Introduction
The level of debate and literature on Rwanda's development since the aftermath of the 1994 genocide is noteworthy for both its volume and its intensely polarised nature. On the one hand, commentators and scholars celebrate the remarkable developmental achievements of this small, landlocked, resource-poor state recovering from the trauma of a horrific genocide (Clark 2010; Ensign and Bertrand 2009; Golooba-Mutebi 2008; Stansell 2009; see also Crisafulli and Redmond 2012; Gourevitch 2009; Kinzer 2008; Zakira 2009 for more popular accounts). There is much substance to these claims. In the last 10-year period, the country has registered average growth rates of 8% per annum and has reduced inflation since 2009 (IMF 2013). It has received widespread praise for its economic governance reforms and low levels of corruption, making it an increasingly attractive location for inward investment (IMF 2013; World Bank 2013, 37–41). On the social front, through a series of ‘fast-track reforms' in agriculture, infrastructural developments and social service provision at local levels, the government appears to have effected a spectacular reversal in poverty and inequality in recent years, reporting a decline in income poverty from 57% in 2006 to 45% in 2011 and a fall in the country's Gini coefficient from 0.52 to 0.49 over the same period (GoR 2013).1 The country is on track to meet two of its Millennium Development Goals (MDGs), in health and education, and has drawn much attention as the global leader on female political representation, surging to 64% in recent parliamentary elections.
These remarkable outcomes notwithstanding, Rwanda is not without its critics. Criticisms are, for the most part, centred around two areas. The first is the country's alleged political interference and criminal activities in eastern Congo (Prunier 2011; Reyntjens 2009; UNSC 2012) – allegations which the government continues to deny but which led to the suspension of development aid in 2004, 2011 and 2012–2013. The second focuses on the authoritarian, top-down style of governance which, critics argue, employs a discourse of national unity and reconciliation to silence political opposition and dissent (Ansoms 2011; Desrosiers and Thomson 2011; Purdeková 2011; Reyntjens 2011, 2004; Straus 2006) while also exercising an excessive control over the media, information and narratives on the Rwandan story more broadly (Beswick 2010; Frère et al. 2010; Ingelaere 2010; Pottier 2002; Reyntjens 2011). At more local levels, while the government, through its much-celebrated decentralisation programme,2 emphasises the participatory nature of local governance with development priorities and actions determined by communities themselves, others have pointed toward strong centralised tendencies in the implementation of certain policies and programmes at these levels (see Waldorf [2011] on gacaca, Newbury [2011] on the umuganda villagisation programme, and Ansoms [2008, 2009] and Ansoms and Rostagno [2012] on rural development initiatives). These contributions raise significant questions in relation to the popular legitimacy of the fast-track reform process, together with the role of local communities within it.
This article aims to make a contribution in this regard. It draws on relevant policy and field research on local governance in Rwanda. Field research was conducted in February–March 2013 across six diverse sites (selected to reflect both geographic and socio-economic diversity and including one of the wealthiest and one of the poorest districts in the country). It was not possible to select for ethnic difference or to explore ethnic issues more broadly within the research as legislation introduced in 2008 now renders any discussion or mention of ethnicity a criminal offence (charges include ‘divisionism’, ‘negationism’ and ‘genocide ideology’). This would have provided a clear rationale for Rwandan authorities to deny the research visa which is now required of all foreign researchers and which proved extremely difficult to secure in this case.3 Semi-structured interviews with 99 randomly selected residents (50 female, 49 male – selected as the third woman/man encountered while walking through villages each day where my freelance translator and I arrived unannounced) and 23 semi-structured interviews with local authorities (nine at district level, eight at sector level and six at cell level) were conducted in total across the six sites. In addition, research included a structured observation of an umuganda local planning session in Kigali, and semi-structured interviews with two national government officials, three non-governmental organisations (NGOs) and eight international donor representatives. Interview transcripts were coded following fieldwork and the data were processed through SPSS software to allow for descriptive statistical analysis.
The findings chart developments in local governance over the last decade and demonstrate an increasing centralisation of deliberation and decision-making on local development in tandem with growing pressures and demands on local communities to invest – physically and financially – in centrally promoted activities and programmes. These findings, which uncover growing levels of popular disquiet and dissent with the centrally driven approach, raise questions regarding the legitimacy of the ruling regime and therefore, the sustainability of its development project. The article is structured as follows.
The following section conceptualises the Rwandan state's fast-track reform project by situating it within contemporary debates on the nature and evolution of African developmental states and the current interest in development studies circles in ‘good enough governance’. Drawing on my fieldwork, the third section examines the two principal community-planning initiatives, ubudehe and umuganda. The transformation of the former in the mid 2000s from a local, participatory, deliberative planning process to a technocratic, socio-economic classification system is traced, while the equation of community participation with individual responsibility and locally negotiable cost-sharing for centrally driven reforms in the latter is examined. The resultant disconnect between community and local authority priorities is also highlighted in this section. With an eye to the ongoing high levels of inequality within Rwanda, the fourth section takes a closer look at the critical question of who pays for these centrally driven reforms. It finds that a disproportionate burden of cost lies with local communities and reveals a growing public disquiet with the escalating costs of arbitrary, locally negotiable taxes. The article concludes with a discussion of the implications of the findings for contemporary thinking in relation to the currently popular ‘good enough governance’ agenda and the African developmental states literature more broadly.
‘A nation in a hurry’: theorising the Rwandan state
The recent confluence of the increased emphasis on state-building within Development Studies and International Relations (see Marquette and Beswick 2011) with the post-good governance move from normative and ultimately unsuccessful ‘best practice’ to ‘best fit’ (Booth 2011, 1), as typified by the range of ‘good enough’ institutional arrangements now advocated by a number of analysts (for example, Menocal's [2011] ‘good enough statebuilding’, the Overseas Development Institute's ‘good enough governance’ [Booth 2011 following Grindle 2004], Hayman's [2011] ‘good enough democracy’ and the World Bank's [2011] ‘inclusive enough settlements'), has focused attention on the nature and performance of the Rwandan state, together with those of African states more broadly. In particular, drawing on the ‘developmental state’ literature, which emerged in the wake of developmental successes in East Asia, there has been a noticeable interest in examining if this concept proves useful in analysing contemporary African political economies4 (see Edigheji 2010; Fritz and Menocal 2007; Meyns and Musamba 2010; Mkandawire 2001).
While the meaning of the term ‘developmental state’ is both disputed and evolving (Evans 1995), two broad fundamental features of such states may be identified (Meyns and Musamba 2010; Routley 2014). The first is that they produce ‘developmental outcomes' – again a contested concept but generally including economic growth promoted by policies for rapid industrialisation; increases in the standard of living for a large proportion of the population through redistribution; and agricultural reform. The Rwandan government's aspirations to transform Rwanda into a middle-income country by the year 2020, combining high growth levels with ‘pro-poor’ policies as set out in its influential Vision 2020 document (GoR 2000, 4), resonate strongly with this feature. In practice however, the focus has been more on growth than redistribution (Ansoms 2008; Ansoms and Rostagno 2012) although, through its decentralisation programme, service provision and agricultural reform have become policy priorities in recent years.
The second feature, focusing on state–society relations, is that the state is not overly influenced by particular social groups (captured) but remains connected enough to society and enjoys sufficient legitimacy to be able to ensure growth and distribution – what Peter Evans (1995) has termed ‘embedded autonomy’. The degree to which this is a feature of the Rwandan state at the national level remains highly contested. On the one hand, Booth and Golooba-Mutebi (2012), following their study of the central role played by the ruling party in the development of the private sector, argue that the party's investment of dividends from its private holding companies back into the economy represents a shift from the neo-patrimonial tendencies of other African regimes towards a more developmental model. The centralised rent management among a narrow group of Kigali-based elites is identified by these authors as the key feature of the country's developmental success. On the other hand, this analysis is vigorously contested by Nilgün Gökgür, who presents evidence to contrarily argue that the close relations between the ruling party and its private companies amount to state capture, with state companies turning into extractive institutions which are draining the scarce resources of both the state and donors and impeding the development of a more diverse private sector (Gökgür 2012).
A related dimension, and one especially pertinent to this study, is whether high levels of authoritarianism are compatible with developmentalism and whether democratic and developmental imperatives should (or even can) go hand in hand. This goes to the heart of the ‘good enough governance’ debate (Grindle 2004), which advocates some trade-off in democratic imperatives against developmental outcomes. Writing from this perspective, Booth (2011) argues that elite-level action is what matters and that efforts at broad-based citizen participation yield poor results. ‘Citizen pressure is at best a weak factor and at worst a distraction from dealing with the main drivers of bad governance’ (2). Thus, for Booth (3–4), ‘sufficiently inclusive coalitions of support’ are enough and ‘the idea of “democratic ownership” … [represents] a step backwards’ or a ‘naïve liberalism’ (Booth and Golooba-Mutebi 2014). A contrary view is provided by Evans (2008), who, looking at the requirements of future developmental states, highlights the necessity for broader, more complex and more democratic forms of social embeddedness. According to Evans, the twenty-first-century developmental state is one which departs from the narrowly embedded East Asian model and derives its policies ‘from democratically organised public deliberation’ (Evans 2010, 43, cited in Routley 2014, 166). Interestingly, this view echoes contributions from recent peacebuilding literature which highlight the importance of local ownership and participation as a key factor in building state legitimacy and support (Berdal 2009; Menocal 2011). Thus, while the ‘good enough governance’ argument, which highlights the dangers in promoting liberal-democratic norms and institutions regardless of context, injects some much-needed realism into debates about contemporary African states and governance, it perhaps fails to appreciate the underlying premise of Evans' future developmental state – that states (particularly resource-poor states) depend on active citizen support and participation in the developmental project. The key question therefore becomes whether it is possible for a strongly authoritarian or, in Jones, Soares de Oliveira, and Verhoeven's (2013) terms, ‘illiberal’ state to be developmental and secure popular legitimacy and support. This is the question examined here through an examination of local decentralised governance mechanisms and their political impact.
‘Participatory’ planning: ubudehe and umuganda
citizen participation in decision making is one of [the] key elements of [the] national decentralization policy adopted in 2000 and revised in 2013. By law local government authorities are required to conform to participatory process[es] in planning and budgeting as well as other processes in their area of jurisdiction … Most importantly, citizens participate in planning process[es] directly at the village and cell levels. (MINALOC 2013, x)
Although the focus of limited academic attention,5 Rwanda's ambitious decentralisation programme is regarded as one of the country's significant successes, both among donors in Kigali and within the region more broadly. Interviewees routinely praised it for its administrative efficiency, its target-driven focus and its ability to engage communities through innovative participatory planning mechanisms. These mechanisms are examined below.
Ubudehe
Ubudehe, described by the Ministry of Local Government as ‘the traditional Rwandan practice and cultural value of working together to solve problems’ (MINALOC n.d., 1), was developed in the early 2000s as a local planning mechanism. Involving a four-step process – household classification (assigning households to wealth/poverty-based ubudehe categories), local resource mapping, problem prioritisation and action planning (MINALOC n.d.) – the process has been hailed as genuinely participative and democratic. In 2008, the Rwandan government was awarded a UN Public Service award for ubudehe and Shah (2013, 53), following ethnographic research on the process in the late 2000s, argues that it ‘has contributed to democratisation in Rwanda through the introduction of elements of village level participatory democracy and through a less tangible impact on a broader shift from an “obedience culture” to a “citizenship culture” in the country’.
Today, however, little remains of the original process. Resource maps, once available for consultation in the offices of local authorities, are now archived in Kigali and deliberations on problem prioritisation and resolution are, for all interviewed for this research, a dim and distant memory. Ubudehe now encompasses the first step only – household classification (on a scale of 1–7 where 1 is the poorest and 7 the most affluent). While communities remain involved in this annual exercise, final decisions on categories now rest with local authorities. Moreover, following the elaboration of a national database of statistics from ubudehe data in December 2010, the government is now linking welfare payments and tax obligations to ubudehe categories. Category 1 and 2 households receive some welfare payments through a donor-funded scheme called VUP6 and their mandatory public health insurance payments are waived. Category 3 households receive no welfare assistance and must pay an annual health insurance premium of RwFr3000 per household member, with this rising to RwFr7000 for Category 4 households.
The consequences of this technocratisation of ubudehe are twofold. First, with resource availability now the guiding factor in household categorisation, interviewees' discussions of changes in their household categorisations reveal a consequent ‘upgrading’ of households to Category 3, where they both lose their VUP assistance and become liable for state taxes. While the most common category among the 99 interviewees (42%) was Category 2 last year, this year has seen a significant increase (up to 64%) of households in Category 3. Overall, 31% of interviewees' households have been moved up a category while just 2% have been moved down one. Second, there is evidence of unease and dissatisfaction with these new classifications. Overall, 49% of interviewees expressed dissatisfaction with their classification. Very often, this view was provided unsolicited with a typical response to an introductory question regarding ubudehe being ‘I am in Category 3, I should be in Category 2.’ Comparing across research sites, a correlation is apparent between levels of satisfaction and lower-level categorisations. For example, in Site A, where 70% of interviewees are in Categories 1 or 2, 80% were happy with their categorisation. Contrarily, in Site E, where 73% of interviewees are in Category 3, 82% of interviewees expressed dissatisfaction with their categorisation. While, given the small sample size, these figures in no way reflect overall ubudehe categorisations in the sites sampled or indeed nationally, the findings do suggest a growing dissatisfaction with the state's increasing financial demands. As to ubudehe's demise, local officials and quite a number of civil society representatives argue that its spirit continues in the dialogue, discussions and planning sessions which take place at monthly umuganda meetings. This is discussed below.
Umuganda
Umuganda refers to the tradition, prevalent in the Great Lakes region since the colonial era, of obligatory communal labour on public projects. In Rwanda, umuganda labour works take place on the last Saturday morning of every month. Following the physical work, a meeting is held where, in theory, the communal work carried out that day is evaluated and plans for further works are discussed and agreed, together with community issues and plans more broadly. The issues raised at these meetings, officials stress, feed upward to cell, sector and district plans with, every five years, a formal upward-planning process taking place for the development of the District Development Plan.
Clearly, for meaningful community participation to take place within any policy or planning sphere, there needs to be a well-worked-out system of ranking of preferences which is open, transparent, and follows a set of clearly defined criteria. Yet none of the local officials interviewed were able to outline any prioritisation process used. At more local levels, authorities suggested that issues raised in meetings in different cells and villages are the same. ‘The priorities are similar. They are almost the same with all the cells so it is not like a big problem how to prioritise. All the cells seem to be on the same level’ (Sector Executive Secretary, Site E). At district level, national (Economic Development and Poverty Reduction Strategy – EDPRS) and international (MDG) frameworks were cited as important in setting development objectives, yet the link to local priorities remains unclear. Indeed, a representative from one of the donor agencies supporting decentralisation claims that the current five-yearly district planning process, tellingly coordinated this time round by the Ministry of Finance rather than the Ministry of Local Government as is the norm, was conducted in considerable haste and constituted merely a desk-based exercise with no field visits or use of local plans of any sort.
Although the link between umuganda discussions and the district planning process appears weak or non-existent therefore, discussions do indeed take place following the completion of communal labour. The form of communication acts (instructing, questioning, proposing, advocating, criticising etc.) and the discourses employed reveal much about local political relations. A structured observation of a monthly umuganda session in Kigali on 23 February 2013 confirmed interviewees' assertions that the main purpose of umuganda meetings is to instruct residents on the latest reform plans and their roles within these. The Kigali meeting took place outside the local government authority's office following several hours of communal work clearing ditches. Seven local officials and approximately 300 community members attended. Over the 50 minutes of the meeting, officials spoke for 40. Their contributions centred on two issues – the announcement of new taxes (for refuse and school maintenance) and fines (for failing to attend or bring a hoe to monthly umuganda sessions). The meeting included interventions from 12 community members (10 men, 2 women), with six of these insisting on the right to speak as the chair tried to bring proceedings to a close. Nine of these community interventions were queries relating to the new taxes, two related to a local theft and the final one was a proposal (rejected as it is not in the District Development Plan) to use next month's umuganda work to repair the local bridge. Overall, the substance, tenor and tone of local authority contributions stifled rather than facilitated debate and revealed paternalistic attitudes and disciplinary intent as officials repeatedly stressed the benefits of government reforms and the responsibility of all to contribute, rebuking those who questioned the announcements. There was, however, a lack of clarity on the precise costs of the new taxes and clearly, on a case-by-case basis, room for manoeuvre. For example, the education tax was initially announced as being a monthly tax of RwFr5000 per household. In response to a number of queries for more clarity (given, as participants noted, that a once-off ‘education tax’ had been paid three months previously), it was then announced that it was to be a RwFr3000 annual payment. The amount appeared, therefore, somewhat arbitrary. Moreover, officials then added that anyone experiencing difficulties paying this should ‘go negotiate with’ their local authority leader. While this perhaps suggests some flexibility in local taxation rates, it also highlights a potential for abuse and clientelism. In response to these announcements, the substance and tenor of individual community contributions demonstrated considerable resistance to the escalating costs of these reforms. More broadly among those present, while announcements of new taxes invoked murmurs of dissent and disquiet, individual participants' queries in this regard were met with applause and good-natured laughter. A telling moment came when the chair reprimanded attendees for only applauding individuals' queries regarding the announced taxes, but not the officials' responses. His request for applause for local officials was met with a stony silence.
These patterns are replicated in accounts of umuganda provided by individual interviewees although a somewhat high number of interviewees claim to have spoken. Overall, 65% of all male interviewees and just under 50% of all females claim to have spoken at an umuganda meeting. They classify their interventions as being either complaints and/or questions regarding certain activities, projects or taxes announced by local leaders. In general, villagers characterise umuganda meetings as being the forum where local leaders communicate to them the latest government programmes and their role within these.
The scale and depth of the disciplined inclusion of communities into these programmes is demonstrated by local officials' frustration with what they term the ‘mindset’ or ‘mentality’ of the population who remain resistant to government programmes. As the following excerpts reveal, the aim is to inculcate new behavioural attributes through both instruction and direction as well as ongoing scrutiny and surveillance to ensure imposed activities are carried out.
When we are in development we need to change the people's mentalities … there is a population who thinks that everything should be carried out by the state. But they also have an important role which they have started to realise, which they did not know before. And so the process is to change their mentality to understand that they also can work, that they can also participate – there is a challenge within it of sensitisation. (District Vice-Mayor, Site D)
The first job we have is to change the comportment and the mentality of the population. To do this is a process.
We do lots of meetings with the population. Through these meetings we tell the population what [how] we wish things to be. We also do home visits to verify that what we have said is being put into practice. We follow what the population does. (Sector Executive Secretary, Site D – my emphasis)
To deal with the population is not an easy thing. It is obvious that there might be some challenges. There are some that have this mindset that is not easy to change.
People who are too old, an old man, or those who have never been to school. They are the ones that we have to keep telling them to make so much effort to make change. We are in a country that is in a hurry. So there is an urgency in implementing government programmes and it can take an old man more time to understand and get the point. (Cell Executive Secretary, Site C – my emphasis)
Overall, although clearly a distance from the broad, open dialogue which some interviewees in Kigali assert takes place, findings relating to umuganda do demonstrate that a space exists for residents to voice their views. Perhaps more importantly, they appear keen to exercise this voice – most particularly in relation to dissent and concern on the growing personal costs of ongoing reforms. There is little evidence to suggest, however, that these concerns are being heard and acted upon in broad policy terms – although, as noted above, individual representation to local officials to plead individual cases appears possible. Further evidence is provided in the next section to demonstrate this.
Community and local official priorities compared
A simple comparison of community and local officials' priorities wherein both parties (in interview) were asked to rank, in order of priority, the most important, pressing issues in their areas, reveals clear differences between both groups. For community members, the top three issues in order of priority – domestic conflict and violence; food security and land access; and access to health services (denied for those who fail to pay their annual health premium) – reflect the multidimensionality of poverty and its effects on psychological as well as physical wellbeing. For local officials, the top three issues were: the expansion and development of the government's ‘land consolidation’ programme (wherein individuals are instructed to pool their land to intensively cultivate specific crops as instructed by local agricultural extension agents); the collection of annual health insurance premiums from all households; and the development of local savings and credit cooperatives. These priorities reflect those of the national government (see GoR 2013). Indeed, local officials note that a failure to meet 80% of their targets (which, drawing from national priorities, include these three issues) means they lose their jobs, so clearly the pressures are immense. These differences between officials' and communities' priorities underscore the poor responsiveness and lack of effectiveness of the government's reforms in meeting local needs. Yet, as we have seen, communities are being forced to share the costs of these reforms. The extent of this cost-sharing and its consequences are explored in the following section.
Revenue redistribution or revenue generation: who pays?
While the percentage of devolved resources to districts is reported to have increased to approximately 33% of the national budget in recent years (MINALOC 2013, 19), district budgets remain complex and opaque and figures for actual (as opposed to budgeted) expenditures are not, unlike in other countries, publicly available. Interviews with district budget officers and other officials revealed considerable confusion regarding both the provenance (direct ministerial transfers vs block grants vs earmarked donor funds – these are all included together in budgets) and the destination (district, sector or cell – this information is not available) of devolved funds. A clear sectoral breakdown of expenditure (salaries, administration costs, education spending etc.) is also unavailable. In this context, it remains unclear how much national revenue is redistributed for local development. What is clear however is that households and communities contribute a substantial amount, both in the form of umuganda voluntary labour and in financial contributions. While umuganda works officially take place just once a month, in rural areas they are far more frequent. In the sites visited for this research, umuganda works were being organised two to three times a week. Many of these involved construction projects – schools, credit and savings facilities7 and offices for cell leaders8 were among the most common activities. Community participation in this context equals free labour for externally imposed programmes and projects and, as a sector leader explains below, it is the community and not the state who pays.
The participation of the population, that is when you can use them well. For [the] construction of the bridge [in the sector centre] for example … For example, the credit and saving building was built by the population. There were no funds from the government or from any organisations. We just needed to mobilise the population and it was built. (Sector Executive Secretary, Site A)
As well as contributing labour, households are also required to financially contribute to a range of projects and programmes including local building projects, local services and the public health insurance scheme – the mutuelle. These payments prove extremely difficult for families and households which are already struggling to meet their daily requirements and they are, as we have seen, a source of growing resentment and frustration. There are a number of reasons for this growing unease.
First, for many households it simply proves too difficult to manage to pay these obligatory contributions. One of the payments consistently highlighted – partly because of cost and partly because of worries over not being able to access health care – is the annual health insurance premium. As we have seen, with an increasing number of households being recategorised upward in the ubudehe scheme, the burden of payment for this scheme has shifted significantly away from the state and towards households. While some interviewees related stories of coercion by local leaders in efforts to bring in these payments in previous years (forcing households to sell livestock, for example), a ‘softer’ form of coercion involving loans appears to be the current strategy. Two interviewees explain how they have managed.
We have made associations [groupings of a number of households] and borrowed money from the bank to pay our mutuelle [health insurance]. All of us here have done this. (Interviewee, male, Site A)
For the health insurance it is mandatory. You must pay this. Last year they brought us books from the bank. The cell leader brought these. So whenever you get [FrRw] 500 you take this to the village chief. He takes the money and he signs that he has deposited this. So by the time of buying a new health insurance, they saw how much we had deposited. If we had not deposited enough, they [cell authorities] would lend us the money. So now some of us are paying back for this. (Interviewee, female, Site B)
For particularly vulnerable households (ubudehe categories 1 and 2), there are reports of VUP welfare payments being used to pay taxes as they are introduced – thus reabsorbing scarce redistributed resources back to the centre. A representative from one of the international agencies supporting one of the social welfare programmes outlines the problem.
I want to talk about the extent to which households have free choice in how they use those [welfare] resources in the current context. For example in our programme, when it was in full pelt, if you talked to people about what they spend their money on, a lot of it was spent on iron sheeting for their roofs [an obligatory purchase last year when thatch roofs were outlawed]. This is a very delicate area and it's a very tricky one but I think there is something there about the extent to which households have full control of the choices they make and about the resources they receive when local space is so controlled by the sector administration in terms of how a household behaves and the contributions they have to make. (Representative of international agency, Kigali)
Second, there is considerable lack of clarity around local budgets. In contrast to decentralised arrangements in neighbouring Burundi and the Democratic Republic of Congo, in Rwanda there is no legal provision for making the district budget public. Moreover, while district meetings themselves are technically public (Article 36, GoR 2006), public notice of these meetings is never provided and the public never attends. None of the 99 villagers interviewed for this research was aware of when district meetings are scheduled and none had ever attended such a meeting. Consequently, community members are not familiar with even the broad contours of local budgets, including sectoral allocations.
Third, it remains unclear to citizens how much of the collected taxes are actually used as they should be. Moreover, taxes can be negotiated on a one-to-one basis with local authorities. As one interviewee notes:
You have to pay for the community policing. Do a simple calculation. You pay 1000 per month. And this is an area of 800 households so it is already 800,000. And imagine, the guy earns maximum 15,000 per month. And they have 5 people. So that's 75,000. So what happens with the 725,000 FrRw? Nobody knows. And this is just one contribution. Then you have to pay here and you have to pay there. And it's always more and more and more. And you don't know what's happening with the money. (Representative of international agency, Kigali)
While President Kagame exercises a strict no-tolerance policy for corruption at national level, there are broad allegations of small-scale corruption or ‘manipulation’ at local levels, in part exacerbated by the reliance on voluntary labour from local leaders at the most local levels and in part by the clientelist nature of local leadership (interviews, Kigali, see also IRDP 2011).
And fourth, as we have seen in the account of the umuganda meeting in Kigali, people complain that the number and amount of these contributions are rapidly escalating. This is borne out by the government's own plans for local revenue generation. The revised Decentralisation Policy notes the low level of revenue generation at local level and is aiming to increase this significantly. ‘It is anticipated that with effective tax administration, within the first year of implementing the LG tax law, local government revenues will increase from the present RwF 16 billion to as much as RwF 90 billion’ (MINALOC 2013, 19). In addition, in the wake of aid cuts at the beginning of 2013, the government introduced a new tax to fund a new Rwanda Development Fund. Under this scheme, public servants must pay a month's salary per annum (or a twelfth of their salary). It has since been expanded for all salaried employees including those in poorly remunerated jobs. While this payment is still called ‘voluntary’, interviewees note that there is no such thing as a voluntary contribution.
It is called voluntary but it is not. Even X, he was told that he had to contribute and that he didn't have a choice. The umudugudu [village] chief came around with a list and said ‘this is what your neighbour paid.’ And he just had to do it. If you don't sign and pay, then you are in trouble. (Representative of international agency, Kigali)
Taken together, the findings indicate that while space is provided for local communities to voice their views, opinions and concerns, little attention is paid to these as local government officials, tied to centrally driven targets and programmes, continue their disciplining and soft coercion of local communities. With the costs of centrally driven reforms shifting to communities, the pressures on and disquiet among these communities are growing. Moreover, while some reforms certainly do address some community needs – for example agricultural production – there is a clear mismatch between the majority of priorities of local authorities and those of communities. As popular disquiet and unease with this situation grow, they raise questions in relation to the long-term impact of local reform politics on government legitimacy. These questions are pertinent to both the ‘good enough governance’ agenda and to the nature and dynamics of evolving developmental states.
Conclusion: When is ‘good enough governance’ good enough?
Reflecting on international donors' acceptance if not tacit approval of the strongly authoritarian nature of the Kigali regime, a number of commentators suggest that donors are more concerned with developmental outcomes and meeting the MDGs than with governance per se in the country (Beswick 2010; Hayman 2011; Zorbas 2011). This is corroborated in interviews with donors conducted for this research, where, although some concerns were expressed in relation to authoritarian governance practices at national level, it was broadly assumed that far more freedom and tolerance are allowed at local levels. It would appear that the international community in Kigali is implicitly, if not explicitly, influenced by the ‘good enough governance’ agenda. A problem here however is that it remains unclear what ‘good enough governance’ really means. Does it mean, as suggested by Booth and colleagues, ignoring local political dynamics and relations altogether? Or is there some middle ground in terms of a prioritisation of focus and resources that should be considered? If the former, while, as noted above, sympathetic to the argument that promoting liberal-democratic norms and institutions regardless of context is unhelpful and ineffective, the findings from this research resonate with the lessons of history in highlighting the dangers of a national elite focus alone which ignores the evolving dynamics of local state–societal relations.
Specifically, this research has demonstrated that the increasing centralisation of deliberation and decision-making on local development coupled with the growing pressures and demands on local communities to invest – physically and financially – in their own development as dictated by the ruling regime's fast-track developmental programme are leading to growing levels of popular disquiet and dissent. This necessarily raises questions regarding the level of embeddedness and legitimacy of the regime and therefore the sustainability of its development project. To return to the questions posed at the outset, the findings suggest that an authoritarian or ‘illiberal’ state can indeed be developmental, but that this comes at not inconsiderable cost to both its citizens and its own legitimacy. Rwanda's neo-developmentalist experiment is certainly ambitious. However, when considered in the broader context of the careful management and negation of historic and ongoing ethnic and centre–periphery animosities within Rwanda, it remains to be seen whether the country's governance culture will prove ‘good enough’ for its nine million citizens who are, in more ways than one, heavily invested in the country's future.