INTRODUCTION

Recent approaches regarding public services and the role of bureaucrats and users have considered that policies result from complex relationships connecting sectors and actors internal and external to the public bureaucracy (Kooiman, 2003; Peters, 2009). This perspective implies rethinking the bureaucrat’s role, as a facilitator or inhibitor of change, especially regarding the users’ involvement in public service delivery (Bovaird, 2007; Boyle & Harris, 2009).

These changes are present in different fields of public administration. In health, this process is quite evident.

Health care, i.e., the direct attention to individuals and their inclusion in the health system, including the definition of responsibilities of those involved in this system, emerged as a major theme of sanitary reforms in the 1960s, in many countries (Scott, Ruef, Mendel, & Caronna, 2000). Despite the differences between national health services in each country, the latest generation of sanitary reforms, still in consolidation, highlights users’ participation in the system (Adinolfi, Starace, & Palumbo, 2016). The assumption is that health professionals working alone, especially in the local context where public policies are implemented, cannot act effectively in complex situations involving other factors beyond clinical or technical issues (Carvalho, 2004). The literature has corroborated this understanding, highlighting aspects such as users’ empowerment and engagement in the co-production of public services (Dunston, Lee, Boud, Brodie, & Chiarella, 2009; Ekman et al., 2011; Dent & Pahor, 2015).

Brazil has witnessed the reorganization of health practices that recognize the collective dimension of producing health care (Araújo & Rocha, 2009). The creation of the Brazilian national health system, Sistema Único de Saúde (SUS), resulted from social actors’ engagement through the sanitary reform movement. The presentation of the system in the Brazilian Constitution of 1988 highlighted the principle of citizen participation. These dynamics led to renewing the roles and the connections between public bureaucracy and users. National health policies and local practices have started to incorporate guidelines aimed at user participation (Rolim, Cruz, & Sampaio, 2013) and the co-production of public services.

In the dynamic and non-linear path between policy formulation and implementation (Sausman, Obborn, & Barrett,2016), the introduction and consolidation of co-production strategies involve citizens and bureaucrats with different profiles and interests. The relationships between different levels of bureaucracy (top, mid, and street-level) often express conflicts instead of mutual support and responsiveness (Lipsky, 2010), contributing to discrepancies between the general guidelines and concrete action in the local context. Also, the relationships between the levels of bureaucracy can happen in different ways, depending on rules, values, and governing mechanisms (Gomide & Pires, 2014), which may or may not be favorable for co-production. According to the coalitions formed, the existing rules, and the relationships established, these patterns change over time, showing advances and setbacks in consolidating co-production practices.

Thus, the state’s administrative structure and bureaucracy can hamper or foster strategies to connect professionals and users (Parks et al., 1981; Kingfischer, 1998; Magno & Cassia, 2015; Yeboah-Assiamah, Asamoah, Bawole, & Buabeng, 2016; Farooqi, 2016).

Therefore, it is possible to infer that co-production is influenced by how the different levels of bureaucracy in local government relate to each other and the context of rules and values around these relations. This influence may inhibit or use co-production to strengthen the state and its local bureaucracy, and, in the opposite sense, promote the citizens’ role as agents of change in a collective dimension.

Given this panorama, this article explores a research gap on how the different relationships between the levels of local bureaucracy influence the practices of co-production of public services. The research describes the co-production arrangements that favor or hinder the consolidation of the user’s role as a direct agent in public services provision, and how they are characterized. The analysis focuses on the Brazilian national health system (SUS), specifically the family health strategy and its implementation in the local context.

CO-PRODUCTION OF PUBLIC SERVICES

Amid debates about the concept of co-production of public services (Brandsen & Honingh, 2016; Nabatchi, Sancino, & Sicilia, 2017), there is a general agreement that it refers to the involvement of users in the design and/or delivery of public services. Through co-production, citizens permeate the structure of public bureaucracy and assume responsibilities – previously exclusive to bureaucrats – in the production of public goods and services (Brudney & England, 1983; Ostrom, 1996; Bovaird, 2007). However, the concept brings nuances and variations in theory and practice.

On the one hand, co-production presents an inclusive nature, increasing users’ power and responsibilities and developing citizenship from a political perspective (Padley, 2013; Rocha, Schommer, Debetir, & Pinheiro, 2019). Those outside the state apparatus have the power to influence the direction of a given action. Their participation goes beyond voice (Pestoff, Osborne, & Brandsen, 2006; Bovaird, 2007; Bovaird & Löeffler, 2012; Osborne & Strokosch, 2013; Bovaird, Stoker, Jones, Löeffler, & Roncancio, 2016). Developing trust in the relationships between bureaucrats and citizens is relevant as a condition to start and sustain the co-production process (Fledderus, 2015). In some contexts, the community is a primary actor, and co-production is similar to community self-organization for public service provision (Levine & Fisher, 1984; Vamstad, 2012; Pestoff, 2014; Poocharoen & Ting, 2015).

On the other hand, it is not uncommon for a government agency to encourage citizen participation and contribution in spaces of co-production, seeking to increase trust in the bureaucrats’ work (Alford, 2009; Ewert & Evers, 2014; Steen, Nabatchi, & Brand 2016) or the cooptation of citizens. Co-production is seen, in these cases, as a way of qualitatively and quantitatively improving the services the bureaucracy delivers and reinforcing the role of the state as a service provider (Kiser & Percy, 1980; Parks et al., 1981; Scavo, 1993; Evans & Sapeha, 2015; Williams, Kang, & Johnson, 2016; Steen, Nabatchi, & Brand, 2016).

The difference in how co-production materializes depends on how citizens interact and how the public bureaucracy is organized (Pammer, 1992).

LEVELS OF BUREAUCRACY AND CONNECTIONS AT THE LOCAL LEVEL

Public bureaucracy is heterogeneous, formed of distinct layers performing activities that offer different degrees and modalities of participation. The literature frequently analyzes public bureaucracy on three major levels – top, mid, and street-level bureaucracy.

The top-level bureaucracy consists of officials with political responsibility. Although the bureaucrats act based on technical capacity, they have connections with elected politicians (Olivieri, 2011), and their performance is based on political will (Aberbach, Putnam, & Rockman, 1981). Responsibility for decision-making and the establishment of guidelines for public policy implementation are inherent to these positions. It is common in the literature to associate the top-level bureaucracy with technical public servants who work in policy-making at the national level (Aberbach, Putnam, & Rockman, 1981; Olivieri, 2011). However, the classification of employees in top, as well as in mid and street-level bureaucracy, also reflect the structure of the bureaucracy at the local level, and this adaptation of the concept is being used in the present paper.

Mid-level bureaucracy gathers civil servants in an intermediary position in the organizational hierarchy (Pires, 2012; Ancarani, Arcidiacono, Di Mauro, & Giammanco, 2020). They connect the phases of policy design and policy implementation, working as a link between the top and the street-level bureaucracy (Cavalcante, Lotta, & Yamada, 2018).

Street-level bureaucracy comprises actors whose central responsibility is to implement public policies and services in direct contact with users (Lipsky, 2010; Gofen, 2014).

Although the bureaucracy levels are presented from a structural perspective, the relationships established between them take on different formats in the organizational and local context (Johansson, 2012), with different combinations of rules and values. The dimension of the territory directly affects the policy implementation and how the levels of bureaucracy are connected in a given context (Lotta & Favareto, 2016). The implementation at the local level considers national parameters together with other factors that significantly affect how relationships are established: (i) the power, resources, capacities, and constraints available to both state and social actors; (ii) interactions and disputes between these actors in formal and informal arenas; and (iii) the main ideas supported by the relevant actors (Bichir, Brettas, & Canato, 2017). This means that the relationships between levels of bureaucracy assume different configurations depending on the context, influencing how a guideline or policy is implemented, including co-production practices.

THE BRAZILIAN NATIONAL HEALTH SYSTEM AND THE FAMILY HEALTH STRATEGY

The family health strategy (FHS) was developed during a period of democratization and reform of public services in Brazil (Silva & Dalmaso, 2002). The engagement of several social actors, especially in the sanitary reform movement, and the debate about promoting health collectively resulted in the construction of the Brazilian national health system, Sistema Único de Saúde (SUS).

Based on Law 8080/1990, the implementation of SUS represented an inflection in the pattern of the health service organization then consolidated in the country (Escorel, Giovanella, Mendonça, & Senna, 2007). The system’s principles express the intended change: universal access to health services, decentralization to states and municipalities, and citizen participation in the health policy definition and implementation. Its implementation prioritized primary health care. The National Primary Care Policy is currently based on the family health model and is conducted within the scope of the FHS (Pinto & Giovanella, 2018).

The FHS is operated at the local level by a multi-professional family health team (FHT), which is formed of a general practitioner or a specialist in family health, a general nurse, or a nurse specialized in family health, a nursing assistant or technician, and community health workers. The professionals in the FHT are street-level bureaucrats working directly with users. Each FHT serves up to 4,000 people in a territory, working out of the basic health units (BHU).

FHTs connect with other levels of the local bureaucracy. The mid-level bureaucracy in this system refers to directors, coordinators, managers, and advisors. The top-level bureaucracy is the municipal secretary of health and their team, who are the closest to the political class of local public administration.

The structure of municipal health and the FHT is the same for all Brazilian municipalities, varying in the number of teams in the territories.

This research was conducted by observing the context of the family health strategy (FHS), users’ first contact with the health system. The strategy’s guidelines include users’ participation in identifying local problems and suggesting solutions. The design of the Brazilian health system offers instruments to implement these guidelines and facilitate co-production. One of these instruments is the Municipal Health Council (MHC), a legally mandatory instance that counts on the participation of users, civil society organizations, and professionals. Two more instruments are Local Health Councils (LHC), a body spontaneously created in the neighborhood served by a BHU, and Health Conferences, which are gatherings to discuss health policy at all government levels (Coelho, 2007).

METHODOLOGY

This research used a qualitative approach (Richardson, 2014). A multiple case study (Stake, 1995) was conducted, examining three municipalities operating within the Brazilian national health system (SUS). The research identified different relationship patterns, observing the particularities of each context to understand the social phenomenon.

The criteria to select the cases were the similarities regarding the size of the municipalities (medium) and the structure of the health service; the heterogeneity of the territories (sociocultural, political, geographic, demographic, and economic – GDP and GDP per capita – aspects), as these differences can influence how the structure operates and each health service is managed; and the trajectory of co-production, or moments when the bureaucracy promoted and facilitated users’ participation in the health service.

The three municipalities were Itajaí and Florianópolis, in the south of Brazil, and Sobral, in the state of Ceará, northeast of the country. Focusing on the FHS, two basic health units were examined in each municipality, totaling six units and territories.

The research was approved by the university ethics committee and by the scientific committee of the Municipal Health Secretaries, following determinations proposed in Resolution 466/2012 of the National Health Council of the Ministry of Health (CNS/MS).

Data collection was conducted between February and June 2019, using documentary research, bibliographic research, exploratory interviews, in-depth semi-structured interviews, group interviews, and direct and participant observation.

The documentary research aimed to understand how the FHS works, map the co-production initiatives in the municipalities, understand their operation, and learn the formal organizational structure in which the FHS is inserted and each bureaucracy level’s roles. The documents analyzed were the National Primary Care Policy; the municipal health policies; laws on the system’s organizational structure; minutes of meetings of bodies where users are represented; norms about the creation, and documents about the operation of co-production strategies.

The narrative and systematic bibliographic reviews on the themes “co-production” and “levels of bureaucracy” were used to map national and international literature in the databases SCOPUS, EBSCO, SPELL, Web of Science, and Scielo until 2019. The literature led to the definition of twelve categories of analysis. Table 1 presents the categories associated with factors that influence bureaucrats’ behavior at each level and their actions, including their participation – inhibiting or promoting – in the co-production of public services.

The interview scripts were prepared based on the categories of analysis. Before the interviews, all respondents signed a consent form.

The exploratory interviews were conducted with three civil servants who participated in the Municipal Health Council of each municipality. They sought to identify the history of the FHS in each municipality and the strategies to engage users in service delivery.

In-depth semi-structured interviews were carried out with 39 civil servants from different bureaucracy levels (top, mid, and street-level). Table 2 presents the number of professionals interviewed and the unit where they work.

Two group interviews were conducted with users involved in co-production practices. The interviews helped identify these actors’ perceptions of the public bureaucracy’s performance in the co-production spaces.

In addition to the interviews, the daily activities of the FHT were observed, particularly their relationships within the bureaucracy levels and the effects of these relations in the analyzed co-production strategies. We followed FHT visits to users’ homes; the work routine of the BHU coordinators; the meeting of a Local Health Council; two pre-conferences; a Municipal Health Conference; and a monthly team meeting of one of the BHU.

The information gathered from documentary research, transcription of interviews, and notes from the field diary were submitted to content analysis (Bardin, 2011); 745 sections were classified within the twelve categories (Table 1). To assist in this process, the qualitative data analysis software Atlas.ti was used.

The analyzed data were triangulated. When analyzing information collected through different techniques and from the point of view of different actors (living different realities), it was possible to identify six co-production arrangements regarding the relationships between the levels of bureaucracy. Each arrangement encourages or hinders co-production differently.

IDENTIFIED CO-PRODUCTION ARRANGEMENTS ACCORDING TO RELATIONSHIPS BETWEEN LEVELS OF LOCAL BUREAUCRACY

All the co-production arrangements share a set of common rules established by the norms and guiding principles of SUS and the National Primary Care Policy – which include users’ participation in the FHS. The contextual differences leading to distinct arrangements are the profiles and the role of the bureaucrats, the role of the community, and the local rules and practices, whether formal or informal. Below, the peculiarities of each co-production arrangement are described.

Arrangement 1 – Co-production in design and implementation, limited to street-level bureaucracy due to its low connection with the other levels

The first identified co-production arrangement comprises top-level bureaucrats with strong party-political ties, unwilling to connect with users. The mid-level is composed of professionals appointed for the positions by party-political criteria, and not open to connect with users. The exception is the BHU Coordinator, who, even if politically appointed, is open to establishing a relationship with users. In contrast, the street-level bureaucracy is open to establishing a relationship with users. The communities show a low degree of engagement in public activities. The few users who actively participate are encouraged by the street-level bureaucracy.

Because the street-level bureaucrats are open, with a history of community engagement, the local health council (LHC) becomes an active instance, a space in which co-production is possible, even if only a few users participate. With good connections between bureaucrats and users, the LHC contributes to responding to local demands, supporting the street-level bureaucracy’s work. Examples of measures taken by LHC include the request for missing drugs and lack of professionals to complete the health teams, actions to reduce local industries’ waste, the installation of a bus stop near the health unit to facilitate users’ access, installation of leisure spaces in the neighborhood, and change in the BHU’s opening hours.

The LHC plans health education actions implemented by the council members in partnership with the FHT, schools, and local churches.

However, the results of co-production achieved by the LHC are limited to the context of street-level bureaucracy and its autonomy. Demands that require resources beyond these actors’ reach are rarely resolved due to the conflict between the levels. Some of these demands were mentioned in the LHC minutes over the years.

The divergence between the characteristics of the top and mid-level bureaucracy and the profile of street-level bureaucracy leads to a relationship permeated by conflicts. In addition to the lack of collaboration and dialogue between the top and mid-level and the LHC, the study collected reports of political persecution of civil servants who participate in the LHC. These persecutions are materialized in transfers of civil servants to other BHU or other instances of the system, weakening the co-production that, in this arrangement, relies on the effort of street-level bureaucrats. Another example is the reports denouncing that top-level bureaucracy prohibited LHC meetings in a health unit in an attempt to demobilize its members.

In short, the performance of the LHC within this first arrangement showed practices of co-production of public services in which decision-making and responsibility for execution are shared (Pestoff, Osborne, & Brandsen, 2006; Bovaird, 2007; Bovaird & Löeffler, 2012; Fledderus, 2015; Rocha et al., 2019), but with limited reach, given the lack of support from the top and mid-level bureaucracy.

Arrangement 2 – Non-existent or manipulative co-production, due to lack of incentive for users and street-level bureaucrats

In the second co-production arrangement, both the top and mid-level bureaucracy have the same profile as the first arrangement - strong party-political ties and unwillingness to connect with citizens. However, the street-level bureaucracy is also not open to working together with users, and the community has a low degree of engagement. Few residents have bonds with the territory. Even the street-level bureaucrats do not hold representativeness.

Professionals who want to promote participation may be present, in a smaller number, and discouraged by the unfavorable scenario of this type of connection. The political focus of the top and mid-level bureaucrats and the conflict with external actors (political opposition) lead street-level bureaucrats who at first tried to develop spaces for dialogue with the community to give up. Some narratives mention that professionals ended up physically sick as a consequence of conflicts.

Discouragement, lack of connection among actors, and lack of incentives from the top and mid-level bureaucracy prevent street-level bureaucracy and users from engaging in co-production. When street-level bureaucrats started creating the LHC, community meetings indicated that users would not adhere to the initiative. The political class took control of these meetings to co-opt the participatory space, discouraging street-level bureaucrats.

Even in spaces required by law – the Municipal Health Council and Conferences– it was observed attempts of cooptation by the dominant power coalition. The relationships overlap the formal rules within this arrangement, creating informal processes related to the dominant political class’s interference. When the political class operates in these spaces, it seeks to transform them into an extension of particular interests. The existing links between users and the public bureaucracy in this arrangement assume a manipulative nature. There is no real sharing of power with users, only demonstrating the state as the main actor (Kiser & Percy, 1980; Williams, Kang, & Johnson, 2016).

Arrangement 3 – Co-production inhibited by technocracy, low connected community, and perception of good service quality

The third co-production arrangement is composed of a technocratic top-level bureaucracy, unwilling to establish connections with users. The mid-level bureaucracy, formed by a qualified technical body, represents one of the main coalitions of power. The nominations for the mid-level bureaucracy in this arrangement follow an understanding – a local rule not formalized in regulation – that they must follow technical criteria.

Combined with this technical profile of the top and mid-level bureaucracy, the street-level bureaucrats do not stimulate participation. The community, in turn, does not show interest in engaging.

The decision-making and information flow between the bureaucracy levels is hierarchical. The upper levels impose guidelines and deadlines for implementing actions to the street-level bureaucracy, based on technical capacity and with no opportunity for questioning.

This combination of technocratic bias in the top and mid-level bureaucracy, the absence of street-level bureaucrats’ openness for participation, and low community engagement prevent co-production. Because the bureaucrats consider they offer high-quality services (and they are high-quality indeed), they do not perceive the LHC as necessary and do not work to make it active. The connection with users is limited to services and individual needs. No spaces were identified for discussion and resolution of collective problems in the territory.

In this arrangement, the municipality maintains mandatory participation instruments, such as the Municipal Health Council and Conference. However, they adopt technical jargon, which may distance users. These spaces are dominated by the professional categories, which seek to defend the corporate interests of bureaucratic groups, oscillating between periods of alignment and conflict with the top and mid-level bureaucracy. The research observed situations in which representatives of professional categories were elected to take seats that were supposed to be filled by community organizations on the Municipal Council.

Notwithstanding, there is some willingness of the bureaucracy to dialogue with users, but the latter cannot penetrate these participatory spaces. The upper levels of bureaucracy realize the absence of the community in these spaces and, based on this, do not believe that co-production is a possibility.

In the existing participatory spaces, the focus is on improving the services provided in a traditional way (Evans & Sapeha, 2015), improving the bureaucracy’s performance and results. The users’ contribution is induced by professionals, given that citizens trust the expert advice and the hierarchy in place (Alford, 2009; Ewert & Evers, 2014).

Arrangement 4 – Co-production based on community self-organization, low connection with the dominant levels of bureaucracy and technocracy

In the fourth co-production arrangement, the top and mid-levels of bureaucracy present a technocrat bias and are unwilling to engage in co-production. The street-level bureaucracy profile does not favor participation. These bureaucrats participate sporadically, seeking personal gains such as time off work in exchange for participating in the council meetings. However, the community is well-connected.

The bureaucracy’s internal structure follows technical and hierarchical decision-making, but co-production still happens through the Local Health Council (LHC), thanks to the strength of community leaders. These actors keep the LHC running and work to respond to community demands.

The effort to establish connections coming from community leaders does not hinder results obtained through co-production. These leaders mobilize key actors, especially politicians, unrelated to Municipal Health Secretary. The results of these connections with the community are expressive, including, for example, the construction of a new BHU.

However, the way co-production takes place in this context highlights a particular model. It is a co-production based on the community, where citizens are central actors in a process that resembles self-organization (Levine & Fisher, 1984; Vamstad, 2012; Pestoff, 2014; Poocharoen & Ting, 2015).

The model includes risks, such as when the LHC is identified with one specific community leader, which may weaken the council.

Arrangement 5 – Co-production with engagement and connections throughout the different levels of bureaucracy

The fifth co-production arrangement has a top-level bureaucracy composed of agents with technical profiles who are open to connecting with users. Mid-level bureaucrats are appointed based on technical criteria. They are professionals – in managerial positions in the health secretary and in the BHU – who perceive users’ participation as crucial in public service delivery. The openness to connect with users observed in the upper-levels is present at the street-level. In addition, the community is willing to work together.

The LHC is active and recognized in the territory, given the community engagement and the street-level bureaucrats’ profile – particularly the community health workers, who have a historical connection with the neighborhood.

Co-production in this arrangement reflects an effective sharing of power and responsibilities among the actors (Pestoff, Osborne, & Brandsen, 2006; Bovaird, 2007; Bergh, 2010; Bovaird & Löeffler, 2012; Osborne & Strokosch, 2013; Bovaird et al., 2016). The LHC discusses and acts on the BHU professionals’ demands, assists in referral to the responsible municipal agencies, and participates in service delivery, such as preventing and combating diseases. In addition, the council acts on community demands, whether individual (when urgent or differentiated referrals are necessary, even according to the regular flow of service delivery) or collective (demands of the territory, such as cleaning areas).

The results achieved, associated with a set of factors covering the relationships between all levels of bureaucracy, contribute to strengthening community engagement.

The community trusts the street-level bureaucracy because the bureaucrats are representative. This trust gains strength in professionals’ daily work with users in delivering services since each step of the service follows a formal flow agreed with the team. Both the top and mid-level bureaucracy support the performance of street-level bureaucrats. The research did not identify interference of the top-level bureaucracy in service delivery, especially the granting of benefits by some party-political linkage. The absence of favoritism helps to maintain and strengthen trust.

Another factor contributing to co-production is the engagement of the BHU coordinator, who is part of the mid-level bureaucracy. The encouragement and involvement of coordinators in the LHC are relevant to street-level bureaucracy engagement, contributing to mobilizing the professionals who initially did not work with the community. Since mid-level bureaucrats have some autonomy – particularly regarding resource management – they can connect the demands and expand the scope of co-production. One example is an LHC that had been inactive and was reactivated due to the initiative of a mid-level professional working at the BHU.

A final factor concerns the support or incentive provided by the top and mid-level bureaucracy of the Municipal Health Secretary. This incentive was observed both in the interview and in the structure to support co-production practices in the LHC. One of the instruments for this support is the institutionalization (formal rule) of a position in the system’s structure. It is a professional linked to the MHC, who is responsible for assisting the operation of LHCs in the municipality. Due to the ideological profile of the political group elected and the top-level bureaucracy, the municipality institutionalized a co-management approach, adopting practices such as the wheel methodology, the administration supported by a committee of managers, or led by a steering committee. These practices facilitate the flow of information and joint decision-making, reducing conflicts between levels of bureaucracy and user engagement. Thus, this arrangement presents a set of formal local rules that favor co-production.

Arrangement 6 – Co-production with connections between the levels of bureaucracy and adaptation of the mid-level bureaucracy to the culture of openness to community engagement

The sixth co-production arrangement includes top, mid (coordination of the primary health care), and street-level bureaucracy open to community engagement and co-production. However, the mid-level bureaucracy in the BHU was unwilling to connect with users.

As in arrangement 5, the community is active, representing a powerful coalition. The community strength, added to street-level bureaucracy’s openness, reinforces co-production practices as the standard. The formal support structure that both the top and mid-level bureaucracy developed to improve user participation instruments – marked by the position created in the system’s structure of a professional responsible for monitoring and supporting LHC and the adoption of co-management methodologies – results in more comprehensive co-production, even without the engagement of the mid-level bureaucrat working at the BHU.

The culture of participation and community coordination developed over the years prevails, and the mid-level bureaucracy adapts to the structure and dynamics that favor user participation. This bureaucracy does not necessarily engage but does not present a barrier to dialogue. The BHU coordinators are hired through a selection process and have a fragile connection with the system’s structure. Therefore, avoiding dialogue with power coalitions formed by both the incumbent political group and the community would be a risk for staying in office.

The LHC operating within this arrangement is one of the pioneers in the municipality and has been active since 1997. From the analysis of minutes and the interviews, it was found that some meetings counted on the participation of more than fifty citizens, an expressive number compared to the average participation. The engagement of users goes beyond the voice, including direct responsibilities for implementation (Bovaird, 2007; Bovaird & Löeffler, 2012).

Some of the co-production practices created in these contexts are the elaboration of a diagnosis of the territory, supporting the discussion of specific public policies; connection with other government sectors to build a new BHU and a Child Education Center; actions to promote adequate solid waste management; implementation of a community garden; and planning and implementing cultural and sports projects for young people. This shows how co-production strategies go beyond the direct provision of health services, touching other aspects of community life indirectly related to health. The LHC works as a space to promote the transparency of the bureaucracy’s actions, like the decisions of the top and mid-level bureaucrats, changes in the services, or drug availability.

A summary of the identified co-production arrangements is presented in Table 3.

Source: Elaborated by the authors.

DISCUSSION

Several aspects are worth discussing after identifying the different co-production arrangements that emerge from the relationships between the levels of bureaucracy and how these relationships affect the consolidation of co-production practices within the family health strategy (FHS).

Firstly, the national policy guideline aimed at citizen participation in the FHS is implemented in different ways. The way services are provided in each territory varies according to the different patterns of relationships between bureaucracy levels, community profiles, and rules and processes based on local political practices and culture. Co-production arrangements make it clear that openness to citizens’ engagement permeates bureaucracy levels, forming different combinations within the organizational structure. Together with the community’s performance and local rules, these combinations indicate how the national participation guidelines will be implemented and how co-production actions will materialize.

In contexts where bureaucrats at different levels are unwilling to connect with users, co-production tends not to occur. In these cases, only the legally established participation instruments are maintained. When bureaucrats are open to connecting with users, at least at the street-level, co-production practices occur. When openness to user engagement appears at other levels of bureaucracy, in a context of an active community and counting on a support structure (norms/rules/resources), other co-production strategies are visible beyond the regular activities and achievements of the LHC. Although rules are important (such as national policies and legislation), the bureaucrats’ profile and the discretion exercised dictate how policies are implemented, including co-production practices (Lipsky, 2010; Dubois, 2014).

Another finding is that when both top and mid-level bureaucracy are open to connecting with users (developing guidelines and structures to facilitate co-production), it is crucial to involve street-level bureaucracy as it is at the street-level that direct contact with users occurs, and trust can be gained (Fledderus, Brandsen, & Honingh, 2014), which is important for co-production (Grissom, Kern, & Rodriguez, 2015).

However, the resistance to users’ engagement in delivering health services may manifest at any level of bureaucracy. It is not determined hierarchically, and bureaucrats co-produce with citizens when they perceive this is relevant. Incentives and support from the top and mid-levels can facilitate co-production, as seen in arrangements 5 and 6. However, these incentives are not the only element conditioning the performance of street-level bureaucracy. For instance, this research found that even with a support structure for co-production, some bureaucrats do not consider the connection with users as part of their work (Dixey & Woodall, 2011) and do not participate in co-production. If the incentive from upper-level bureaucracy is limited to narratives and does not evolve into practices, the co-production, is less likely to thrive. In these cases, bureaucracy engagement in co-production strategies is in no way related to hierarchical determination.

The arrangements identified demonstrate that the street-level bureaucracy can conduct co-production strategies even when the top and mid-level bureaucrats are unwilling to connect with users and do not perceive user engagement as relevant to the services. This dynamic was clear when the top and mid-level bureaucrats tried to discourage co-production, as in arrangement 1. The LHC surveyed were mostly created from the street-level bureaucracy’s initiative with the community, albeit in half of the cases, the upper-level bureaucracy did not support these councils.

Although the creation and operation of co-production strategies are not explained by the top and mid-level bureaucracy’s performance, it is crucial to recognize their importance in how comprehensive co-production can be. Even when top and mid-level bureaucracy manage a large part of the financial and non-financial resources necessary to implement the services and the co-production.

Another finding refers to how the relationships between the bureaucracy levels and between them and users result in different co-production practices. Co-production can vary from a manipulative approach to community self-organization, depending on how the bureaucracy relates internally and interacts with citizens. In some cases, there is sharing of power and in others, power is not shared, but there is a connection between public bureaucracy and citizens.

This research does not discuss the possible advantages of one kind of co-production over another. Nor does it suggest that the co-production of public services is an ideal approach or a solution to problems associated with health or the improvement of public service delivery. One of the municipalities analyzed, for instance, is recognized for its high health indicators even though the bureaucracy is not open to user engagement, and the scenario does not favor co-production. What stands out is that, when seeking connection with users in the planning and delivery of public services, the relashionship between levels of local bureaucracy may facilitate or hinder achieving this goal.

CONCLUSION

This article sought to analyze how the relationship between the different levels of public bureaucracy – top, mid, and street-level – and the context of rules and values they operate affect the consolidation of co-production of health services. The study sought to understand the co-production arrangements that favor or hinder the consolidation of the user’s role as a direct agent in public service provision.

The analysis was based on the Brazilian family health strategy and focused on three medium-sized municipalities. Different patterns of relationship between levels of bureaucracy affected in different ways the emergence of co-production practices and six distinct co-production arrangements were identified.

As for the theoretical contribution, this study helps to elucidate aspects of the public bureaucracy’s performance in the co-production of public services. Even with the development of the field, few studies on co-production investigate the phenomenon considering the state’s perspective and the bureaucratic elements that influence co-production.

The research findings confirm and detail the theoretical assumption that co-production depends not only on citizens’ willingness to collaborate but also on the extent to which the public bureaucracy is open to this collaboration (Pammer, 1992; Boyle & Harris, 2009; Ryan, 2012). The public bureaucracy’s openness toward co-production is associated with the rules, values, and patterns of the connections formed in each context. The co-production arrangements have different i) relationship patterns between levels of bureaucracy, ii) bureaucracy profiles and roles, iii) rules (formal and informal), and iv) community characteristics. Different arrangements and contexts influence co-production. Therefore, the analyses of co-production practices must consider elements of such arrangements..

The research results also contribute to the field of organizational studies and with studies on bureaucracy and its connections with the implementation of public policies. Studies on levels of bureaucracy have focused on each level individually rather than on the relationships between the levels and their impact on public organizations and the interface with citizens. To understand the role of the bureaucratic body from the relational approach, it is necessary to consider the relationships established between the levels of bureaucracy and with other actors, such as citizens. Although street-level bureaucracy plays a central role in the co-production of public services, the influence of the top and mid-level cannot be ignored.

Due to the scarcity of studies that explore the levels of bureaucracy (and the relationships between them) and their effects on co-production, the research results highlight points for theoretical development. These points might be used as hypotheses for future studies to consolidate the findings and the resulting theory, analyzing other co-production arrangements:

• Any level of bureaucracy can offer resistance or openness to co-production, given that the involvement of bureaucrats is not hierarchically determined.

• A structure of support and incentive for co-production by the top and mid-level bureaucracy does not guarantee the engagement of street-level bureaucracy, although this engagement is necessary for co-production strategies to be more effective.

• The collaboration of top and mid-level bureaucrats is essential to expand co-production results, especially as they have access to resources and are closer to politicians and public managers.

• Different relationship patterns between (and within) levels of bureaucracy and users result in different co-production practices, with greater or lesser power-sharing.

As a contribution to practitioners, the arrangements show aspects that, connected or isolated, facilitate or hinder co-production in the public health service. When recognizing these different aspects, bureaucracy and users engaged in co-production can identify how to work better and address the issues in the organizational routine to promote such practices.

The co-production of public services is a complex phenomenon, influenced by a diversity of elements and by citizens and public bureaucracy. If one is willing to co-produce, it is worth considering the structure and performance of the public bureaucracy, even when it is not interested in opening dialogues and sharing.

REFERENCES

Aberbach, J. D., Putnam, R. D., & Rockman, B. A. (1981). Bureaucrats and politicians in western democracies. Massachusets: Harvard University Press.

Adinolfi, P., Starace, F., & Palumbo, R. (2016). Health Outcomes and Patient Empowerment: The Case of Health Budgets in Italy. Journal of Health Management, 18(1), 117-133.

Alford, J. (2009). Engaging public sector clients: from service-delivery to co-production. Basingstoke: Palgrave Macmillan.

Ancarani, A., Arcidiacono, F., Mauro, C. D., & Giammanco, M. D. (2020). Promoting work engagement in public administrations: the role of middle managers’ leadership. Public Management Review, 23(8), 1234-1263.

Araújo, M. B. de S., & Rocha, P. de M. (2009). Saúde da família: mudando práticas? Estudo de caso no município de Natal (RN). Ciência & Saúde Coletiva, 14, 1439-1452.

Bardin, L. (2011). Content analysis. São Paulo: Edições.

Bergh, S. I. (2010). Assessing the scope for partnerships between local governments and community-based organizations: Findings from rural Morocco. International Journal of Public Administration, 33(12), 740-751.

Bichir, R., Brettas, G. H., & Canato, P. (2017). Multi-level governance in federal contexts: the Social Assistance Policy in the City of São Paulo. Brazilian Political Science Review, 11(2), 1-28.

Bovaird, T. (2007). Beyond engagement and participation: User and community co-production of public services. Public Administration Review, 67(5), 846-860.

Bovaird, T., & Loeffler, E. (2012). From Engagement to Co-production: The Contribution of Users and Communities to Outcomes and Public Value. Voluntas, 23(4), 1119-1138.

Bovaird, T., Stoker, G., Jones, T., Löeffler, E., & Roncancio, M. P. (2016). Activating collective co-production of public services: influencing citizens to participate in complex governance mechanisms in the UK. International Review of Administrative Sciences, 82(1), 47-68.

Boyle, D., & Harris, M. (2009). The challenge of co-production. London: New Economics Foundation.

Brandsen, T., & Honingh, M. (2016). Distinguishing Different Types of Co-production: A Conceptual Analysis Based on the Classical Definitions. Public Administration Review, 76(3), 427-435.

Brudney, J. L., & England, R. E. (1983). Toward a definition of the co-production concept. Public Administration Review, 43(1), 59-65.

Carvalho, S. R. (2004). Os múltiplos sentidos da categoria “empowerment” no projeto de Promoção à Saúde. Cad. Saúde Pública, 20(4), 1088-1095.

Cavalcante, P. L., Lotta, G. S., & Yamada, E. M. K. (2018). O desempenho dos burocratas de médio escalão: determinantes do relacionamento e das suas atividades. Cadernos EBAPE.BR, 16(1), 14-34.

Coelho, V. S. P. (2007). A democratização dos Conselhos de Saúde: o paradoxo de atrair não aliados. Novos estudos CEBRAP, 78, 77-92.

Dent, M., & Pahor, M. (2015). Patient involvement in Europe - a comparative framework. Journal of Health Organization and Management, 29(5), 546-555.

Dixey, R., & Woodall, J. (2011). Prison staff and the health promoting prison. International Journal of Prisoner Health, 7(4), 8-16.

Dubois, V. (2014). The state, legal rigor, and the poor: The daily practice of welfare control. Social Analysis, 58(3), 38-55.

Dunston, R., Lee, A., Boud, D., Brodie, P., & Chiarella, M. (2009). Co-production and health system reform - From re-imagining to re-making. Australian Journal of Public Administration, 68(1), 39-52.

Ekman, I., Swedberg, K., Taft, C., Lindseth, A., Norberg, A., & Brink, E. (2011). Person-centered care — ready for prime time. European Journal of Cardiovascular Nursing, 10(4), 248–251.

Escorel, S., Giovanella, L., Mendonça, M. H. M., & Senna, M. C. M.(2007). O Programa de Saúde da Família e a construção de um novo modelo para a atenção básica no Brasil. Revista Panamericana de Salud Pública, 21, 164-176.

Evans, B., & Sapeha, H. (2015). Are non-government policy actors being heard? Assessing New Public Governance in three Canadian provinces. Canadian Public Administration, 58(2), 249-270.

Ewert, B., & Evers, A. (2014). An Ambiguous Concept: On the Meanings of Co-production for Health Care Users and User Organizations? Voluntas, 25(2), 425-442.

Farooqi, S. A. (2016). Co-production: what makes co-production work? Evidence from Pakistan. International Journal of Public Sector Management, 29(4), 381-395.

Fledderus, J. (2015). Does User Co-Production of Public Service Delivery Increase Satisfaction and Trust? Evidence From a Vignette Experiment. International Journal of Public Administration, 38(9), 642-653.

Fledderus, J., Brandsen, T., & Honingh, M. (2014). Restoring Trust Through the Co-Production of Public Services: A theoretical elaboration. Public Management Review, 16(3), 424-443.

Gofen, A. (2014). Mind the Gap: Dimensions and Influence of Street-Level Divergence. Journal of Public Administration Research and Theory, 24(2), 473-493.

Gomide, A. A., & Pires, R. R. C. (2014). Capacidades estatais e democracia: arranjos institucionais de políticas públicas. Brasília: IPEA.

Grissom, J. A., Kern, E. C., & Rodriguez, L. A. (2015). The “Representative Bureaucracy” in Education: Educator Workforce Diversity, Policy Outputs, and Outcomes for Disadvantaged Students. Educational Researcher, 44(3), 185-192.

Johansson, V. (2012). Negotiating bureaucrats. Public Administration, 90(4), 1032-1046.

Kingfisher, C. P. (1998). How providers make policy: An analysis of everyday conversation in a welfare office. Journal of Community and Applied Social Psychology, 8(2), 119-136.

Kiser, L. L., & Percy, S. L. (1980). The concept of co-production and its implications for public service delivery. Annual meeting of the American Society for Public Administration, San Francisco, March 13-16.

Kooiman, J. (2003). Governing as Governance. London: Sage.

Levine, C. H., & Fisher, G. (1984). Citizenship and service delivery - the promise of co-production. Public Administration Review, 44, 178-187.

Lipsky, M. (2010). Street-level bureaucracy: dilemmas of the individual in public services. New York: Russel Sage Foundation.

Lotta, G., & Favareto, A. (2016). Desafios da integração nos novos arranjos institucionais de políticas públicas no Brasil. Revista de Sociologia e Política, 24, 49-65.

Magno, F., & Cassia, F. (2015). Public administrators’ engagement in services co-creation: factors that foster and hinder organisational learning about citizens. Total Quality Management & Business Excellence, 26(11-12), 1161-1172.

Nabatchi, T, Sancino, A., & Sicilia, M. (2017). Varieties of participation in public services: The who, when, and what of co-production. Public Administration Review, 77(5), 766-776.

Olivieri, C. (2011). Os controles políticos sobre a burocracia. Revista de Administração Pública, 45(5), 1395–1424.

Osborne, S. P., & Strokosch, K. (2013). It takes two to tango? Understanding the co-production of public services by integrating the services management and public administration perspectives. British Journal of Management, 24(S3), S31-S47.

Ostrom, E. (1996). Crossing the great divide: Co-production, synergy, and development. World Development, 24(6), 1073-1087.

Padley, M. (2013). Delivering localism: The critical role of trust and collaboration. Social Policy and Society, 12(3), 343-354.

Pammer, W. J. (1992). Administrative norms and the co-production of municipal services. Social Science Quarterly, 73(4), 920-929.

Parks, R. B., Baker, P. C., Kiser, L., Oakerson, R., Ostrom, E., Ostrom, V., Percy, S., Vandivort, M. B., Whitaker, G. P., & Wilson, R. (1981). Consumers as coproducers of public services. Policy Studies Journal, 9(7), 1001-1010.

Pestoff, V. (2012). Co-production and Third Sector Social Services in Europe: Some Concepts and Evidence. Voluntas, 23(4), 1102-1118.

Pestoff, V. (2014). Hybridity, Co-production, and Third Sector Social Services in Europe. American Behavioral Scientist, 58(11), 1412-1424.

Pestoff, V., Osborne, S. P., & Brandsen, T. (2006). Patterns of co-production in public services. Some concluding thoughts. Public Management Review, 8(4), 591-595.

Peters, B. G. (2009). Still the century of bureaucracy? The roles of public servants. Public Policy and Administration, 1(30), 7-21.

Pinto, L. F., & Giovanella, L. (2018). Do Programa à Estratégia Saúde da Família: expansão do acesso e redução das internações por condições sensíveis à atenção básica (ICSAB). Ciência & Saúde Coletiva, 23, 1903-1914.

Poocharoen, O., and Ting, B. (2015). Collaboration, Co-Production, Networks: Convergence of theories. Public Management Review, 17(4), 587-614.

Richardson, R.J. (2014). Pesquisa Social: métodos e técnicas. São Paulo: Atlas.

Rocha, A. C., Schommer, P. C., Debetir, E., & Pinheiro, D. M. (2019). Transparência como elemento da coprodução na pavimentação de vias públicas. Cadernos Gestão Pública e Cidadania, 24(78), 1-22.

Rolim, L. B., Cruz, R. S. B. L. C., & Sampaio, K. J. A. J. (2013). Participação popular e o controle social como diretriz do SUS: uma revisão narrativa. Saúde em Debate, 37(96), 139-147.

Ryan, B. (2012). Co-production: Option or Obligation? Australian Journal of Public Administration, 71(3), 314-324.

Sausman, C., Oborn, E., & Barrett. M. (2016). Policy translation through localization: implementing national policy in the UK. Policy and Politics, 44(4), 563-589.

Scavo, C. (1993). The use of participative mechanisms by large US cities. Journal of Urban Affairs, 15(1), 93-109.

Scott, R. W., Ruef, M., Mendel, P. J., & Caronna, C. A. (2000). Institutional change and healthcare organizations: From professional dominance to managed care. Chicago: The University of Chicago Press.

Silva, J. A., & Dalmaso, A. S. W. (2002). Agente Comunitário de Saúde: o ser, o saber, o fazer. Rio de Janeiro: Editora Fiocruz.

Stake, R. E. (1995). The Art of Case Study Research. Thousand Oaks: Sage.

Steen, T., Nabatchi, T., & Brand, D. (2016). Introduction: Special issue on the co-production of public services. International Review of Administrative Sciences, 82(1), 3-7.

Vamstad, J. (2012). Co-production and Service Quality: The Case of Cooperative Childcare in Sweden. Voluntas, 23(4), 1173-1188.

Williams, B. N., Kang, S., & Johnson, J. (2016). (Co)-Contamination as the Dark Side of Co-Production: Public value failures in co-production processes. Public Management Review, 18(5), 692-717.

Yeboah-Assiamah, E., Asamoah, K., Bawole, J. N., & Buabeng, T. (2016). Public sector leadership-subordinate ethical diffusion conundrum: perspectives from developing African countries. Journal of Public Affairs, 16(4), 320-330.