Effective responses: Protestants, Catholics and the provision of health care in the post-war Kivus

Introduction

The eastern Democratic Republic of Congo (DRC) provides a useful setting in which to examine the task of non-state actors as substitute providers of services that are traditionally the purview of the state. The Congolese state's authority has not extended to its eastern territories in any meaningful way since 1994. Meanwhile, 15 years of local, civil and international conflicts further contributed to the weakening of state institutions already in severe decline. In this situation, non-state actors often serve as the creators and maintainers of social order in what would otherwise be a chaotic region.

However, in a fragile state in which the authority of the central government is extremely limited and local government struggles to maintain basic public order, all is not chaos. Surviving civil society organisations (CSOs) – along with an influx of external intervening forces and humanitarian agencies – operate schools and hospitals, and provide security in at least some areas, particularly major cities. In the process, they often function as de facto legitimate authorities with power to make decisions about policy, employ personnel, and manage institutions. Because of their capacity in the social-service sector, these organisations often have far more influence on public life than state actors. While certainly not a desirable situation, this assumption of authority by non-state actors does allow for some children to be educated, some of the sick to be made well, and for participants in the informal economy to maintain at least a minimal level of activity. But there is considerable variation in the ability of different CSOs to deliver social services.

This article seeks to explain why, in a situation of extreme state fragility, some CSOs are more successful at building social order than others by examining the health-care sector in the Kivu provinces of the eastern DRC. Based on data collected during field research in Bukavu and Goma between July 2005 and August 2007, and in a follow-up study in Butembo, Beni and Bunia in July 2010,¹ it presents a theory of internal organisational cohesion to explain why some churches and other CSOs are more successful at providing health care than others. This theory suggests that a CSO's ability to successfully organise social services is directly related to the degree of cohesion or fragmentation within the organisation. A number of historical, demographic and institutional factors cause some groups to develop stronger social capital networks than others, which in turn affects the degree of effectiveness an organisation will have in providing social services in the state's absence. It introduces a measure by which the effectiveness of social-service provision by non-state actors can be evaluated, and then applies that model to the provision of health care in the eastern DRC. Taking into account the fact that standard measures of effective service provision such as maternal mortality or disease prevalence rates may not be most applicable in a situation of extreme violence and poverty, a measurement system based on 15 indicators of the effectiveness of health-care service providers is used. Finally, the article questions how civil society groups that substitute for the state affect authority structures and long-term prospects for state reconstruction.

Social services in a fragile state

When states weaken, social services are among the first government-provided services to decline. Health care is expensive, and leaders who lack revenue streams and authority are typically more concerned with paying the army or shifting funds into their personal bank accounts than purchasing medication. If the state becomes fragile, government often becomes completely incapable of providing social services and human suffering becomes widespread (Milliken and Krause 2002, p. 764, Brennan and Nandy 2001, p. 148). Professionals in the public-health sector often have economic incentives to seek employment in the private sector, where they can at least be guaranteed a salary. This further contributes to the decline of the public sector. Since the state no longer has the ability to fully secure its territory, conflict almost always disrupts public-goods provision. Failed states become vulnerable to predation by outside forces and can pose a danger to neighbouring states (Rotberg 2002). Finally, in the downward spiral, elites who have learned to profit from the state's absence may have little incentive in re-establishing state authority and order (Menkhaus 2003).

Yet local populations still need basic social services. Where government is incapable of delivering them, local CSOs fill the gap. Often operating in de jure partnerships with a weak state, these organisations typically have almost full control over their own activities and the ‘state’ apparatuses they manage. Many establish partnerships with international donors and non-governmental organisations (NGOs) to support their work, but final say over the organisation's social-service programmes belongs to the CSO and its leaders. In contrast to most instances of civil-society-based service provision in which CSOs are in partnership with the state, in many African states, as Clayton et al. note, CSOs ‘often … operate with little reference to state providers’ (Clayton et al. 2000, p. 5). In such cases, elements of civil society often serve as the only social-service providers. They effectively substitute for state regulation, management and authority. As Hagmann and Péclard (2010) note, the reality of the state that results is based on ‘negotiation, contestation, and bricolage’ between the variety of state and non-state actors involved in providing services and governing.

In the DRC, these effects are complicated by the state's continuing existence that paradoxically exists alongside its weakness. Vlassenroot and Raeymaekers (2004a) show that the Congolese state's fragility opens new opportunities for forms of political engagement by local elites. Englebert (2003) examines the weakness of the Congolese state that nonetheless ‘persists’, and concludes that the state continues to legally exist because local elites and international actors continue to want it to. Tull (2005) concludes that while the events of the last two decades significantly changed political life in North Kivu, it is too early to declare the death of the Congolese state. Following Englebert, he argues that there are too many stakeholders in the existing weak Congolese state to allow it to go away completely. Lund's (2006) notion of ‘twilight institutions’ eloquently captures the way in which post-conflict state institutions persist in a hybridised form that can be described as ‘non-state’ but also ‘non-private’. State institutions in the DRC do exist, but in the east they generally operate not as agents of government authority, but rather as institutions that are, in Lund's words, ‘not the state, but [ones that] exercise public authority’ nonetheless. Individuals in positions of ‘state’ authority may try to administer programmes and attempt to regulate activities, but they have typically not received a government salary for 10 to 20 years and act according to their own self-interest rather than as agents of the state. In most of eastern DRC, ‘state’ officials are actually independent agents who collect rents from the population and use it for their own benefit, enforcing regulations and collecting taxes only when it is convenient to do so. Every ‘state’ service has a price, be it a building permit, a medical examination, or a trial. The fees paid for services are collected in lieu of taxes, but those fees are deposited directly into the pockets of the service providers, not into the state's coffers. However, those fees may or may not lead to the effective provision of the type of service paid for, as state capacity is so limited. In the health-care sector, a patient may pay a government-run hospital for services, but receive ineffective and inadequate treatment as the hospital will likely not have adequate supplies with which to treat the problem. Likewise, while government health employees accepting fees for services may attempt to treat an illness, they often cannot effectively do so due to a lack of expertise and access to high-quality continuing professional education.

The church and the state in DRC

Congolese churches have long played a leading role in the country's civil society, and the Catholic Church's dominant role in Congolese political affairs is well-documented (Oyatambwe 1997, Makiobo 2004, Okitembo 1998). In the social-service sector, Congolese CSOs generally cooperate with local authorities, but they in many ways function as the state by operating institutions, managing disputes, collecting user service fees in lieu of taxes, choosing personnel (who are often technically state employees, but who are actually hired and paid by the CSO), providing security and maintaining a basic level of public order. As one subject noted, since instability began, ‘Civil society has grown stronger and stronger, while the state has become weaker and weaker’ (author's interviews 2005–2010). Officially, churches in the eastern DRC simply partner with the government to provide health care; in reality, they substantially take over for the government and are often subject to very limited substantive government oversight.²

Since the management of most of the Kivu's ‘state’ social-service institutions is essentially contracted out to local CSOs – primarily to churches and the Islamic community – in most cases, true authority in the social-service delivery domain lies with those who operate and finance those structures. Despite the state's weakness and civil society's role as a de facto service provider, both entities do maintain a formal relationship based on contractual agreements. These agreements require the state to undertake certain functions (hiring and firing personnel, constructing and rehabilitating buildings and paying staff salaries) and CSOs to perform the day-to-day tasks of operational management. As one subject stated, ‘Theoretically, it is the state’ that bears responsibility for oversight and management of Congolese civil society (author's interviews 2005–2010), but the reality is that while decisions about hiring personnel, constructing and rehabilitating facilities, or salary levels need nominal approval from titular state authorities, actual decisions are often made by the church bureaucracies, and what remains of government typically endorses decisions that have already been made. While religious authorities do not have absolute power to do whatever they want (and are largely constrained by donor preferences), they usually have more capacity than Congolese state authorities, who often lack even the most basic elements required to conduct regular inspections of the facilities for which they are responsible. For example, most health care professionals interviewed in 2007 did not have petrol to fuel their vehicles and were thus unable to conduct their required inspections. One interview subject (along with others) noted that some health professionals with 30 years' employment by the government have never been paid official salaries (author's interviews 2005–2010). Much lip service is paid to state/non-state collaboration, and important efforts at state capacity-building in the health-care sector (namely the European Union's 9eme Fond) are underway, but in the period of this study, religious actors in many ways had the final say over the means by which they deliver health care.

Religious institutions are the primary health-care providers in the DRC today (Trefon 2011, pp. 127–128). This does not mean that political processes, disputes and international assistance do not affect health-care delivery in the DRC; to the contrary, health-care delivery is as much a political process in the DRC as it is anywhere else. One example of the potential for politics to cause confusion in health-care delivery is the creation of stronger provincial governments that was introduced in the DRC's 2005 post-war constitution. The new constitution provides for a strong decentralisation programme and allocates 40% of the national budget to provincial governments. However, as of mid 2012, this money has yet to be allocated to the provincial governments, leaving most provincial ministers of health idly waiting for something to do and money with which to do it. It is not entirely clear how responsibilities for management and oversight of the health-care system are to be divided between the new provincial authorities and the old state health officials. Meanwhile, those pre-existing state health authorities, the provincial health inspectors, continue their administrative work in close collaboration with those who pay their salaries, namely international donors, NGOs and local religious authorities.

Catholic organisations

The Catholic Church has a long history of involvement in social-service provision throughout the DRC. Catholic schools are widely regarded as the best in the region, such that any parent who can afford to do so sends his or her children to the elite Catholic schools in Bukavu and Goma (author's interviews 2005–2010, Tull 2005, p. 206). Likewise, the Catholic Bureau Diocésain des Oeuvres Médicales (BDOM) is a major provider of health services in North Kivu and South Kivu, serving as the primary partner of one health zone in Bukavu and another in Goma.

Why are Congolese Catholic social-service institutions so effective? First, in Bukavu and Goma, the Catholic Church's history helped to strengthen the organisation. As part of the Belgian Congo's ‘colonial trinity’, the church has always held a privileged position in Congolese public life. The work of the church and its missionaries was necessary for the colonial authority's ‘civilising mission’. Church schools inculcated values of hard work, loyalty and basic literacy skills in the territory's future labouring class, a service that the colonial state rewarded with large state subsidies for the Catholic schools and support for the missionaries who ran the stations on which those schools were established (Young 1965, pp. 10–20). State subsidies were limited to Catholic schools until after World War II, which made establishing and maintaining schools at a reasonable cost to parents significantly easier for the Catholics. In addition, unlike many other expatriates living in the Congo at the time of independence and the subsequent five years of civil war and rebellion, many foreign Catholic priests, monks and nuns returned to the Congo to serve in churches and hospitals after the independence era. They provided expertise in the social-service sectors and helped build capacity among Congolese Catholics who eventually took over the operation of their own institutions. While certainly there was competition among Catholic missions in the colonial and post-colonial context, and while there have been many power struggles within the church, the church has nonetheless maintained its position as the strongest institution in Congolese society. After independence, the status quo of health-care provision by religious actors continued largely uninterrupted until Mobutu committed in the early 1970s to a nationalisation programme known as Zairianisation. Relations between the Catholic Church and the Zairian state became increasingly complex in this period, reflecting the renegotiation of the relationship between civil society and the state that occurred in that decade (Young and Turner 1985). By mid decade, the state's primary concern was co-optation of civil society into its ranks. The Catholic Church was one of a very small number of institutions able to partially withstand Mobutu's attempt to ‘Zairianise’ (and personalise his control over) every aspect of Congolese life. When in 1973 Mobutu nationalised church social-service systems, the disastrous effects on the health care and other social-service systems were so immediately apparent that contractual arrangements returned the management of most of these structures to church hands in less than five years.

Also contributing to the institutional strength of the Catholic Church is the fact that, unlike their Protestant counterparts, Catholic churches cannot divide over theological or social disputes. Protestant churches in the Kivu region are prone to splitting along ethnic lines. But splitting on the basis of ethnicity is not an option for members of the one holy, universal and apostolic Catholic Church. This social cohesion helps to mitigate difficulties that could arise as a result of disputes that divide members on ethnic lines. That the Catholic Church cannot divide along ethnic lines means that politicians and civil society leaders who are Catholic are able to draw on a broader range of societal support for their institutions than their Protestant counterparts. This finding stands in contrast to a large body of existing literature on the relationship between ethnic homogeneity and successful service delivery which argues that less diverse communities are typically more successful at providing public goods, a hypothesis that seems to hold true across a wide variety of regime types (Banerjee et al. 2005, Alesina et al. 1999, Easterly and Levine 1997, Miguel and Gugerty 2005, Poterba 1997, Goldin and Katz 1997). While the sample size in this study is too small to draw definitive conclusions about this question, these findings do suggest that Habyarimana et al. (2007) are correct that certain institutional arrangements and social norms can make it possible to overcome barriers to cooperation in ethnically diverse communities. However, further study is needed to determine whether ethnic homogeneity plays a role in determining the effectiveness of social-service provision among Congolese Catholic archdioceses, specifically with reference to the question of relationships between bishops of one ethnicity and parishioners who are largely of another ethnicity.

The Catholic Church is one of the few institutions in Congolese society that never collapsed. While by the mid 1980s, national and local governments had lost almost all of their capacity to accomplish even the most basic tasks of meeting payroll or maintaining public order, the formal economy ceased to function. Informal means of wealth generation became the only way to survive (MacGaffey 1991), transportation infrastructure rotted, the banking system collapsed, and the postal service stopped delivering mail. But the church's ability to organise its members, accomplish tasks and maintain a basic structure of public order never disappeared. Except in areas of open fighting, Catholic institutions even stayed operational, and throughout the turmoil of the two Congo wars (1998–2003), leaders in the Catholic church maintained the ability to actively engage in politics and to influence national political outcomes as well as to maintain social-service structures.

Additionally, Catholic schools and hospitals can still depend to some extent on a global network of financial support from other Catholics worldwide. This is particularly true of the BDOM's health services, which continue to benefit from the generosity and assistance of the global Catholic Church through agencies like Catholic Relief Services and Caritas. As Catholic social teaching requires the faithful to care for the poor, it is not surprising that Catholic social-service institutions can call on the support of other Catholics worldwide for assistance in times of great need. This high level of external support is further evidence of the high degree of internal organisational cohesion in the Catholic Church in the Kivus, particularly in Bukavu. Despite the turbulence and conflict that have affected the Kivu provinces for more than 15 years, Catholic organisations have largely managed to maintain or rebuild their social-service infrastructures. Even with the disruptions of war and the poverty of the vast majority of their clientele, Catholic schools and hospitals in Bukavu are generally in better shape than their public counterparts and many of those structures supported by other civil society institutions. Catholic institutions in Goma are often less strong, but relative to competing public institutions, Catholic schools and hospitals in the city work fairly well (author's interviews 2005–2010). This success is partly due to the support from international donors in the post-war reconstruction efforts, but the long history of Catholic involvement in the region and early financial support from the colonial authorities also meant that infrastructures were generally in better shape before the wars. The high degree of ethnic diversity in the Goma archdiocese may explain why the group is slightly less effective than its counterparts in other archdioceses at providing services.

Finally, the role of Mobutu's patronage networks also influences the degree to which a CSO is cohesive, generally in a negative manner. It is well-known that former Zairian President Mobutu Sese Seko used the state's resources to maintain his control over the territory and its institutions. Mobutu actively maintained patronage networks in the Kivus, and part of his activity there was providing land grants in the rural areas (particularly in North Kivu) to the Catholic Church. The effects of this relationship were actually favourable to the church's organisational cohesion during the Mobutist period, especially in the North Kivu archdioceses, as it gave its institutions a financial advantage over its competitors in the social-service sector. However, because these issues are so closely related to the disputes over land that drove North Kivu to conflict in the early 1990s (Vlassenroot and Huggins 2005), and because those disputes had such a heavy anti-Tutsi ethnic taint, what was beneficial during the Mobutist era became in some ways a drawback in building internal organisational cohesion during the war and post-conflict periods (author's interviews 2005–2010). The Bukavu archdiocese is significantly more ethnically homogeneous, meaning that social pressure and close social ties lead to increased cooperation, which leads to more effective service provision over the long run.

It is difficult to overstate the importance of internal cohesion in the Catholic Church's success as a CSO and in operating schools and health facilities in the state's absence. Although the historical trajectory and ethnic diversity of Goma makes its Catholic structures slightly less successful than those run by the archdiocese of Bukavu, both archdioceses still provide much better social services than do many of their Protestant counterparts.

Protestant organisations

Like the Catholics, Protestants have a long tradition of missionary and church involvement in providing health care to citizens of the Kivus. While they did not have the same level of access to subsidies as did the Catholic schools for most of the colonial era, Protestant missionaries could depend on support from congregations in the United States and Europe to start schools, and they worked to make their service structures and churches self-sufficient from the beginning. When subsidies became an option for Protestants after political changes in Belgium in the late 1940s and early 1950s allowed for broader state support to religious organisations, at least one Kivu religious organisation, the Mission Baptiste du Kivu, divided into two churches in part over the question of whether its schools should accept state subsidies. The organisation now known as the 3eme CBCA (Communauté des Baptistes au Centre d'Afrique) accepted subsidies, a decision that made it more cohesive, while the 55eme CEBCE's (Communauté des Églises Baptistes au Congo-Est's) refusal to accept subsidies and longer-term reliance on missionary support eventually contributed to an institutional history that is not associated with a high degree of internal organisational cohesion (Nelson 1992).³

The long-standing presence of some Protestant churches in the region means that some have been able to organise effective networks of support for their activities. As with the Catholic Church, the ability to raise external and internal financial support, an aspect of internal organisational cohesion in a CSO, has made some of the Protestant churches a strong force in social-service delivery. In North Kivu particularly, contributions from wealthy businessmen have furthered the success and strength of many Baptist institutions. In both provinces contributions from churches, other faith-based organisations and international NGOs have helped many Protestant churches to operate long-standing health-care programmes (author's interviews 2005–2010, Nelson 1992, p. 68).

As with the Catholic Church, Protestant churches did not collapse along with the Zairian state. They were also among the only social institutions to survive the political, economic and social chaos of the war period, and provided stability and a basis for order when neither the state nor the economy could do so. Although their operations in rural areas were and continue to be interrupted during periods of fighting in areas of intense conflict, such as Masisi, Walikale and Rutshuru in North Kivu and Shabunda and the Haut Plateau of South Kivu, by and large, the Protestant churches stand with the Catholics in providing the only functioning institutions in what are highly insecure and sometimes anarchic environments. They also gained a high degree of trust among the population. As one church official in Bukavu noted, while ‘the population has problems under the government, the church has a lot of credibility’ (author's interviews 2005–2010).

A key difference between the Catholic and Protestant churches, however, involves the role of ethnicity. Unlike the Catholics, Congolese Protestants have no external hierarchy to impose order or to force them to stay together, especially in those forms of Protestantism that are in the Free Church tradition. Because local churches in the Free Church tradition are almost completely autonomous, if there is a dispute among a church's members, dissenters are generally free to leave to start another church. In a region in which questions of ethnicity are often paramount, particularly with respect to the place of Kinyarwanda-speakers in Congolese society, Protestant organisations can and do split, largely along ethnic lines (Nelson 1992). Part of this is due to early mission efforts. Faced with the task of learning unfamiliar languages in a difficult environment, the first missionaries generally initially worked among one regional ethnic group, which created missions and then churches that were largely ethnically homogeneous. Later disputes over such issues as school subsidies caused other missions to split into groups that were, if not completely ethnically homogenous, typically comprised of a significant majority from one ethnic group.

A theory of internal organisational cohesion

Many Congolese CSOs are attempting to deliver social services, but there is considerable variation in their ability to accomplish this task. Some Congolese health-care facilities are well-stocked with drugs and medical supplies and employ adequate numbers of health-care professionals who have the knowledge and skills necessary to treat patients. Others do not. What accounts for this variation? What factors influence a CSO's ability to provide social services in the state's absence? The central argument of this article is that a CSO's degree of internal cohesion strongly influences its ability to successfully deliver health care. The independent variable, ‘internal organisational cohesion’, refers to the degree to which the members of an organisation share a common history and identity and belong to unified social networks. These networks enable members of the organisation to work more closely together and increase the likelihood of attaining their goals. A high degree of internal organisational cohesion strongly correlates with a higher rate of success at organising social services.

Internal organisational cohesion can be assessed according to several indicators. These include institutional history (whereby the CSOs presence in the country during the colonial period contributes to cohesion, division along ethnic lines diminishes cohesion, and receiving subsidies from the colonial regime contributes to cohesion). The more a CSO's membership is ethnically homogeneous, the more likely it is to be cohesive and the more likely it is to successfully provide social services. Likewise, the greater its ability to obtain external funding, the more likely it is to be cohesive. Finally, whether the organisation was a beneficiary of government patronage networks during the Mobutu era (1967–1997) is of particular concern. Being a beneficiary of Mobutu's patronage diminishes cohesion.

Each indicator of institutional organisational cohesion is assigned a value of ‘1’ or ‘0’ based on information gathered during interviews and from archival research. An indicator is coded as ‘1’ if the organisation has a long-time presence in the region (30 or more years), has no history of division on ethnic lines, received subsidies from the colonial regime, has a high degree of ethnic homogeneity, received a high level of support from external sources and avoided being tainted by Mobutu's patronage networks, and vice versa for null values. A simple, additive scale then determines the level of internal organisational cohesion. CSOs with scores of 5 or 6 are considered highly cohesive, those with scores of 3 or 4 are considered moderately cohesive, and those with scores of 2 or less are considered to be highly fragmented. Thus, a high degree of internal organisational cohesion is the product of an institutional history of presence in the region during the colonial period, no experience of division, receipt of subsidies from the colonial authorities, an ability to obtain strong external support, and not having been tainted by association with patronage networks in the independence period. By contrast, moderate and low levels of internal organisational cohesion are associated with a shorter presence in the region, experiences of division, not having received subsidies, not obtaining external support, and having been closely associated with Mobutist patronage networks.

The dependent variable, ‘the successful organisation of social services’, is a measure of how effectively a CSO is able to operate health clinics, hospitals and other structures that continued to operate during and after a period of state weakness and war. Measuring the quality of social services is difficult in fragile states due to missing data and other problems. It is virtually impossible, for example, to know whether low usage rates at a hospital or high infant mortality rates in a neighbourhood indicate poor quality health care or simply that the local population is too poor to afford the services. Likewise, the expense and distance of travel to some health facilities means that the popularity or extent to which individuals seek services at a particular facility is not a reliable indicator of the provider's effectiveness. Therefore, this study does not focus on the quality or popularity of social services provided by different CSOs. Instead, it relies on a series of 15 operational indicators of the successful ability to organise social-service delivery, including physical infrastructure, organisational hierarchy and organisational scale. An additive scale rates each organisation according to the number of attributes each has, and use that rating as a basis for determining whether a civil society group successfully organises social services. These attributes are outlined as a series of questions in Table 1.

For each of the indicators in Table 1, values are assigned to the answers as follows. For ‘yes/no’ questions (nos. 1–14), a ‘yes’ answer is assigned a value of 1. For a ‘no’ answer, it is assigned a value of -1. Values were determined based on data gathered in interviews, observation and archival research. When data are not available or the question does not apply to an organisation, no value is assigned. For indicator 15, organisations are scaled on the basis of the total proportion of a city's social-service structures that each operates and whether the church operates a major hospital in the city. If a group operates less than 10% of the city's service structures, it receives a 0. For 11–30%, it receives a value of 1. For 31–50%, it receives a value of 2. For values greater than 50%, it receives a 3.

The level of success in organising social services is determined by adding the values coded for each indicator. The maximum score an organisation can receive is 17. An organisation with a value of 15–17 is highly successful at organising social services, those with values of 12–14 are moderately successful, and those with values of 11 or below have a low level of success. High levels of internal organisational cohesion are expected to correlate with high levels of success in organising social services, and vice versa.

Internal organisational cohesion in the DRC

In order to answer the question of why some CSOs are more successful than others at organising social-service delivery, six religious CSOs in two cities, Goma and Bukavu, the provincial capitals of North Kivu and South Kivu, are analysed here. All of the organisations are churches (the Catholic archdiocese of Bukavu, the Catholic archdiocese of North Kivu I (Goma)), the 3eme CBCA in Goma, the 55eme CEBCE in Goma, the 5eme (Communauté des Églises Libres Pentecostistes en Afrique) in Bukavu and the 8eme CEPAC (Communauté des Églises du Pentecôte en Afrique Centrale) in Bukavu).

As Table 2 illustrates, the relationships between factors that produce cohesion are multiple and complex. Existence during the colonial period always correlates with a high degree of internal organisational cohesion. Receiving external support for operations also generally produces higher degrees of internal organisational cohesion. Other factors, including avoiding entanglement with Mobutu's patronage networks and not having divided along ethnic lines, suggest a higher degree of internal cohesion as well. This sample also suggests that a high degree of ethnic heterogeneity is not an absolute impediment to successful service delivery if specific institutional constraints to force inter-ethnic cooperation are in place. The following sections provide an overview of this framework as applied to Catholic and Protestant CSOs in Goma and Bukavu. As primary research for this study was undertaken during the post-war transitional period and immediately after the 2006 election of a new government (2005–07), this study primarily reflects the situation in the Kivus during that time. However, multiple subjects confirmed that the nature of the Congolese health-care delivery system and the role of religious organisations in operating it have changed little since well before the 1997–2002 wars.

Successful service organisation and internal cohesion in the DRC

As outlined in Table 1, 15 indicators are used to determine the effective organisation of social services. Table 3 summarises these findings with the scores attained by each CSO and corresponding level of successful organisation of social services.

As the above data indicates, a pattern emerges with respect to the relationship between levels of internal organisational cohesion and the successful delivery of social services. Table 4 outlines findings in this regard:

As the results in Table 4 suggest, there is a strong correlation between the extent of internal organisational cohesion in a CSO and its ability to successfully organise social services. Organisations such as the archdiocese of Bukavu and the 8eme CEPAC, which enjoy long institutional histories and high degrees of external support are the most successful at maintaining social-service provision over the long term. CSOs that are more fragmented, evidenced in shorter or divided institutional histories (such as the 55eme CEBCE), have a lower ability to raise external support and are ethnically heterogeneous, are less successful at organising social services. The variation in cohesion across CSOs is historically contingent. However, the broad pattern is clear: a high degree of internal organisational cohesion is necessary for successful social-service provision over the long run by non-state actors in fragile states.

Conclusion: substitution and reconstruction

This study has implications for and raises questions about the nature of authority in fragile states where CSOs provide public goods in the state's absence. When local and national authorities are extremely weak and unable to enforce the will of the state, CSOs are largely left to their own devices. They manage, administer and operate schools and hospitals independently, generally get to hire and fire personnel of their own accord, and are subject to only nominal regulation by state authorities who lack enforcement power. The locus of authority is blurred, and different organisations and individuals have a great deal of authority in specific sectors. There is no entity that truly exercises legitimate authority over all sectors of society.

Although it is impossible to reach definitive conclusions on the long-term effect of such arrangements, it seems clear that the displacement of the state in public-goods provision will have far-reaching implications if the state is ever able to reassert its authority. The trust necessary for government to have legitimacy among the population is largely absent in DRC. Instead, trust and reciprocity relationships are primarily among individuals and ethnic groups alone (Vlassenroot and Raeymakers 2004a, p. 27, Vlassenroot 2004, p. 39). In the DRC, local populations tend to look directly to service providers for help with problems, rather than the state. Even when problems are outside the purview of their mission, CSOs are often asked to take on additional responsibilities. However, unlike officials in a functioning state, CSOs are only accountable to their donors and themselves. The need for accountability to service users is a key reason that the reassertion of state authority over – at a bare minimum – the full regulation of health-care provision is desirable. Likewise, it is likely that state-regulated and/or controlled health-care provision will be more uniform and of similar quality than the widely varying levels of quality available today.

It is entirely logical for citizens of weak states to look to non-state entities to provide public goods that the state would normally provide or regulate. But when populations ascribe more legitimacy to CSOs than to the government over the course of three decades, how can the government ever reassert its authority? As Englebert (2003) notes, the idea of the Congolese state ‘persists’ in the eastern portions of the country despite its actual absence from the territory (Engelbert 2003). Anecdotal observations and interviews confirm that the vast majority of the people of the eastern DRC claim to want the government to resume its role as the governing power in the territory and to provide and regulate social services. (In particular, most Congolese would be glad for anyone to exercise a monopoly on the legitimate use of physical force.) But how this will happen in a place in which other organisations are widely credited with providing far superior public goods than the state has been able to provide in 30 years is very unclear. CSOs substitute for the state as social-service providers, and they are better at it than the state has ever been. Despite all the problems with access and quality of Congolese health care, the systems currently in place are still better than anything seen in the territory since the colonial era. Why would the population – or CSO leaders – return control of the social-service systems to a corrupt, incompetent and predatory state?

Public goods provision is a key task of the state. A state that cannot provide a basic level of services and regulation is not a state at all. The international community's vision of DRC's future involves building the state's capacity until it is completely reconstructed, meaning that it can police its own territory, defend its borders, regulate its economy and provide and regulate public services. There is also an expectation that the DRC's culture of entrenched corruption will somehow disappear once the government functions and the economy is put on track. How to transfer de facto authority out of the hands of church and other CSO officials to move back to a ‘partnership’ with the state is a complicated question that may not have an easy solution (Seay 2010).

More research is needed to fully understand the implications of hybrid or non-state models of social-service delivery in extremely fragile states. In particular, research not simply on the ability of civil society actors to provide services, but also on their health outcomes and factors affecting those outcomes would help to improve our theoretical understanding of the situation while pointing towards means of creating tangible public health benefits for the Congolese population and populations in other fragile states. More research is needed of patronage in health-care delivery systems and the relationships between patronage systems, successful health-care provision and health outcomes.

More research is also needed on the increasingly diverse landscape of Congolese religious life and the role that new forms of religious expression play in the health care and other service delivery sectors. Leinweber (2011) has demonstrated how the Islamic community in Maniema province came to fill a role in providing educational services, but there is virtually no work on Muslim service provision in other parts of the DRC. There are extensive studies on the rise of Pentecostalism and charismatic Christianity and its effects in Kinshasa and elsewhere (Pype 2006, 2009a, 2009b, de Boeck 2005). While some of the Église du Christ au Congo (Church of Christ in Congo – ECC) churches profiled in this study are in the charismatic tradition (the 5eme CELPA and 8eme CEPAC), the DRC's newer charismatic churches have thus far made few attempts at or inroads into the service-delivery sector, particularly in the east.

Finally, researchers can build on the existing literature through studies of what precisely – if anything – Congolese citizens expect from their government. Survey-based research and ethnographies should yield more specific data and insights into this question, and would provide scholars with a basis to theorise more soundly on the future of the Congolese state.

Note on contributor

Laura Seay is Assistant Professor of Political Science at Morehouse College in Atlanta, USA.