The Police and Citizens as Co-Producers of Crime Prevention in Johannesburg

: The city of Johannesburg is the largest metropolitan municipality out of eight in South Africa. It is notorious for its high levels of crime, with a crime index of 80.72 and a safety index of 19.28. This article examines crime prevention co-production in the Johannesburg metropolitan area. The purpose of the study was to analyse how co-production of crime prevention is organised in Johannesburg and to understand the effectiveness of the practice there. Collaborations between the regular producers (police) and citizens were assessed to fully determine the potential of crime prevention co-production. This led to qualitative interviews with 20 police ofﬁcers from 10 police stations in the Johannesburg metropolitan area. A citizen survey was conducted with a selected sample from the population (n = 400) residing in the neighbourhoods covered by the police stations investigated. The ﬁndings indicate that police ofﬁcers have devised practices and programmes to co-produce crime prevention. These include the sharing of pamphlets and crime prevention education and awareness during school visitations, the use of Zello technology, community police forums, and information sharing at the youth desk and in WhatsApp groups. Nevertheless, these practices do not seem to be popular among the citizens. There is a contradiction in the perception of successful crime prevention co-production between citizens and the police. This is mainly due to citizens’ lack of trust in the police. The study acknowledged the importance of the work carried out by the police in crime prevention co-production but also observed the citizens’ lack of trust in the police as a potential threat to crime prevention co-production. Trust is crucial in managing information sharing on crime prevention. Front line professionals, such as the police, will not be able to meet future crime challenges if there is a trust deﬁcit. Hence, it is important to restore trust in the work carried out by the police. Professionals in public administration could learn about the importance of trust in their crucial role of implementing policies


Introduction
The challenge of crime in South Africa remains a pertinent issue and is exemplified by the World Population Review ranking the country as having the third-highest crime rate globally-its crime index is 77.29% (United Nations 2021). The City of Johannesburg Metropolitan Municipality (hereafter referred to as Johannesburg) contributes significantly to this rate.
In 2021, the crime index in Johannesburg was rated at 80.54% (Numbeo 2021). The profound sense of vulnerability among the people and the challenges citizens face regarding their safety validate these statistics. According to the Constitution of the Republic of South Africa, No. 108 of 1996, Chapter 7, Section 152 (d) and (e), respectively, the objectives of local government are to "promote a safe and healthy environment" (The Republic of South Africa 1996) and to "encourage the involvement of communities in community matters of local government" (The Republic of South Africa 1996). Some argue that reducing inequality in violent cities and neighbourhoods helps to reduce violence (Gleixner et al. 2021). Neighbourhoods exhibiting higher levels of inequality and concentrated disadvantage Adm. Sci. 2023, 13, 138 2 of 19 often experience significant mistrust, social disorganisation, and crime. The relationships between these communities and the government, especially law enforcement, are often troubled. With the high crime rates in South Africa, there is a clear indication that law enforcement agencies, such as the police, need to find ways to meaningfully co-produce safety and crime prevention with citizens. Involving the citizens in such prevalent issues in South Africa should be more than merely desired.
The concept of involvement would promote organisational strategies that support partnerships between the police and communities to proactively address issues such as social disorder, crime, and violence (U.S. Department of Justice 2014). In support of this, the shared view of the local government is that collaboration is necessary for the successful performance of public institutions. Collaboration can be used not only as a problemsolving mechanism for multifaceted security issues but also to enhance the institution's performance in dealing with the issues (Lira 2016). Inequality and highly organised crime in Johannesburg necessitate solutions to enhance police collaboration with citizens and improve community safety in the long haul.
This study investigated the perceptions of regular producers (the police) and citizens on crime prevention co-production to contribute towards a safer and more secure Johannesburg. Citizens were also asked about their involvement in crime prevention strategies because they are considered co-producers of safety and crime prevention (Nabatchi et al. 2017).
Interest in co-production has existed since the 1970s and early 1980s. At its inception, co-production focused on the explanations and theoretical foundations of practices that focused on the involvement of citizens in the provisioning of public services (Ostrom 1972(Ostrom , 1996Ostrom and Ostrom 1977;Parks et al. 1981;Percy 1978). The concept is currently one of the high-research areas in public administration and management due to the revival of interest in citizen involvement in public policy and service delivery (Brandsen et al. 2018). The research questions guiding this article are: (1) How is co-production of crime prevention organised in Johannesburg? (2) What are the factors that influence the co-production of crime prevention in Johannesburg?
The following section reviews the theoretical framework reinforcing the relationship between regular producers, citizens, and security provisions. Observations obtained through the textual analysis of books, journals, and articles were used to explore the current bodies of knowledge contributing to the existing gaps in neighbourhood security. Consequently, this research article used both quantitative and qualitative approaches, and the methods are fully described in the Materials and Methods section. The qualitative research utilised a case study design, and a content analysis of the transcripts from unstructured interviews conducted with various police officers was employed. A logistic regression analysis was conducted from the data collected during the Johannesburg citizen survey. Finally, the results are discussed, and potential recommendations are offered.

The Concept of Co-Production
Alford (1998, p. 128), defines co-production as "The involvement of citizens, clients, consumers, volunteers, and/or community organisations in producing public services, as well as consuming or otherwise benefiting from them". Similarly, Brandsen et al. (2018, p. 1) suggest that co-production is "the mix of activities that both public service agents and citizens contribute to providing public services. The former is involved as professionals, or 'regular producers', while 'citizen production' is based on voluntary efforts by individuals and groups to enhance the quality of the services they use".
Co-production, in the context of this article, can be defined as an inclusive approach adopted by the police to actively involve citizens in the processes and protocols aimed at reducing crime and violence. Nikolakis (2020) and Bell and Pahl (2018) argue that co-production is suitable for dealing with crime prevention because it has been viewed in a variety of ways, including as a way to solve complex societal problems, reveal fundamental agendas for the future that Adm. Sci. 2023, 13, 138 3 of 19 have previously been obscured by "cognitive exclusion", and provide a "utopian space" for imagining futures that are, for example, "free of structural inequalities".
Central to the definitions of co-production is that citizens and government institutions are involved. This distinguishes it both from government services without active citizen involvement and from citizen self-organization. Bovaird (2007) suggests that full co-production occurs only when individuals and the community are highly involved in commissioning and delivery functions. Citizens are co-producers endowed with knowledge, resources, assets, and capabilities that can be used to create more public value.
Furthermore, the Trade Union Congress (2013) identifies co-production as the cocommissioning of services, including co-planning, co-prioritisation, co-design, and codelivery, which covers co-managing, co-performing, and co-assessment. In this context, co-commissioning refers to those activities that are aimed at strategically identifying and prioritizing needed services, outcomes, and users. Co-design refers to activities that incorporate "the experience of users and their communities" into the creation, planning, or arrangements of public services. Co-delivery refers to joint activities between state and lay actors that are used to directly provide public services and/or to improve the provision of public services. Lastly, co-assessment focuses on monitoring and evaluating public services.
Authorities such as Needham andCarr (2009), Bovaird (2007), and Sicilia et al. (2016) contend that co-production is applied across the phases of the service cycle, with regular producers and communities working together at any stage to "produce" something of value. Jakobsen and Andersen (2013) argue that problems have become increasingly complex and solutions less evident in today's society. They argue that, in such a scenario, co-producing public services could be the way forward.
Furthermore, Boyle and Harris (2009) argue that co-production is meant to improve service quality, responsiveness to customer needs, and client satisfaction and make the services more responsive and cost-effective. According to Sicilia et al. (2016, p. 11), the development of co-production is based on the following: "the attempt to improve public service quality by bringing in the expertise of users and their networks; the need to provide public services that are better targeted and more responsive to users; the possibility of using co-production as a way of cutting costs; the opportunity to create synergies between government and civil society with a positive impact on social capital". Pacheco et al. (2017) expounded on the numerous benefits of co-production. These include the fact that it promotes a more remarkable ability among service providers to reach the root of citizens' need to develop citizen-focused solutions; facilitates better user satisfaction; enhances the creation of more cohesive communities with a greater sense of local ownership; builds the confidence and capacity of individuals and communities; enhances the better use of public resources; and empowers citizens to take control of their lives and the areas where they live. Brandsen and Honingh (2016) state that co-production focusing on implementation and design should directly involve citizens in producing the core services of an organisation. This implies that co-production may be an integral part of the successful implementation of the productive improvement of public services. Osborne and Strokosch (2013) state that co-production could be a way to deliver public services with the maximum feasible community participation. Boyle and Harris (2009) suggest that co-production can rebuild and redevelop the core economy, realise its potential, and rediscover human resources by recreating social networks. Therefore, co-production aims to shift the balance of power by involving people in the subsequent service delivery.
In terms of encouraging collaborations, international best practices have been or are currently implemented to obtain better public safety and order outcomes. According to Loeffler (2018), restorative justice in countries such as the United States of America, Canada, and Australia involves previous offenders and community public service volunteers in co-producing solutions to resurrect relationships with local citizens affected by illegal or incongruous behaviour.
Additionally, in the Netherlands, the police have worked on a new system called the Burgernet that draws citizens into active engagement with police work. This system works because citizens sign up and give their home or work addresses, enabling the police to contact them immediately after a crime has occurred. Citizens may receive a message from the police regarding help attending to a crime or offering reassurance that a matter has been resolved (Loeffler 2018). In other jurisdictions, co-production is valuable because it contributes to making public service delivery more democratic, efficient, and effective. It has been proven to foster satisfaction with service delivery and trust in the government (Fledderus 2015).
While co-production is widely considered a plausible alternative solution to service delivery, there have been concerns about the dark side of co-production. Mangai and de Vries (2019) indicate that many factors, such as governance failures, inadequate human and financial resources, and poverty in some African countries, have constrained public officials to co-produce public services together with citizens.
A further concern with co-production lies in the possibility that regular co-producers are taking advantage of citizens. This was revealed by Matosse (2013), who equated this dark side of co-production with tokenism, where citizens are consulted but their views are not adopted. Mbhele (2017) argued that co-production has only provided symbolic participation.
Further, Mbhele (2017) contends that, in some instances, public officials are incompetent and/or unwilling to be held accountable for the services provided to the citizens. Moreover, according to Molaba (2016), government officials are often unwilling to give up control over directions as they do not acknowledge the importance of citizens' views because they consider themselves to be the experts in their fields.

The Concept of Security
With this article's objective of analysing co-production aiming to enhance safety and security provision, a definition of security should be applied. According to Krahman (2008), security is defined as "the endurance of a threat that has the potential to become a reality". Krahman (2008) suggested that the security service is more likely to shift from an individual level to a collective level, allowing citizens within communities to eradicate crime and violence actively.
Westermeier and Nolte (2020) indicated a comprehensive range of contributions to critical security studies, revealing several ways in which security provisions depend on both public and private actors. They (2020) further postulated that complex security networks that bind public and local stakeholders should be understood as developing co-production when studying public-private security engagements.
However, it is essential to acknowledge that security and a constant police presence help sustain capital accumulation. This gives rise to the commodification of security. Providing security and security practices should thus lie with the state and be a public good rather than a citizen commodity.
According to Africa (2015), human security has yielded various results in South Africa. Applying a wider definition of security would further expand the interpretation of where security practices should be considered to frame national security. The use of the National Intelligence Agency to reflect issues of crime and social protest is a worthy area of interest. Despite these developments, arguments have suggested that the South African government has been heavy handed over the past few years. Africa (2015) alluded to issues such as neglecting the demands of citizens, the unjust use of force, lack of trust, cases of excessive police brutality, diminishing democratic principles, and the decline of parliamentary oversight structures. Africa (2015) further added that violence is prevalent in poorer and marginalised communities, which has exposed several problems in South Africa. These include the inequality that still exists, the structural violence that continues to lead to violence among citizens, and how secluded South Africans remain from the rest of Africa.
To bring these issues into the context of Johannesburg, the practice of law enforcement should consider the Batho Pele principles. The Batho Pele principles encapsulate the idea of "people first". They are described in a white paper promoting the transformation of public service delivery. These principles echo the values of co-production and aim to foster law enforcement through partnerships with the community (Roberts and Hemson 2008).
The most notable Batho Pele principle that speaks directly to co-production is consultation. This principle entails consultation among stakeholders about service delivery. In the context of security, consultation and collaboration with communities can occur through several avenues. Firstly, customer surveys that enquire about the current state of security within Johannesburg can be used. Then, campaigns aimed at addressing crime and violence issues can be implemented. Finally, workshops that could potentially offer alternative solutions to deal with a crime can be introduced (Department of Public Service and Administration 2014).
Lessons can be learned from Japanese private security companies (PSCs) and their approach to dealing with security. There is encouragement from both the state and police for PSCs to engage in security practices for which the state is not held responsible. Therefore, the main priority for PSCs is to establish connections with the police and communities and develop trust to sustain safer living conditions. This essentially promotes cooperation from all government stakeholders in ensuring that safety remains a priority in Japan. Specific target groups are recognised depending on the crime level in Japan. For example, almost every school in Japan has measures to safeguard its pupils and staff against any potential perpetrators of crime (Polak-Rottmann 2018).

Results
This section presents and discusses the study's results. Firstly, we present the key results from the Johannesburg citizen survey. These include citizens' perceptions of the levels of crime and violence, policing services, and their involvement in crime prevention co-production.
Secondly, the qualitative results present the experiences and views of police officers in Johannesburg on how the co-production of crime prevention is organised or should be organised to realise safer neighbourhoods.

Citizens' Perceptions of the Level of Crime and Violence
In many security studies, the questions related to the perceptions of crime and violence are often generic. We asked the respondents: "How do you think the levels of crime and violence in your area have changed in the last 12 months?" We expected the responses to reflect our assumption that respondents may have experienced crime first-hand or witnessed or perceived certain levels of crime and violence in their area. The responses show the levels of crime and violence in the different neighbourhoods. Figure 1 shows the distribution of the respondents' perceptions of the levels of crime and violence.
Based on Figure 1, the responses were almost evenly distributed, as 36.36% perceived crime and violence to be on the increase, 38.71% perceived them to be constant, and 24.93% thought they were decreasing. This shows that many respondents had yet to feel the impact of policing in their neighbourhoods.
As revealed above, neighbourhoods such as Alexandra, Cleveland, and Langlaagte, which are geographically disadvantaged as a result of the apartheid regime, continue to witness increased levels of crime and violence. The informal settlements in these neighbourhoods are a breeding ground for criminals, since it is difficult to track them without an official address, and their poor infrastructure and lack of community make co-production difficult. As revealed above, neighbourhoods such as Alexandra, Cleveland, and Langlaagte, which are geographically disadvantaged as a result of the apartheid regime, continue to witness increased levels of crime and violence. The informal settlements in these neighbourhoods are a breeding ground for criminals, since it is difficult to track them without an official address, and their poor infrastructure and lack of community make co-production difficult.
One would expect crime to decrease in Norwood, which is considered an affluent neighbourhood with a multiplicity of actors in the co-production of security. In South Africa, many such neighbourhoods are largely responsible for their own security. Residents of such areas use private security companies and neighbourhood watches to protect their lives and properties. The increased crime and violence in Norwood needs attention and investigation. In a nutshell, the data depict crime and violence as increasing or remaining the same in many of the investigated neighbourhoods, which should be a matter of concern.
In view of the above perceptions, we studied the respondents' satisfaction with the South African Police Services (SAPS). Respondents were asked if they were satisfied with the SAPS in their area. Table 1 shows 98 cases of satisfied respondents, representing 28% of the responses, and 227 cases of unsatisfied responses, which means that most respondents were generally not satisfied with the service provided by the police.  One would expect crime to decrease in Norwood, which is considered an affluent neighbourhood with a multiplicity of actors in the co-production of security. In South Africa, many such neighbourhoods are largely responsible for their own security. Residents of such areas use private security companies and neighbourhood watches to protect their lives and properties. The increased crime and violence in Norwood needs attention and investigation. In a nutshell, the data depict crime and violence as increasing or remaining the same in many of the investigated neighbourhoods, which should be a matter of concern.
In view of the above perceptions, we studied the respondents' satisfaction with the South African Police Services (SAPS). Respondents were asked if they were satisfied with the SAPS in their area. Table 1 shows 98 cases of satisfied respondents, representing 28% of the responses, and 227 cases of unsatisfied responses, which means that most respondents were generally not satisfied with the service provided by the police. Those who said that they were not satisfied with the SAPS in their area were further asked to give their reasons. Their responses were used to conduct the stepwise logistic regression analysis below. The listed reasons were as follows: (1) they do not have enough resources, (2) they are lazy, (3) they are corrupt, (4) they do not come to the area, (5) they release suspects early, (6) they cooperate with thieves/criminals, (7) they are harsh towards victims, (8) they never recover goods, (9) they do not respond on time, (10) they are gender-insensitive/intolerant, and (11) they are disability-insensitive/intolerant.
Using the stepwise logistic regression analysis, the important reasons were selected as the main reasons why the respondents were not satisfied with the SAPS in their area. Table 2 shows the results of the stepwise regression analysis of the relationship between satisfaction with the SAPS and factors informing the respondents' lack of satisfaction. A stepwise regression keeps only statistically significantly descriptive or independent variables. Thus, the three variables listed in the table below were statistically significant. Corruption among the police was the main reason why the respondents were unsatisfied. Community members were 5.0 times more likely to be unsatisfied when the police became involved in corruption than when they did not. A unit incident of corruption among the police resulted in 1.6018 cases of dissatisfaction among residents. Visible policing was positively related to community satisfaction, and releasing suspects early slightly increased dissatisfaction with the police. The research revealed that, in relation to bribery and extortion, police officers solicit bribes from members of the public when ordinary persons seek the protection of the police or when criminals seek ill-gotten freedom. Further analysis could be conducted to determine the satisfaction levels in each neighbourhood, as there may be varying differences in community satisfaction per police station neighbourhood. However, the sample size did not allow us to analyse neighbourhood-level satisfaction.

Citizens' Involvement in Crime Prevention Co-Production
There must be a strong relationship and mutual trust between communities and policing agencies to maintain public safety and effective policing. Lately, South Africans have witnessed widespread looting and the destruction of economic infrastructures, including the burning of malls and trucks, vigilantism, and the failure of the South African Police Services to contain these situations. This section explores the relationship between the community and the police in terms of knowledge about police services. Table 3 shows that the majority of the community know where the nearest police station is located. This means that the community has an idea of how far police help is when needed. Some may be discouraged from calling for police help if the nearest police station is perceived to be too far. A critical component of co-production is active citizen involvement in public services. As shown in Figure 2 below, when respondents were asked, "are you actively involved (i.e., rendering any form of help) to ensure that your neighbourhood is safe?". most respondents from all of the ten selected police station neighbourhoods admitted that they did not actively involve themselves in ensuring that their neighbourhood is safe. In Bramley, about 23.68% of the respondents said that they actively involved themselves in ensuring that their neighbourhood is safe, and 39.47% did not know whether they actively involved themselves in ensuring that their neighbourhood is safe. Bramley is an affluent neighbourhood, and many private security companies are situated there, which may explain these responses. Respondents from the remaining neighbourhoods were largely not involved in policing their neighbourhoods. The following section examines the regular co-producer perspectives on crime prevention. Adm. Sci. 2023, 13, x FOR PEER REVIEW 8 of 20 needed. Some may be discouraged from calling for police help if the nearest police station is perceived to be too far. A critical component of co-production is active citizen involvement in public services. As shown in Figure 2 below, when respondents were asked, "are you actively involved (i.e., rendering any form of help) to ensure that your neighbourhood is safe?". most respondents from all of the ten selected police station neighbourhoods admitted that they did not actively involve themselves in ensuring that their neighbourhood is safe. In Bramley, about 23.68% of the respondents said that they actively involved themselves in ensuring that their neighbourhood is safe, and 39.47% did not know whether they actively involved themselves in ensuring that their neighbourhood is safe. Bramley is an affluent neighbourhood, and many private security companies are situated there, which may explain these responses. Respondents from the remaining neighbourhoods were largely not involved in policing their neighbourhoods. The following section examines the regular co-producer perspectives on crime prevention.

Regular Producers' Experiences and Perceptions of Crime Prevention Co-Production in Johannesburg
This section presents the experiences and perceptions of the police regarding the involvement of citizens in crime prevention co-production. The police enumerated several programmes and initiatives for preventing crime and violence. Figures 3-5 provide an elaboration of the practices used by the police to involve citizens in combating crime and violence in Johannesburg.

Regular Producers' Experiences and Perceptions of Crime Prevention Co-Production in Johannesburg
This section presents the experiences and perceptions of the police regarding the involvement of citizens in crime prevention co-production. The police enumerated several programmes and initiatives for preventing crime and violence. Several existing programmes and initiatives were identified based on the responses from the police officers in the unstructured interviews. According to the police officers interviewed, these programmes and initiatives are the practices implemented by the police to reduce the crime levels in Johannesburg.
A common denominator in the transcripts was the emphasis placed on education and collaboration with the community. Education currently exists in the form of awareness programmes aiming to help the community take better precautions, such as not walking alone at night, alerting the community of any suspicious behaviour in the area, and constant communication with various crime and safety forums. Several existing programmes and initiatives were identified based on the responses from the police officers in the unstructured interviews. According to the police officers interviewed, these programmes and initiatives are the practices implemented by the police to reduce the crime levels in Johannesburg.
A common denominator in the transcripts was the emphasis placed on education and collaboration with the community. Education currently exists in the form of awareness programmes aiming to help the community take better precautions, such as not walking alone at night, alerting the community of any suspicious behaviour in the area, and constant communication with various crime and safety forums. Progress has been made in certain areas, such as community policing forums, meetings held with the citizens, and the development of awareness programmes. Such awareness programmes include "the youth desk", which informs young citizens about the issues of crime and violence in Johannesburg.
The youth desk initiative is a volunteer-based structure located at different police stations in Johannesburg. This initiative was created to enable and encourage young people to identify youth-related crime and violence. In addition, it also allows the youth to collaborate in developing social crime prevention strategies for the communities of Johannesburg. As gathered from the transcripts, this initiative was carried out through crime awareness campaigns in the form of public meetings, school talks, and debates.
The police officers also referred to their involvement with the Community Police Forums (CPFs). They claimed that they collaborate with the CPFs to further inform them on the state of security throughout the neighbourhoods in Johannesburg. The CPFs consist of individuals from the community who partner with the police on matters relating to safety and security. The police have monthly meetings with the CPFs to receive reports and feedback and to discuss challenges and solutions to crime and violence. Another partnership the police referred to was that with the Johannesburg Metropolitan Police Department, with whom they share information and cooperate to mitigate crime and violence. Progress has been made in certain areas, such as community policing forums, meetings held with the citizens, and the development of awareness programmes. Such awareness programmes include "the youth desk", which informs young citizens about the issues of crime and violence in Johannesburg.
The youth desk initiative is a volunteer-based structure located at different police stations in Johannesburg. This initiative was created to enable and encourage young people to identify youth-related crime and violence. In addition, it also allows the youth to collaborate in developing social crime prevention strategies for the communities of Johannesburg. As gathered from the transcripts, this initiative was carried out through crime awareness campaigns in the form of public meetings, school talks, and debates.
The police officers also referred to their involvement with the Community Police Forums (CPFs). They claimed that they collaborate with the CPFs to further inform them on the state of security throughout the neighbourhoods in Johannesburg. The CPFs consist of individuals from the community who partner with the police on matters relating to safety and security. The police have monthly meetings with the CPFs to receive reports and feedback and to discuss challenges and solutions to crime and violence. Another partnership the police referred to was that with the Johannesburg Metropolitan Police Department, with whom they share information and cooperate to mitigate crime and violence. According to some of the uniformed police officers involved in these partnerships, some of the partnerships urgently require reinforcement to function effectively.
According to some of the uniformed police officers involved in these partnerships, some of the partnerships urgently require reinforcement to function effectively. Based on the interviews conducted with the police officers, three main communication tools were identified: WhatsApp groups, meetings, and pamphlet distribution. According to the police officers, WhatsApp groups are used to notify patrolling officers and citizens of any suspicious behaviour in the area. The officers indicated that the WhatsApp groups were also used to coordinate CPF meetings. In one of the stations, an officer suggested that meetings were usually held weekly or monthly. This communication tool worked with the training and collaborative approaches (Figure 3) to make citizens aware of how co-production aims to combat crime and violence in Johannesburg. Another police officer in another station indicated that they held meetings with property owners during the Greater Maboneng Security Campaign. The purpose of these meetings was to identify different strategies to improve security. Pamphlets were distributed to invite citizens to participate in a programme on security enhancement in their neighbourhoods. Based on the interviews conducted with the police officers, three main communication tools were identified: WhatsApp groups, meetings, and pamphlet distribution. According to the police officers, WhatsApp groups are used to notify patrolling officers and citizens of any suspicious behaviour in the area. The officers indicated that the WhatsApp groups were also used to coordinate CPF meetings. In one of the stations, an officer suggested that meetings were usually held weekly or monthly. This communication tool worked with the training and collaborative approaches (Figure 3) to make citizens aware of how co-production aims to combat crime and violence in Johannesburg. Another police officer in another station indicated that they held meetings with property owners during the Greater Maboneng Security Campaign. The purpose of these meetings was to identify different strategies to improve security. Pamphlets were distributed to invite citizens to participate in a programme on security enhancement in their neighbourhoods.
Although these meetings and programmes are reasonable, the police officers professed that they do not receive information about crime as they should. Citizens are afraid to report crime issues in their neighbourhoods as they fear becoming victims once the criminals are released from police custody. The police also admitted that information about the whistle-blowers is often released to the perpetrators of crimes, thereby endangering the lives of the citizens. As indicated in the stepwise logistic regression model in Table 4, the police confirmed that the trust between the police and citizens is eroding due to police corruption and betrayal of confidence. Trust appears to be a major challenge in crime prevention co-production since the relationship is meant to be an equal partnership. More trust leads to more crime prevention co-production, while less trust results in less crime prevention co-production. A similar study (Fledderus 2015) found that trust is correlated to co-production. Figure 5 refers to the existing and potential crime prevention co-production programmes implemented, including visible policing, patrolling, and neighbourhood watches. It illustrates practices leading to the co-production of crime prevention, including community training, creating high-tech security companies, and harnessing sustainable partnerships between the police and citizens. Existing initiatives and programmes that could be refined to achieve greater security are also illustrated. These initiatives and programmes include refining community forums, neighbourhood watches, patrolling, and visible policing.
According to a participating police officer, members from the police stations patrol different areas in partnership with the community policing forums. These areas are segmented into what they refer to as sectors. A police officer is responsible for each sector to ensure effective patrolling. Another officer indicated that community members occupy the role of patrollers, and they open lines of communication with the police stations to offer updates on crime-related behaviour.
Since this article focuses on co-production, analysing the personal links to an individual in terms of profiling during patrol is necessary. For example, in Lyon, when the police close in on identifying a culprit and realise it is someone familiar to them who is not a criminal, they will stop their search immediately. This approach of stopping and searching citizens is more prevalent in Germany, and ethical profiling is even more pronounced in the West (Maillard et al. 2016).
Although this practice may be viewed as a human rights infringement, the backing power of conducting ID checks is based on the principle of reasonable suspicion. Therefore, in the context of Johannesburg, this could be highly effective, especially with patrolling and visible policing highlighting any suspicious criminal behaviour. This course of action naturally requires legal refinement and legislative development, but with such a practice, identifying potential or current perpetrators of crime and violence can be made much easier. If police officers in Johannesburg engage in a more cooperative approach with citizens, then a certain degree of trust and loyalty can be developed.
To further elucidate crime prevention co-production efforts, an officer in one of the participating stations indicated that one of the highlights of crime prevention co-production was the Women's Day forum addressing issues of gender-based violence. They emphasised that such initiatives go a long way in strengthening the relationship with the citizens.
According to one police officer, a training programme is offered to unemployed youths to train them to become certified security guards. The security guards receive stipends for their services, but their tenure is not sustainable due to a lack of financial resources. As a result, many have been disengaged, and the police fear that they could contribute to crime and violence, having obtained the necessary training and tactics while being unemployed at the same time. In some instances, these security guards have gained employment in the police service as opportunities open up for more recruitment into the police service.
The police officers expressed their views on potential strategies for improving security. They indicated that creating high-tech security companies is a viable option, especially with the movement toward the Fifth Industrial Revolution. This ties in with investment in security to improve communication with and protection of communities in Johannesburg. In conjunction with this, a police officer in one of the participating stations referred to an application (app) called Zello. This app is an alert tool for the police, CPFs, and private security guards to report any suspicious behaviour observed during patrol. This app is not widely accessible, especially to the public, and it was used by only one of the police stations in the sample. The extension of the app's use to other stations could yield meaningful progress in policing Johannesburg and other parts of South Africa. The app could be further enhanced and programmed to be data-free so that every citizen could access it.
Despite some of these laudable initiatives, the police officers thought that there was inadequate involvement and participation among community, government, and nongovernmental organisations (NGOs) in crime prevention co-production. Based on this, increased partnerships among these stakeholders are required to achieve the desired results in crime prevention. Essentially, attention should shift towards developing more inclusive policies and programmes to create a larger policing network with citizens to combat crime and violence.

Materials and Methods
Mixed research methods were adopted to present the outcomes of this study. The study employed a qualitative approach using unstructured interviews with police officers at various stations within Johannesburg and a quantitative approach using a citizen survey.

Research Design and Data Collection
A case study research design was used to answer the main research question. The study sample included relevant members from various police stations in Johannesburg, such as station commanders and uniformed police officers. The key interviewees were drawn from ten police stations around the Johannesburg metropolitan area.
The qualitative study used the purposive sampling technique to allow the researchers to choose respondents with knowledge and experience in crime prevention (Campbell et al. 2020). Station commanders are the custodians of the police stations and could be considered key decision makers, whereas the visible policing department is responsible for face-to-face and social media interaction with citizens. The interaction seeks to explore the visual culture of contemporary policing and its impact on citizen-focused programmes and partnerships. The visible policing programme includes crime prevention, specialised response services, intelligence gathering, social development, school education, and awareness campaigns.
Semantic links are presented in the Results section to show the relationship between the different categories in answering the research questions. Different themes that relate to co-production were coded and developed from the transcripts to answer the research questions. Each interview took an hour or slightly longer to complete, and it was transcribed verbatim and uploaded in Atlas.ti 23 software for analysis. The generation of codes, categories, annotations, and networks was easily facilitated using the software compared to the manual way of analysing qualitative data. Open and list coding were used to produce a network of relevant codes and quotations, and the network was used to visualise the results. The themes were developed to break down the current relationship in implementing crime prevention co-production in Johannesburg, what could be refined, and the potential solutions to enhance the co-production of safer neighbourhoods in Johannesburg.
From the quantitative study, 10 police station neighbourhoods in the city of Johannesburg were purposively sampled, and 40 respondents per police station neighbourhood were interviewed (n = 400) to study citizen crime prevention co-production in their neighbourhood. The purposive selection of the neighbourhoods was employed to mitigate the existing geographical inequality created during the apartheid regime. This means that there are rich and poor suburbs, and co-production is less likely to occur in the rich suburbs because the rich have access to private security, while the poor depend on the police for their safety.
Research assistants familiar with the terrain and involved in other citizen surveys were recruited and trained in how to administer the questionnaire. To ensure the adequate representation of respondents, the research assistants were provided with residential maps of the neighbourhoods with the police stations identified on the map. The starting point to administer the questionnaire was the first building to the north of the sampled station. The first adult in a household to approach the research assistant was expected to respond to the questionnaire. If the adult objected to being interviewed, the research assistant would ask for the adult with the most recent birthday to be interviewed-this method was selected to ensure that the respondents were randomly selected. The research assistant would then count ten houses away from the first house before administering another questionnaire. This process was repeated until the 40th questionnaire was administered. Each questionnaire took an hour to complete. A pilot study was necessary for the research assistants to familiarise themselves with the questionnaire and to ensure that the questions were not too complex. Several topics were considered, including citizens' experience of or opinions on crime and violence levels, citizens' input on crime prevention in their neighbourhood, the interaction between citizens and the police, and the involvement of citizens in the crime prevention co-production in their neighbourhood.

Data Analysis
As mentioned earlier, we analysed quantitative data focusing on responses to questions about citizens' experience of neighbourhood security in the selected police stations' neighbourhoods. The analysis used both descriptive and inferential statistics to understand the factors informing citizens' experiences of or opinions about their neighbourhood security. A stepwise logistic regression model on citizen satisfaction with police services was employed to further understand the odds of people's satisfaction with the police services, which would be indicative of the performance of the police. With regard to the qualitative data, a content analysis of the transcripts from the unstructured interviews conducted with various police officers was completed. Table 5 shows a sample size of 400 respondents stratified into 10 police station neighbourhoods with a unit response rate of 358 out of 400, with two strata in the form of Cleveland and Yeoville having the lowest response rate of 5.13. Among the units, there were further non-responses and varying response rates. Structural edits were performed on the data to ensure reliability.  Table 4 provides the descriptive statistics for the respondents from the population in the study. This information is important as it sheds light on the demographic composition of the respondents and, thus, the section of society represented by them. The table indicates that 75% of respondents were male and 25% were female. About 71% of these respondents were single, with 59% having matric/high school as the highest level of education. An important variable in the demographics described how the respondents considered the living conditions in their neighborhoods. Almost half of the respondents (41%) said that their neighborhood was neither good nor bad. This was indicative of the inclination citizens had about where they reside. Due to the city's apartheid history, the spatial sorting of the population is by income, and this correlates with the economic resources and opportunities.

Limitations
The study could have benefitted from a larger population sample and a more balanced gender grouping, but the stratified randomization would mitigated selection bias in some way. That said, the citizen survey was specific to the selected police station neighborhoods and may not be fully applicable to other contexts. Studies that focus on a specific context or region may not be applicable to other contexts, and the findings may not be applicable to other populations. Despite these potential critiques, empirical studies are essential for advancing our understanding of successful co-production and the practices involved.

Discussion
The study found that police officers have devised practices and programmes to coproduce crime prevention. These include the sharing of pamphlets and crime prevention education during school visitations, the use of Zello technology, community police forums, and information sharing at the youth desk and in WhatsApp groups. Nevertheless, these practices do not seem to be popular among the citizens. There is a contradiction in the perception of successful crime prevention co-production between citizens and the police.
Most citizen respondents from all ten of the selected police station neighbourhoods, when asked about their level of involvement, said that they were not actively involved in making their neighbourhoods safe-consequently, co-production of security services is limited in the view of the citizens. This contradicts the core of co-production as described by Bovaird (2007), who suggested that full co-production occurs only when individuals and the community are highly involved in commissioning and delivery functions.
Additionally, the findings of the stepwise regression analysis revealed that the respondents' lack of satisfaction with police services was due to the fact that the police released suspects early, were invisible to the community, and engaged in other corrupt practices, despite the fact that several existing programmes and initiatives were identified based on the responses from the police officers in the unstructured interviews. This study showed several existing protocols that could be revisited and redeveloped to enhance security in the city of Johannesburg. The plausible solutions gained from the police officers' responses suggest that community training, creating high-tech security companies, and forming partnerships should be seen as potential solutions for crime prevention co-production. Other practices associated with co-production include visible policing, patrolling, partnerships, and community policing forums.
Trust was a critical variable identified in the data obtained from both citizens and police officers. The police also confirmed that the trust between the police and citizens is eroding due to police corruption and betrayal of confidence. Lira (2016) discussed the benefits of co-production, including developing a higher level of trust among citizens and greater resource contributions, based on the evidence from the province of Gauteng. It is important to prioritise trust if crime prevention co-production is to be successful in Johannesburg. The police need the cooperation of the citizens to achieve better policing, as the citizens will feel safer and collaborate more if they are able to trust the police with information. The greater the diversity of co-production is, the more effective collaboration processes can be in reducing transactional expenses. The employment of co-production will also yield success in enhancing efficiency, stimulating growth in relationships between stakeholders, and encouraging participant learning. Therefore, co-production should be further explored in policing Johannesburg.
As outlined in the Joburg 2040 Growth and Development Strategy paradigm, there is a desire to confront Johannesburg's realities. This strategy defines clear outcomes against which progress can be measured and attained. With Johannesburg being excessively unsafe and having excessive inequality, transformation is urgently needed. A driving force in alleviating this would be reducing crime through citizens and police officials working together. Another avenue for the city to develop greater security would be for city officials to become closer to their communities. Collaboration between communities and city officials within the metropolitan area would encourage trust, as expressed earlier. The current collaboration systems involving communities and city officials in Johannesburg are neither sufficient nor deliberative in practice (City of Johannesburg Metropolitan Municipality 2000).
Based on the practices and programmes indicated by the various police stations in Johannesburg, the state of co-production in Johannesburg can be described as progressive. Although many programmes and initiatives have been developed, the citizens do not see enough tangible outcomes. The results of the citizen study contradict those obtained from the police officers. There is room for improvement, and refinements can still be introduced to enhance co-production with citizens. The onus also falls on the government to strengthen the resources and infrastructure provided to the Johannesburg police officers so that current initiatives and programmes can be refined to obtain cohesive crime prevention co-production.

Conclusions
The research identified crime prevention co-production practices among the regular producers as including: the sharing of pamphlets and crime prevention education during school visitations, the use of Zello technology, community police forums, and information sharing at the youth desk and in WhatsApp groups. These practices would have been considered as successful co-production but they fell short of the actual meaning of coproduction, which acknowledges citizens or groups of citizens as lay actors on the "co-" side of co-production (Nabatchi et al. 2017;Brandsen et al. 2018;Brandsen and Honingh 2016).
The research also revealed that trust between the police and citizens is low due to police corruption and betrayal of confidence. Trust was found to be a major threat to crime prevention co-production. A level of trust is required to maintain police-community relations. Police must co-design and co-deliver their policies and programmes with the community to enhance transparency. A number of studies have found trust to be an influential factor for successful co-production, and building trust is necessary in order to engage the public in co-production (Lira 2016;Fledderus 2015). Trust is crucial in managing information sharing on crime prevention. Frontline professionals, such as the police, will not be able to meet future crime challenges if there is a trust deficit; hence, it is important to restore trust in the work of the police. Professionals in public administration could learn about the importance of trust in their crucial role implementing government policies and programmes and service delivery.
Other avenues could employ community immersion, including enhancing policing policies to strengthen community relations and a cadet programme to educate and train Johannesburg's youth in dealing with crime and violence, while also making the appropriate budgetary allocations. These programmes could offer shadowing and on-the-job training for citizens who may want to enter the police force in the future.
In conclusion, specific conditions must be addressed for co-production to exist in crime prevention. The first condition is for co-production to generate new forms of organisation and structure that are democratic and far removed from the traditional authority and order found in many security clusters. This entails developing a new, flexible, citizen-centric approach to security provision. The second condition is that learning from successful security networks is essential, and this involves looking at existing mechanisms to address intractable issues. Finally, the third condition is to understand innovation better, implying that relevant stakeholders must observe how innovative ideas could evolve into good practices (Boyle and Harris 2009). Co-producing crime prevention in these ways could present opportunities and benefits, creating safer communities in Johannesburg.
Further, adopting community policing can also promote the creation of peaceful neighbourhoods. Community policing involves collaboration between the police, citizens, and private and non-profit community groups. The premise of community policing is vested in service provision; the reduction of fear; community mobilisation; and crime prevention, reduction, and control (Lira 2016). This is the ideal route for developing safer and more secure communities in Johannesburg.
There is some evidence to suggest that co-production is an avenue for realising safer neighbourhoods in Johannesburg. Nevertheless, data on the opinions and views of CPF members, neighbourhood watches, and private security guards are required in further studies to substantiate the claims made by the police officers on crime prevention co-production in Johannesburg. References were made to collaborations with these groups, and their story should also be told. Future research should extend our work into understanding the degree to which co-production predictors could influence successful crime prevention.
Finally, this study only focused on selected Johannesburg neighborhoods and may not fully explore the country's general perception on crime prevention co-production. Future studies could conduct similar empirical studies in other notorious regions around the country. Funding: This research was funded by the National Research Foundation (NRF) under N01872/120789RUN. We also thank the participant of a workshop held in Ghent who provided feedback on an earlier draft of this article. This workshop was co-financed by the inGOV project, funded by the European Commission within the H2020 Programme under grant agreement 962563 (https://ingov-project.eu/), and the INTERLINK project, funded by the European Commission within the H2020 Programme under grant agreement 959201 (https://interlink-project.eu/).

Institutional Review Board Statement:
The study was conducted in accordance with the Declaration of the University of Pretoria and the South African Police Service (SAPS) and approved by the Ethics Committee of the Faculty of Economic and Management Sciences (protocol code: EMS 188/19 and date of approval: 15 November 2019).

Informed Consent Statement:
Informed consent was obtained from all subjects involved in the study.
Data Availability Statement: Data is unavailable due to privacy/ethical restrictions.