THE SPLM'S STATE-BUILDING PROJECT: A HISTORY OF EXTRAVERSION

Extraversion can be defined as the capacity of individuals or groups to profit from their situation of dependence upon external resources, strategically using these resources for the pursuit of their own goals (Bayart Reference Bayart2000; Tull Reference Tull2011). In other words, it is the process by which individuals or groups ‘employ their dependent relationship with the external world to appropriate resources and authority’ (Peiffer & Englebert Reference Peiffer and Englebert2012: 361). In opposition to the idea of the marginalisation of the African continent, Bayart argued that Africa has always been part of the world system. Criticising dependency theorists, he showed that Africa was not a victim of its subaltern position vis-à-vis the world economy: in many instances, ruling elites took advantage of their dependence, turning it into an asset to pursue their own ends and to strengthen their grip on power. Extraversion strategies evolved over time: if the colonial and immediate post-colonial years were dominated by trade relations between Africa and the rest of the world, development aid and the discourse of democracy gained increasing importance in the 1990s together with migration flows (Bayart Reference Bayart2000), followed by state-building, security and counterterrorism (Hagmann Reference Hagmann2016).

South Sudan was no exception: its incorporation into the political economy and the territorial jurisdiction of the colonial and post-colonial state happened through strategies of extraversion that enabled the southern political elite to play on its marginalised position to attract external support and to extract rents from the struggle to reduce Southern marginalisation (Johnson Reference Johnson2003; Leonardi Reference Leonardi2013). This section illustrates how the SPLM/A harvested this heritage of extraversion and managed to intercept external resource flows not only to sustain the war effort but also to effectively occupy power nodes and establish a state-like structure, strictly controlled by the movement itself, to rule over the ‘liberated areas’.

After eight years of socialist sympathies relying on the support of the Derg regime in Ethiopia, the movement changed its orientation. In the 1990s, it undertook a process of internal reorganisation, motivated partly by the need to seek new allies after the fall of Mengistu and partly by the split of the movement caused by divergences in the leadership (Johnson Reference Johnson2003; Rolandsen Reference Rolandsen2005). Between 1994 and 1996, the SPLM established the Civil Administration of the New Sudan (CANS). The CANS increasingly took the shape of a ‘state-within-a-state’: an entity that ‘imposed effective control over a territory within a larger state and [had] an impressive array of institutional structures that, among other things, allow taxes to be collected, services to be provided, and business with other international actors to be conducted’ (Spears Reference Spears, Kingston and Spears2004: 16). In spite of the extreme geographic variation in the establishment and functioning of local governance structures, the SPLM managed to collect taxes, provide limited services to the civil population – including security and, to some extent, education and primary healthcare – and to conduct ‘business’ with international actors, primarily in the form of negotiation with donors. It was a state-like form of control of the territory that required more than mere coercion, though the latter was never completely absent. Through mechanisms limiting the use of violence, the establishment of administrative practices and the delivery of some public goods and services, the SPLM also came to acquire a certain degree of legitimacy in the eyes of the local population. It is not the aim of this paper to discuss in detail the mechanisms through which the SPLM was able to have its authority recognised by people in the areas it controlled through means that went beyond coercion: this has been done elsewhere (Rolandsen Reference Rolandsen2005; Mampilly Reference Mampilly2011). Rather, this paper will focus on the resources that were used to establish the conditions for the legitimisation of the rebel movement, which remained relatively strong throughout the interim period between the signing of the peace agreement and the referendum for independence.

The emergence of the SPLM/A as the hegemonic actor on the Southern Sudanese scene was only made possible thanks to the support of the international donor community. Since the mid-1990s, its relations with international actors strengthened as Western countries increasingly sought strategies to destabilise the government of Sudan, accused of having ties with global terrorist networks. For its part, the SPLM leadership showed an impressive capacity to appropriate a donor-friendly lexicon revolving around democracy, decentralisation, good governance and service delivery. Peace through Development, a pamphlet published by the SPLM in 2000, represents the most comprehensive endorsement of this kind of discourse, not only supporting democratic institutions but also linking democratic governance with development and peace. Indeed, aid to the Sudan increased dramatically towards the end of the 1990s, rising from US$138 million in 1997 to US$243 million in 1999. Of these funds, over 70% consisted of emergency aid, much of which was directed towards the southern region (Lehtinen Reference Lehtinen2001).

In his essay, Bayart identifies six modes of extraversion: appropriation, coercion, trickery, flight, intermediation and rejection (Bayart Reference Bayart2000). The SPLM/A was able to employ many of these in the construction of a stable relationship with Western countries and international organisations. Appropriation and trickery, understood as what ‘allows individuals and groups to make a living by circumventing the law, policies or rules imposed by foreign authority’ (Hagmann Reference Hagmann2016: 41) were the primary strategy of the SPLM in dealing with material assets delivery, as has been widely documented by several authors (African Rights 1997; Johnson Reference Johnson2003; James Reference James2009). Another example of trickery was the way in which the SPLM leadership convinced the donor community, at least partly moved by genuine humanitarian concerns, to directly target the movement's local government structures.

International relief programmes thus contributed – both willingly and unconsciously – to the creation of ‘proto-government’ structures, as the SPLM Governance Cluster defined them in its final report (SPLM Governance Cluster 2005), within the rebel movement's controlled areas. This process was characterised by a fundamentally different understanding of the aim of creating such structures. The donor community had a Weberian legal-rational ideal type of state in mind, characterised by functioning institutions capable of maximising efficacy and effectiveness for the sake of good governance. The SPLM's political leadership, for its part, readily adopted the reforms and policies that were required to secure a stable inflow of foreign aid resources (Awet Akot 2013 Int.), while the movement started developing its own state-building project aimed at maintaining control of these resources (Achol Deng 2013 Int.).

The two state-building projects shared the language of the modern democratic state, with its charge of universalism and legitimacy and its emphasis on development and service delivery. Whether or not this commitment was only a façade, it allowed the movement's politico-military elite to be increasingly involved in the distribution of aid resources to the areas under the SPLM's control, allowing it to extract considerable rents from the political economy of state-building. In the 1990s–2000s, the SPLM had access to two kinds of rent: one was the flow of aid resources in the form of humanitarian aid and direct support to local governance; the other was formed by state-levied resources such as taxes, licences and revenues of various kinds (including oil revenues since 2005). The SPLM's capacity of appropriating the second kind of rent grew in time (Reno Reference Reno2010) together with its legitimacy in doing so (Mampilly Reference Mampilly2011; Leonardi Reference Leonardi2013). Indeed, the state structure emerging from the donor-sponsored state-building effort in SPLM-controlled areas concealed a fundamental bias: it was the rebel movement, rather than an abstract state object, acting behind the legitimate mask of statehood.

This produced a process of ‘dysfunctional state-building’, as Hagmann (Reference Hagmann2016) calls it in the case of Somalia: a process based on the strengthening of unaccountable SPLM structures that made it extremely profitable to occupy government institutions. Humanitarian aid, and the aid that was delivered with the aim of strengthening local governance during the war and immediately after, fed into the political economy of state-building, ultimately contributing to this process.

HUMANITARIAN AID AND POLITICAL LEGITIMISATION OF THE SPLM/A

Operation Lifeline Sudan (OLS), the biggest UN-coordinated humanitarian operation ever, was negotiated between 1988 and 1989, following the devastating famine in Bahr el Ghazal. It was signed as a tripartite agreement between the UN and donor governments, the government in Khartoum and the SPLM/A. Despite the efforts of the Southern Sector coordinator, UNICEF's executive director James P. Grant, to make it clear that the UN did not intend to give any official recognition to the SPLM/A while dealing with it only for humanitarian purposes, it was the first time that a UN-coordinated operation openly engaged with a rebel movement to access areas under its control. This was extremely innovative at the time and has been described as a ‘pragmatic victory’ (Pegg Reference Pegg, Kingston and Spears2004), preventing – or at least reducing – mass starvation and gaining access to many areas that would otherwise have remained completely off-limits (Scaramella 2015 Int.).Footnote ³

OLS has been strongly criticised for several reasons, ranging from its influence on local power balances providing resources for patronage (African Rights 1997; Duffield Reference Duffield2002) to its lack of neutrality when supporting capacity-building and institution-building projects for the SPLM, providing a non-state armed rebel movement with diplomatic recognition (Bradbury et al. Reference Bradbury, Laeder and Mackintosh2000; Washburne Reference Washburne2010; Maxwell et al. Reference Maxwell, Santschi and Gordon2014). Indeed, from the very beginning of the operation, ‘humanitarian recognition’ resulted in a de facto political recognition of the movement, enabling it to negotiate with international actors for the sake of civilians’ protection. This section illustrates how its direct involvement in OLS and in consequent humanitarian agreements contributed to the political legitimisation of the SPLM as a local counterpart for international actors and as a government-like actor in the eyes of the local population (Akol Reference Akol2005: 54), and how this strengthened the movement's capacity for performing state-like tasks.

The SPLM's internal reforms of the 1990s and its rapprochement with the West coincided with the strengthening of its claim to control the relief effort in liberated areas through its relief wing, the Sudan Relief and Rehabilitation Association (SRRA). When OLS began, the SRRA became its major southern counterpart in the field, even though it was under strict military control: due to a lack of expatriates and staff in many areas, data collection and food distribution relied upon the SRRA, and were therefore often biased by military concerns (Riehl Reference Riehl2001). While appropriation and diversion of relief aid was a constant feature of all parties’ war strategies, it became particularly intrinsic to sustaining the rebel movement. This tactic figures prominently as a form of trickery among Bayart's modes of extraversion, but the SPLM/A was also able, to some extent, to control food distribution and to take part of it in the form of taxation (Johnson Reference Johnson2003; James Reference James2009).

Donors were at least partially aware of these forms of trickery and appropriation. Nevertheless, the SPLM/A's donor-friendly rhetoric made them turn a blind eye to the movement's actual practices, its poor human rights records, its diversion of relief aid, and even its scarce commitment to ensuring protection to humanitarian workers on the ground. In 1994, the killing of four humanitarian workers led OLS to negotiate the Agreement on the Ground Rules with the SPLM/A, the SRRA and two other rebel movements. The agreement marked another step forward in the international credibility of the SPLM leadership (Bradbury et al. Reference Bradbury, Laeder and Mackintosh2000) while increasing its capacity to officially perform state-like tasks such as taxation and work permit releases (Mampilly Reference Mampilly2011).

The Ground Rules Agreement has been described as a form of ‘humanitarian governance’, an expression employed to indicate the use of humanitarian principles to influence the behaviour of state and non-state actors (Maxwell et al. Reference Maxwell, Santschi and Gordon2014). Its seven sections determined the reciprocal property rights, responsibilities and obligations of the INGOs working within OLS as well as the rebel movements and their humanitarian wings. The effectiveness of the Ground Rules Agreement has been debated (Rone Reference Rone1999; Autesserre Reference Autesserre2002); nevertheless, there is some evidence that the SPLM/A's attitude towards civilians did improve (Rone Reference Rone1999). This might also have been a consequence of the growing number of opportunities to control the territory. These included the creation of local NGOs as a channel of service provision to the population and the strengthening of bureaucratic procedures of administration within the CANS.

In the early 1990s, the ‘civil society-building’ approach encouraged John Garang to form the first secular indigenous NGO in SPLM/A-held areas, the Cush Relief and Rehabilitation Society (CRRS). It was the first of a number of local NGOs created around the mid-1990s, providing new channels for relief aid and international funds. Most of them were ‘briefcase NGOs’ run by former SPLM/A members based in Nairobi or Kampala. Capacity building workshops for local NGO members were organised within the framework of OLS, who gave the SRRA the authority of selecting the participants and, consequently, those who would be eligible for foreign funding (Reno Reference Reno2010).

Local NGOs were not only supported within the framework of the humanitarian operation for relief supplies distribution. From the mid-1990s, USAID, one of the most generous donors behind OLS, claimed that it was time to start ‘doing relief developmentally’ (Dembowski Reference Dembowski2002: 3), and started supporting development projects such as seed production, road rehabilitation and market establishment. In 1999, the US Secretary of State Madeleine Albright expressed the US's solidarity with the movement's objectives and the will to provide it with direct support. With a total of US$159 million in aid assistance to Sudan, USAID decided to channel US$28.6 million to more flexible non-OLS NGOs such as Norwegian People's Aid (Autesserre Reference Autesserre2002).

The continued flow of aid funds and relief items to the rebel-held areas through the SRRA and the web of local NGOs allowed the rebels to easily access supplies without distraction from the fighting. At the same time, it had the effect of fulfilling one of the basic functions to which the CANS committed itself: the provision of basic services to the population under its control. Moreover, continued interaction with international NGOs brought a decrease in the level of violence for extractive purposes towards the local population, and an increasing bureaucratisation of the civil administration established in 1994. For example, taxation practices became less violent towards the end of the 1990s. Reports from WFP quoted in Food and Power in Sudan suggest that households included in follow-up visits after food deliveries often reported SPLA taxation on the food delivered (African Rights 1997), rather than violent appropriation from the military.

At the same time, however, the relationship between relief organisations and the SPLM/A was contentious, and always characterised by suspicion from the rebels, who saw international agencies as too complacent about government conditions and too independent from the movement's directives (Man et al. Reference Man, Diu, Aburas and Riak2004). The increasing demands of coordinating and controlling the relief aid and the activities of aid agencies led the movement to sometimes adopt extreme measures such as the expulsion of the French NGO Action Contre la Faim in 1997 upon vague allegations that it was threatening security in rebel-held areas (Mampilly Reference Mampilly2011). In 2000, the SPLA forced OLS and international NGOs to sign a Memorandum of Understanding, declaring it could no longer guarantee the security of those deciding not to sign it. The MOU was considered an unacceptable and illegitimate imposition, and provoked the withdrawal of many organisations (Mampilly Reference Mampilly2011). Despite never officially claiming to be a separate government – though it was de facto referred to as such by many local chiefs and the general population – the MOU contained typical ‘state’ demands: payment of fees to the SRRA for issuing work permits, payment of taxes on NGO assets, permission to enter SPLM-held areas. In William Reno's words, ‘Travelling to rebel-held parts of Sudan at that time was like travelling to a new country, with SPLA travel permits, registries, and other administrative paraphernalia typical of a sovereign state’ (Reno Reference Reno2010: 117). Not dissimilar to recent attempts at imposing a greater degree of control over international NGO operations and consequent tensions with the donor community (Hamsik Reference Hamsik2017), the MOU was an attempt at enforcing the movement's decision-making capacity over external resource providers.

In spite of contradictory analyses of the actual impact of the MOU (Riehl Reference Riehl2001; Mampilly Reference Mampilly2011), it seems likely that it contributed at least in part to the empowerment of local rebel administrative structures that were used to distribute relief and development funds. The channelling of at least part of aid funds through the movement's structures also allowed it to nurture its political project of winning hearts and minds in non-Dinka areas, where its presence was often contested despite the common goal of defeating the ‘Arabs’ (Mampilly Reference Mampilly2011).

PREVENTING GOVERNANCE FAILURES: DIRECT SUPPORT TO THE SPLM'S GOVERNMENT STRUCTURES

In the late 1990s, the SPLM was not only an active partner in relief distribution, but also became the major target of donor-funded capacity-building and institution-building efforts. Several projects aimed at strengthening the SPLM government capacity were implemented, justified by the need to make local institutions more effective in managing relief flows. This section will provide an overview of the first of such projects, showing how it provided both material and symbolic resources that fed into the SPLM/A leadership's strategy of extraversion.

The Sudan Transitional Assistance and Rehabilitation (STAR) project was the first capacity-building project explicitly targeting the SPLM civil administration.

It was not the first time that the SPLM/A had been directly supported and involved in capacity building: in the second half of the 1990s, UNICEF funded organisational development workshops, office equipment and even the construction of office facilities for the SRRA (African Rights 1997; Riehl Reference Riehl2001).Footnote ⁴ These activities were always justified as part of the work needed for the sake of humanitarian effectiveness: it was in the interest of the humanitarian operation to increase the organisational and coordination capacity of the SRRA, because it was de facto controlling territory and the actual delivery depended on it.

The STAR project was conceived immediately after the first visit of the US Secretary of State to John Garang in Kampala in 1997, when US–Sudan relations were deteriorating, and represented exactly the kind of ‘non-lethal support’ that Washington was willing to offer to the SPLM/A besides relief food (Autesserre Reference Autesserre2002). The democratic and developmental ideals to which the rebel movement was increasingly paying lip service in the second half of the decade looked like steps in the right direction to improve the movement's poor human rights records. A report commissioned by USAID on its activities in Southern Sudan between 1993 and 1999 praised the establishment of the CANS and maintained the need to acknowledge its control of wide regions, while at the same time strengthening its democratic nature and its capacity of protecting human rights. The US Congress earmarked US$7 million for the three-year programme, with the overall goal of increasing participatory democracy and good governance practices in opposition-held areas of Sudan while reducing heavy reliance on relief (Salinas & D'Silva Reference Salinas and D'Silva1999). More specifically, it aimed to expand participation in community-level administration, rehabilitating dwellings and infrastructures, promoting local economic development and increasing accountability, transparency and respect for human rights among civilian authorities.

The STAR project had three major components: a national-level component providing training to the National Democratic Alliance (NDA), the umbrella organisation of political opposition parties to the Khartoum regime; a county/regional-level component providing training to local administrators, delegated to UNICEF-OLS; and a local/community-level component to promote economic recovery and development, targeting Sudanese civil society organisations and delegated to Catholic Relief Services (CRS).Footnote ⁵ While the first was not implemented, the second and third components can be considered part of the donor-sponsored state-building project. They targeted two fundamental aspects of state functioning: the creation and strengthening of a non-violent institutional apparatus working through bureaucratised procedures and practices, and the delivery of basic services.

The Civil Administration Training (CAT) component only absorbed a small percentage of the total STAR funds (US$1 million out of US$7 million), yet it was remarkable at least from a symbolic point of view: for the first time, it provided direct support to SPLM members, helping them to develop local administrative capacities and structures. This component was delegated to UNICEF and implemented within the framework of OLS until 2001, despite UNICEF's scepticism over the training of rebel administrators as part of a humanitarian operation. The organisation repeatedly made clear that its role could not go beyond the ‘empowerment of grassroots level communities and … the promotion of efficient administration in local governance’ through the strengthening of universal good governance principles such as the recognition of ‘grassroots communities [as] the legitimate holders of political rights and entitlements and … local government administrators [as their] representatives’ (UNICEF 1999: 8).

UNICEF's work under the STAR project consisted of the organisation of workshops and conferences with various purposes. Besides several workshops for the dissemination of the humanitarian principles contained in the Ground Rules Agreement, between May and June 1999 a civil society conference in Mapel was also held to discuss the root causes of the famine that had hit Bahr el Ghazal the previous year. A preliminary document prepared by Mario Muor Muor, a senior SPLM official, identified the causes of the famine as ‘chronic insecurity’ and ‘lack of basic services’ (Muor Reference Muor1999). The unreliability of food supplies, he argued on behalf of the SPLM, was due both to Khartoum's war strategy of starving the ‘New Sudan’, and food being diverted by SPLA individuals beyond the control of the central command. The individualisation and criminalisation of what was, in fact, a tactic of the rebel movement as a whole allowed him to make a further point: looting of food and other relief supplies happened because of weaknesses in local governance.

To establish law and order was thus ‘imperative’ for the SPLM/A:

With this apparent mea culpa, the SPLM was thus asking for its government and judicial structures to be reinforced, perfectly in line with the purpose of the STAR project. Muor further suggested that it was time for the ‘southern people's friends’ to shift from an insufficient relief provision to development, if they were to ‘tackle the root causes of famine’ rather than simply ‘cure its symptoms’. This point is reaffirmed in a conference document written by another SPLM official, complaining about the lack of sustainability of relief operations, the risk of ‘emergency-dependency syndrome’ and the fact that the systematic bypassing of ‘local structures, institutions, staff’ neglected local ownership and participation in processes controlled by NGO expatriates and the government in Khartoum (Leek Reference Leek1999).

Consistent with the STAR plan of activities and with the SPLM's expressed needs, in July–September 1999 UNICEF and the SRRA organised the first Civil Administration Training Course in Akot, Lakes State. While initially planned for 25 payam and county administrators, the number of participants was doubled following the high demand of attendance that it generated. Participants were selected by county Commissioners under the supervision of the Secretariat for Interior and Public Administration, later renamed Secretariat of Local Government, based on broad guidelines provided by the UNICEF project implementation team. The workshop, as did similar other workshops following suit, aimed to improve democratic governance in the administration, increase civilian participation, accountability, transparency and respect of human rights by civil authorities with particular regard to property rights and the rights of children and women. Besides the general rules of behaviour of a ‘good administrator’, the training also tackled technical aspects aimed at developing the civil administrators’ capacity to deliver services effectively: keeping financial accounts, conducting general meetings, organising public elections of various popular organs, and maintaining law and order in civil society were some of the topics addressed. In a training course held shortly thereafter in Rumbek, administrators were even divided into categories in order to cover the different fields of social life they had to deal with (UNICEF/OLS 1999).Footnote ⁶ By 2001, an indefinite number – ranging from 150 to 290 – of payam administrators and deputy administrators in Western Equatoria, Lakes and Bahr el Ghazal had been trained, covering approximately half of the payams in the three regions (Dembowski Reference Dembowski2002).

The third component was designed to target the ‘community level’ through the Grant Making/Capacity Building scheme (GB/CM), managed by Catholic Relief Services (CRS). Besides improving people's living conditions through the provision of grants for the start-up of income generation activities and loans for the purchase of capital equipment and supplies, the GM/CB component assigned a central role to County Development Committees (CDC). The CDC came into existence in 1999 as a liaison between INGOs, the SRRA and the population, a core idea of US development cooperation, considered as an outstanding means of achieving good governance and peace. They were tasked with formulating development strategies and with implementing development programmes in their respective counties (Dembowski Reference Dembowski2002). CDCs were established in 17 counties; they drew 30% of their membership from the civil authorities and 70% from civil society institutions such as women's organisations, cooperatives, farmers, traders and disabled persons associations, though one could question how genuine all these forms of community organisations were. CDCs should have managed revolving funds when loans to selected beneficiaries for capital supply were repaid, but, according to the STAR programme evaluation conducted in 2002, almost none of the loans were repaid. CDCs thus functioned as a channel to distribute external resources to local organisations and the local people, increasing the decision-making power of the civil administrators involved, who were also in charge of identifying the beneficiaries. In fact, the evaluation team noticed that the relationship between CDCs and the county administrators remained fairly ambiguous, being described as ranging from independent to consultative to a direct reporting responsibility (Dembowski Reference Dembowski2002). Despite being presented as a ‘major achievement’ of the STAR project for bringing together representatives from the private and public sectors with local residents and thus being worth donors’ technical assistance, CDCs proved not to be sustainable and disappeared shortly after (Anzai 2013 Int.).

If we consider the objectives of the programme, STAR was not a success. Besides the lack of sustainability of CDCs, the extremely low rate of loan repayment and the short life of many of its initial achievements, many of the expected results of the CAT component – such as the creation of legislative bodies in counties and payams, preparation and approval of budgets, and keeping record of local administrators’ activities – were not achieved. Accountability and transparency were not improved: planning, budgeting, revenue collection and expenditure records continued to be poorly managed and kept secret if they existed at all. The judiciary remained firmly under the control of the military, and it was unlikely to act independently when events such as rapes or requisitions of property occurred. The SPLM expressed support for women's participation in the public sphere, but the 25% of seats for women in Liberation Councils at every level of the administration went unfulfilled. Some women interviewed by the evaluation team even claimed that they had been invited to public meetings only to do the cooking. Despite specific training and the provision of some infrastructure, no service delivery from local authorities was in place after STAR had ended (Dembowski Reference Dembowski2002).

Nevertheless, even if the commitment of the SPLM to good governance and democratic principles was only rhetorical, the ‘capture’ of these programmes enabled the movement to acquire an increasing coordinating and controlling capacity over goods and services from external providers. On the one hand, this allowed it to position itself as a credible political interlocutor both for the internationally sponsored peace process and internally as the only force with governing skills and the capacity to attract external resources.Footnote ⁷ On the other hand, several observers provided evidence that, since 1998–9, living conditions in SPLM-controlled territory improved, with tax collection becoming less violent and NGO presence turning increasingly developmental (Dembowski Reference Dembowski2002; Johnson Reference Johnson2003; Rolandsen Reference Rolandsen2005).

Since the mid-1990s, we can thus argue that the SPLM established a state-like structure under its control, projecting its own authority in its ‘liberated areas’. War-making and its rents, mostly in the form of relief aid flows, became a form of state-making in Tilly's terms (Tilly Reference Tilly, Evans, Rueschemeyer and Skocpol1985).

As the end of the war became increasingly likely with the signing of the Machakos Protocol in 2002, efforts at creating proper state structures, laws and bylaws increased. The ‘peace brigade’ – as John Young called it 10 years later (Sudan Tribune 2015) – became more involved in the organisation of workshops and training for civil servants and would-be government officials. A large part of this engagement was – and still is, as suggested by an interview with a senior staff member of a USAID subcontractor (Dawson 2013 Int.) – based on the assumption that the main reason for formally existing institutions to remain on paper was a lack of funding and capacity. While it is indisputable that South Sudan suffers from historical weaknesses in its education system, the ‘lack-of-capacity’ and the ‘lack-of-resources’ have become mantras for justifying the SPLM elite's continued requests of support from the donor community to strengthen the movement's structures and presence on the ground. These structures tended to conflate with state and government ones in the post-2005 period.

STATE-BUILDING THE SPLM

The signing of the ceasefire in January 2002 marked a turning point for Southern Sudan: the decrease in actual fighting left room to concentrate efforts on strengthening the structures of local civil administration (Mampilly Reference Mampilly2011). The Machakos Protocol made prospects for peace between the SPLM/A and the government of Sudan more and more tangible, drawing a number of splinter militias back to the mainstream rebel movement. The SPLM/A's control of the southern territory consequently increased through the co-option of isolated military commanders (Young Reference Young2006).

While the peace negotiations tackled political arrangements at national level and the creation of the government of Southern Sudan, the establishment of local government structures was left to negotiations within the southern elite and treated as an overall technical issue. Organisational structures (number of seats/positions in the civil service, number and kind of ministries, departments, commissions, committees, organisational charts, etc.), infrastructural and equipment needs for newly established institutions and financial aspects (both resource-raising and expenditures) were discussed as priorities. The SPLM expressed a commitment to decentralisation, and replicated each institution at all levels of the local government through decentralised or deconcentrated branches. This section addresses the synergy between the SPLM and international development agencies in the creation – and capture – of South Sudanese local state structures.

References to the state as a whole are very rare in SPLM documents of the early and mid-2000s, while those to local government and local administration abound. The Governance Cluster, one of the SPLM working groups created in 2004, was tasked with the formulation of recommendations on local government structures. These recommendations, together with the SPLM's pre-existing local administrative structures, represented the basis upon which the South Sudanese state was built. In the words of the Acting SPLM Chairman for Unity State in early 2013:

The creation of state structures in South Sudan was treated as a technical matter of finding the right institutional formula, but it resulted in an extremely exclusive exercise in which those who had proven to be capable of exercising local control and of keeping relations with the international community received resources, training and, ultimately, political power.

‘Workshop’ is a word that became very common in the vernacular languages of South Sudan to refer to public meetings in which issues of public interest were discussed and some sort of decision was taken – at least in the form of the distribution of tasks or the creation of ad hoc committees to address a specific problem. The frequent use of the word ‘workshop’ comes from the extensive organisation of ‘consultative’ workshops by development agencies not only for assessing local needs but also to engage local administrative and traditional authorities in the creation of government institutions.

SPLM local government institutions were indeed involved in a wide range of workshops organised with the support of several international actors. UNDP, particularly, played a major role in the process of consolidation and development of SPLM local state-like institutions between 2002 and 2005. This process received a major thrust in 2003, when the Secretariat of Local Government produced a number of ‘Laws of the New Sudan’ addressing virtually every aspect of social, economic and political life.Footnote ⁸ In 2003, the chairperson of the Secretariat of Local Government appointed a Nairobi-based Focal Point on Local Government and Civil Administration with the specific purpose of coordinating with donor agencies – mostly based in the Kenyan capital. UNDP promptly engaged with the Focal Point, not only providing it with office space in Nairobi (Focal Point on Local Government and Public Administration 2003), but also discussing the way forward and the formulation of a Local Government Framework and Act to provide the forthcoming government of Southern Sudan with strong local roots. By the end of 2003, the Focal Point had produced a draft Framework for the Development of the Capacity of the Secretariat of Local Government and Public Administration in South Sudan, containing the core ideas upon which the SPLM wanted to develop its government structures, as well as detailed requests for the donor community, addressing them in particular to UNDP.

Successive drafts of the framework were submitted for scrutiny to a number of international experts on governance and the public sector (Focal Point on Local Government and Public Administration 2003, 2004) and discussed in workshops organised by international donor agencies. The Local Government Technical Team, established in early 2004, was charged with further revisions, while other members of the Secretariat of Local Government were charged with assessing local administration in SPLM-controlled areas. They produced thematic papers on subjects such as service delivery, democracy and participation, natural resource management, traditional authorities, fiscal decentralisation, and food and agriculture.Footnote ⁹ These topics went far beyond the scope of establishing the structure of the local government and, together with the process of constitutional formulation, contributed to setting the scope and limits of would-be state policies. UNDP also contributed to the assessment effort, commissioning a study on the situation of the local government in three southern garrison towns, Juba, Wau and Renk. Though its findings were limited to the (very few) government-controlled areas in the southern region, they reflected the image of fragility that the SPLM's documents had been advancing since the end of the 1990s (Kot Riak Reference Kot Riak2004) and strengthened the movement's demands for training and capacity building.

A comprehensive mapping of all the workshops, meetings and training courses realised in the period 2002–5 and immediately thereafter, as well as a detailed study of the types of discourses they promoted, would be extremely useful in tracing the influences of international approaches to post-conflict governance on the SPLM's internal discourses and institutional development. Indeed, others have looked at capacity building initiatives as ‘points of interaction and spaces for dialogue between donors and recipient governments’ (Bergamaschi Reference Bergamaschi, Olivier de Sardan and Bierschenk2014: 274). However, the fragmentary nature of the available documentation and the poor institutional memory of the international organisations involved in such processes make it extremely challenging to produce a comprehensive picture of all the ‘negotiating tables’ (Hagmann & Péclard Reference Hagmann and Péclard2010) initiated in these years. With no pretence of providing a definitive and complete analysis, it is possible to advance some general reflections on how the SPLM co-opted the material and symbolic resources provided by these workshops into its strategy of extraversion in the establishment of local state and government structures.

First, from the analysis of the lists of participants to both institutional and policy-design workshops and administrative training, an astonishingly complex pre-existing state-like structure emerges at the level of local branches of the SPLM. When signing up as participants in a certain workshop, people introduced themselves with very specific titles and positions following the institutional structure of a state in all its aspects: agriculture extension officers, child protection officers, finance officers, syndicated organisations representatives, economic commission, legal affairs and constitutional development secretariats, etc. These are only a few of the locally existing variety of SPLM-related offices and departments, with the clear ambition of regulating every aspect of life not only at the central level, but also in more peripheral areas, where nuclei of statehood were reproduced loosely based on directives from the headquarters. Of course, one could argue that there is very limited evidence that these distinctions in the roles of civil servants and public security forces were anything more than superficial or that they were more than local initiatives. The variance in titles and groupings that people used to register themselves in the workshops may indeed account for the difficulties in circulating information and for a lack of actual coordination from the centre. Nevertheless, the idea of how a civil service should be structured and of the division of organised forces between the army and other ‘civil’ security forces (such as police, wildlife, prison forces) was appropriated by the movement and clearly penetrated all areas under its control where interaction with resource-rich international actors occurred.

This leads to the second point: these training programmes and workshops provided a valuable meeting opportunity for the SPLM's people in distant areas, and thus a tool for spreading the SPLM's vision and modes of governance. This would have been difficult otherwise, in a context where both telecommunication and transport remain a challenge even 10 years after the CPA. Provision of transport in the form of flight tickets, fuel and road rehabilitation was always one of the main concerns emerging from planning documents in preparation for these workshops.

Third, the involvement of SPLM civil administrators in the process of establishing the local government and other state institutions, as well as in training, meant the legitimisation of people appointed to public functions often with no merit other than being a loyal ex-combatant in the movement. These people were trained in successive workshops organised by several development agencies on the basis that they lacked capacity to implement good governance, and therefore acquired an aura of respectability as skilled local government officers in a society that places great value on education and expertise (Ayod 2013 Int.).

Considered singularly, workshops constituted negotiation tables in which every move of the SPLM leadership regarding local government establishment was shared and discussed with donors, organisers of the workshop, NGOs and any other concerned ‘stakeholder’ – a word very much in vogue in the framework of participatory approaches to development. However, the outcome produced by this process of discussion and negotiation taken as a whole goes beyond that of each workshop and meeting, suggesting the existence of a ‘negotiation arena’ in the sense that Hagmann & Péclard (Reference Hagmann and Péclard2010) attribute to the term. The outcome of this negotiation arena involved, on one hand, the strengthening of SPLM international legitimacy as a source of state-like power, and, on the other, the consolidation of modes of governance based on external resource-raising ultimately leading to the SPLM's control of the state.

TODAY'S LEGACIES OF STATE-BUILDING

When in 2013 the South Sudan success story of state-building collapsed, the SPLM had been ruling for over eight years with the strong support of the international community, which invested over US$4 billion in international aid between 2005 and 2013.Footnote ¹⁰ This period was characterised by constant tensions in the leadership and among the lower ranks of commanders, who were co-opted into the politico-military elite through a complex system of patronage, often organised along ethnic lines (Schomerus & Allen Reference Schomerus and Allen2010; de Waal Reference de Waal2014). The SPLM leadership's capture of state institutions guaranteed its strict control over state resources, which mostly derived from oil revenues and foreign aid. Because of its firm control of state institutions and of the legitimacy deriving from statehood, the former rebel movement could count on an unchallenged primacy on the political scene, since all its possible competitors had no option other than being co-opted into the ruling elite or going to the bush. The continued control over state rents made it less urgent to work on more genuinely inclusive governance systems, as well as on the actual separation – sometimes claimed on paper – between party and state structures. Instead, keeping this overlap was instrumental to maintaining control over resources.

With over 3 million displaced people and an unknown number of victims, the outbreak of war in 2013 regained the humanitarian community's attention, mobilising approximately US$1.5 billion per year from 2014 to 2016Footnote ¹¹ – a figure that increased following the declaration of famine in parts of the country (de Simone Reference de Simone2017). There has been some acknowledgment of the state-building roots of the current situation – for example, in an opinion article published in the Boston Review (28.7.2016), Mahmood Mamdani stated that: ‘The problem in South Sudan did not spring from the society. Its genesis lay in the proto-state, created in a hot-house fashion in the throes of the War on Terror by the troika.’ However, there has been little in-depth reflection on these legacies of state-building.

Such reflection is urgently needed, especially since many observers – including Mamdani himself – have been calling for greater involvement of the international community in South Sudan's crisis (Sudan Tribune 11.1.2014; Newsweek 19.7.2016; Kindersley & Rolandsen Reference Kindersley and Rolandsen2016; Noel Reference Noel2016; The New York Times 8.11.2017). Even though the type of intervention advocated is primarily a diplomatic one, aimed at directly influencing the security situation, the donor community is again heavily involved in providing support to the population affected by the war as well as to government institutions that would otherwise arguably collapse. A recent paper published by the Humanitarian Practice Network of the Overseas Development Institute analyses the features of the current humanitarian intervention in South Sudan, drawing some comparisons with OLS. The paper shows that South Sudanese authorities are engaging in ‘predatory rent-seeking activities’ diverting humanitarian resources such as vehicles, fuel, cash and phone credit (Hamsik Reference Hamsik2017: 25). Even though there are notable differences between OLS and the current humanitarian operations – starting with the humanitarian community's decision not to negotiate access with the rebels – this kind of behaviour is nothing new among the southern elite.

Reminiscent of the 1990s and 2000s, the government of South Sudan is also imposing increasingly restrictive regulations and forms of control over international NGOs. The 2016 NGO Bill gives the Relief and Reconciliation Commission the power to develop policies and procedures to monitor NGOs, ranging from registration procedures, hiring policies, movement clearance and asset ownership. Humanitarian assistance is one of the major sources of revenue in the country, especially considering the insecurity around the oilfields that makes oil revenues unreliable; it is therefore important to impose a strict control over aid flows. Moreover, control over humanitarian actors gives the government a powerful tool to direct relief only towards ‘loyal areas’, excluding rebel-controlled zones (Hamsik Reference Hamsik2017).

The diversion of humanitarian aid and the narrowing of NGO liberties are not the only ways in which South Sudan's ruling elite is repeating its strategy of extraversion and taking advantage of the extremely fragile situation – which it arguably played a part in creating. Playing on its fragility, and on international fears of state collapse, since 2016 the government of South Sudan has been asking aid programmes to pay salaries or bonuses to civil servants to cope with inflation and currency devaluation. This request was for the most part accepted for fear of worsening the population's living conditions with the total collapse of service provision (Irin News 27.2.2017), and much of its acceptability has to do with the fact that the SPLM now acts as the government of South Sudan, a legitimate state actor.

The available analyses of current interactions between South Sudan's ruling elite and the international donor community thus confirm the SPLM's use of extraversion as a strategy to maintain power and hegemony over its adversaries.