Introduction
After a failed political transition in the 1990s and ensuing armed conflicts lasting eight years, in 2003 the Democratic Republic of Congo (DRC) embarked on an ambitious reconstruction effort under the auspices of the international community. Western statebuilding ideals were written into the constitution that was ratified in 2005. Among these ideals, democratisation and decentralisation figured prominently. They took shape through presidential and parliamentary elections held in 2006. The new constitution moreover foresaw the expansion from 11 to 26 provinces and the creation of decentralised territorial entities (Englebert 2012). However, the hope that was kindled through these events was soon dashed through ongoing armed conflict, increasing authoritarianism and persisting low investment in public goods (De Herdt and Titeca 2016; Larmer, Laudati, and Clark 2013).
This, however, did not lead to an abandonment of reform implementation. In fact, accommodating reform failure (Moshonas 2013) and continuously seeking ‘improvement of improvement’ (Li 2007b, 274) has become part and parcel of donor activities in the DRC. For these reasons, various authors have started to look ‘beyond reform failure’ (Moshonas 2013) and proposed to ask ‘what do these reforms do?’ (Li 2005). Investigating the actual impacts of reforms can be a fruitful endeavour. Reforms are hardly ever implemented according to pre-conceived design (Long 2001). In reality, they are embedded in ‘processes of political compromise and accommodation’ (Conteh 2017, 31) and a vast array of reasons can lead to their adaptation. For instance, governments might lack political will and fail to accompany a reform with the required financial and organisational inputs. Most importantly for this article, reforms are never facing a tabula rasa (Li 2007b). Scott’s (1998) seminal work stresses how centrally organised states create simplistic representations of reality in order to facilitate classifications and interventions. Li (2007b) builds on Scott and proposes to use ethnographic methods to unpack how these reforms are negotiated on the ground. She underlines that reactions to and adaptions of these reforms are inherent in the reform process. Implementation faces a complex reality. Schematic attempts to represent reality cannot prevent dynamic social factors from unfolding. Specifying her understanding of these processes, Li mentions four inherent limits to government practices. Among them, ‘population’ is the most relevant for the article at hand. Li writes that:
The sets of relations and processes with which government is concerned present intrinsic limits to the capacity of governmental interventions to rearrange things. There is inevitably an excess. There are processes and interactions, histories, solidarities and attachments, that cannot be reconfigured according to plan. (Ibid., 17)
Despite the state’s quasi-retreat from funding public education, there is a high demand for schools (De Herdt and Titeca 2016). Similar to other sectors, people demand the state’s services and presence (Trefon 2009). Since the state does not ensure full remuneration of all teachers, parents have to fund approximately three-quarters of the entire primary and secondary education budget (World Bank 2015). Nevertheless, demand for schooling has skyrocketed: the number of public schools has risen from 24,108 in 2001–02 to 61,789 in 2013–14 (MEPSP 2014; World Bank 2005). This expansion corresponds to a demographic process where, in the same period, the number of school-aged children increased by about 50% (De Herdt, Marivoet, and Muhigirwa 2015). Furthermore, it reflects a process of recovery and longing for education after almost two decades at a standstill.1 However, trying to accredit a school via official procedures is futile. Instead, educational communities must seek the help of politicians and administrators to obtain an accreditation decree for their school.
Indeed, it has already been suggested that obtaining accreditation decrees from the national Minister of Education is prone to clientelism (De Herdt and Titeca 2016). Official procedures regulating the accreditation procedure are circumvented by people with direct access to the Minister of Education and his entourage (Ibid.). These people are primarily politicians and educational administrators. In this article, I consider them as brokers (Bierschenk, Chauveau, and Olivier de Sardan 2000; Mosse and Lewis 2006). I identified three processes that have led to the multiplication of positions with a potential for brokerage: first, democratic elections; second, administrative unit proliferation; and third, the (unplanned) fragmentation of faith-based educational administration networks.2 Given existing knowledge on the everyday functioning of African bureaucracies (Bierschenk and Olivier de Sardan 2014a), it would be naive to expect an administrator or politician to strictly limit their actions to formally ascribed functions. Instead, the multiplication of administrators and politicians is likely to further strengthen pre-existing patterns of accessing public resources, such as clientelism (Abdulai and Hickey 2016; Boone 2003; Grossman and Lewis 2014). Therefore, in this article I explore how statebuilding interacted with existing forms of governance and opened up the space for new brokers to enter and engage in obtaining accreditation decrees for public schools from the National Ministry of Education.
Through this analysis I wish to make the following three points: First, my research sustains that the implementation of democratisation and decentralisation reforms can be a negotiated process that is likely to reproduce pre-existing governance modalities, such as clientelism. Second, I nuance these findings by showing that a reproduction of clientelism is not necessarily equal to a reproduction of socio-spatial patterns in accessing state resources. Third, in the conclusion I elaborate how this brokered expansion is related to ongoing efforts to regulate the growth of the education sector and the governance of teachers. Taking into account the longue durée of administrative structures and educational governance, I propose to understand this brokered expansion as one element in a much wider process that is best described as a ‘permanent provocation’ (Li 2007b).
Research for this study was carried out in different parts of the DRC for 14 months between 2013 and 2016. The educational province (called Division, after the French term Division éducationnelle) under study is made up of several educational sub-provinces (called Sous-Division, after the French term Sous-division éducationnelle).3 Some of these are largely urban with private schools accounting for the majority of new openings. Since private schools are subject to different accreditation procedures, the remaining six educational sub-provinces were deemed particularly relevant for this article. The Division under study is relevant for four reasons. First, it has experienced a steep expansion of the number of public schools since 2004. Second, cross-regional differences in the numbers of schools between educational sub-provinces point to some low performers and some high performers. These differences suggest either varying demand patterns or unequal access to accreditation decrees. Third, four of the six educational sub-provinces have been created since 2008, which allows an exploration of the particular role of government educational administrators as brokers. Fourth, two sub-provinces have been affected by conflict for many years. These sub-provinces have historically been neglected in terms of access to schools.
Through original sets of complementary quantitative and qualitative data, this article offers a fruitful approach of discussing both the outcomes of brokers’ activities and the underlying relationships and processes. The quantitative data set consists of 23 accreditation decrees issued between 2004 and 2013 for the Division.4 The 23 decrees include 776 school accreditations for the six educational sub-provinces. The decision to start with the year 2004 is due to the fact that 2004 was a turning point when large numbers of decrees started to be issued. 2013 was the last year for which I could find accreditation decrees. Insights yielded through this quantitative analysis are triangulated and enriched through the analysis of semi-structured and informal interviews.5
In the following section I discuss the multiplication of brokers through the three processes of democratisation, decentralisation and fragmentation. Next, I turn to a specific educational province as the case study in order to explore the practice of obtaining accreditation decrees. Finally, I discuss the positive and negative outcomes of these practices.
Statebuilding and the multiplication of brokers
Instead of treating certain state administrations as deviant from an ideal Weberian type of rational bureaucracies, scholars have started to investigate how state administrations and reforms are really functioning (Bierschenk and Olivier de Sardan 2014b; De Herdt and Olivier de Sardan 2015b). Such analyses require an unpacking of the notion of the state. There are three aspects that are most relevant for this article: first, the state is not a homogeneous entity. Instead, it is made up of a vast array of actors, institutions and symbols. According to Hagmann and Péclard (2010), statehood is subject to persistent negotiations. Jessop (2015) considers the state as a relation that is permanently shaped through interactions and conflicts between state and non-state actors, the line between the two being utterly blurry. Second, administrative processes do not simply function according to official scripts and norms (Bierschenk and Olivier de Sardan 2014b). Third, despite the immense importance of non-state actors, the state remains a pivotal reference point in people’s lives (De Herdt and Titeca 2016) and people are likely to seek its services (Trefon 2009).
With these insights in mind, it would be unrealistic to assume that statebuilding functions exclusively according to formal guidelines and objectives. Boone (2003, 356) argues that ‘decentralisation has not necessarily empowered local citizens and can simply strengthen local power brokers or state agents’. Along similar lines, Englebert and Kasongo (2016) have shown how decentralisation reforms in the DRC (re)produce local systems of a predatory national state. Thus, the reproduction of existing structures and modes of governance is a potential consequence of statebuilding reforms. Patron–client relationships are a frequently occurring mode of governance (Abdulai and Hickey 2016; Bierschenk and Olivier de Sardan 2003; Boone 2003). Corrales (2005, 18) defines clientelism as the ‘the distribution of valued resources … according to political criteria … from a strong actor toward a weak actor’. He points out that clientelism is among ‘the most intense political forces that push states to expand education’ (18). Since democratisation and decentralisation might expand the potential number of clientelist relationships, we might therefore expect an increase in the number of schools. My study affirms this expectation but also sheds light on more nuanced outcomes. With Abdulai and Hickey (2016, 45) I propose an analysis that addresses ‘the more complex workings of elite behaviour and elite–popular dynamics that underpin the interaction of political clientelism and democratic processes’. I do so by looking at interactions between politicians and bureaucrats.
Bureaucrats take up an important role in these dynamics as they broker and mediate access to state resources (Berenschot 2010). Drawing on Mosse and Lewis (2006, 12), I understand brokers as a ‘specific group of social actors who specialise in the acquisition, control, and redistribution of development [in this article: state] revenue’. Broker can furthermore highlight ‘the ways in which social actors operate as active agents building social, political, and economic roles rather than simply following normative scripts’ (Ibid., 11). The following section reveals how processes of democratisation, decentralisation and fragmentation can lead to a multiplication of brokers.
Multiplication through democratisation; or, emerging high-level brokers
After the formal end of the second Congo war (1998–2003) and the ensuing transitory government, the first and second democratic presidential and parliamentary elections in the DRC took place in 2006 and 2011. Nationwide, 500 members of parliament were elected. The educational sub-provinces under study each correspond to a district (territoire in official nomenclature) that each have between one and three MPs representing them in the national assembly. Whereas Congo’s political institutions perform poorly (Englebert 2016), the emergence of elected politicians had consequences beyond formal stipulations. I follow Bierschenk and Olivier de Sardan’s (2003, 146f.) analysis of the Beninese context where they claim that the ‘democratisation process has reinforced the pre-existing, hybrid and composite form of local government’. What counts for this article is the role of national MPs in obtaining accreditation decrees. Officially, MPs do not occupy any role in the accreditation process. In reality, they are the link between provinces/districts/villages and the national level – frequently through interactions with educational administrators.
Multiplication through decentralisation; or, a proliferation of administrative brokers
Administrators represent and enact the state on the ground. Administrators are not mere indifferent implementers but brokers with multiple roles well beyond distortion and corruption. They have essential roles in the everyday workings of the state (Bierschenk and Olivier de Sardan 2014b). Moreover, they enjoy a certain margin of manoeuvre to ‘construct their everyday relationships with service users’ (Ibid., 4) and to interpret and enact their own roles and responsibilities (Berenschot 2010). In the DRC, many of them have served throughout political turmoil and remained in office during changing government constellations. The number of administrators increases with the ongoing proliferation of administrative units (Grossman and Lewis 2014). Starting around 2004, the educational administration in the DRC preceded the general territorial decentralisation by several years.6 In Katanga, for instance, the number of educational provinces increased from one to three. Nationwide, in 2007, there were 28 educational provinces and 191 educational sub-provinces (MEPSP 2007). In 2014 this number had increased to 30 and 258 respectively (MEPSP 2015). In the Division under study, four new educational sub-provinces have been created since 2008.
Owing to the particularities of the Congolese education system, educational administrators have specific incentives for a school to exist and to be accredited: they receive money from each school that exists through their share of monthly fees paid by parents (Verhaghe 2007). Also, many administrators used to be teachers themselves and might have an intrinsic interest to provide education. Furthermore, educational administrators in conflict-affected regions often see schools as a major tool to curb enrolment in armed groups. For these reasons, they are eager to seek direct relationships with MPs in order to obtain accreditation decrees. The statement that administrative offices ‘are not able to play a significant role in educational management’ (World Bank 2005, 15) must therefore be scrutinised in the light of their actual achievements.
Multiplication through concessions; or, fragmentation of faith-based organisation brokers
In the DRC, in general, public goods and services are co-produced by state and non-state actors (Seay 2013). In certain sectors, such as education, this co-production is rooted in the ‘concessionary state’ (Poncelet, André, and De Herdt 2010); the Belgian colonial government entrusted the provision of education largely to faith-based organisations (FBOs). This co-management was written into the so-called convention from 1977. The convention was signed by the government, Catholics, Protestants, Kimbanguists and Muslims. For FBOs, schools have always served as a vehicle for evangelisation and missionary purposes (De Herdt, Marivoet, and Muhigirwa 2015, 58). With regard to FBOs, three drivers underlying the multiplication of brokers need to be looked at: first, traditionally, the Catholic Church has been the dominant provider of education. Despite its vertical organisational structure in religious matters, the Catholic educational administration is not a top-down affair (Titeca and De Herdt 2011, 230). Decisions on school constructions are taken at the local level. Second, since the early 2000s, the Protestant network has overtaken the Catholic Church in terms of the number of schools it manages. Protestant communities in the DRC are organised under the banner of the Church of Christ in the Congo (Église du Christ au Congo, ECC). The ECC is a conglomerate of about one hundred Protestant faiths and denominations (Baptist, Methodist, Lutheran etc.). The Protestant educational administration is organised in national and provincial offices that oversee the practices of the various communities. These communities have the real decision-making power when it comes to opening schools. This heterogeneous bottom-up structure facilitates dynamic and unregulated growth. Third, in 2006–08 new religious networks7 signed a management mandate (mandat de gestion) with the government allowing them to manage schools. In sum, FBO networks are highly dynamic and not subject to coherent practices or top-down coordination.
Altogether, the two drivers linked to statebuilding (democratisation and decentralisation) and the existing co-management of educational delivery facilitate a multiplication and fragmentation of providers of education with a legitimate interest and leverage in opening schools. The following section provides an analysis of the influence of these three dynamics on the practice of obtaining school accreditation decrees.
Brokered educational expansion: outcomes and processes
Democratisation, decentralisation and fragmentation have led to a multiplication of brokers who compete with each other while pursuing the same goal: for various reasons all of them try to obtain accreditation decrees for public schools from the national Ministry of Education. A major rule of the game is the fact that the Minister of Education is able to accredit schools without prior agreement of the ministries of Budget and Finance. The following analysis consists of two parts. First, by drawing on quantitative data I present the outcomes of the brokered educational expansion in a particular educational province. Second, I use qualitative data to complement the insights yielded through the quantitative analysis in order to strengthen the understanding of brokers’ roles in obtaining accreditation decrees.
Expansion of the education sector
The following analysis focuses on the years 2004–13 for six rural educational sub-provinces. As a reference, the following table8 shows the national and provincial educational expansion over recent decades.
Table 1 shows a clear trend of educational expansion. The first post-war national annual educational yearbook funded by UNESCO was issued for the school year 2006–07. Unfortunately, these national yearbooks only disaggregate data between the 11 provinces. Thus, no intra-provincial data are available. This makes it impossible to establish an exact baseline of the numbers of schools in the educational sub-provinces under study for the year 2004, and to compare their growth over time. However, other data sources I collected allow me to provide estimations. Drawing on unpublished annual reports established in every educational sub-province (Rapports des travaux de la commission sous-provinciale de promotion scolaire) and the provincial annual educational yearbook – which was first published in 2010 – I can provide the following data.
| 1986–87 | 2001–02 | 2006–07 | 2007–08 | 2008–09 | 2009–10 | 2010–11 | 2011–12 | 2012–13 | 2013–14 | |
|---|---|---|---|---|---|---|---|---|---|---|
| DRC | 14,864 | 24,108 | 37,893 | 41,609 | 45,714 | 46,744 | 49,490 | 53,875 | 57,848 | 61,789 |
| Division | n.a. | n.a. | 4049 | 4344 | 5001 | 5036 | 5120 | 6164 | 7153 | 7070 |
Although these data might be somewhat incomplete, they allow several conclusions: first, the trend of educational expansion observed in Table 1 is sustained through these data. The Division has experienced a growth in the numbers of public primary and secondary schools over recent years. Second, there seem to be some low performers (e.g. Sous-Division 1) and some high performers (e.g. Sous-Division 6). Third, the rapid increase of Sous-Division 5 between 2010–11 and 2013–14 suggests that important events occurred in this time frame that influenced the number of accredited schools. Finally, I added more recent data for Sous-Division 4 and Sous-Division 6 as both sub-divisions have been affected by conflict. Schools in these sub-divisions had difficulties in responding to the surveys sent out for the educational yearbook in 2013–14.
These data allow me to confirm the occurrence of educational expansion at national, provincial and sub-provincial levels. By examining school accreditation decrees, the analysis now turns to the question of how this growth occurred. Table 3 displays the annual number of accreditation decrees (at the top) and the annual number of accredited public primary and secondary schools per educational sub-division between 2004 and 2013.
It is important to note that the number of school accreditations does not equal the growth of schools as presented in Table 2. A decree per se represents the potential of a new school, but not all decrees are used immediately. As the qualitative data below will sustain, some administrators indeed have repertoires of unused accreditation decrees. Overall, from 2004 to 2013, at least 23 accreditation decrees for the districts under study were issued from the national Minister of Education. One accreditation decree can list schools for one or several districts. Together, these 23 decrees contain a total of 776 schools.
| 2002–03 | 2007–08 | 2010–11 | 2013–14 | 2015–16 | |
|---|---|---|---|---|---|
| Sous-Division 1 | n.a. | n.a. | 48 | 51 | n.a. |
| Sous-Division 2 | n.a. | n.a. | 63 | 72 | n.a. |
| Sous-Division 3 | 96 | n.a. | 162 | 181 | n.a. |
| Sous-Division 4 | 67 | n.a. | 121 | 116 | 175 |
| Sous-Division 5 | n.a. | 71 | 73 | 136 | n.a. |
| Sous-Division 6 | 141 | n.a. | 241 | 239 | 289 |
Let me take a closer look at spatial and temporal differences that can be seen in Table 3. First, how do different sub-divisions perform? There is a stark difference in the number of accreditations between sub-divisions, which can be distinguished as low performers (Sous-Division 1, Sous-Division 2 and Sous-Division 3 in the middle) and high performers (Sous-Division 4, Sous-Division 5 and Sous-Division 6). Second, which temporal differences or extraordinary trends exist? Upon the official end of the armed conflicts, 2004 is the year in which the accreditation decree-issuing machinery was switched on again. Accreditation decrees in the past (e.g. in 1986, 1992 and 1995) included only very few schools per district. From 2004 onwards, accreditation decrees started to include larger number of schools per district. An even stronger increase can be observed since 2008. Finally, 2011 has the highest number of school accreditations and accreditation decrees. It was a year of presidential and parliamentary elections in the DRC. Almost one-quarter of all 776 accreditations and almost one-third of all accreditation decrees were issued in 2011. These patterns suggest that elections and thus MPs indeed play a decisive role in obtaining accreditation decrees. However, since for example 2008 and 2013 also display high numbers of accreditations although no elections took place, there must be other factors affecting how accreditation decrees are obtained. From the discussions above, it is likely that these factors relate to administrative unit proliferation. In fact, since 2008, four new educational sub-provinces have been created in the province under study. Whereas the older educational sub-province of Sous-Division 6 has continuously received accreditations since 2004, Sous-Division 4 started to take off in 2008, and the development of Sous-Division 5 really took off only in 2011. Qualitative data will explore this issue further. To allow for a better comparison between networks, Table 4 shows the number of accreditations per network.
| 2004 | 2005 | 2006 | 2007 | 2008 | 2009 | 2010 | 2011 | 2012 | 2013 | ∑ | |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Total number of accreditation decrees | 2 | 2 | 2 | 0 | 1 | 2 | 3 | 7 | 1 | 3 | 23 |
| Accredited schools in … | |||||||||||
| Sous-Division 1 | 0 | 1 | 0 | 0 | 4 | 0 | 7 | 9 | 0 | 0 | 21 |
| Sous-Division 2 | 0 | 0 | 2 | 0 | 6 | 1 | 13 | 12 | 0 | 15 | 49 |
| Sous-Division 3 | 25 | 2 | 4 | 0 | 15 | 23 | 0 | 8 | 1 | 38 | 116 |
| Sous-Division 4 | 4 | 1 | 0 | 0 | 55 | 0 | 41 | 58 | 0 | 25 | 184 |
| Sous-Division 5 | 0 | 5 | 12 | 0 | 14 | 22 | 2 | 39 | 0 | 94 | 188 |
| Sous-Division 6 | 19 | 17 | 31 | 0 | 44 | 38 | 16 | 53 | 0 | 0 | 218 |
| Total number of school accreditations | 48 | 26 | 49 | 0 | 138 | 84 | 79 | 179 | 1 | 172 | 776 |
Sources: Accreditation decrees.
| Network | Government | Catholic | Protestant (except 10th & 11th) | 10th (Protestant) | 11th (Protestant) | 30th (Protestant) | Other | Total |
|---|---|---|---|---|---|---|---|---|
| Number of accreditations (total) | 184 | 271 | 107 | 101 | 50 | 44 | 19 | 776 |
| Number of accreditations (%) | 23.7 | 34.9 | 13.8 | 13 | 6.4 | 5.7 | 2.4 | 100 |
Source: Accreditation decrees.
The government network received 23.7% of all accreditations, the Catholic network 34.9% and the Protestant network alone received 38.9% of all school accreditations. All FBO networks combined obtained 76.3% of all accreditations. These shares reflect the nationwide pattern of school numbers per network and underline the increasing quantitative weight of the Protestant network compared with earlier decades (Titeca and De Herdt 2011; World Bank 2015). As mentioned above, the networks are not homogeneous entities. The Catholic network has provincial representatives but its daily operations are managed by diocesan administrators. Even more fragmented and with spatially overlapping administrations, the Protestant network consists of several different denominations.9 In the educational province, there are 18 Protestant denominations that manage schools.
According to the data provided in Table 5 the minimum number of accreditations per Protestant denomination is two and the maximum is 101. Fifteen denominations obtained 22 school accreditations or fewer and the average number of school accreditations is 16.8 per denomination. Three denominations together (30th, 11th and 10th) received almost twice as many school accreditations (195) as the remaining 15 denominations (107). One single Protestant denomination (10th) stands out as it acquired one-quarter of all accreditations. These numbers show the large discrepancies between different Protestant networks, the fragmentation of the Protestant administration, and point to some crucial networks with extraordinary success in obtaining accreditation decrees: the 30th, the 10th and the 11th Protestant networks. As a final step in the quantitative analysis, I look at the accreditations per sub-division and network in order to complement the tables above.
| Denomination | Number of school accreditations |
|---|---|
| 22nd | 2 |
| 55th | 2 |
| 8th | 2 |
| 9th | 2 |
| 40th | 2 |
| 41st | 3 |
| 43rd | 4 |
| 46th | 4 |
| 31th | 6 |
| 37th | 7 |
| 39th | 8 |
| 3rd | 13 |
| 60th | 13 |
| 38th | 17 |
| 9th | 22 |
| 30th | 44 |
| 11th | 50 |
| 10th | 101 |
| Total | 302 |
Source: Accreditation decrees.
Table 6 reveals four aspects: first, government administrators were particularly effective in obtaining accreditation decrees in Sous-Division 4 and Sous-Division 5. Second, the Catholic network obtained high numbers for Sous-Division 3, Sous-Division 4 and Sous-Division 6. Third, the table shows a high number of accreditations obtained by the 10th denomination for Sous-Division 6. Fourth, the 11th Protestant network received a similar number of accreditations for all sub-divisions.
| Catholic | Protestant (except 10th & 11th) | 10th (Protestant) | 11th (Protestant) | Government | Other | ∑ | |
|---|---|---|---|---|---|---|---|
| Sous-Division 1 | 13 | 1 | 1 | 5 | 1 | 0 | 21 |
| Sous-Division 2 | 11 | 12 | 0 | 6 | 20 | 0 | 49 |
| Sous-Division 3 | 71 | 11 | 18 | 8 | 8 | 0 | 116 |
| Sous-Division 4 | 74 | 11 | 20 | 12 | 62 | 5 | 184 |
| Sous-Division 5 | 34 | 57 | 8 | 10 | 68 | 11 | 188 |
| Sous-Division 6 | 68 | 59 | 54 | 9 | 25 | 3 | 218 |
Source: Accreditation decrees.
In sum, the quantitative data suggest three overarching insights. First, the data confirm prior insights that accreditation decrees have been issued in high numbers since 2004. Elections, especially in 2011, seem to have triggered a huge amount of new school accreditations. There is a total of 22 networks/denominations that have obtained accreditations since 2004. This fragmentation is contrasted by a strong accumulation of accreditation decrees among several crucial networks. Second, different networks seem to specialise in specific sub-provinces. Third, apparently not all school accreditations have been distributed as some are kept as clientelistic capital instead. Taken together, these three points suggest a strong impact of democratisation, administrative unit proliferation and fragmentation. In the following section, these impressions will be corroborated by qualitative data. Drawing on qualitative data, I will explain these trends by looking at the interactions between educational administrators and MPs. This allows me to explore how they construct their own roles and responsibilities. In other words, I will shed light on everyday state formation and an underlying game of the rules in which they all participate.
Low performers
Three educational sub-divisions can be seen as low performers according to the number of school accreditations they obtained: Sous-Division 1, Sous-Division 2 and Sous-Division 3 with 21, 49 and 116 school accreditations respectively. In these three sub-divisions, the contact between administrators and MPs is not very strong. One of the administrators complained that: ‘They have never come to my office’ (personal communication, April 4, 2016). In fact, the last time his sub-division received an accreditation decree was before he even took office in 2011. ‘Since I started’, he said, ‘there have not been any decrees for us’ (Ibid.). With regard to MPs, he and a teacher colleague were very eager to point out that some of them simply have no passion for education: ‘They are not concerned with schools and put their time and energy into other matters’ (Ibid.). These utterances suggest that administrative unit proliferation – this educational sub-province was created in 2011 – does not automatically lead to clientelistic bonds.
When I approached another administrator with names of schools that showed up on recently issued accreditation decrees, he told me that: ‘They don’t exist. Today it’s the opposite: first the decree and then the construction’ (personal communication, March 31, 2016). He added that MPs visit villages and collect demands without always taking into account his advice. At other times, the MP calls the administrator who then recommends some villages that still need a school. In contrast to the administrator in Sous-Division 1, this senior administrator adopts a more active role. He is clearly aware of the circumvention of the official procedures to obtain accreditation decrees, and collaborates with MPs when possible. However, despite his authority, MPs might still open schools behind his back and without his authorisation.
Regarding the third low-performing sub-division, I encountered a similar situation. The administrator complained that: ‘We are not very lucky with our MP. The MP of Sous-Division 6 is very dynamic … . But he [MP of Sous-Division 6] really abandoned us’ (personal communication, April 25, 2016). In the past, until 2008, Sous-Division 6 and Sous-Division 3 used to form one common educational sub-division. At that time, the ‘dynamic’ MP of Sous-Division 6 also ‘looked for decrees for Sous-Division 3’ (Ibid.). However, as the administrator reported, he was soon told by ‘those people’ (meaning other MPs) not to engage in that activity anymore (Ibid.). Indeed, the last time Sous-Division 3 obtained decrees through the 10th Protestant denomination – to which the MP is affiliated – was shortly after the separation of the two sub-provinces. Thus, administrative unit proliferation and democratisation do not only carry potential for the creation of new clientelistic ties, but might also jeopardise the existence of well-established ones.
High performers
On the other side of the performance spectrum, I find high-performing educational networks. I first discuss the relationship between government administrators and MPs, then FBO administrators and MPs, and finish with a focus on dynamics particularly related to FBOs. Whereas government administrators in Sous-Division 1 and Sous-Division 3 seemed rather passive and complained about the lack of obtained accreditation decrees, the educational administrator in Sous-Division 5 positions himself very differently. Responding to my surprise about the recent expansion of public primary and secondary schools, he affirms: ‘That’s me. When we got here, there were no schools’ (personal communication, April 23, 2016). The Sous-Division 5 was created in 2011 – previously, it used to belong to another sub-division – and he became the first administrator. Before, there was reportedly little contact between the MP of Sous-Division 5 and the responsible educational administrator based in another district. The administrator’s claims are substantiated by my quantitative data: 70% of all accreditations of Sous-Division 5 were obtained after he arrived in 2011. What might be the reasons for his success? The administrator explains that he does not accept the straitjacket of a passive and waiting bureaucrat but actively constructs his roles and assumes responsibilities. Furthermore, through a particular interpretation of the MP’s role he has put himself much more closely alongside the MP:
It is part of their responsibilities. We have to exchange. As a technician, you can be like a counsellor in a specific area for the MP. An MP doesn’t have knowledge in these areas. When an MP comes he/she observes the bureaucratic services, that’s the role of an MP. When the educational administrator doesn’t know them, he needs to ask. It is about knowing the role of an MP at the national level. That’s all. (Ibid.)
The educational administrator of Sous-Division 4 strongly underlines the MP’s central role. When asked ‘Do you have the impression that the reports you write for Promo-Scolaire are read?’ he answered: ‘Absolutely not. If it were not for our MP we wouldn’t receive anything’ (personal communication, April 22, 2015). In fact, it was reported that no school in Sous-Division 4 has been accredited via the official mechanism over the last couple of years. In Sous-Division 4, 45% of all decrees issued between 2004 and 2013 have been obtained since 2011, that is, since the creation of the sub-division. Reportedly, the educational administrator had already campaigned for the MP during the administrator’s prior role as a school principal. Now, as administrator, he said that: ‘When our MP obtains a decree and I can attribute it to a school, I tell them who to thank for the decree’ (Ibid.).
I now turn towards the relationship between FBO administrators and MPs. During discussions it became clear that Sous-Division 6 had one of the most active MPs. A former teacher himself, the national representative obtained numerous accreditations, especially for his former network (10th network, see Table 6). His network obtained the record number of 101 accreditations – about one-third of all Protestant accreditations (see Table 4). Curiously, the respective Protestant network administrator did not praise this intervention. He complained about the lack of inclusion and communication: ‘No MP ever comes here. I think it [the request] passes through Promo-Scolaire. If it is indeed thanks to the MP, we only see the results, the copies of decrees’ (personal communication, March 23, 2016). This resonates with an observation made in the case of Sous-Division 2: the network administrator is not necessarily strongly included. Another government administrator even mentioned this exact MP without my raising the issue: ‘The MP could arrive in a village, pose questions, everywhere, to find out what the people need. He even short-cuts the network administrator, and directly sends the decrees to the schools’ (personal communication, March 29, 2016). His colleague from Sous-Division 6 added that: ‘I receive copies of decrees for schools where I never conducted a viability report. Once they arrive, I visit the schools to conduct the viability report. If I notice that the school does not respect the criteria, I file a report to my superiors’ (personal communication, March 19, 2015). In these cases, MPs become ‘street-level politicians’ (Berenschot 2010, 888) who take note of the population’s needs and desires in direct encounters.
Finally, FBOs display certain specific dynamics. The 11th Protestant network received school accreditations for 50 schools (see Table 4) but only one school has been opened (Personal communication, March 23, 2016). One accreditation decree from 2011 contains school accreditations exclusively for the 11th Protestant network. Decrees with accreditations exclusively for one network are rare but not unheard of. What is more striking is that it is the only decree in which all districts are represented. It includes a total of 49 schools, more or less equally distributed between the districts (five schools per district at least, nine at most). However, none of these schools had been opened in 2016, as the 11th Protestant network is currently managing precisely one school in the provincial capital. The 11th Protestant network does not even have an educational office or counsellor in the province. This underlines that accreditation decrees are not granted after a thorough evaluation of educational needs but suggests an important role of political ties. The government educational administrator of Sous-Division 5 comments on the activities of the 11th Protestant network as follows: ‘There are some partners who first obtain decrees through other channels but who do not build. There are unused decrees lying around’ (personal communication, April 23, 2016). If, however, he saw the need to use one of their school accreditations, he assumes that he could talk to someone in charge of the 11th Protestant network and attribute it to a new school. In general, the government educational administrator claims to have a tight grip on newly opened schools and that he is always aware when a school opens.
Conclusion
In this article, I combined quantitative and qualitative data to shed light on two aspects: first, the multiplication of brokers as an unintended outcome of statebuilding and second, brokers’ involvement in obtaining accreditation decrees for schools from the national Minister of Education. I focused on the distribution of outcomes and the underlying processes, interactions and negotiations. In sum, massive educational expansion has been facilitated through a multiplication of brokers since 2004. This multiplication was facilitated by three drivers: first, through democratisation in the form of parliamentary elections. Second, through a proliferation of administrative units. Third, via concessions granted to FBOs. In terms of brokers’ success in obtaining decrees, the analysis revealed low performers and high performers among the administrative networks and sub-provinces. The quantitative distinction was affirmed by qualitative findings.
So far, literature dealing with the impact of democratisation and decentralisation in the DRC put a focus on the reproduction of the predatory Congolese state: research suggests that reform processes rarely follow official scripts, but are adapted and appropriated by a range of actors. These studies suggest that reforms reinforce opportunities for local bureaucrats and elected politicians to extract resources from the population (Englebert 2012; Englebert and Kasongo 2016). My research underlines that statebuilding and the implementation of reforms are negotiated processes. Instead of simply playing by the rules, actors constantly engage in a game of the rules: they adapt norms and rules according to their interests, obligations and possibilities (De Herdt and Olivier de Sardan 2015a). Thus, my research sustains that statebuilding can reinforce pre-existing governance modalities, such as clientelism. However, my findings further point to more nuanced impacts of clientelism.
I find that certain historically neglected and conflict-affected districts have caught up on the numbers of schools through patron–client relationships. Moreover, some well-established patron–client ties were weakened through the mentioned reforms. Finally, an MP might simply not be interested in providing decrees. Therefore, reproduction of clientelism is not necessarily equal to a reproduction of socio-spatial patterns in accessing state resources. Establishing clientelist ties is highly dynamic process with undetermined outcomes. Clientelism can have positive impacts for certain parts of the population, but is a highly unequal and unaccountable way of governing. I believe that it is necessary to simultaneously recognise positive impacts without ascribing labels such as ‘developmental patrimonialism’ (Hickey 2013) to the DRC. Therefore, I would like to finish with a reflection on the broader meaning of this mode of governing.
I argue above that I consider it useful to go beyond a focus on reform failure and instead focus on the real impacts of a given set of reforms. In a next step, it can be revealing to position the real impacts of one set of reforms to related sets of reforms. More concretely, the expansion of the number of schools takes place vis-à-vis three key intermediate objectives of educational governance in the DRC: the ongoing effort to regulate the growth of the education sector, the identification and management of teachers, and the abolition of school fees (De Herdt, Marivoet, and Muhigirwa 2015). Controlling growth and identifying teachers have been on donor agendas since the 1980s. Since they are crucial for budget and educational planning, they have taken up an immense amount of time and resources. The most prominent proposed solutions to achieve these objectives have been the following: the creation of a specialised ministerial department in the 1980s, several censuses, the reform of the aforementioned department in the 2000s and the recent bancarisation reform. This latter reform envisages public servants being paid individually via bank accounts, thus establishing a reliable overview. Until today, however, no full overview has been established. The unregulated growth of the education system is an enormous hindrance to this objective. In a similar vein, the objective to abolish school fees is equally contested (Ibid.). The Congolese education sector is still largely funded by parents. Whereas parents financed 90% in 2005, today their contributions still amount to 73% of all expenditures for education (World Bank 2015). The practice of allowing schools to be opened without a guaranteed government salary for teachers reinforces this funding mechanism.
Therefore, I propose to position the brokered educational expansion as one element in a much wider process that Li (2007b) conceptualises as a ‘permanent provocation’. Building on Foucault (Dreyfus and Rabinow 1983, 222), Li (2007a, 276) argues that power can never be total but always incites and provokes responses. Along similar lines, discussing the relation between state and Church in educational governance in the DRC, Titeca, De Herdt, and Wagemakers (2013, 129) also turned to the notion of permanent provocation. They characterised it as a process ‘where each step forward invites responses from different sides, the only certainty being that the process will never come to rest’. In Li’s words, brokers can be seen as part of an inherent ‘limit’ of reform processes: people, their particular ways of thinking and acting, and their relations with each other ‘cannot be reconfigured according to plan’ (Li 2007b, 277). The massive, unplanned and personalised expansion of the education sector is the consequence of this limit: the expansion stands diametrically opposed to the objective of better managing the teaching workforce and abolishing school fees. The permanent provocation is likely to continue owing to an ongoing multiplication of brokers (new Protestant communities, creation and autonomisation of provincial ministries, elections of newly decentralised territorial entities etc.). These events will continue to multiply the number of brokers with an interest in providing education to their electorates.10 Through this permanent provocation, the authority of the state is constantly affirmed and the state reaches a deeper level or territorialisation. People’s actions and perceptions reproduce the game configuration and thereby end up reproducing the underlying system (De Herdt and Titeca 2016). Given the state’s abysmal performance, this seems to me to be a more crucial impact than the positive benefits achieved through school openings. In sum, state formation is what happens while experts are busy making other plans.