus, the bivariate results support the assumptions of performance theory, procedural justice, and community policing.
Further, trust in police has a positive relationship with a neighborhood crime problem, fear of crime, informal social control, and social cohesion.However, the association between trust in the police and crime and the components of collective efficacy-social cohesion and informal social control-is weaker than the relationship between faith in the police and police effectiveness and procedural justice characteristics.Moreover, crime victimization has a negative and weak relationship with trust in the police, with increasing crime victimization leading to lower trust in the police.As a result, we can argue that the performance and procedural justice theory and community policing assumptions can be used over social and criminogenic assumptions to explain the policepublic trust relationship in Ethiopia and, more specifically, in Addis Ababa.

The effects of police performance, procedural justice and police-community relation on trust in the police
To substantiate the above line of reasoning, ordinal regression was undertaken to estimate the effects of police effectiveness in controlling crime and providing service, procedural justice, and police-community relations on trust in police.Table 4. depicts the results of the ordinary regression coefficient results.As shown in the right-hand column, controlling other variables, police effectiveness, procedural fairness, and police-community relations were statistically significant predictors of trust in the police in the first model.The regression analysis shows that a 50.2% (R 2 = 0.502) change in trust in police is due to variations in police effectiveness, procedural justice, and policecommunity relations.Therefore, the three predictor variables explained approximately 50% of variance in public trust in the police.Three of the same predictors remained statistically significant in the second model even after the addition of the control variables.About 1.4 percent more variance was explained by control factors, however not all of the increases were statistically significant.
Further, the ordinary logistics regression results show that, with the model fitting the data (χ 2 (3) = 194.7,p < .001.), police effectiveness in controlling crime and providing services is the most significant and positive predictor of trust in the police than other factors (odds ratio = 5.869; p < 0.01).Accordingly, those who evaluated the local police as effective in controlling crime and providing essential services are nearly six times more likely to trust the police to protect their lives and properties.The statistical evidence indicated that citizens' perceptions of police effectiveness in controlling crime and providing services substantially and significantly affect trust in the police.
Similarly, procedural justice was found to have a significant causal relationship with trust in the police (odds ratio = 1.667; p < 0.05); those who evaluated the local police as procedurally fair are more likely to trust the police to protect life and property than the respondent who did not perceive the police as procedurally fair.Further, police-community relations are also a significant positive predictor of trust in the police, where those who rated the local police as having a productive relationship with the community are nearly twice as likely (odds ratio = 1.899; p < 0.05) to trust the police to protect life and property than those less rated the police in terms of its relationship with the community.Three of the same predictors remained statistically significant in the second model even after adding the control variables.In addition, social cohesion and crime victimization were discovered to be essential determinants of police trust.Control factors contributed an additional 1.0 percent of the variation, which was statistically significant.To compare the relative impact of each predictor variable on the outcome of the regression model, β weights (standardized regression coefficients) were also calculated.Based on these coefficients, the strongest predictor for public trust in the police was police effectiveness (β = 4.299), followed by police-community relations (β = 1 .535),procedural justice (β = 1.257), and then prior crime victimization (β =.71).These results showed that respondents with a higher level of public trust in the police were those who believed that police officers effectively prevent crime, respond to calls for service, build positive relationships with the community, and apply the law equitably to all residents.
Overall, at all levels of statistical analysis, the association between police performance, police procedural fairness, and police-community relations is strong and statistically significant.Hence, changes in trust in the police are significantly predicted by the police's efficiency, procedural fairness, and community relations.In addition to the effects of the independent variables, three control variables, including crime victimization, neighborhood crime problem, and social cohesion, were also found to predict residents' trust in the police.

Discussion
To further knowledge on public trust in the police, the current study examined how police effectiveness in crime control, procedural justice, and police-community relations affects trust in police using representative data collected in 11 neighborhoods located in three sub-cities of Addis Ababa, Ethiopia.The finding of the study revealed that citizens overall trust in the police-conceptualized as the police's ability to protect persons and property-was found below the average, casting doubts about the local police's ability to protect persons and property.Further, the study finding demonstrated that police effectiveness in controlling crime and providing services, procedural justice, and police-community relations had significant direct effects on public trust in the police.
On the other hand, the public's trust in the police was not significantly correlated with crime victimization, fear of crime, neighborhood perception of crime, or informal social control.These results show that the performance, expressive, and community policing models can be used to explain public trust in the police in Ethiopia.
In particular, the results of the study demonstrated that police effectiveness in controlling crime and providing services is the strongest predictor of trust in the police followed by procedural justice and police-community relations respectively.The relative effect of police effectiveness on public trust in the police was shown to be three times stronger than procedural justice, which is consistent with earlier studies (Boateng, 2017;Lim & Kwak, 2022;Tankebe, 2008).In other words, people who see the police as being more effective are three times more likely to report having positive attitudes and trust in the police.This result was in line with performance theory's premises towards trust in the police, which claims that when police officers efficiently carry out their tasks, citizens are more likely to be satisfied with the police (Bouckaert et al., 2002;Van Craen, 2013).According to the performance theoretical framework, residents' assessments of the efficiency of the police are based on their abilities to combat crime, lessen public fear of crime, and improve neighborhood safety and security (Jackson & Bradford, 2009;Sun et al., 2014).To enhance the public's trust in the police, they must therefore better carry out their fundamental duties of preventing and reducing crime.
This study also provides empirical evidence for the relationship between public perceptions of procedural justice and public trust in the police, as well as the fact that procedural justice is a solid predictor of public trust in the police.In other words, people are more inclined to trust police officers if they see that they are being treated fairly and equitably by the police.Further, the current study established that public perceptions of procedural justice play a crucial part in the normative evaluation process of public trust in the police.Thus, consistent with earlier research in Western societies (Mazerolle et al., 2013;Nix et al., 2015;T. C. O. O'brien & Tyler, 2019), this finding could extend the present understanding of public trust in the police to non-Western nations, including Ethiopia.Moreover, the results of the study are consistent with the procedural justice theoretical framework that people who think the police are procedurally just or fair, that is, police treat people with respect, equally and fairly, and carry out their duties following the law and minimum use of force, also say they trust the police (Tyler & Huo, 2003;Tyler, 2000).
To enhance public trust in the police and be perceived as procedurally fair, police officers should correctly understand and apply the law, make decisions based on facts rather than personal biases or opinions, give honest and legal justifications for their actions, use only the amount of force necessary to complete their tasks, treat everyone equally, and involve citizens in decision-making.Police officers should be aware that while performance and competence are important to buy in public trust, so too are the quality of interactions and the recognition of citizens in fair procedures and treatment.If citizens are treated fairly, with dignity, and respect, fighting crime will be more effective, cost-effective, and moral.When the police are effective, procedurally fair, and concerned with local interests, this would not only make the police more responsible, but would also enhance moral relationships between citizens and the police, and promote active citizen engagement in community safety as well (Jackson & Bradford, 2010).
The current study also presents an empirical basis for the correlation between policecommunity ties and public trust in the police.Those who responded that local police have good relations with the community also indicated they trust in the police.In particular, trust in the police was specifically connected with police misconduct, whether the police consider community feedback when handling situations, and whether they use community policing strategies to reduce crime.Consistent with the results of previous research (Gill et al., 2014;Roberts & Plesnicˇar, 2015), improving police-community relations can improve public trust in the police and their legitimacy.Thus, more investment in police-community relations can enhance public trust in the police.In sum, the study's findings show that more effective and procedurally fair policing, coupled with improved police-community relations, results in higher levels of public trust in the police.
Based on the study's findings, we recommend the following programmatic and practical implications to improve public trust in the police in Addis Ababa specifically and throughout Ethiopia generally.To begin, citizens' trust in the police is not at the level required for healthy policing, which prevents the police from taking advantage of the positive ramifications of trust in police, such as willingness to obey the law, cooperation, legitimacy, and empowerment (Lim & Kwak, 2022;Nix et al., 2015;Tyler, 2000).Thus, police should enhance public trust and confidence by improving their effectiveness in reducing crime and providing services, ensuring procedural justice, and fostering better relationships with the local community.In particular, police should strive to improve effectiveness by controlling violent crime, arresting suspects of theft and burglary crimes and responding to emergency calls for assistance.Police should be well aware that improving actual crime control efficacy alone cannot guarantee public trust in the police; rather, it is a required condition, and one of the fundamental aspects of how the public views police.As Goldsmith (2005:15) noted actual incompetence is not necessary to undermine trust; instead, believing that the police are ineffective or performing poorly is a necessary condition for citizens' negative ratings of the police.Thus, police should also improve public perception of their efficiency by effectively communicating outcomes of crime prevention, arrest rates, and emergency response.
The public's trust in the police was found to be influenced by procedural justice.Therefore, police officers must correctly comprehend and apply the law, base their decisions on the facts rather than their personal biases or opinions, and give truthful and legal justifications for their actions.Police should also treat everyone fairly, use only the amount of force necessary to complete their responsibilities, and consult with the public before making decisions.Studies have shown that fair supervision, strong disciplinary procedures, and training in procedural justice principles can support the maintenance of internal procedural fairness among officers (Van Craen, 2016;Wang et al., 2020).Whereas, Van Craen and Skogan (2017) make a bold proposal to restructure the police organization to support external procedural justice or fair policing practices.In conclusion, police organizations should promote procedural fairness principles among line officers in to improve the level of fairness in police practice and their interactions with citizens (Van Craen & Skogan, 2017).
Consistent with previous research (Gill et al., 2014), this study provides empirical evidence that the relationship between local police and citizens influences perceptions of trust in the police.Thus, to enhance public trust in the police, police should work together with residents in solving local problems, handle victims of crime and complaints with care and courtesy, and involve the community in discussing crime issues, identifying causes and providing solutions.Further, police can also increase public trust by promoting social cohesiveness and enabling the community to participate in various community safety projects (Lim & Kwak, 2022;Sargeant, 2015).This study also recommends the importance of considering police viewpoints on the issues affecting trust relations between the public and the police, as this has implications for the police and the community they serve.

Conclusion
The current study considerably extends our understanding by providing empirical evidence that public trust in the police is affected by police effectiveness, procedural justice, and police-community relations.Thereby, the study established that the assumptions of the performance model, procedural justice and community policing models are applicable in explaining trust in the police in the context of Ethiopia.Despite the contribution, there are several drawbacks that future research should address.First, due to the small sample size and the fact that data was only drawn from Addis Ababa City, Ethiopia, the current study's conclusions can only apply to urban contexts.Further, this study used cross-sectional data, a nonexperimental setting; thus, it could not control and account for temporal variations in the public's faith in the police.Finally, the public's trust in the police is precarious, especially in a nation where the police are firmly linked to the political system and have become an instrument of the regime.Therefore, variables other than police behavior affect how the public perceives the police.
Effective policing depends on the public's trust, which ensures public compliance and collaboration, and empowers police to execute their jobs effectively (Nix et al., 2015;Sunshine & Tyler, 2003;Tankebe, 2008).It has been determined, with considerable regularity, that the public's support, respect, and mutual trust are essential for cooperation and effective law enforcement.Thus, the trust relationship should be dualistic and mutual, with the police needing the public's trust as much as the other way around.Against this backdrop, most studies in the domain of trust relationships are one sided-focusing only on public trust in the police while overlooking police trust in the public and factors that are central to the police that would help us to understand trust comprehensively (see Sherman, 2002;Tyler & Huo, 2003).Research shows that the excessive emphasis placed by academics on public trust in the police runs counter to the fundamental tenet of policing, which holds that the community is the police and the police are the community (Renauer, 2007).Thus, police assessment of public trust in the police and factors that affect this trust relationship should be the focus of future research.
In addition, the present study did not include the factors such as democracy or police misconduct, including police corruption, or political impartiality and their influence on trust relationships.Thus, future research should incorporate the perception of democratic variables and the perception of police corruption as explanatory factors of public trust in the police.Moreover, future studies should employ multilevel analyses to examine cross-level interaction effects of neighborhood and individual factors on public trust in the police.Finally, future study needs to examine the potential ramifications of the public's trust in police officers and police officer's perception of the public.That is, future studies should examine if increased public trust in the police will increase cooperation and compliance with them.In addition, future studies must combine different research environments, such as rural and urban locations, and representative samples with longitudinal data.
The Ethiopian Federal Police, as well as regional police forces, should keep up their organizational reform efforts and give high priority to improving police performance, procedural justice, and relations with citizens if they want to win the hearts and minds of citizens.In essence, trust in the police and public collaboration with the police go hand-in-hand (Jackson et al., 2012;Tankebe, 2008;Tyler & Huo, 2003;Tyler, 2005).

Table 3 . Descriptive statistics and correlations for trust in police and the independent variables
**P <0.01.* P< 0.05.