Development assistance and the lasting legacies of rebellion in Burundi and Rwanda

Abstract

Rwanda and Burundi have both emerged from civil wars over the past 20 years and foreign donors have provided significant contributions to post-conflict reconstruction and development in the two countries. Yet, although Rwanda and Burundi share several important characteristics, their post-conflict social, political and economic trajectories have been different. This article argues that the nature of the ruling parties in Rwanda and Burundi is key to understanding the countries’ relationships with donors. Rather than seeing aid as an exogenous factor, causing particular development outcomes, it shows how local party elites exert considerable agency over the aid relationship. This agency is influenced by a number of different local contextual factors, including how the parties were established, how they evolved and the ways in which their civil wars ended. Thus, the article provides an analysis of how local context matters in understanding donor–recipient aid relationships, and how the ruling party in Rwanda (the RPF) and in Burundi (the CNDD–FDD) emerged from their respective conflicts with different relationships with international donors.

Keywords:

• Rwanda
• Burundi
• development assistance
• rebel movement transitions
• local agency

It has long been acknowledged that the effects of overseas development assistance are contingent upon the domestic context of recipient countries. Thus understanding development outcomes cannot be limited to analyses of the volume of aid, type of aid and identity of the donors, but must also consider the relationship between aid and political, economic and social structures in the recipient country. Likewise, in seeking to explain the relationship between donors and recipients, several scholars have argued that well-established international relations and development theories have underestimated the role of local agency, meaning the agency of elites and other actors in aid-recipient countries. While local agents certainly face structural constraints, there is a large and diverse literature, particularly in African politics and development, which focuses on how African agents exert their influence in the international and domestic spheres.¹

Yet, even though there is some agreement in the literature on the importance of local political context and local agency in understanding donor–recipient aid relationships, there is no consensus on precisely how they matter. This article focuses on international donor relations with Rwanda and Burundi, two countries that share important social, political and economic characteristics. Rwanda and Burundi have been called false twins, because of their similar ethnic cleavages, colonial histories and experiences of political violence.² Nonetheless, there are important differences as well (hence the name false twins). Perhaps most notably, the majority Hutu dominated the postcolonial political, economic and military landscape until the 1990s in Rwanda, whereas in Burundi it was the minority Tutsi that dominated until the early 1990s.³ There was civil war in Rwanda between 1990 and 1994 and in Burundi between 1993 and 2003, which resulted in two former rebel movements coming to power.

Rwanda and Burundi have different relationships with donors, even though development assistance has been hugely significant in both countries across a range of sectors, both before the civil wars and afterwards. The article explains how national elites in the two countries have influenced aid relationships. It traces the development of the two ruling parties, the Rwandan Patriotic Front (RPF) in Rwanda and the National Council for the Defence of Democracy–Forces for the Defence of Democracy (CNDD–FDD) in Burundi. It shows how the RPF and the CNDD–FDD were influenced by their particular contexts and by the ways their civil wars ended, leading to divergent relations with donors. The trajectory and dominance of the RPF in Rwanda have meant that the Rwandan regime has been better able to manage donors, whereas Burundi’s more fractured political space has made it more difficult for the CNDD–FDD regime to influence donors.

Aid is therefore not an exogenous factor leading to particular development outcomes. Rather, aid is the product of particular histories, relations and interactions. Local elites in Burundi and Rwanda are not entirely ‘free agents’, as they are shaped and constrained by structures and histories, but they are not mere pawns of powerful donors and their agendas either. The contrast between Rwanda and Burundi is thus a contrast in how local agents negotiate the aid relationship.

The article thus argues that the nature of the RPF in Rwanda and the CNDD–FDD in Burundi is key to understanding aid relationships. It begins by briefly outlining key aid patterns. It then shows how the two rebel movements emerged from particular political, economic and social structures, and examines the impact of the respective civil war endings: military victory by the RPF in 1994 and prolonged negotiated settlements in Burundi, culminating in a democratic electoral victory by the CNDD–FDD in 2005. Finally, the article shows how this affected the two ruling parties and their relations with donors.

Development assistance in Rwanda and Burundi

Rwanda and Burundi are frequently compared since they are neighbouring states with historical, geographical, social and political commonalities.⁴ As suggested in Table 1, however, economic and human development indicators are stronger in Rwanda than in Burundi. Rwanda’s GDP growth rates, for instance, averaged more than 10% per year between 1994 and 2004 and have been greater than 5% a year since then, even after the global financial crisis.⁵

Table 1. General indicators in Rwanda and Burundi.

	Rwanda	Burundi
Population (million), 2013	11.78	10.16
Size (km^2)	26 338	27 834
Population density (per km^2)	447	365
Human Development Index 2013 (rank out of 187 countries)	151	180
GDP 2013 in US$ billion	7.52	2.71
Growth 2012	7.3% (2013 est. 5%)	4% (2013 est. 4.5%)
GNI per capita 2013	$630	$260

Sources: Based on data from the World Bank at http://data.worldbank.org/country/burundi and http://data.worldbank.org/country/rwanda and from UNDP Human Development Report 2014, 162.

Thus the two countries have had divergent post-war economic trajectories, despite their low-income status and similar (though not identical) histories and social structures. This has led some authors to conclude that aid has played a central role in post-war development in Rwanda and Burundi.⁶ Holmes et al make a similar point about the advantages Rwanda has had thanks to substantial budget support.⁷

Rwanda has been called an ‘aid darling’ and Burundi an ‘aid orphan’,⁸ although levels of aid have fluctuated over time in both countries. As shown in Figure 1, Rwanda receives a greater volume of aid than Burundi, whereas the volume of aid in the two countries was more similar in the 1980s. Both countries remain highly dependent on foreign aid.

Figure 1. ODA in Rwanda and Burundi (all sectors, all donors).

Source: Data from OECD-DAC, http://www.oecd.org/dac/stats.

As seen in Figure 1, aid levels to Rwanda rose sharply in 1993 with the signing of a peace agreement and the deployment of a UN peacekeeping mission. Aid dropped dramatically in the aftermath of the 1994 genocide, rising steadily again from 1997. In 1998 Rwanda re-established relations with the IMF through its Enhanced Structural Adjustment Facility Programme. There was a drop in development assistance from 2011 in response to concerns over the Rwandan government’s activities in the Democratic Republic of the Congo (DRC).

In Burundi official development assistance (ODA) decreased during the war and especially in 1996 as a result of the sanctions that were placed on the country following a military coup. The Arusha Peace Agreement was signed in 2000 and a transitional government was instigated in 2001. Burundi re-established relations with the IMF in 2001 and ODA steadily increased, especially after the 2005 elections. In 2006 Burundi was one of the first countries on the agenda of the new United Nations Peacebuilding Commission. From 2007 onwards Burundi therefore received funds from the Peacebuilding Fund, which was set up to help fund quick impact projects in the aftermath of conflict.⁹ Aid dropped in 2010 because of donor concerns over financial scandals and governance after the elections that year.

These trends suggest that donors do have leverage, but this does not capture the full story in donor–recipient relations. Donor decisions about aid commitments and disbursements are not separate from the politics of recipient countries. Thus, to understand differences in Rwandan and Burundian post-war donor relationships, it is helpful to consider how the two ruling parties manage and influence donors. The Burundian CNDD–FDD is a weaker party than the RPF in Rwanda, and has less leverage vis-à-vis donors, despite the CNDD–FDD’s greater post-war domestic democratic legitimacy. The RPF is better able to coordinate, manage and manipulate donors, and to channel assistance into priority areas. These differences can in part be explained by how the two movements were created, how they evolved and how their respective civil wars ended.

The emergence and evolution of the RPF and the CNDD–FDD

The CNDD–FDD and the RPF are both former rebel movements but they emerged from different systems with different constraints. A number of authors have argued that the organisational structures of a rebel movement continue to influence its transition to a political party.¹⁰ As Sara Dorman explains, rebellions that aim to ‘liberate’ populations from oppressive rule often have a well-articulated ideology to attract recruits and civilian supporters and to present to the media, academics and donors. These legacies and institutional practices tend to play themselves out in post-liberation years.¹¹ There may also be the lingering influence of a ‘mentalité du maquis’, or a way of thinking that develops from spending years in an armed struggle.

The RPF and the CNDD–FDD emerged in different political circumstances but there were important parallels in the histories of the two countries that influenced later governance. First, unlike many other African states, neither Rwanda nor Burundi is an artificial creation of colonial rule. By the time they were absorbed into German East Africa in 1898–99 most of the territory had already been incorporated into two kingdoms, with centralised state structures. The kingship was the focus of popular loyalties and factional struggles. So both states are strongly socially anchored, with current boundaries that closely mirror pre-colonial political boundaries. Compared with Rwanda, however, the Burundian Kingdom was less centralised and the Mwami (King) was more dependent on popular support.¹²

Second, following Independence on 1 July 1962, both states were strongly authoritarian. After a coup in 1973, when President Habyarimana took power in Rwanda, the National Revolutionary Party for Development (MRND) became the only party. The state was centralised and hierarchical and had an extensive presence at the local level, controlling many aspects of Rwandan social life.¹³ Similarly before the 1990s there was one party in Burundi, the Union for National Progress (UPRONA), and social mobility depended upon participation in the party. The party was linked to the country’s only women’s movement, youth organisation, radio station and newspaper.¹⁴ The state was centralised and influenced all levels of society.¹⁵ The economies of both Rwanda and Burundi were organised through generalised patrimonial rent-seeking. Small business entrepreneurs depended upon having patrons within the administration or the military.

A third feature in both countries is that power structures were ethnically exclusive and regionally divided. Whether the Hutu/Tutsi distinction was historically a racial or a social class difference is contested, but ethnic categories became rigid during and after colonial rule.¹⁶ While the ethnic composition of the two countries was similar, the basis for post-Independence exclusion was different. In Rwanda, following the 1959 Revolution and the overthrow of the Tutsi monarchy, many Tutsi fled to refugee camps in neighbouring Uganda and the majority Hutu dominated post-Independence governments.¹⁷ In Burundi it was the minority Tutsi who dominated post-Independence political, military and economic structures.¹⁸

Fourth, development assistance played an important role in both countries before their civil wars. As Peter Uvin has noted, Rwanda was seen as a ‘model of development in Africa’ before the genocide, with strong development indicators such as high GNP growth and growing industrial production. It also had strong human development indicators, such as high vaccination rates and a vibrant civil society.¹⁹ In Burundi the Tutsi ruling class used the language of national unity and development to garner legitimacy, particularly under the presidency of Bagaza from 1976. Unifying symbols such as the independence hero Prince Louis Rwagasore were propagated, and there was a denial of ethnicity.²⁰ At times there were tensions with donors but Bagaza was successful in attracting development assistance. There are indeed interesting parallels between the Bagaza regime in Burundi and the current Rwandan regime.

Thus the RPF and the CNDD–FDD emerged in countries with a history and tradition of strong statehood. There was greater political accommodation in pre-colonial Burundi, but the colonial and post-colonial elites in the two countries were ethnically exclusive. Post-Independence governments in the two countries were characterised by high levels of authoritarianism and social control, and also high levels of development assistance.

Significantly, however, the RPF has its origins outside Rwanda, whereas the CNDD–FDD was a breakaway group from a political party within Burundi. The RPF can be traced to Tutsi refugees who fled to Uganda following the Rwandan revolution of 1959. The Tutsi refugee community formed the Rwanda Refugees Welfare Association, which was later renamed the Rwandese Alliance for National Unity (RANU) in 1979. Persecution of the refugees grew increasingly severe after the fall of Idi Amin in 1979 and many young Rwandan Tutsi men joined Yoweri Museveni’s National Resistance Army (NRA) guerrilla movement. In January 1986, when Museveni and the NRA captured Kampala, his force included an estimated 3000–4000 Tutsi Rwandan fighters. In December 1987 RANU changed its name to the Rwandan Patriotic Front (RPF). In October 1990 the RPF invaded Rwanda, saying that they wanted to return to their homeland.²¹ This invasion occurred at the same time as the Rwandan regime faced economic vulnerabilities resulting from a drop in food production and coffee prices, as well as political vulnerabilities resulting from pressures from donors to open political space in the context of the post-cold war enthusiasm for multiparty democracy.

Initially, therefore, RPF members were recruited from the armed forces of a foreign country. The RPF was highly influenced by the ideas and structures of the NRM/NRA, as well as by other revolutionary movements such as Mozambique’s Frelimo party. The RPF fighters were well-trained and highly disciplined. The first two leaders of the RPF, Fred Rwigyema (killed in battle in 1990) and Paul Kagame (the current president of Rwanda) were senior figures in the NRA.²² The RPF claimed to be liberating Rwanda, yet in the early stages of war the RPF could not recruit young Tutsi men from the local population in classic guerilla fashion, since the RPF did not have the support of the Rwandan population in whose name they claimed to be fighting.²³ Instead, young Tutsi exiles from all over the Rwandan diaspora, but especially from Uganda, joined to fight in a country that they did not know. This began to change in late 1992, when the RPF started recruiting from the Tutsi population within Rwanda as the Rwandan government’s anti-Tutsi rhetoric increased.

Unlike the RPF, the Burundian CNDD–FDD has its origins within the country as a breakaway faction of the political party that had won the 1993 elections. These elections led to a veritable reversal of power, with a mostly Hutu party, FRODEBU, winning over the incumbent mostly Tutsi party, UPRONA, which had dominated since Independence. The new FRODEBU president, Melchior Ndadaye was assassinated by members of the Tutsi-dominated army three months after taking office, leading to widespread massacres across the country. Unable to control the country, FRODEBU members were divided over whether or not to share power with the former ruling party UPRONA. FRODEBU reluctantly agreed to share power but some members refused, led by Leonard Nyangoma. Nyangoma and his associates created a military wing called the CNDD, which aimed to recapture political power by force and to instigate army reform.²⁴ The CNDD was able to unite some of the different pockets of armed resistance that had emerged around the country in response to President Ndadaye’s assassination, although some people joined other Hutu armed movements such as Palipehutu and FROLINA.²⁵

The CNDD thus emerged as a more radical wing of an established political party in Burundi, even though at various times during the subsequent civil war the CNDD had rear bases in Tanzania and in the DRC.²⁶ Many of the CNDD leaders were former heads or deputies in FRODEBU, and it was only in 1996 that the executive committee of the CNDD became independent from FRODEBU.²⁷

As a result of shifting alliances, regional dynamics and financial considerations, the CNDD split several times. There were frequent clashes with the other main Hutu armed movement, the Palipehutu–FNL, and tensions within the CNDD between members from different regions and religions.²⁸ In 1998 Nyangoma was ousted by Jean-Bosco Ndayikengurukiye and the movement was renamed the CNDD–FDD. It split again in 2001 and Pierre Nkurunziza, the current president of Burundi, became the leader. Nkurunziza and his faction of the CNDD–FDD stayed out of the regionally sponsored Arusha peace negotiations to end the Burundian civil war. Even once the Arusha peace agreement was signed in 2000 and the Burundian transitional institutions were established in 2001, the CNDD–FDD continued to wage war, claiming that the real issue, Burundian army reform, had been inadequately addressed.

So while the CNDD–FDD and the RPF were both rebel movements seeking to influence and capture state power, they were very different kinds of organisations. The RPF was highly trained and disciplined, with a strong focus on political education. The movement was divided into different zones and sectors, and was very well-organised and coherent. In contrast, the CNDD–FDD was constantly affected by desertions and realignments, depending on alternative opportunities for leaders and fighters. The CNDD–FDD leadership had profound disagreements over whether and when to join peace negotiations, whether to share power and how to engage with international and regional mediators. Furthermore, the CNDD–FDD was competing with other Hutu rebel movements in Burundi, particularly the Palipehutu–FNL.

Thus the RPF had weak links to the Rwandan population inside the country, but they were an internally coherent, disciplined, well-structured organisation and they did not face competition from other Tutsi groups. The CNDD–FDD had more domestic popular support, but it was more internally fragmented and faced intense competition from other Hutu groups. As shown below, these differences had an effect on donor relations.

Military victory and negotiated settlement

The Rwandan conflict and genocide ended with the military victory of the RPF in July 1994, whereas the Burundian conflict ended through protracted negotiated settlements, leading to democratic elections won by the CNDD–FDD in 2005. These different conflict endings have important ramifications on governance and institutions, and subsequent relations with donors.²⁹

When the RPF took over as ruling party in Rwanda, many members of the former regime had fled to refugee camps in Zaire, in other neighbouring countries, or overseas. The RPF reaffirmed its commitment to the terms and the spirit of the internationally sponsored 1993 Arusha Accords. The former single-party the MRND and the extremist Hutu party CDR were banned, but other political parties took up their seats in a National Unity government and parliament as set out by the Arusha Accords. Nonetheless, the decisive military victory meant that the RPF had tight military control over the country, and this enabled the RPF to establish social and political control relatively quickly.

Relations between the RPF and some donors became close in a reasonably short space of time. Aside from France, which had provided extensive support to the previous Rwandan regime, most Western donors and diplomats were reluctant to question the good faith of the new RPF rulers, even as a number of human rights organisations began to outline some of the questionable practices of the new regime.³⁰ The USA, the UK and The Netherlands were relative newcomers to Rwanda and became important donors.³¹ Western governments had been criticised for their inaction during the genocide, but they were now committed to assisting reconstruction through development.³²

Thus the way in which conflict ended in Rwanda led to increased centralised control by the RPF, and ‘guilt’ over the inadequacy of international response to genocide gave the RPF greater legitimacy and policy independence vis-à-vis donors. Donors, especially those with little previous experience in Rwanda, were eager to contribute to reconstruction efforts and were receptive to the narratives being articulated by the RPF.³³

In Burundi the negotiated settlement to civil war brought about a very different dynamic. The war in Burundi and the protracted peace negotiations contributed to the fragmentation of the state.³⁴ The Arusha Agreement was signed by 19 parties in 2000 but the CNDD–FDD was not a signatory, as it had stayed out of the internationally and regionally brokered negotiations.³⁵ Burundian transitional institutions were set up from 2001, but the CNDD–FDD continued to fight, claiming that the peace process was not legitimate. The Burundian transitional leadership, regional mediators and international diplomats tried to entice the CNDD–FDD leadership into the peace process and transitional institutions through a mixture of carrots and sticks. In late 2003 a ceasefire agreement was reached which gave key CNDD–FDD leaders important cabinet portfolios in the transitional government, and guaranteed that CNDD–FDD commanders would be given positions within the army and police.³⁶

In 2005 the CNDD–FDD won multiparty democratic elections by a significant margin and former rebel leader, Pierre Nkurunziza, became president. One of the reasons for the CNDD–FDD’s popularity was that it was not associated with the extended Arusha peace process, which many Burundians viewed as an elite-driven exercise that had enriched politicians in the capital Bujumbura. Furthermore, many Burundians across the country believed that the CNDD–FDD was the party that could bring security to a war weary population. Also the movement had extensive shadow administrative structures in many parts of the country during the war, which it was able to mobilise during the election campaign. Nonetheless, despite the electoral victory, Burundi remained highly divided after the 2005 elections, with elites scrambling to strategically reposition themselves to take advantage of the new political landscape and the power-sharing requirements in the new Constitution.³⁷

The extensive and prolonged negotiations to end the Burundian conflict therefore contributed to the further factionalisation of Burundian political space but also had important consequences for CNDD–FDD relations with the donor community. International and regional facilitators were more accustomed to working with other Burundian parties, particularly UPRONA and FRODEBU, since the CNDD-FDD had stayed out of the peace process for such a long time. There were thus strong personal and professional connections between UPRONA and FRODEBU and international donors and diplomats, whereas the CNDD–FDD representatives were relatively unknown.³⁸

In contrast to the RPF, then, the CNDD–FDD gained domestic legitimacy through popular support expressed in what was largely regarded as a fair democratic contest in 2005. Yet the very nature of these negotiated settlements, particularly of the carefully balanced Burundian power-sharing settlement, meant that politics and power in Burundi were much more fractured and diffuse than in Rwanda. In Rwanda, military victory, a receptive international audience because of guilt over the genocide, and the RPF’s tight internal hierarchical structure meant that it was easier for the party to exert effective social and political control and to articulate a consistent narrative to donors.

Local agency and donor relations

The nature of the two ruling parties and the way in which the civil wars ended have important implications in terms of donor relations. In Rwanda the post-genocide political record of the RPF is a matter of enormous disagreement. Critics focus on human rights abuses committed by the RPF against political opposition figures, as well as in the DRC. It is difficult to express one’s views openly in Rwanda, or to assert political identities that fall outside officially circumscribed categories.³⁹ Critics also say that Rwanda’s impressive growth rates hide large and growing vertical and horizontal inequalities in the country. Government policy is to concentrate land-holdings and modernise agriculture through intensification, leading to large gaps between the urban elites in Kigali and small-scale farmers.⁴⁰ There are ambitious, top-down rural development schemes designed to encourage villagisation, to commercialise production and to promote regional crop specialisation, with harsh penalties for non-compliance.⁴¹ Supporters of the regime, on the other hand, point to economic success and notable progress on a range of social and economic indicators, despite a very difficult history.⁴² They praise the RPF for being able to (re)build a strong state.⁴³

Both critics and supporters point to the ability of the RPF to exert considerable agency in the donor relationship. Despite its heavy reliance on donors, the RPF has achieved a high level of political autonomy.⁴⁴ There are several factors that have increased the RPF’s ability to direct and manage donor relations. First, military victory by the RPF helped it gain a near monopoly over information and the production of knowledge about the history of the genocide, and enabled relations with new donors. Aid workers, international (mainly anglophone) journalists, scholars and donors who arrived in Rwanda in the immediate aftermath of genocide often knew very little about the country and were willing to accept the new government’s version of the truth.⁴⁵ The RPF actively cultivated close diplomatic, aid and intelligence relations with the USA and the UK.⁴⁶ Thanks to the RPF’s origins in Uganda, many of its officials speak English very well and have emphasised some donor priorities such as gender equality and effective governance, while simultaneously pursuing their own political agenda.⁴⁷ For instance, Rwanda has comparatively high percentages of women in the national parliament and in the labour force.⁴⁸

Second, the RPF used the language of unity and reconstruction, which resonated with donors.⁴⁹ In this ‘transformed’ Rwanda ethnic identities no longer matter, but are always in danger of resurfacing (thus requiring heavy-handed governance tactics).⁵⁰ Instead of ethnic categories, the categories in Rwandan legal and administrative documents are derived from the official RPF interpretation of the genocide: survivor, old caseload returnee, new caseload returnee, suspected genocidaire.⁵¹ In the absence of popular support within Rwanda, the RPF uses memory and memorialisation of the genocide as a means to justify its power.⁵² Furthermore, the RPF presents itself as an essential bulwark against the risk of future genocide, and has used this justification to pass legislation, such as the genocide ideology law of 2008, which removed voices of dissent or opposition.⁵³ The risk of future violence has also been used to justify Rwanda’s military presence in the DRC.

Third, the RPF has defined itself as a party offering modernity, development and progress. While the RPF cannot draw upon democratic legitimacy, it derives its legitimacy from economic progress, or what is sometimes called ‘performance legitimation’. Interestingly this is reminiscent of earlier government strategies. In the 1970s, despite exclusionary and authoritarian governance, the Rwandan government garnered legitimacy through the discourse of successful ‘development’. In 1974 the Rwandan parliament was renamed the National Development Council and the president announced that Rwanda’s 143 communes would be the ‘motors of development’.⁵⁴ Today some authors say that Rwanda should be characterised as a developmental patrimonial state, albeit with some qualifications.⁵⁵ The RPF wholly owns the private sector holding company Tri-Star Investments/Crystal Ventures Ltd (CVL) and has majority stakes in many other leading Rwandan companies.⁵⁶ Profits from the operation of subsidiaries are taxed and either reinvested or returned as dividends to the RPF.⁵⁷ Marysse et al say that, when making decisions about aid allocation, donors look at the technocratic elements of governance such as government effectiveness, the regulatory burden, corruption and rule of law, instead of political sensitivities, meaning that Rwanda is an attractive recipient.⁵⁸ In 2010 Rwanda was ranked 67th out of 143 countries in the Doing Business Report, and was called the world’s top reformer.⁵⁹

Fourth, the RPF has positioned itself as being central to efforts to bring security to unstable parts of Africa.⁶⁰ It has a highly trained military and is currently the sixth largest troop and police contributor to the UN.⁶¹ Rwanda deployed its first peacekeepers to the African Union Mission in Sudan in 2004. Since then it has contributed, among others, to the AU–UN Hybrid Operation in Darfur (UNAMID), the UN Mission in South Sudan (UNMISS), the UN Stabilization Mission in Haiti (MINUSTAH), the UN Mission in Liberia (UNMIL), the UN Interim Security Force in Abyei (UNISFA), the Operation in Côte d’Ivoire (UNOCI), the UN Peacebuilding Office in Guinea-Bissau (UNIOGBIS), and the African-led International Support Mission in the Central African Republic (MISCA).⁶² These contributions should not be underestimated. When Rwanda was criticised in a 2010 draft UN report the Rwandan government warned that it could withdraw its peacekeepers from Darfur.⁶³

Fifth, RPF rhetoric mirrors the language of local ownership and African solutions. For instance, RPF officials often use the language of the 2005 Paris Principles on Aid Effectiveness.⁶⁴ The Rwandan government developed its ‘Vision 2020’ policy guidelines, outlining targets and goals to transform Rwanda into a middle-income country by 2020.⁶⁵ Given the RPF’s skill at using ownership language that appeals to its most important donors, the international donor community has been inclined to follow the priorities and plans set by the government.

Thus the RPF has been successful in managing donors and controlling the narrative about its post-conflict trajectory. Nevertheless, since 2010, criticism of Rwanda among many donors has increased. In 2011 and 2012 key donors, including the USA, UK, Germany and the Netherlands delayed disbursement or withdrew part of their aid because of renewed allegations by the UN that Rwanda was supporting M23 rebel activity in the DRC.⁶⁶ It is too early to say if this marks a fundamental shift in donor relations. There are also some signs that the internal coherence of the RPF regime has increasingly come under stress, although most critics are in exile.

Burundi, on the other hand, has been consistently unable to exert the same kind of leverage over international donors. The negotiated settlement to the conflict in Burundi, the politics of accommodation and the divided power-sharing political landscape have led to competing poles of power within Burundi. In Rwanda the RPF has maintained tight discipline within the party, with detractors forced to go into exile. In Burundi constitutional provisions for power sharing have meant that the CNDD–FDD is required to include dissenting voices within the party, and it has recruited many prominent Tutsi within its ranks. As described above, the CNDD–FDD never had the same level of discipline as the RPF, because of the different conditions it faced in its armed struggle. Indeed, Willy Nindorera points out that the same neo-patrimonial and factionalised practices that currently characterise the Burundian government also characterised the earlier CNDD.⁶⁷

One might have expected that a Burundian emphasis on democracy, inclusivity and shared power would appeal to donors, especially since international and regional actors were instrumental in the establishment of the Burundian power-sharing system. Yet the CNDD–FDD has been unable to articulate a narrative that resonates with donors. At the time of its creation, throughout its armed struggle, as well as during the post-conflict period, the CNDD–FDD has been a fractured organisation with competing voices that have hindered the emergence of such a narrative. Even though the post-war Burundian political environment – with its multiparty elections and more inclusive political space – stood in marked contrast to the Rwandan one, it was not possible for the CNDD–FDD to put forward a compelling narrative based on democracy, rights and inclusivity. In part this is because the CNDD–FDD has increasingly turned to more coercive and authoritarian tactics and governance strategies, particularly since the 2010 electoral campaign. The CNDD–FDD has tried to tighten its control of state companies, provincial governorships and the court system. It has tried to emulate the RPF’s political governance,⁶⁸ and has sought to assert itself as the only legitimate decision maker that can distribute the spoils of office in return for loyalty. The youth wing of the CNDD–FDD, the Imbonerakure, has been accused of violence and intimidation against opposition members in the countryside.⁶⁹ The large protests in Bujumbura in 2015 against President Nkurunziza's decision to run for a third presidential mandate were met with further violence and repression against the independent media, civil society and the opposition. Thus a narrative of democracy and inclusivity is not credible.

The CNDD–FDD was also unable to put forward a narrative based on economic performance and technocratic management. In general, the aid system in Burundi has supported and even strengthened the neo-patrimonial system that benefits the ruling elite.⁷⁰ Economic relations are largely neo-patrimonial, as in Rwanda, but since structures are more fractured, patrimonialism is less likely to be developmental. In other words, the CNDD–FDD is not able to be directive in the same way as the RPF. Decentralisation was reintroduced in 2005 but this has reinforced non-developmental neo-patrimonial political structures.⁷¹ The financial sector in Burundi is used as a source of rents rather than development finance and thus has not had a significant effect on economic development.⁷² Institutions are weak and Burundi has a limited capacity to mobilise revenues.⁷³ Weak capacity has also affected monitoring ability, which further affects accountability.⁷⁴ In 2011 a Vision 2025 document was released by the government, which articulated a vision of goals for the country to 2025, including an annual GDP growth of 10% and a reduction in the poverty rate.⁷⁵ It is very unlikely, however, that these goals will be met.

Furthermore, as outlined above, when the CNDD–FDD won the 2005 elections, relations with the UN mission were strained, in part because the donor community was accustomed to dealing with the two other main political parties.⁷⁶ CNDD–FDD officials perceived the UN and other international actors to be biased towards the political opposition and civil society, and felt that, since their party had won the elections, the UN mission should leave governance tasks to the new democratically elected government. As relations grew worse, international donors described the CNDD–FDD as inexperienced, intransigent, authoritarian and in need of ‘training’.⁷⁷ CNDD–FDD officials used the language of sovereignty, legitimacy and autonomy in an attempt to loosen donor influence and to exert their control and agency. Several of the most senior UN officials in the country, including two United Nations Special Representatives of the Secretary-General, were seen as being too close to the opposition and were asked to leave the country by the government.

Despite these attempts to assert its autonomy, the CNDD–FDD has not been able to coordinate or direct donors in the same way as the RPF. A number of institutions have been set up to manage relations, such as the National Committee for the Coordination of Aid, established in 2005. A permanent secretariat for this National Committee was established to monitor aid flows and increase coherence and coordination, to support the government’s implementation of the Paris Principles, and to serve as the secretariat for the Partners Coordination Group.⁷⁸ However, in practice, while the government has embraced the aid effectiveness agenda, it has not been able to implement many of the principles.⁷⁹ Since donors largely mistrust Burundian institutions, they often work outside these national structures.

The CNDD–FDD has also tried to exert agency through security narratives. Like Rwanda, Burundi has contributed troops to peacekeeping missions, most notably to the African Union Mission to Somalia (AMISOM) as well as to the MISCA, which became a UN force in September 2014 (MINUSCA). The CNDD–FDD has tried to justify its harsh and unlawful treatment of opposition members with language echoing the Rwandan security discourse, but this has not significantly mitigated donor criticism.

Conclusions

Despite its authoritarianism and human rights abuses, the RPF has largely maintained its image as an effective moderniser in the eyes of many donors, consistently emphasising the country’s economic accomplishments as an alternative (African) source of internal legitimacy. Paradoxically the more fractured but more inclusive and democratic Burundian state has had less leverage vis-à-vis donors. This has worrying implications, as it suggests that, under some conditions, more monolithic, authoritarian and repressive regimes can better manage and influence donors. Indeed, the CNDD–FDD learned a lesson through watching the RPF, and has increasingly limited political space and governed in a heavy-handed, coercive, and exclusive manner. The rising tension and violence in the lead-up to the 2015 Burundian elections, and the constitutional controversy over President Nkurunziza’s candidacy, present a sobering view of Burundi’s political trajectory.⁸⁰ Many regional leaders and international donors strongly criticised President Nkurunziza and withdrew support for the 2015 elections. The crisis has highlighted serious divisions within the CNDD-FDD and the vulnerability of Burundi's post-war institutions, with further violence and increased authoritarianism as two likely outcomes.

This article has shown that international donor involvement in Rwanda and Burundi is, in part, a product of very different war-time trajectories. Aid is a consequence of particular relationships, not only a cause, and ruling elites in recipient countries play a critical role in constructing those relationships. Part of the difference between Rwandan and Burundian aid relationships can be explained by the different political contexts leading to the emergence of the RPF and CNDD–FDD, their different internal structures, and the fact that the Rwandan civil war ended in military victory, while the Burundian civil war ended in protracted negotiated settlements. Thus the RPF has been better able to articulate a single narrative vis-à-vis donors, and has more policy autonomy and a stronger ability to direct its international donor engagements. The CNDD–FDD operates within a more fractured political context and it has had much more difficulty managing donor relations. Yet in both countries it is the wider population that continues to suffer the consequences of the gap between the rhetoric and actions of both national elites and international donors.

Notes on contributor

Devon E.A. Curtis is a Senior Lecturer in the Department of Politics and International Studies at the University of Cambridge and a research associate in the Department of Politics at the University of Johannesburg. Her main research interests focus on power-sharing and governance arrangements following conflict, UN peacebuilding, non-state armed actors in the Great Lakes region, and critical perspectives on conflict, peace and development. She is the co-editor of Peacebuilding, Power and Politics in Africa (Ohio University Press, 2012).

Acknowledgments

This article is a revised version of WIDER Working Paper 2014/128. Special thanks to Rachel Gisselquist, the editor of this special issue, as well as the participants in the UNU-WIDER workshop held in Helsinki. Thanks also to Marie-Eve Desrosiers, René Lemarchand, Katrin Wittig and two anonymous reviewers for their helpful comments on an earlier version of this paper.

Notes

1. See for instance, Bayart, “Africa in the World”; Clapham, Africa and the International System; Brown, “A Question of Agency”; Whitfield, The Politics of Aid; Beswick and Hammerstad, “African Agency”; and Chabal and Daloz, Africa Works.

2. For instance, Lemarchand, “Part II: Rwanda and Burundi.”

3. In both countries the figures that are typically cited put the Tutsi at 14% of the population, the Hutu at 85%, and the Twa at 1%, but these figures are disputed and doubtful.

4. See Lemarchand, Rwanda and Burundi; Hintjens, “Rwanda and Burundi”; and Chrétien, Le défi de l’ethnisme.

5. IMF, Rwanda.

6. Marysse et al., “The Aid Darlings.”

7. Holmes et al., For State and Citizen.

8. Marysse et al., “The Aid Darlings.”

9. See Campbell et al., Independent External Evaluation, for an analysis of these projects.

10. Bøås and Dunn, African Guerrillas; and Manning, “Armed Opposition Groups.”

11. Dorman, “Post-liberation Politics,” 1086.

12. Lemarchand, Rwanda and Burundi, 19–39.

13. Newbury, “Rwanda.”

14. Lemarchand, Burundi.

15. Ndikumana, “Distributional Conflict.”

16. See Mamdani, When Victims become Killers; and Lemarchand, Burundi.

17. The first post-Independence president, Gregoire Kayibanda (1962–73), drew his support mainly from the south, whereas President Juvenal Habyarimana (1973–94) drew his support mainly from northern Hutu.

18. Post-Independence presidents Micombero (1966–76), Bagaza (1976–87) and Buyoya (1987–93, 1996–2003) were all Tutsi from the same clan and commune in Bururi province.

19. Uvin, Aiding Violence, 40–50. Uvin argues that the development aid system in Rwanda interacted with political and social processes that made genocide possible. See also Andersen, “How Multilateral Development Assistance.”

20. Lemarchand, Burundi, 108–110; and Vandeginste, “Governing Ethnicity.”

21. Otunnu, “An Historical Analysis.”

22. Rwigyema had been Deputy Commander of the NRA and Major Paul Kagame had been Head of Intelligence.

23. Prunier, “The Rwandan Patriotic Front,” 131.

24. Another Hutu rebel movement, called the Palipehutu, had formed earlier in Tanzanian refugee camps in 1980. Many Palipehutu members were refugees who had fled Burundi during the 1972 genocide. See Malkki, Purity and Exile.

25. Many thanks to Katrin Wittig for highlighting this point.

26. For an excellent discussion of the origins and evolution of the CNDD–FDD, see Burihabwa, “Unpublished PhD dissertation.”

27. Nindorera, The CNDD–FDD, 15.

28. Ibid., 16.

29. See also Uvin, “Structural Causes.”

30. See, for instance, Amnesty International, Reports of Killings; and Human Rights Watch, Rwanda.

31. By 2011–12 the top four donors to Rwanda were the World Bank (International Development Association), the USA, the Global Fund to Fight AIDS, Tuberculosis and Malaria, and the UK. The top four donors to Burundi that same year were the World Bank (International Development Association), EU institutions, Belgium and the USA. See OECD-DAC, www.oecd.org/dac/stats.

32. Hayman, “Abandoned Orphan, Wayward Child,” 347; and Marriage, “Defining Morality,” 480.

33. Hayman, “Rwanda,” 164.

34. Uvin and Bayer, “The Political Economy,” 266.

35. For a discussion of the peace processes, see Daley, Gender and Genocide.

36. For details of the various agreements, see Vandeginste, “Power-sharing.”

37. Curtis, “The International”; and Vandeginste, “Power-sharing.”

38. Jackson, “The United Nations.”

39. Reyntjens, “Constructing the Truth.” Because of this repression, challenges to RPF authority and hegemony are largely expressed outside Rwanda by the Rwandan diaspora.

40. Ansoms, “Re-engineering Rural Society.”

41. Newbury, “High Modernism.”

42. For an articulation of the RPF view of history and politics, see Murigande, “The Rwandan Genocide.” Murigande was the Rwandan Foreign Affairs Minister when he wrote this piece.

43. McDonough, “From Guerrillas to Government,” 358.

44. Hayman, “Rwanda,” 168–178; and Ansoms and Rostagno, “Rwanda’s Vision 2020,” 430.

45. Pottier, Re-imagining Rwanda.

46. Hintjens, “Post-genocide Identity Politics”; Beswick, “Managing Dissent”; and Fisher, “Structure, Agency.”

47. Reyntjens, “Constructing the Truth.”

48. UNDP, Human Development Report 2011.

49. Rwandan peasants’ ‘hidden’ resistance to this narrative is outlined in Thomson, “Whispering Truth to Power.”

50. See Straus and Waldorf, “Introduction,” 8–10.

51. Hintjens, “Post-genocide Identity Politics,” 14.

52. There has been extensive international support for the Rwandan government’s project of memorialisation. See Ibreck, “International Constructions.”

53. Waldorf, “Revisiting Hotel Rwanda.”

54. Uvin, Aiding Violence, 24.

55. Booth and Golooba-Mutebi, “Developmental Patrimonialism?”

56. Tri-Star emerged from the RPF’s war-time ‘Production Unit’, which generated funds for the RPF as a rebel movement, and helped fund the government in the immediate aftermath of the genocide.

57. Booth and Golooba-Mutebi, “Developmental Patrimonialism?” 389.

58. Marysse et al., “The Aid Darlings,” 449–450.

59. World Bank, Doing Business 2010.

60. Beswick, “The Risks.”

61. From http://www.un.org/en/peacekeeping/resources/statistics/contributors.shtml.

62. From http://www.gov.rw/Rwanda-deploys-peacekeeping-troops-to-help-stabilise-the-Central-African-Republic.

63. Gettleman and Kron, “Rwanda threatens to Pull Peacekeepers.”

64. Zorbas, “Aid Dependence,” 108.

65. Government of Rwanda, Rwanda Vision 2020.

66. The UK decided not to release £21 million to Rwanda because of the Rwandan government’s support of the M23 rebel movement in eastern DRC. In March 2013 £16 million of this aid was reprogrammed, channelled through projects rather than as direct budget support. See DFID Ministerial Statements, November 30, 2012, and March 1, 2013.

67. Nindorera, “The CNDD–FDD,” 15.

68. Many thanks to an anonymous reviewer for pointing this out.

69. Report of the Secretary-General; and Bouka, “Status and Dynamics.”

70. Uvin and Bayer, “The Political Economy,” 272.

71. Gaynor, “Bringing the Citizen Back In.”

72. Nkurunziza, Why is the Financial Sector? Nkurunziza et al., The Financial Sector.

73. African Development Bank, Domestic Resource Mobilization.

74. Desrosiers and Muringa, “Effectiveness under Fragile Conditions?” 513–514.

75. Ministry of Planning and Communal Development, Complete Vision 2025.

76. Jackson, “The United Nations.”

77. See, for instance, ICG, Burundi.

78. Desrosiers and Muringa, “Effectiveness under Fragile Conditions?” 507.

79. Ibid., 511.

80. For a discussion of the controversy surrounding President Nkurunziza’s eligibility for a third presidential mandate, see Vandeginste, La limitation constitutionnelle.