Heath Sector Network Governance and State-building in South Kivu, Democratic Republic of Congo

Abstract

Longstanding patterns of interaction exist between state and non-state actors seeking to improve public health in Democratic Republic of Congo (DRC). DRC is a weak state, and, in many cases, private actors have stepped in to fill the void created by the lack of state health care provision. However, the role of these interactions in creating a governance network in the health sector has been underexplored. Using data from 18 months of qualitative field research, this study aimed to explore governance networks in DRC’s health sector, examining how multiple stakeholders work to manage the health system and how the resulting governance network has been relevant for the state-building process. The findings demonstrate that the health sector in South Kivu is emerging as an arena of networked governance based on active partnerships between state institutions and non-state actors. Interactions between state and non-state actors account for the persistence of the health sector in a setting characterized by state weakness. However, networked governance does not function optimally, because, although non-state interventions fill the void where the state falls short, the DRC state has faced the challenge of interacting with partners with fragmented and horizontally competing agendas. Although weak, the shadow of state authority is present in the arena of stakeholders’ interactions, as the state plays a determining role by providing a regulatory framework. Overall, the findings show that the interactive engagement of non-state actors contributes to improving institutional capacity through these actors’ engagement with state institutions for health system management and institutional development. However, although networked health sector governance does contribute to state capacity, it is difficult to assess the real influence of these interactions on the state-building process in a context of critical fragility, where coordination and alignment have been problematic.

Introduction

Both state and non-state actors are involved in the operation of the health system in Democratic Republic of Congo (DRC).These actors fulfil multiple, and sometimes overlapping, roles. Little is known about the nature of interactions between these actors or the resulting impact on the health system and state-building. This study addressed this gap in the past research. The article begins with a description of the history and context of public governance and the health sector in DRC, building a frame for our findings on DRC’s health sector as a governance network based on state/non-state partnerships, state-building outcomes of multilevel interactions in this governance network, and arenas of interaction and methods of engagement in DRC health sector governance network.

State-building, state legitimacy and the provision of services in fragile states

The rise and character of the contemporary state-building agenda have bearing on the post-Cold War world (Hehir and Robinson 2007). Commonly understood as the creation of new governmental institutions and the strengthening of existing ones, state-building is a crucial issue for the world community today, because, since 9/11, weak states have been regarded as close to the root of the world’s most serious problems (Fukuyama 2004). State-building consists of ‘purposeful actions to develop the capacity, institutions and legitimacy of the state in relation to an effective political process’ (Bruce and Chandran 2008, p. 14). State-building thus includes strengthening three dimensions of statehood: state authority, state capacity and state legitimacy (Grävingholt et al. 2012; World Bank 2012).

However, in fragile and/or conflict-affected states, there is policy tension between prioritizing state-building and the urgent provision of basic services to the population (Batley and Mcloughlin 2010). In fragile states, the inputs of non-state actors enhance the capacity of state authority structures and contribute to improving service delivery (Krasner and Risse 2014). State-building and service provision are linked, as service delivery supports the building of state effectiveness, legitimacy and resilience (Batley and Mcloughlin 2010). In the same vein, authors have identified the delivery of social services as a dimension of state-building from the public health perspective (Witter et al. 2015).

However, little is known about how service delivery contributes to state-building where the state is fragile and—as in DRC—mostly reliant on non-state providers through the governance network. This article focuses on state-building processes and outcomes within DRC’s health sector, exploring how the governance network in this sector contributes to building state capacity and effectiveness through the delivery of social services.

Conceptualizing governance networks

Governance involves state–society problem solving in public arenas (Brinkerhoff and Bossert 2014). Within the network literature, ‘governance refers to the horizontal interactions by which various public and private actors at various levels of government coordinate their interdependencies in order to realize public policies and deliver public services’ (Klijn and Koppenjan 2012, p. 8). In contrast, the institutionalist perspective sees governance as concerning the rules that distribute authority, roles and responsibilities among societal actors (Brinkerhoff and Bossert 2014). Governance can be carried out both with and through networks, which are sets of generally stable, non-hierarchical and interdependent resource-sharing relationships among actors with a common interest (Rhodes 2007). The concept of governance networks is useful for analysing the interactions among health service providers in DRC. Although constitutionally owned by the state, the DRC health sector is in effect a common-pool resource that is owned by everyone in community, where all of the key stakeholders are potential governance practitioners (Abimbola et al. 2014).

Although governance networks necessarily relate to the networks of actors engaged in public governance (Torfing 2005), scholars disagree on the specific definition (Dedeurwaerdere 2005; Esty 2006; Torfing and Sørensen 2014). Here, we adopt the ‘governance school’ definition, which ‘conceives policy networks as a specific form of governance, as a mechanism to mobilize political resources in situations where these resources are widely dispersed between public and private actors’ (Börzel 1997).

Highly relevant for DRC’s health system is the idea that governance networks may play an important role when governments fail: When non-state actors step in to compensate for the state’s failure, they may also contribute to the state’s capacity development and help to improve state policies (Börzel 2011). However, governance networks in weak states face challenges related to effectiveness, accountability and legitimacy, and critics caution that governance networks should not be seen as ‘cure-alls’ in fragile contexts (Torfing and Sørensen 2009; Börzel 2011). Weak states have difficulty assuming a regulatory role, ensuring harmonization and integration of non-state interventions and coordinating the stakeholders taking over state functions (Provan and Milward 2001).

To this point, there has been little evidence-based research on how the interdependent stakeholders in DRC’s health sector interact, or on how they contribute to health care delivery and state-building. This study aimed to explore governance networks in DRC’s health sector and to investigate the implications of these governance networks for the health system and the state-building process in the fragile context of DRC. Our findings map the actors involved as well as their methods of working and demonstrate how governance networks work in practice in a context of fragility.

History and context: public governance and the health sector in DRC

The health status of DRC’s population is alarming in terms of population access and coverage of services. As a result, the overall life expectancy at birth is estimated at 52 years, with healthy life expectancy being 44 years—both under the average values for the region and for the World Bank income group (World Health Organization 2015a). The World Health Organization’s Country Cooperation Strategy for DRC identifies the ‘fragmented health system’ and inadequate access to health services as critical problems in the country (World Health Organization 2015b). Indeed, the health sector is plagued by underfinancing, dysfunctional human resource management, the fragmented organization of the health zones (HZs) and a lack of leadership in the Ministry of Health (MoH) (Carlson et al. 2009). Some have argued that inefficiencies in the DRC health sector originate in the collapse of the state and the economy after three decades of non-governance (Ntembwa and Lerberghe 2015).

The health system in DRC has been progressively made and remade, and understanding the current system requires an overview of its trajectory of institutionalization, as well as the public policy-making process and patterns of inter-organizational interaction.

Evolution and institutionalization of public health in DRC

The health sector has been the state’s flagship division for public policy making since the colonial era. However, private actors and churches were also involved in public health management during the colonial period (Lyons 1994). After World War II, the Belgian Congo was divided into districts, with each district functioning as an operational unit with a medical centre and lower-level satellite health centres and dispensaries (Bukonda et al. 2012).

DRC became independent in 1960, inheriting a health system that relied on hospitals and clinics backed up by mobile teams for controlling major endemic diseases (RDC/MINISANTE 2006a). In the first decade following independence, the national health sector hardly existed, and non-state actors took charge of most health services (Lyons 1994). There was, however, an awareness of the need to reorganize the health system (RDC/MINISANTE 2006b). In 1973, the National Council for Health and Well-being was set up to design, direct and monitor the national health policy (NHP) (RDC/MINISANTE 2006b). In addition to major political developments, the 1970 s were marked by experiments with community medicine in DRC. These experiments gave rise to HZs, the country’s first decentralized health facilities (RDC/MINISANTE 2006a).

In 1975, the first national conference on community medicine adopted the principles of integrated care and proposed the creation of geographically defined HZs and the decentralized management of health care services. In 1984, DRC completed its health policy and strategy (RDC/MINISANTE 2006a). After the restructuring of the health sector in 1985, the country was to be subdivided into 306 HZs (World Bank 2005). This increased to 515 in 2001 (Waldman 2006).

Beginning in the 1990s, the need for a revised NHP was considered imperative. The government understood that setting up a NHP was essential for achieving sustainable socioeconomic development. The NHP was revised in 2001, following DRC’s subscription to a number of international declarations, such as the Alma Ata Declaration on Primary Health Care (1978), the African Charter for Health Development (1980) and the Bamako Initiative (1987). The NHP aims to promote the population’s health by providing a package of high-quality, globally integrated health care and continuously calling for community participation (RDC/MINISANTE 2006a).

In 2006, along with development partners, the DRC government adopted the first version of a Health System Strengthening Strategy, which was revised in 2010. The Health System Strengthening Strategy was intended ‘to impose better controls, licensing and regulation in the private sector through developing the capacities of Health Zonal, Provincial and Central health teams’ (Carlson et al. 2009: 18–19). The strategy was operationalized through a 5-year implementation plan covering 2011–2015. Together, the strategy document and the 5-year implementation plan constituted a turning point in the process of reforming public health in DRC regarding health care delivery and system management.

Health care delivery, system design and management in DRC

Following the Alma Ata Conference, DRC further developed its hierarchical structure of health care provision, organized at the three levels: central (national), intermediate (provincial) and operational (HZ) (Bukonda et al. 2012). At the central level, the MoH is responsible for general sector policy and system regulation, national programmes and tertiary hospitals (Waldman 2006).

The intermediate level has considerable administrative power over the HZs. Health workers, for instance, are supposed to be appointed by the state, even when an HZ is managed by a non-state organization. However, because the state is too weak to assert itself, these decisions are often made by non-state organizations such as churches; the government then simply endorses the decisions (Seay 2013).

Methods

This study explored interactions between relevant state and non-state actors operating at each level of the health system and also investigated the effects of these interactions on the health system and the development of the state. The study is based on 18 months of qualitative data collection conducted from September 2013 to April 2015 at the national, provincial and operational levels.

In South Kivu, one of the 26 provinces in DRC, we collected provincial-level data for the province as a whole and operational-level data in three HZs (Katana, Uvira and Idjwi). National-level data were collected in Kinshasa in January 2014. The study drew upon multiple methods of data collection; including open interviews, semi-structured interviews, focus groups and both direct and indirect observation (see Supplementary Table S1).

Interview and focus group participants included key informants from the MoH, representatives of international and local NGOs, health care providers and community members. Focus groups were conducted mostly with community members and frontline providers. Interviews were conducted with local community organization representatives, frontline providers, state officials and INGO representatives. Observation focused on state and non-state actors’ interactions. Interviews and focus groups with community members primarily covered their experiences with and perceptions of the state and non-state actors regarding health service provision. Focus groups with frontline providers covered interactive engagement of state and non-state providers in the local health system. Interviews with state officials and INGO representatives concerned health sector multi-actor management and its outcomes for capacity-building and service delivery.

Public officials participating in the study were mostly from the MoH in Bukavu and Kinshasa, and the Management Boards in 10 HZs. At the national MoH, interviews were conducted at the Study and Planning Department, the Primary Health Care Department, the Department of Partnership, the Cellule Technique du Financement Basé sur les Résultats and the Projet d’Appui au Renforcement du System de Santé. In Bukavu, interviews were conducted with provincial MoH workers and officials employed by the Health Inspectorate, the Partnership Office, the Primary Health Care and Data Analysis Office, the Study and Planning Office, the Office of Legislation and Management of Health Structures, and the Human Resource Management and Administration Office.

Non-state actors participating in the interviews and focus groups included representatives of the World Health Organization, UNICEF, the Dutch NGO Cordaid, Médecins Sans Frontiers Holland and France, Malteser, the Belgian NGO Louvain Coopération au Développement, the United Kingdom’s Department for International Development (DFID), the International Rescue Committee (IRC), Projet de Santé Intégrée (PROSANI), Coopération Suisse, Caritas Congo, Eglise du Christ au Congo, community-based health insurance offices and the Civil Society Office in Bukavu.

Community-related data were collected in Katana, Uvira, Bukavu and Idjwi through interviews and focus groups from a total of 500 ordinary community members. Community members were selected using convenience sampling. They were recruited at health facilities, community-based health insurance offices, marketplaces and schools, as well as in the villages and after church services. Selection criteria for these research participants were experience with local health care delivery or having an opinion on how different stakeholders engage in community health governance. Recruited regular community members were mostly from the rural (Katana and Idjwi) and semi-rural (Uvira) HZs (n = 470). Women were overrepresented in the sample (n = 351), as they attend health facilities more frequently compared with men, especially in the villages.

NVivo qualitative data analysis software was used for thematic data coding and analysis. The empirical data were transcribed and then initially coded according to the research questions. NVivo was used to conduct a thematic analysis, grouping relevant evidence on related thematic concepts. This process seeks to unearth and structure salient themes in the data (Attride-Stirling 2001).

Results and discussion

This section revolves around four points. First, we discuss the nature of state/non-state partnerships, mapping the system of health partners and their engagement. Second, we present evidence on the DRC health system as a governance network, analysing the DRC health system’s architectural set-up and operational functioning. Third, we present the findings on multi-level interactions for state-building outcomes. Here, the interactive engagement of state and non-state actors at different levels are presented as contributing to institutional capacity in health governance. Fourth, we discuss the arenas of interaction that show networked health governance.

State/non-state partnerships in the health sector

DRC’s public health policy identifies several types of partners: the state, private non-profit organizations, private for-profit organizations, the population and external funders (RDC/MINISANTE 2003). Among non-state partners, the government recognizes two sub-types: external and internal partners. External partners provide the sector with financial and technical support, whereas internal partners contribute to the sector with their expertise and act as service providers.

Based on negotiations, the government signs different kinds of agreements, contracts and memoranda of understanding based on typologies of interventions, as well as the scope and nature of activities. These include bilateral and multilateral agreements, frameworks and specific agreements, and protocols with a wide spectrum of actors. With internal partners such as churches, framework agreements are signed for health care provision and the co-management of public structures in critical situations.

The state engages in different ways with external partners. In operational partnerships, INGOs or donor organizations engage at the grassroots level by providing or supporting health services (e.g. Swiss Cooperation, IRC, Cordaid, Louvain Coopération). In bilateral cooperation, external partners engage with the state at central level (e.g. Belgian Cooperation, French Cooperation, Canadian Cooperation, USAID and DFID). In multilateral cooperation, external partners interact with both the state and their internal partners (e.g. World Bank, African Development Bank, International Monetary Fund, Global Alliance for Vaccine and Immunization, UNICEF and World Health Organization). Some external partners act at the interface between international institutions and national partners (e.g. USAID, UNICEF, Cordaid, USAID and IRC).

The full implementation of the Paris Principles on Aid Effectiveness, such as alignment and coordination, has been a contentious issue between the state and its partners. Most INGOs focus on vertical programmes, whereas the state wants to transition to system-building. In speaking about the real world of INGOs and their everyday interactions with the state, one top state official in the sector voiced a complaint: ‘They take advantage of the weakness of the state to implement their personal agenda’ (Interview, Kinshasa, January 25, 2014). Conversely, representatives of INGOs interviewed in Kinshasa and Bukavu asserted that mistrust between the state and external partners results from allegations of corruption.

In this study, health service providers’ and public health officials’ opinions converged concerning the beneficial impact of donors on health care, but their opinions on the strengthening of the health system differed. Until recently, donor interventions have been mostly vertical, hardly aligned, scarcely harmonized and uncoordinated. Parallel structures at the intermediate level are blamed for much of the limited progress in terms of sector building. Health service providers also reported that the health sector is weakened by the lack of state leadership and the absence of a shared vision for the health system.

Based on intensive interviews with public health officials at the provincial health department in Bukavu, state officials are cognisant of the effects of the enduring absence of state leadership in the health sector. To restore state leadership and stakeholders’ alignment, the MoH set up a consultation framework consistent with both strategic and operational needs. This framework [Comité National Pilotage Santé (CNPS)], chaired by the MoH, is composed of key stakeholders engaged in the health sector. The CNPS is duplicated at the provincial level [Comité Provincial de Pilotage Santé (CPPS)] under the leadership of the governor and is represented at operational level by the HZ coordination offices. For many informants from the provincial MoH, intervention harmonization and the coordination of INGOs have been a weak spot in health sector governance. However, this is being improved through the CPPS. Accounts from the provincial MoH showed that the CNPS/CPPS is a network governance arena where negotiations and horizontal interactions take place. The CPPS is an open forum for key stakeholders’ interactions, but the leadership role is played by the state. Therefore, the CPPS has characteristics of both a participant-governed network and a lead organization-governed network. In a focus group of experts regarding health sector governance, one of the participants made the following remark:

If there is something promising regarding the health sector governance, it is the introduction of the CPPS, which can so far be regarded as encouraging.

(Focus group, Bukavu, December 18, 2013)

Though much is still needed in terms of the MoH’s leadership and institution building, the CNPS provides a venue for consultations and intervention coordination.

The DRC health system as a governance network

Understanding the governance network operating within DRC’s health system necessitated an in-depth analysis of the system’s architectural set-up and operational functioning. In the DRC health system, state and non-state service providers have a long history of interaction and joint management. In addition to participating in service provision, non-state actors are involved in organizational management and the government’s national-level sectoral policy making. In South Kivu, for example, 13 of the 34 HZs are managed or co-managed by the Catholic Church, and seven are managed or co-managed by the Protestant Church. In many remote areas where the state is virtually absent, churches or NGOs are the only health service providers. Their activities are, in principle, regulated by the NHP. In this vein, the representative of BDOM asserted that ‘non-state actors engage as auxiliaries of the state’ (Interview, Bukavu, October 16, 2013). Similar views were expressed by representatives of INGOs such as IRC: ‘In their interdependent relations with public institutions, the state provides a legal framework for non-state actors’ (Interview, Bukavu, October 24, 2013).

Since the 1990s, literature on governance has focused on non-hierarchical modes of coordination and the role of non-state actors in creating and carrying out public policy (Börzel and Risse 2010). Interactions among multiple actors in the network may be characterized as either goal-oriented or serendipitous (Provan and Kenis 2008). Goal-oriented networks are extremely important as formal mechanisms for achieving multi-organizational outcomes, where collective action is often required. Serendipitous interactions develop more opportunistically, often within goal-oriented networks.

We found that stakeholders in DRC’s health sector interact in both goal-oriented and serendipitous ways. Goal-oriented interactions arise from the NHP recognizing the state’s long-term partnership with non-state actors (RDC/MINISANTE 2003). These interactions can also be seen as serendipitous, because they have often developed opportunistically throughout the state-building trajectory; this can be seen especially in the increased engagement of NGOs in the 1990 s in response to state fragility.

The DRC public health policy recognizes two categories of non-state partners: traditional partners such as churches, and situational partners, including INGOs operating in the sphere of humanitarian emergencies or health system development. The traditional partners have a long record of involvement, and some state functions have been formally delegated to them (Seay 2013). Situational partnerships are characterized by more dynamic processes, and interactions between actors are not imposed through authority structures or legal contracts (Jones et al. 1997).These different histories and characteristics of partnerships may have implications for policy-making processes and the effectiveness of multi-actor health service provision.

Exploring DRC’s health system through the lens of governance networks implies considering the system to be the product of an interactive process in which the locus, rules of the game and levels of interaction may have multiple forms. Previous work has distinguished three basic forms of governance networks: participant-governed, lead organization-governed and network-administrative organization (Provan and Kenis 2008). In these networks, decision making is carried out jointly (participant-governed), by one of the actors (lead organization-governed) or by a separate administrative entity set up for this purpose (network-administrative organization) (Provan and Kenis 2008).

Our analyses revealed that health system management in DRC is characterized by a blend of participant-governed and lead organization-governed networks. It embodies participant-governed networks because its functioning depends on the involvement of all committed partners, who interact in different arenas and on a relatively equal basis.

The health system management in DRC also has characteristics of a lead organization-governed network, as all major network-level activities and key decisions are supposed to be coordinated through and by the MoH. Although horizontality is promoted in sector-related decision-making processes, the shadow of state dominance in this regard is permanent. An example from our field work illustrates the point: During a March 2015 meeting at the provincial department of health, two INGOs presented findings from a survey on sexual and reproductive health. At the end of the meeting, the NGO representatives requested the Ministry’s endorsement for the validation of the findings. The request was accepted. This example shows that, although the state’s existence and value is sometimes disputed, its nominal shadow is ever-present in internal multi-stakeholder interactions.

Despite evidence of the state taking up some roles in health sector management, the community members involved in this study largely perceived the state as absent in the sector. This was illustrated by statements made during a focus group discussion with 20 people from Lwiro, South Kivu (Focus group, Lwiro, April 27, 2014). These participants were asked about their views of the state’s role in the health sector. A young man stood up and said, ‘With regard to the state, I have a feeling of profound and mortal regret [‘hasira ya kufa’ in Swahili]. I feel driven to war against the so-called state, but, alas, as I have no means to face the state by war, then I can but pray.’ A more senior man, of whom one would culturally expect a more nuanced view, added, ‘There is no state. It is completely absent. Be it in the health sector or in public administration, there is no state at all.’ The above responses reveal the population’s frustration, but also their insufficient knowledge concerning the administrative role of the state in health sector governance.

Perceptions of the absence of the state in the health sector appear to have contributed to an overall negative view regarding the state. However, despite community members’ perception that the state plays no role in the health system, the findings presented in this section reveal that the state does take up some limited roles in this system. This initial involvement might be seen as the early stages of a fuller participation of the state in the health sector governance network.

State-building outcomes of multilevel interactions in the health sector governance network

Interactive engagement of non-state actors at different levels of the health system contributes to improving institutional capacity. In addition to their funding role, non-state actors engage with state institutions for health system management and institutional development. For this reason, DFID’s health officer in Kinshasa noted that donors are committed to strengthening state institutions, noting that, ‘with this goal, the DRC MoH and its international partners have set up a Management Support Unit [Cellule d’Appui et de Gestion] and a Finance Agency [Agence de Gestion Fiduciaire] for management capacity strengthening’ (Interview, Kinshasa, January 24, 2014). Both agencies are multi-actor mechanisms for donor interventions coordination at national level (ADE, 2014, p. 99).

Based on our observations of the DRC health sector, we realized that the involvement of non-state actors at the central, intermediate and operational levels indicates the presence of a governance network because of the patterns of multi-level interdependencies and collaboration between the state and its non-state partners.

The DRC health system is de-concentrated at provincial level through the HZ coordination offices representing the MoH and decentralized at operational level by means of the HZs’ operational autonomy. The governance network including the state and its development partners is multilevel in terms of interaction arenas but also multidimensional in terms of the nature and content of interventions. Interactive governance in the health sector is embedded in, and possibly affected by, structural arrangements made within the institutional landscape.

Interactions among state and non-state actors in the health system reflect the architectural shaping of this system. Networked governance operates in various arenas (national, provincial and HZ) of the health sector, with either traditional non-state partners (faith- and community-based organizations) or development NGOs interactively engaging with the state in the management of the sector. A representative of the MoH maintained that ‘The NHP encourages the participation of non-state actors in the management of health system’ (Interview, Kinshasa, January 15, 2014).

Impaired by advanced state disruption, state and non-state actors interactions take different forms—the embodiment of integrative network governance for sector strengthening, the reflection of an aggregative body with disparate interests or even quintessential manifestations of statelessness and chaos. With integrative actors such as churches and traditional bilateral/multilateral cooperation partners, it is possible to build an interactive ‘policy community’. However, with occasional actors—mainly humanitarian INGOs and private for-profit organizations—interactions are generally difficult, because these occasional actors pursue their own agendas. In an individual interview, a top official from the MoH’s National Partnership Department commented on this point:

Donors come with their own agenda according to their respective interests, and as a consequence there is always a priority-setting problem between the state and donors.

(Interview, Kinshasa, January 24, 2014).

Nonetheless, for many interviewed representatives from the state civil service, INGOs, churches and frontline providers, interactive governance including state/non-state partnership accounts for the persistence of the health sector in DRC. A representative of the MoH asserted that ‘Interactive collaboration with non-state actors contributes to the rationalization of health care delivery processes and strengthens the health governance’ (Interview, Kinshasa, January 23, 2014). During an interview, a health clinic administrator in South Kivu praised the work of donors:

International and local NGOs have been working and reaching out where the state is absent. NGOs have a better understanding of the health territory than the governor and even the provincial health department. The latter learn from and refer to international and local NGOs’ demographic health records to get an image of what the situation looks like.

(Interview, Bukavu, March 9, 2015).

Our findings demonstrate that partnership with non-state actors was often regarded by state actors as the only way they could improve the governance of the sector and thus live up to the population’s expectations. In one of the focus group sessions with state officials at the Provincial Inspectorate of Health, a participant posed a rhetorical question: ‘What would have become of the public health sector without the involvement of non-state actors?’ (Interview, Bukavu, May 17, 2014). This opinion, which our results indicate is widely shared by Congolese actors of all outlooks, points to the relevance of non-state actors for health sector building as well as the management of health services in DRC.

Arenas of interaction and methods of engagement in the health sector

The pyramidal configuration of the multilevel governance of DRC’s health system requires an analysis of the different methods of engagement used by the actors involved that considers three levels of interaction.

Macro-level networked governance

At the central level, the health system is an arena of top-level interactions for horizontal discourse among stakeholders for the purpose of creating a policy community. Interactions take the form of negotiations, which lead to collaboration and cooperation around frameworks and specific agreements. The MoH makes up this central level and has a standard-setting and regulatory responsibility. The central level is an arena where multilateral and bilateral contracts and framework agreements are negotiated and signed. Mechanisms such as CNPS, the Management Support Unit and the Finance Agency facilitate interactive collaboration for national level policy coalition building. In this vein, Cordaid’s national health officer maintained that ‘high level multi-actor arrangements are handled at national level’ (Interview, Kinshasa, January 26, 2014).

The macro level is the recommended entry point for international engagement, but, through de-concentration and decentralization, structures at the provincial and HZ levels have become entitled to handle certain issues. Still, for partners whose interventions require framework agreements such as multilateral or bilateral contracts, the current configuration implies direct involvement with the national MoH. At this level, negotiations usually result in signed contracts and conventions, frameworks and specific agreements of national scope. The NHP, the Paris Declaration and the International Health Partnership are reference frameworks for signing bilateral and multilateral agreements.

In interviews, state officials noted that negotiation is not always smooth between the stakeholders, and power dynamics often determine the outcome. In the negotiation process, donors enjoy expert legitimacy and financial leverage for policy making and enforcement; the government, although it is weak, brandishes its juridical sovereignty. State officials use the sovereignty argument politically as a way of securing non-state actors’ alignment and defending against critics of the failed state. This was conveyed during an interview, when a provincial state official made the following remark about donors’ engagement in the health sector:

Donors’ inputs in the management of the health sector are laudable, but this does not give them leeway to do whatever they like, as some do; they should not be taking policy decisions on behalf of the state.

(Interview, Kinshasa, January 14, 2014).

From state officials’ point of view, donors have the upper hand because of the weakness of the state, combined with financial resources that influence national politics and claims of expert knowledge. Agreement frameworks nevertheless include non-negotiable matters related to sovereignty and the NHP.

The intermediate arena of interactive governance

The intermediate level is the location for managing national policy at provincial level, regulating health care provision processes and accommodating population expectations. The intermediate level is normally responsible for technical support and NHP implementation supervision at the operational level. To fulfil its public mission, the state’s health department engages with development partners to support state interventions. This includes signing different deeds of collaboration at the intermediate level of the multilevel governance network. At the intermediate level, protocols and addendums may be signed, depending on the projects. The intermediate level hosts the signing of memoranda of understanding and conventions related to packages of activities for programmes with framework agreements already signed at the central level.

Depending on the type of interventions being carried out, external interveners can be classified either as developmental and working to provide institution-building support, or as humanitarian, with emergency-based interventions. Interviews with provincial health department representatives indicated that, of over 100 INGOs engaged in the health sector in South Kivu in 2014, only four engaged in sector strengthening and took a developmental approach. However, informants from the national and provincial MoH conveyed the state’s aspiration for an intervention paradigm shift to the institution-building—or integrated—approach. For the provincial Inspectorate of Health, the organizations engaged in development were IRC/PROSANI, Malteser, the Belgian NGO Louvain Coopération au Développément and the Dutch NGO Cordaid (Interview, Bukavu, October 30, 2013). These INGOs provide comprehensive support to the government for system strengthening.

At the provincial level, there are also sub-arenas of interaction where stakeholders play interactive games, including the Cluster Santé thematic group for non-state actors and the CPPS. Both venues duplicate and reflect the workings of the two frameworks at central level. The CPPS as an inclusive framework is par excellence typical of networked governance, where stakeholders directly interact, negotiate and use their capital leverage in rationalistic trade-offs. For a provincial representative of the Partnership Office, ‘The CPPS is a mechanism through which the state is taking over its leadership role regarding the coordination of health interventions’ (Interview, Bukavu, November 17, 2013). In the CPPS, all interventions are supposed to be rationally oriented according to both national policy and context priorities.

However, for many informants, the CPPS is a new process, and its implementation is still difficult, because policy alignment has been a significant problem in interactions between the government and external partners. Some HZs, especially near Bukavu, are reportedly saturated with local and INGOs. In contrast, several HZ directors in remote zones, which are also the most war-affected areas, said that their HZs suffered from the absence or insufficient presence of these organizations.

The operational arena of interaction

The NHP is operationalized in the HZ operational arena. This is the arena where interactions may take place among state representatives, service providers, local and international NGO interfaces, local faith-based organization representatives and community leaders. The HZ central office is normatively the setting for interactive operational planning, monitoring, evaluation and activity reporting for primary health care. HZ business meetings develop an operational anatomy of the needs of the population that is then submitted to the intermediate level—or, for the luckier HZs, to local or international NGO partners. As is the case for other levels, non-state actors such as churches and local and INGOs bargain with one another based on their respective expectations and power positions. They also proceed with operational planning and possibly with operational contracting for health output performance.

During the field work, local health system managers reported rivalry and competition between state and non-state actors—especially churches—in some HZs regarding staffing. These informants described instances where churches or INGOs imposed their individual preferences regarding personnel appointments. There were also reports that the HZ has sometimes been the setting for ‘turf wars’, especially between Catholic and Protestant churches tacitly competing for influence. However, it is not only churches that engage in competition; informants reported horizontal competition among local and INGOs with overlapping programmes, especially in areas where many such organizations were present—mostly in the northern part of the province. The HZ is therefore not only an arena of interaction, but also an open ground for factional influence and competing struggles between state and non-state actors, among local and INGOs or their interfaces, and sometimes between these NGOs and the HZ coordination offices.

At the operational level, there can be conventions/protocols with the HZ management office along with NGOs providing health care to the population. Some agreements and memoranda of understanding can be signed at this level, depending on the partner’s agenda, flexibility and contextual needs.

Conclusions

Health sector management in DRC is the embodiment of a burgeoning governance network in which a wide range of stakeholders interact to solve public and community health-related problems. Interactions between the state and non-state actors in DRC is not a new phenomenon; what is quite new are the new patterns of relationships that are evolving into networked health sector governance after decades of disorganized management of the sector. Four themes emerged from our analysis.

First, the health sector in South Kivu is emerging as an arena of networked governance. The health sector governance network is not yet fully developed in DRC, but there are indications of state and non-state service providers negotiating and cooperating through different sets of institutions, arenas, games and processes.

Second, interactions between state and non-state actors through this governance network contribute to explaining the persistence of the health sector in a setting characterized by state weakness. Longstanding patterns of negotiation-based interactions between state and non-state actors in the health sector are enacted at the three levels of the health system. The concrete value of state/non-state interactions and partnerships for the maintenance and development of the health sector was acknowledged by a wide range of actors participating in this study.

Third, during interactions, power relations are skewed in donors’ favour because of their resource dominance, a situation that causes the state to play a limited role in managing the health sector. It is clear that non-state interventions fill the void where the state falls short, and non-state actors have undoubtedly contributed to meeting population needs. In addition to being hampered by its weaknesses in terms of a lack of capacity for governance and allegations of management misconduct, research we conducted in South Kivu revealed that the DRC state has also faced the challenge of interacting with partners with fragmented and horizontally competing agendas. Consequently the governance network does not function optimally in DRC’s health sector.

Fourth, though it remains weak, the shadow of statehood authority is present in the arena of stakeholders’ interactions. The state plays a determining role by providing a regulatory framework and hence in managing the formal room for manoeuvre available to non-state actors. However, the state has not yet become stronger through this process, resulting in the persistence of its lack of effectiveness.

Ultimately, although the health sector governance network enacted in South Kivu cannot fully address state weaknesses, networked governance facilitates the management of population needs regarding health welfare. However, strengthening the role of the state in networked governance is a requisite for reinforcing the stewardship role, which is crucial for the state-building process. Therefore, the issue of networked governance effectivity raises normative concerns about the nature, model and priorities of engagement for policy community interventions in weak states. These emerging concerns, though beyond the scope of this study, are important for understanding the effectiveness of networked governance in fragile states and should be explored in future research.

Supplementary Data

Supplementary data are available at HEAPOL online.

Acknowledgements

The author is indebted to Prof. Hilhorst Thea, Prof. Mashanda Murhega and Dr Jacobs Carolien for guidance and field research follow-up. The author is also grateful to the south Kivu provincial minister of public health, Mr Mwanze Nyangunya, for his moral support; and to Dr Jennifer Barret for heartily editing work.

Funding

This work was supported by UK Aid from the UK government [P05112-Secure Livelihoods Research Consortium] and by Irish Aid [Secure Livelihoods Research Consortium: Proposal to Irish Aid for a Funding Agreement 2014–17].

Conflict of interest statement. None declared.

References

Author notes