Caitlin Blaser 1 September 2007 Balancing Opportunity and Conflict: The Impact of a Refugee Influx on the Decentralisation Process Caitlin Blaser Submitted in partial fulfilment of an MA in Forced Migration Studies 1 September 2007 1 Caitlin Blaser 1 September 2007 Table of Contents 1 Introduction: Theoretical and Methodological Foundations 3 Defining Decentralisation 5 Decentralisation and Democracy 7 Empirical Additions to Decentralisation Theory 10 Forced Migration Theory 14 Research Design 19 2 Setting the Stage 25 Introduction to Loulouni 29 Establishment of the Refugee Camp 32 Historical Overview of Decentralisation in Mali 38 Contemporary Decentralisation in Mali 42 Actors 50 3 Decentralisation in Loulouni 55 Local Government Structures 57 Accountability and Transparency 61 Fiscal Elements 64 Service Provision 69 4 The Refugee Experience 73 Refugee Community Relationship 75 Access to Common Pool Resources and Services 76 Economic Patterns 83 Perception, Identity, and Trends 86 5 The Impact of Refugee Arrival on Decentralisation 91 Administrative Impact 92 Political Impact 97 Developmental Impact 102 6 Conclusion 104 Bibliography Annexes 2 Caitlin Blaser 1 September 2007 WHEREAS DECONCENTRATION SEEKS EFFICIENCY, THE MAJOR VALUES OF POLITICAL DECENTRALISATION INCLUDE PARTICIPATION, EFFECTIVENESS, AND RESPONSIVENESS?IN SHORT, DEMOCRACY. NKRUMAH, 2000 1. Introduction: Theoretical and Methodological Foundations Decentralisation is so central to Mali's development and democratisation strategy that the country's constitution explicitly recognises local governing bodies, and articulates a pro- decentralisation agenda (Republique du Mali, 1995). Mali is far from alone; seven other African nations enshrine local governing bodies in their constitutions,1 and virtually every country on the continent is employing decentralisation as a strategy to strengthen governance at all levels, and improve service provision (Oyugi, 2000). International organisations claim decentralisation is the centrepiece of current policy reforms in Africa (Ribot, 2002; UNCDF, 2000; World Bank, 2000). Despite widespread agreement about the importance of decentralisation, the debate on its promises and limits is far from over. Some claim that as powers are "redefined by the state through the process of decentralisation, contradictions, paradoxes, and unintended consequences increase, and struggles over meaning and power intensify (Benjaminsen, 2001 p15)." Others herald decentralisation as ?a silent revolution in public sector governance....to move decision making for local public services closer to the people (Shah and Thomson, 2004 p5).? Empirical data emerging from these ongoing decentralisation experiences shows that decentralisation is a useful tool for some, but not all development challenges it has been employed to address (Manor, 1997). Decentralisation's goals are generally categorised as administrative, political, and developmental, all of which will be defined in the following section, and applied to the Malian context in Chapter 2. The current decentralition debate currently centres on two issues. The first question is why decentralisation is successful at achieving some of the political and developmental impacts it should theoretically be able to deliver, but not others. The second question is why decentralisation is able to achieve some goals in certain situations, but not others. These are questions that are only answerable by emerging empirical evidence, which is refining the theory of decentralisation. It is both providing a more robust discussion of the potential and limitations of decentralisation, as well how contextual factors affect the implementation of the process. One contextual factor that is contentuous is migration; its political impacts are particularly important to 1 Ethiopia, Ghana, Namibia, Nigeria, Senegal, South Africa, and Uganda 3 Caitlin Blaser 1 September 2007 understand given the unique questions it raises about social inclusion and access to services that are highlighted when political decisions are made at the local level. These debates are often all the more applicable in the case of forced migration. This study contributes to this discussion adding the impact of refugees as an explanatory variable. This thesis will present evidence from Loulouni to argue that, based on this particular case, in places with a refugee presence it is easier for the political dimensions of decentralisation to be achieved. This includes increased interaction between community members and local government officials, and increased allocative efficiency of resources. The creation of the downwardly accountable mayor fostered communication between local government and the residents of Loulouni. However, a lack of resources and capacity in the office of the mayor threatened to undermine this relationship, as the mayor was unable to deliver on the commune's development goals. With the arrival of the refugeesand the resources that accompanied them, the demands of the residents of Loulouni changed from service provision to negotiating rights and access to services. Since the mayor was able to respond to these needs of the community, this relationship was strengthened. However, this thesis will also argue that a refugee presence makes it more difficult to achieve the administrative aspects of decentralisation. This includes certain aspects of implementing downward accountability structures, transparency, and participatory systems of decision making. The relief effort was accompanied by a lack of clarity of roles, and consequently aggravated existing power struggles at the local level. In terms of the developmental aspects of decentralisation, there is a balance between more efficient allocation of resources stemming from a more constructive relationship between citizens and local government officials, and a lack of clarity of mandate between local government and other service providers. Exploring the impact a refugee presence has on decentralisation is compelling for numerous reasons. First, in an attempt to understand decentralisation's successes and failures, there is a call in the literature to look at decentralisation in the context of global economic and political trends (Litvack, 1998; Work, 2002). The connection to state transformation that is nascent in decentralisation literature is already well established in the literature surrounding refugees. Forced migration is widely considered to be "a product of wider processes of social and economic change...normally referred to as 'globalisation'....It follows that forced migration....can provide a kind of window on these processes (Turton, 2003 p7-8)." It is equally compelling because decentralisation and forced migration are both central to understanding the process of state transformation in Africa. While Africans constitute only 12 per cent of the global population, around 28 per cent (i.e. 3.2 million) of the world?s 11.5 million refugees and just under 50 per cent (i.e. 9.5 million) of the world?s 20 million internally displaced persons are to be found in Africa.....Of the 20 top 4 Caitlin Blaser 1 September 2007 ?refugee-producing? countries around the world, nine are to be found in Africa. Twenty-five African states have refugee populations in excess of 10,000... (Crisp, 2000 p158). Mali is among these twenty-five states (UNHCR, 2003), and certainly not an isolated case where interaction must occur between the decentralisation process and a refugee presence. While this study on one hand responds to questions posed by decentralisation literature, there is also a need to bring decentralisation into refugee studies literature. There is growing evidence that refugees have more than a material affect on their host communities; they are also transforming the host country's governmental practices, and the expectations citizens have of their elected officials (Landau, 2003). As it becomes increasingly clear that refugees have an impact on the very foundations of democracy at a local level, it is a necessary step to bring decentralisation into the debate, since it is the central player in Africa's democratic reforms. This study focuses on the places where these literatures coincide, exploring how decentralisation and a refugee influx are mutually shaping the impacts of the other. To look at the impact a refugee presence has on decentralisation, this study takes the case of Loulouni, a rural commune capital in southern Mali implementing the national decentralisation initiative. Soon after fighting broke out in Cote d'Ivoire in 2001, the town also hosted a refugee camp. The refugee presence is both a symptom and a cause of the current process of state transformation in Loulouni and, as such, is simultaneously shaping and being shaped by the decentralisation process. It is apparent that the refugee presence is enabling decentralisation to accomplish some of its goals, while inhibiting it from accomplishing others. Understanding how the refugee presence is affecting decentralisation in Loulouni will accomplish the following three goals: 1. Build on the current understanding linking refugees to state transformation by contextualising it in a very specific way - the case of decentralisation. 2. Contribute to a technical understanding of how a specific factor, refugee presence, may shape decentralisation's successes and failures. 3. Link decentralisation to issues of state transformation through a case study in which refugees shape the local context. In the following sections, the theoretical and methodological bases of this study will be examined. The trends in both decentralisation and forced migration studies literature will be presented as they apply to each other, to bring to the forefront places of intersection in the literatures. Comments are also made about the structure and methodology of this study. Chapter 2 looks at how decentralisation is being practiced in Mali today. It first presents a brief overview of the history of 5 Caitlin Blaser 1 September 2007 political and administrative decentralisation movements in Mali, and then a snapshot of the current decentralisation effort's accomplishments, motivation, articulation, and challenges. This is done first on a generalised national level, and then Loulouni is introduced, along with how the various aspects of decentralisation are playing out in one specific context. Chapter 3 discusses the refugee dimension of Loulouni, with an overview of migration trends between Mali and Cote d'Ivoire, the establishment of the camp in Loulouni, and then the perceptions and relationships between the host community of Loulouni and the refugee population. Chapter 4 interrogates the impact this refugee influx has had on the decentralisation process in Loulouni. It looks at the impact in terms of the three generalisable aspects of decentralisation: political, administrative, and developmental. The conclusion places these findings in a broader political context, exploring the extent to which lessons from Loulouni can be applied to different contexts, as well as addressing where the limitations of this study demand further research. 6 Caitlin Blaser 1 September 2007 Defining Decentralisation Defining decentralisation for this study is only necessary as far as it affects the implementation and measurements of success of decentralisation in Mali. As such, it is necessary to understand two aspects of decentralisation. This section will explore the political rationale for decentralisation, which is ultimately what determines not just the process of implementation, but also the indicators of success of the decentralsiation process. The following section will frame decentralisation historically, since it is the history of decentralisation in a global economic context that has shaped the successes and failures of implementing decentralisation in Loulouni. One challenge to studying decentralisation is that it has been used to serve such a wide variety of objectives. Some claim it is a response to economic crises (Therkildsen, 2001), others look at it as a strategy central governments employ to offload burdens of service delivery and administrative capacity (Litvack, 1998). It is commonly described as a knee-jerk reaction to the failure of large centralised governments to promote development in many countries (Wunsch and Olowu, 1995; Faguet, 1997). In some places it has been the outcome of populist political movements (Olowu, 2001), while perhaps more frequently it is more the outcome of external pressures, either from donors, or as a condition in structural adjustment packages (World Bank, 2000; Therkildsen, 2001). For still others, it is framed as a response to politicised identity, or a pandering to local elites who have various problematic relationships with the central government (Crook and Sverrisson, 2001). In Mali, a small part of all of these views applies, and each one impacts differently on the specific successes and failures of the process in Loulouni. Only since the mid 1990s has the discussions about decentralisation come from a history of implementation, comparison, and empirical evidence. Before beginning an analysis of the debate on decentralisation, it is important to define the process. The decentralisation literature has developed a large vocabulary to differentiate between different processes and structures involved in decentralisation, but since this is not a comparative study and only one decentralisation process is being examined, it is only necessary to establish working definitions of applicable concepts in order to proceed. One of the best and most widely accepted definitions of decentralisation was developed by Rodenelli, et al. (1984) as: the transfer of responsibility for planning, management, and resource- raising and allocation from the central government to (a) field units of central government ministries or agencies; (b) subordinate units or levels of government; (c) semi-autonomous public authorities or corporations; (d) area-wide regional or functional authorities; or (e) NGOs/PVOs. Private firms are often added to (e), and while there is a lively debate about the acceptability of this, 7 Caitlin Blaser 1 September 2007 the limited role of the private sector in decentralisation in Mali does not make it terribly important to take a stance in this case. Bennet (1990) makes a useful distinction between intergovernmental decentralisation, which is the transfer of authority and resources downward among different actors within the public sector, and market-based decentralisation, which transfers these powers from governments to NGOs or private sector players. These two processes are happening in tandem in Mali, as will be explored in more detail in the following chapter. While the above definition is theoretically informative, it is not adequate for more empirical studies of decentralisation. Since there is rarely a higher level than the central government, nearly every restructuring undertaken by the central government could then be described as decentralisation. Further articulations of decentralisation is commonly used by practitioners, particularly in fields of natural resource management, distinguish between different decentralisation processes based on how powers are being reallocated. This study draws heavily from Ribot's (2002) framing of decentralisation, simply because for the concepts relevant to the processes in Mali, his terms proved to be the most mutually exclusive and collectively exhaustive. He distinguishes between the following four kinds of decentralisation: political, administrative, fiscal, and delegation. All four are important for understanding decentralisation in Mali, but the first two are particularly key. Political decentralisation is when powers and resources are transferred from the central government to downwardly accountable, lower level political institutions. (Manor, 1999; Crook and Manor, 1998; Agrawal and Ribot, 1999; Ribot, 2002) Because of the element of accountability to a local population, this form of decentralisation is also often called democratic decentralisation. Most contemporary advocates of decentralisation argue that this is the 'strongest' form of decentralisation, and offers democratic advantages such as greater participation, and through greater participation more effective service delivery and maximised use of public resources. Administrative decentralisation2 is often contrasted with political decentralisation. This is the transfer of powers and resources to localised braches of the central state, whether on a territorial basis (from a national to state or district level), or through technical line ministries. In this form of decentralisation, while power does move from the centre out, those exercising it are still accountable to the central government, and as such it is considered a weaker form of decentralisation than political decentralisation. Its advantage is articulated in terms of efficiency, rather than effectiveness. Fiscal decentralisation is often described as a separate process from political or administrative decentralisation, although is usually a feature of both of those decentralisation processes, and never stands alone. Fiscal decentralisation is the transfer of fiscal resources and revenue generating 2 Also commonly called deconcentration in decentralisation literature 8 Caitlin Blaser 1 September 2007 powers from the central government to local authorities, whether they are upwardly or downwardly accountable. While fiscal decentralisation cross-cuts political and administrative decentralisation to such an extent that it is not necessarily a distinct process in itself, the practical mismatch between political/administrative decentralisation and their fiscal aspects in Mali make it a useful concept. Delegation is when functions that are the responsibility of the central government are transferred to lower levels of government, public corporations or institutions, technical line ministries, or authorities outside the government's political-administrative structure to implement activities on behalf of the central government (Ostrom et al., 1993; Ribot, 2002). Because the central government generally maintains decision making power when delegation is taking place, it is considered a relatively weak form of decentralisation. Interestingly, this form of decentralisation is not an inherent feature of Mali's decentralisation process, but was called upon in the refugee relief effort. These kinds of decentralisation rarely take place in isolation of one another, and any decentralisation process generally has political, administrative, and fiscal elements. The current decentralisation process in Mali is primarily political in nature, but for a number of reasons that will be explored in the following chapter, the administrative elements have been more easily adopted than the political elements. Additionally, there are challenges to the fiscal aspects of decentralisation that have implications for the rest of the process. Academics are split about the relative importance of each aspect of the decentralisation process. Prud'homme (2001) places a great deal of importance on the administrative aspects of decentralisation, arguing that the structures established for formation and functioning of government are central to the shape the political aspects of the decentralised society will take. Ribot (2002) argues that administrative decentralisation is merely a reflection of the objectives of decentralisation, and that to learn about the true functioning of decentralisation, it is necessary to look at the political aspects. For example, whether the electorate's values are truly reflected in the actions of local government officials. While many authors sideline fiscal decentralisation as simply a supporting element of political or administrative decentralisation, others (de Mello, 2004; Ebel and Yilmaz, 2001) argue that it is often the most crucial determinant in the other aspects of decentralisation. The claim that fiscal powers offer a more accurate reflection of where power truly lies in a decentralised structure than the voting or vetoing powers emphasised by focussing on the political or administrative aspects respectively. Since this paper aims at targeting one influence on a specific decentralising system, it takes a more descriptive than prescriptive approach. The result is that the relative importance of each aspect is only determined by the impact it has on governance in Loulouni. 9 Caitlin Blaser 1 September 2007 Decentralisation and Democracy Nothing about democracy inherently implies decentralisation; in fact, decentralisation in Mali was also used as a development tool of socialist regimes. However, the current decentralisation process in Loulouni is integrated into the process of democratisation in the country. In fact, the form of decentralisation taking place in Mali is sometimes referred to in the literature as democratic decentralisation, due to the shift from upward to downward accountability at the local level. Because of this, certain democratic values become indicators for the success of the decentralisation process in Loulouni, which will be discussed in more detail in future sections. More generally than the case of Loulouni, there are a number of reasons, both theoretical and empirical, to tie the two processes. The origins of decentralisation stem from fundamental arguments about democracy and the relationship between citizens and the state. Much of the classical cannon of 17th and 18th century political philosophy advocated decentralisation on the basis that the liberty of free men could best be preserved through small democratic units, which should share fundamental values and collective objectives. Madison argues in the Federalist Papers that tyranny can be prevented by balancing powers between central, regional, and local governments. Tocqueville echoes his sentiments, saying Decentralisation has not only an administrative value, but also a civic dimension, since it increases the opportunities for citizens to take interest in public affairs; it makes them get accustomed to using freedom. And from the accumulation of these local, active, persnickety freedoms, is born the most efficient counterweight against the claims of the central government, even if it were supported by an impersonal, collective will. It may be surprising how little the argument has evolved in two centuries. Wolman (in Bennet, 1990) presents a very similar modern case for decentralisation. Decentralisation, by placing government closer to the people, fosters greater responsiveness of policy-makers to the will of the citizenry and, it is argued, results in a closer congruence between public preferences and public policy....because decentralisation permits these decision-makers to be held directly accountable to the local citizenry through local elections. While there are no intuitive problems with these claims, the argument is a priori - as are many on the other side of the decentralisation debate. Both sides of the debate have a vast reservoir of real world anecdotes and examples to support their points, but there is relatively little systematic or rigorous analysis in the decentralisation discussion. Additionally, the academic literature has concentrated on what should happen under decentralisation, from a political theory point of view. With the recent emergence of empirical evidence indicating that the reality is not always exactly as 10 Caitlin Blaser 1 September 2007 the theory would suggest, there is little in the literature that could be used to address why this is the case. This study tries to compensate for this weakness in the decentralisation literature both by taking a historiographical approach, and by juxtaposing decentralisation literature with forced migration literature, since forced migration literature has a history of grappling with the irrelevance of theoretical expectations that have little empirical justification (Misago, 2005). Decentralisation in Africa particularly is an old phenomenon, having gone through many waves - at least three identifiable throughout francophone West Africa, and several more limited to specific locales. During colonialism, discourse around decentralisation was centred around penetrating and managing rural populations and resources efficiently. Of course, this led to the establishment of upwardly accountable local officials, and a deliberately non-democratic system often hiding behind 'tradition.' Post-independence articulations of decentralisation focussed on national cohesion and the engineered development of local populations. This was often a thinly-veiled method for corrupt central governments to expropriate rural resources. The current wave of decentralisation, however, is a discourse focused on participatory democratic processes and rights-based approaches to development. This new discourse does not mean the success of decentralisation is inevitable, but it also cannot be dismissed as repeating the errors of the past. The arguments for and against decentralisation have been reiterated many times, each time altered slightly to reflect the economic or political context of the day. The arguments themselves, however, have remained remarkably unchanged. Simplifying for the sake of brevity, advocates of decentralisation generally claim that it is an avenue for the disempowered to take part in the political process, since it brings the institutions that deliver public services closer to the recipients. This proximity also makes the accountability of public officials more likely. Furthermore, it localises unwieldy state apparati, making service delivery more effective.3 This is often inherently linked to democracy, as the democratic system that helps local governments respond to the needs of their constituents and provide a means for these constituents to hold officials accountable is the foundation underpinning these benefits (Smoke, 2000). Many critics of decentralisation argue that the goal of participation is undermined by the fact that decentralisation effectively gives power and legitimacy to power structures already in place on local levels which are often unequal and exclusionary, having given rise to the marginalisation of certain sectors of the population to begin with. Central governments, with superior human resources and less to gain from oppression (in theory), are thought to be better placed to protect minority rights. 3 While proponents of decentralisation do not always ? although do occasionally ? claim that it is a more economically efficient way of delivering services in aggregate, it is very often claimed to be more effective in that the available resources are used to more accurately address the desires of local populations. 11 Caitlin Blaser 1 September 2007 They also argue that efficiency of service delivery is undermined as decentralisation often simply adds another level of bureaucracy without the necessary oversight and coordinating bodies. A second, rapidly growing branch of critics of decentralisation let by Oyugi (2000), Olowu (2001), Ribot (2002), Shah (2004) and others, argue not that there is a structural or theoretical problem with decentralisation itself, but that the way it is currently being implemented in the developing world is flawed. Oyugi (2000, 10) goes so far as to argue that ?the legal-political design of local government in Africa tends to weaken the cultivation of a democratic culture at the local level as well as weaken the ability of local authorities to take initiative in the field of service provision.? As the arguments about the implementation of decentralisation tend to be situation-specific, they will be discussed in more detail in the next chapter, which will speak specifically to the articulations and implementation of decentralisation in Mali. 12 Caitlin Blaser 1 September 2007 Empirical additions to decentralisation theory Recent empirical studies are finally adding some substance to the debate on decentralisation. With so many conflicting theoretical aims, and so many divergent examples of decentralisation's practice, many of the outstanding debates in the literature can only be taken forward with the emergence of empirical evidence. While in rhetoric decentralisation is still often heralded as either the harbinger or destroyer of democratic institutions, a consensus is emerging among theorists and practitioners that it is a useful tool to achieving some goals while blunt in other arenas. This study supports such an assertion in several concrete ways. Empirical evidence gathered in Loulouni indicates that decentralisation has been quite successful in accomplishing its political goals, but less successful in accomplishing its adminsitrative goals. However, by looking at the way the refugee influx shaped this outcome, other empirical trends emerge. For example, it is clear that in Loulouni, a lack of resources at the local level threatened to undermine the successes of decentralisation before the arrival of the refugees. Manor, a leading decentralisation theorist, has led several large comparative studies of decentralisation often connected to the World Bank, and is a reserved supporter of decentralisation. He argues in ?Promises and Limits of Decentralisation? that decentralisation, while generally a positive step in transforming governance, has often suffered from misplaced expectations and therefore misapplications. Several of the expectations that Manor suggests are misplaced are true of the decentralisation process in Mali, and understanding the roots of these expectations is essential to defining how to measure the success of this political transformation process. First, Manor argues that some conditions are crucial for the success of decentralisation, while others are quite helpful. Topping the list as the most essential element is the existence of reliable accountability mechanisms, both to ensure that elected politicians are held accountable to citizens, and that bureaucrats are accountable to elected officials. This is problematic in cases including Mali where a central goal of decentralisation is the creation of these very same accountability mechanisms. Other crucial conditions are sufficient financial resources, and administrative capacity. Arguably, Mali has neither. However, a great deal of the effort going into the decentralisation process in Mali has been directed towards compensating for the shortcomings in these two areas. Manor further cites as helpful a history of democracy before the decentralisation process, and social capital. He emphasises the fact that there is no formula for whether or not decentralisation will work. While the above factors can provide some guidance, a complex combination of history, geography, sociology, economic structures, international support, and internal political dynamics all interplay to 13 Caitlin Blaser 1 September 2007 determine how decentralisation will play out. Based on the list of factors Manor lays out, Bolivia in the early 1990s would not have been a good candidate, and yet it is heralded as one of the most successful decentralisation processes in the developing world. One anthropologist argues, ?Bolivia is the poorest, most backward country in South America. It has dozens of spoken languages, a ruinous geography, and almost no infrastructure. If we can make decentralisation work here, it can work anywhere? (Armando Godinez, quoted in Faguet, 1997, p136). Faguet and other analysis (Alfonso, 1999) give much of the credit for Bolivia's successful implementation of decentralisation to the ability of Bolivia's civil society at a grassroots level to hold political leaders to account. An effective system of accountability tops Manor's list of necessary conditions for decentralisation. While Bolivia had few other characteristics to recommend it for decentralisation, effective accountability mechanisms alone may have compensated for Bolivia's lack of economic development and democratic history. Similarly, many countries with more attributes of a ?good candidate? have experienced far less success in decentralising. Manor's laundry list just begins the conversation about the specific considerations for any country embarking on decentralisation. Given favourable conditions for a successful decentralisation process, Manor claims that it is then often able to promote collective action through association building, enhance the responsiveness of government institutions to the needs of the community, and make government function more transparently and accountably. These are all central goals of decentralisation in Mali. However, Manor goes on to argue that even if successfully implemented, decentralisation is not well equipped to alleviate poverty that is characterised by local inequality, reducing overall government expenditure, mobilising local resources, and providing services delegated by the central government, some of which are also expectations of decentralisation in Mali. Looking at resource mobilisation in some detail is useful in uncovering some of the political dynamics that often accompany the implementation of decentralisation. The frequent inability of local governments to successfully mobilise local resources is an oft-cited frustration of central governments with decentralisation efforts, and many theorists and practitioners say it indicates the lack of capacity of local governing officials. Fageut (1997), in an extensive study of decentralisation in Bolivia, strongly argues against those who criticise decentralisation on the basis of a lack of local capacity. By analysing quantitative data on changes in government expenditure, Fageut found that local government patterns in both tax collection and spending do change substantially with decentralisation. Furthermore, he found that these changes were better reflecting the desires of the community. This directly contradicts claims that local government is too poor, too ignorant, or too prone to interest-group capture to operate efficiently, necessitating the guiding hand of national government which is technocratic, capable, 14 Caitlin Blaser 1 September 2007 and generally knows what to do. Here we begin to see evidence of the opposite: local government has a deep understanding of its task, and has the capability and the incentive structure to produce the public outputs that people want. Manor and Ribot also reframe the criticism of local government capacity as one of incentive structures. First, local governments are often actually given disincentives to collect taxes. This can be for numerous reasons, whether central governments hesitate to devolve important tax-raising powers to local officials, economic structures make the administrative costs of collecting taxes nearly prohibitive, there is strong political undesirability to raise taxes, etc. Secondly, in evaluating the performance of local governments, standards developed at the central government are still used. For example, if the promotion of judicial authority is a priority of the central government, but the constituents of one district are instead concerned with access to basic services, when the elected official acts in line with the interests of his constituency, he will likely be accused of either lacking the capacity or will to strengthen judicial authority in his district. Finally, neither Manor nor Faguet discount claims that human resources of local governments may be lacking, or that capacity is a genuine problem. Rather, they argue that it is not often a limiting factor, and can be managed through effective planning. Other studies also point out that decentralisation is often characterised by poorly defined administrative boundaries (Smoke, 2000; Oyugi, 2000; Crook and Manor, 1998). Some excuse this lack of clarity as likely to be a feature of any large-scale political transition. Most, however, grant that decentralisation uniquely lends itself to certain administrative ambiguities, although not always due to any undesirable features it may have as a political system. Decentralisation involves more levels to which powers must be delegated and clarified, for example, and the inclusive ideal of democratic decentralisation mean that more role players with more agendas become a part of the political process. The consequences of this can include accusations of a lack of capacity (because it is not clear who is supposed to be doing what), or slow implementation that is frustrated by power struggles. Similarly, a lack of capacity can be caused by the power struggles themselves, as they then prompt a reluctance to coordinate and share resources. This is a central issue in this study as decentralisation has been effectively implemented in Mali before ongoing legislation establishing the rights and responsibilities of political role-players has been solidified. While the literature on decentralisation is vast and somewhat diffuse, it is grappling with a wide variety of political changes that have taken place in a wide variety of social, economic, and historic contexts. Political theory literature is useful in defining the decentralisation debate. Embedded in this literature are the fundamental questions of governance to which decentralisation hopes to provide answers, as well as certain critiques to using decentralisation as the response to these challenges. 15 Caitlin Blaser 1 September 2007 The nascent empirical decentralisation literature adds substance to these theoretical debates by providing evidence about where decentralisation has proven effective, and where the criticism has proved to be correct. However, evidence is still emerging, and the body of literature is not yet large enough to provide definitive answers to many of the decentralisation debates. Forced migration studies literature is in a position to contribute to both the theoretical and empirical decentralisation literature. As discussed, much of the contemporary decentralisation theory deals with implementing decentralisation in a specific socioeconomic context. This process of contextualising theoretical ideals is one that forced migration literature has grappled with from its founding. Since the legal definition of a refugee emerged with the 1950 UN Convention, there has been a discourse of international standards that has often been in sharp contrast to the reality on the ground. Bringing these two divergent conversations closer together has often been an explicit role of forced migration studies. Additionally, forced migration literature, while perhaps not the most intuitive source for answers to fundamental political questions such as the relationship between citizen and state, in fact represents an excellent departure for such inquiries, since it is these 'grey areas' that must be drawn on to challenge assumptions, or precisely delineate the limits of claims. Furthermore, as an interdisciplinary area of study, forced migration has almost by definition looked at the same issue through a number of different lenses. Theory in forced migration has often evolved because one specific case is looked at from anthopological, sociological, political, economic, environmental, and other points of view. As such, the literature provides solid critiques for decentralisation theory, since it is a process that has social, economic, and political dimensions, but which is theoretically nested in one particular field. Looking purely at the political science literature, the empirical decentralisation literature, while rapidly growing, lacks texture due both to the limited number of cases studied, and the limited number of 'lenses' used in understanding its impacts. Very practically, linking this to forced migration literature first builds on it by contributing to a technical understanding of how a refugee presence shapes decentralisation, which may also suggest inquiries for related contexts, such as situations of labour migration, or areas of limited state control. However, forced migration literature also provides a bridge from empirical decentralisation studies to broader issues of state transformation. Currently, there is a gap between the theoretical discussions of decentralisation, which discuss, often very broadly, the kind of relationship between citizens and state institutions such a political system can foster, and the practical discussions of decentralisation, which often take the form of 'lessons learned' from specific cases where decentralisation has been implemented. Theoretical contextualisations of practical cases, or a search to link the results to the theory is much needed yet largely absent from decentralisation literature. This process of contextualisation and linkage is currently mushrooming in 16 Caitlin Blaser 1 September 2007 forced migration, as the field asserts itself as more theoretically substantive than its rather recent beginnings on the periphery of the social sciences. 17 Caitlin Blaser 1 September 2007 Forced Migration Literature Forced migration literature was born from refugee studies, which initially focused nearly exclusively on the plight of refugees themselves, framing them as victims of the international system. The field began expanding when a group of scholars (spurred in part by Chamber?s ?Hidden Losers? article in 1993) began exploring the impact refugees have on their host communities. By now, a growing body of literature is developing on the costs and benefits refugees impose on their host communities. Looking at the impact refugees have on their environment, rather than the way their environment affects them, demonstrates a change from framing refugees as victims to agents of an international system, affected by and simultaneously shaping this system. This is particularly apparent in this study, since refugees in Loulouni were both affected by the decentralisation process in terms of their access to public services and economic opportunities, and at the same time affecting the decentralisation process through the changes in the local political and economic landscapes their arrival prompted. Whitaker (2002) looks at the impact a large number of refugees from neighbouring countries had on Tanzania in the late 1990s. She sums up that ?refugees generally impose a burden on local infrastructure, environment, and resources, but they also provide cheap labour, expand consumer markets, and justify increased foreign aid.? She also argues, however, that the aggregate impact of refugees on the host community as a whole is not relevant. She then goes on to disaggregate the impact these changes are having on different groups in the host community. For example, women may be disproportionately affected by deforestation, as firewood collection is their responsibility. She joins Chambers (1993) and Waters (1999) in the disaggregation of hosts using the economic patterns of both refugees and hosts to argue that some parts of the population are benefiting from the influx, while others are being marginalised. The stance of all of these authors is essentially that a refugee influx exacerbates certain inequalities through the economic characteristics shared by refugee situations, such as an increase in cheap labour and increased pressure on common pool resources the poor are being further marginalised, while the increased labour pool and markets, and a select few high paying jobs with the relief effort benefit the wealthy. Landau (2003) uses a comparative study to question the causal relationship assumed between economic changes and the refugee influx. He argues that while the influx he studied did in fact change economic patterns, these changes were not transformatory, and moreover did not represent a shift towards integration into the cash economy. Rather, they reflect the continued application by both refugees and hosts of diverse and creative livelihood strategies. Profound economic changes, 18 Caitlin Blaser 1 September 2007 he argues, are more the result of macro-economic policy than the refugee influx. Moving away from the immediate questions about who was losing and who was gaining from a refugee presence, several authors are now looking towards longer term, fundamental transformations refugee impacts could have on host populations (Waters 1999, Jacobsen 2002, Whitaker 2002 and Landau 2003 and 2004). Waters (1999) begins this discussion by speculating that these economic patterns are creating fundamental social and political changes. Whitaker (2002) has taken short term material indicators and, by putting them in the context of economic liberalisation, drawn conclusions about the possible longer term material and social affects they may have. For example, by looking at the development projects of international organisations in refugee hosting communities in the context of declining government budgets for service provision, Whitaker predicts that the added expenditure on services despite the lack of government funds prevented frustration with the government. Jacobsen (2002) advances this line of inquiry, arguing that refugee resources could be tapped into the help with the state building process. In her argument, however, she does not move beyond the economic potential refugees come with, meaning the discussion on how a refugee influx could best be tied to state building could be taken further. Whitaker (2002) suggests the refugee presence in Tanzania seems to increase in ?mwanko? in the host population ('awareness,' largely in the socio-political sense). This is a key intersection of decentralisation and forced migration literature. Landau (2004) builds on this speculation about political transformation, finding that the refugee influx is in fact fundamentally changing the relationship between citizens and the state in Tanzania. Bluntly stated, Kasulu4 residents have affirmed their normative affiliations to a Tanzanian nation, population, territory, and Government, while material and functional expectations for the national administration have declined. Supra-domestic actors' insinuation into logics of cause and responsibility - what I term the transnationalisation of governmental practice - goes a long way in explaining this counter-intuitive outcome. (p37) This finding is crucial when looking at decentralisation - a specific transformation of government institutional form and practice. The affirmation of normative affiliations Landau discusses is closely connected to the national unity articulation of decentralisation. Proponents of decentralisation argue that this is accomplished as decentralisation creates downwardly accountable structures that citizens are then more willing to invest in. However, the impact of the refugee influx as described by Landau simultaneously also creates a threat to accountability structures. The 'transnationalisation of government practice' is essentially an introduction of resources used for things that were traditionally the responsibility of governments. Accountability is threatened not only because there is a significant 4 a refugee hosting district in Tanzania 19 Caitlin Blaser 1 September 2007 injection of resources that are not a part of downwardly accountable, local government structures, but also because with the decline in expectations from government that accompanies these resources, the population will be less likely to hold the government accountable as expectations are lowered. While this could be seen as desirable confidence building in the short term, its long term dangers are clear. A common starting point when looking at host country political processes is that refugees serve as an ?other,? or opposing social force that will create solidarity in the community. ?The significance of outsides threats in rousing national allegiance is...almost too much of a truism to deserve to deserve further comment? (Landau, 2002 p20) and many forced migration theorists positing that the refugee presence provides just such an ?other? identity to rally against, this presence ought to help decentralisation achieve its goals. While the idea that refugees are a distinct group often in opposition to the local population is a dominant one in forced migration literature, it has not gone unchallenged. Bakewell (2000) argued that it was necessary to put down the lens of the emergency, which portrayed the world starkly in terms of refugees and hosts, and use a more fuzzy lens which allowed boundaries to be blurred between refugees and hosts, or emergency and normality. (p2) When this was done, he claims that the discourse of aid agencies that defines practice is quite inapplicable to the local situation. This study has found that in the case of Loulouni, the assumption that ?othering? is a characteristic that can be assumed of all refugee situations is off base. While an extensive discussion of identity is beyond the scope of this study, one possible explanation for the different perspective is methodological, as this study has adopted Bakewell's grounding in the local population rather than the relief effort. Another possible explanation is historical; the construction of ethnic and nationalist identities in different areas, as well as the dynamics of the conflict and refugee movement itself (whether refugees are self settled or in camps, for example) certainly affects the articulations and power dynamics of a refugee/host relationship. It is important, however, not to push forward a framework, no matter how applicable in some situations, that could prove an undesirable self-fulfilling prophesy in others. Regardless of its merits, the ?fuzzy lens? adopted by Bakewell, while arguably more accurate in the case of Loulouni, is another place of clear intersection with decentralisation processes, which already involve a lack of clarity in mandate. These are not the only goals of decentralisation that could be affected by the transformative influence of refugees. By bringing the decentralisation literature and forced migration literature together, this study initially hypothesised that the social goals of decentralisation such as increased collective action and increased interaction with government officials would be helped by the refugee 20 Caitlin Blaser 1 September 2007 presence, because it would introduce an ?other? around which to mobilise, building and consolidating a sense of community. Political goals, however, such as increased accountability, resource mobilisation, and responsiveness to the needs of the community would be harmed by the introduction of resources outside the control of the government. Upon analysing the data, however, this was found not to be the case in Loulouni. This points to certain important gaps in the decentralisation literature. It is no less true that perceptions and expectations of the government are shifted by the refugee presence. While it will be discussed in more detail in later chapters, the refugee influx in Loulouni did not bring in an ?other? around which all citizens could mobilise. Instead, refugees integrated very well with the majority of the population, causing tension only with the wealthy elite with whom the Ivorian refugees competed for certain kinds of employment and resources. While this can, to an extent, be attributed to the history of migration and class expression in Mali and Cote d'Ivoire, it is also reflecting a tension between the mayor and sous prefet, local political authorities who use the refugee influx as a battlefield for their own power struggles. Regardless of other dynamics finding expression in the arrival of the relief effort, the fact that the integration of Ivorian refugees took place largely along class lines has implications for the implementation of decentralisation Specifically, the wealthy are more likely than other sectors of the population to turn to governmental authorities with their concerns. With responsiveness to the community being central to the mandate of the newly created political authorities in Loulouni, the elite bias already acknowledged as part of the decentralisation process is increased by the arrival of refugees. This study has found that when the refugees arrived, accompanied by relief resources and services, access to, rather than the establishment of infrastructure and services became the priority of those interacting with local government. This means that the host population effectively changed demands of local government officials from those related to service provision to those related to service management and conflict resolution. The evidence for and implications of this are discussed in Chapter 4. These empirical findings have implications for both decentralisation literature and forced migration literature. Most significantly, the intersection of these two literatures currently suggests that the injection of resources by the humanitarian intervention, often for the provision of public services and yet outside the control of state officials, should have the most significant impact on decentralisation. This study found that this is not the case. First, the state has never had the capacity to be responsible for most service provision in practice, so the arrival of the humanitarian 21 Caitlin Blaser 1 September 2007 intervention does not mark a change from the norm in that respect, nor did it have a significant transformative affect on decentralisation. It was actually the control state officials did have over refugee influx related resources that shaped the administrative decentralisation process. Administrative roles at a local level are already poorly defined under decentralisation. Tensions existed, but for the most part, a working balance had been established before the arrival of the refugees. The refugee influx introduced new resources, and therefore the power to control them, into the local administrative system. Because these powers were not clearly defined, as they are not almost by definition in emergency situations, significant barriers to successful administrative decentralisation were created. The support the refugee influx offered to the political decentralisation process, however, has many lessons for the empirical literature on decentralisation. Greater collaboration and community involvement in the political process was an outcome of the refugee influx. In the case of Loulouni, this was not, as the literature suggested, because a collective enemy was brought in uniting the community. Rather, it was caused by the very same poorly defined powers and access to resources that posed a barrier to the administrative decentralisation process. This ambiguity changed community demands on the local government from those of service provision to conflict resolution. The local government was largely unable to fulfil requests for service provision due to lack of resources. Resolving conflicts is something local government was much better equipped to do. Because of its success responding to these new demands relative to its failures in its former service delivery role, relationships with the community were strengthened, and participation in local governance increased. 22 Caitlin Blaser 1 September 2007 Research Design As stated in the introduction, the objectives of this study are threefold. The first is to contribute to the current literature linking forced migration to state transformation in the specific case of decentralisation. The second is to add to the growing empirical decentralisation literature by understanding how one specific factor ? a refugee influx ? can shape the decentralisation process. The third objective is to use an understanding of the forced migration context to link decentralisation to broader issues of state transformation. Developing a methodology that could serve all of these objectives was particularly challenging since they include both normative and empirical questions, demanding both enough quantitative data to provide a reliable description of both decentralisation and refugee influx-related issues, and enough qualitative data to contextualise these phenomena accurately and suggest causal linkages in the correlations. While these challenges are certainly formidable, this study had the advantages of a great deal of time and access. As such, it was possible to take an inductive approach, using multiple stages of inquiry that built on each other to put together as complete a picture as possible. Answering the research question of this study required identifying the impact the refugee presence had on decentralisation. The first thing this requires is an assessment of the state of the decentralisation process. This was done first by examining Malian legislation on decentralisation, and then drawing heavily on the literature introduced earlier in this chapter to identify the goals of decentralisation in Mali, and the indicators of their success (appendix 4). First, information about the status of these goals both before and after the refugee influx was needed. Then, the impact of the humanitarian influx needed to be measured by comparing the two, and then including concrete examples of any way refugees may have impacted on the achievement on the goals. One further tool was used to measure the impact of the humanitarian influx. Geographic, historic, economic, and political information about Loulouni and Mali more generally was gathered, and brought to the decentralisation literature, which suggests a likely path for the decentralisation process given the generic information. After the empirical data on the decentralisation process was gathered, this picture was compared back to what the literature indicated. Places of divergence were flagged as possibly impacted by the influx, to be explored in greater detail. Ideally such a question would be answered by a comparative study of Loulouni, and a Loulouni existing in a parallel universe that experienced no refugee influx (King, et. Al 1994). Finding an alternative methodological approach to answer the research question was preferable to looking for such a parallel universe. Comparative studies with a location of similar geographic/ economic/ 23 Caitlin Blaser 1 September 2007 demographic characteristics often provide such an alternative, but this was inappropriate for this study two reasons. The first was simply logistical ? no appropriate control town existed within a reasonable distance from Loulouni, and arranging for a similar study further afield was impractical. The second problem, however, eliminated a comparative study from consideration. Because of the nature of decentralisation, ?micro?-level studies of decentralisation processes such as this simply do not lend themselves to comparative analysis. Local political dynamics are by definition shaped by individuals, local trends and events. Furthermore, because of the geographic, cultural, etc., diversity in Mali, any reasonably comparable town to Loulouni would have had so many intervening differences such an approach would have been futile. Another likely choice would have been to create a comparison between Loulouni and various national averages or trends, derived from census and other secondary data. Unfortunately, such data proved inadequate in both quality and quantity to provide for any real comparative study. Additionally, the problems of diversity in Mali and subsequent intervening variables would be exacerbated rather than solved by such an approach. Longitudinal studies that looked at Loulouni before and after decentralisation and the refugee influx were also out, since both had already happened, and little extensive or reliable data on Loulouni before the refugee influx existed. Reaching the best compromise available, a case study approach was taken, incorporating longitudinal elements, as well as archival data for some level of contextualisation, if not quite comparison. A longitudinal element was incorporated into both surveys and interviews by having many questions looked not only at the current situation, but whether it had changed in the past five years, and, in the opinion of the respondent, why. This allowed for information on the perceptions of change, if not the change itself. Because of the large number of interviews conducted with a wide range of community leaders, and this was combined with archival data, it was possible to construct a reasonable picture of ?real? changes, common perceptions, etc. Finally, in making causal inferences, instead of looking exclusively at relationships within the data, certain ?smoking guns? were identified in the interviews and open ended questions ? places where the refugee influx had a direct impact on some element of decentralisation. Survey data was then called upon to investigate the extent of the suggested relationship. King, Keohane, and Verba (1994) argue that theoretical leverage is maximised when multiple measures and methodologies are used over multiple data sets. While no one method employed by this study would alone be adequate to uncover the impact refugees have on decentralisation, each analysis reinforces the other on a number of different levels of analysis, across all methodologies. This approach does not allow for generalisability beyond what is happening in Loulouni, but it does gaurantee a high level of accuracy of the overall claim. Additionally, that does not mean it holds no 24 Caitlin Blaser 1 September 2007 value for those with no specific interest in southern Mali. By providing a comprehensive description of the Loulouni case, certain assertions and assumptions in the literature are challenged. It is because of this that this study has a value that extends beyond its degree of generalisability. For the eighteen month period between February 2005 and September 2006, I conducted research in Loulouni to interrogate issues at the intersection of decentralisation and the refugee presence. This research roughly took the six following forms, although at times there was overlap: ? Structured and semi structured individual interviews with seventy-five community leaders, selected based on their exposure to and role in the decentralisation process, their interaction with the refugee presence, and their ability to place Loulouni in either a spatial or temporal comparative context; eighteen additional individuals from Sikasso, Bamako, or Faragouaran were interviewed who could speak to the situation in Loulouni. When 'interview' is cited, it is in reference to these ninety-three interactions. Since interviewees were selected on the basis of their professional capacity, anonymity was not expected unless specific comments were designated as 'off the record.' Despite this, interviewees are referred to by their position rather than their name. This is first in an effort not to 'personalise' any issues, but also to contextualise responses as much as possible, and to avoid the confusion that would be caused by repeated names (six Madou Ouattaras count among those interviewed, as do four Maryam Traores). These interviews allowed for a deep exploration of the political, social, and economic transformations Loulouni has experienced, as well as a more theoretical approach to topics like decentralisation and social transformation. About half the interviews took place before the community survey, which allowed me to identify certain key issues to focus on in the survey. The other half took place after the survey had been conducted, so I could use the interviews to address certain questions raised by the data. Key individuals, such as the mayor and former mayors, were interviewed several times throughout the research period. ? A survey of 100 community members including both open and closed ended questions covering a variety of topics including a profile of the respondent, their views on issues surrounding decentralisation and the refugee presence, their political involvement, and more general issues about the economic, social, and political life in Loulouni. When the survey was finished, many respondents expressed an interest in continuing the conversation, and the interaction then developed into a semi structured interview borrowing from issues raised in the 25 Caitlin Blaser 1 September 2007 survey and previously conducted interviews. The sample was selected randomly from a residential map of Loulouni drawn by a Peace Corps volunteer in 2003. Responses were anonymous, and are cited by number. Surveys were conducted either at the respondent's house or place of work (depending on their preference), with the respondent and myself being the only individuals present. Occasionally during the open-ended responses at the end of the survey, respondents asked to invite friends or family into informal discussions prompted by the questions. In such a case, contributions by friends or family outside the sample are taken as casual discussions (see point 6). When 'respondent' is cited, it is in reference to these 100 surveys. ? Over the duration of this research, fifty members of the community of Loulouni asked to participate upon hearing the topic of research, usually on the basis of specific information they had to contribute. The majority, although not all of these individuals had been migrant workers in Cote d'Ivoire, many of whose return to Loulouni was prompted by the fighting that caused the refugee influx. As this is an important point of view for this study, this information has been included. However, because they do not fit into the random sample they are excluded from statistics cited, and their contributions are considered informal discussions. ? Structured interviews with sixty-five refugees at the camp. This was initially planned to be a survey comparable to that of the community members, but pre- testing proved a survey format inappropriate, and information could be more suitably gathered from interviews. Sampling the refugee population proved challenging, due to the high rate of migration; errors were minimised as much as possible by conducting all of the interviews in an intense two week period. A random sample of fifty individuals was chosen based on a map drawn up by AVNU officials, but over the course of the study fifteen additional individuals were added to the sample using semi-structured interviews because they themselves requested to take part based on either positions of leadership in the community or specific interactions with residents or the government in Loulouni. When 'refugee interview' is cited, it is in reference to these sixty-five interactions. When statistics about refugee opinions or practices are cited, the fifteen individuals are omitted in order to maintain the random sample. ? Archival research conducted at the levels of Loulouni, Sikasso, and Bamako both at official government archives, and through inquiries at the offices of technical line 26 Caitlin Blaser 1 September 2007 ministries. Information includes decentralisation legislation, tax records, rainfall changes, court cases, country- and region-wide economic trends, national public sector expenditure, etc. In this case, sources are cited individually. ? Unstructured, informal interactions with acquaintances and friends in both surrounding villages in the commune of Loulouni, and Ivorians or Malians originally from the commune of Loulouni living in Sikasso. These people were often in a position to offer a different perspective on local government and the refugee influx than was available in Loulouni. They frequently served as test groups for my survey and interview questions, discussed cultural and linguistic differences that could cause bias in my responses, suggested new lines of inquiry, and explained power dynamics and interpersonal relationships in Loulouni from the safe distance of relative uninvolvement. While not a part of the original project design, and not subject to any rigorous methodological framework, these individuals still ended up providing anecdotes and information useful to this study. Interviews and surveys were conducted in either Jula (a dialect of Bambara), French, or Samago. They represent the first language of a majority of individuals involved, and I have functional fluency in all three. While there were respondents whose first language was not represented, fluency in Jula by all members of the community meant that translation difficulties were not encountered. All translations are my own, unless otherwise noted. Because of the number and linguistic variety of surveys and interview guides used, they have not been included as an appendix. For copies of any of the surveys or data, contact me. It is important to mention that I served as Peace Corps volunteer for the duration of the research period in Sieouba, a village on the periphery of the commune of Loulouni, approximately twenty-five kilometres from the town itself. Conventional wisdom recommends against doing research while simultaneously serving as a 'development' worker. My experience doing fieldwork while serving in the Peace Corps challenges this assumption. To begin with, the first two goals of Peace Corps are related to cultural exchange, with the development of human resources being only the third goal on the Peace Corps agenda. While Peace Corps does have an element of development work to it, it also has a long history of sending volunteers to Mali, and it is generally recognised in communities that have historically hosted a volunteer (including Loulouni and various surrounding communities) that far less success is achieved towards its development aspirations than the elements of cultural exchange. The focus on integration into the host community allowed me a cultural and linguistic level of fluency that informed my research immensely in terms of contextualisation, as well ensuring support 27 Caitlin Blaser 1 September 2007 of my research by all relevant authorities. Living in an area similar to, and yet separate from the town of Loulouni allowed me to maintain a great deal of independence from the authorities of Loulouni through whom it would normally be necessary to work for reasons of logistics and protocol, but a dependence upon whom would have created a bias in responses. Additionally, the emphasis of Peace Corps on development work based on education and human resource capacity building rather than funding and external financial support meant that I did not face the same perceptions of wealth of a traditional development worker, or even possibly a traditional researcher. This was supported by the fact that for the entirety of my research I lived in a village considered remote and economically marginal by residents of Loulouni, at a standard in certain ways more modest than they themselves. This is not to imply that power dynamics were absent from my interactions with participants in the research process. The response of one woman to the question ?Are you involved in politics?? sums it up: I am not a member of a political party, no. I have never voted. You will not see me at the mayor's office when he holds meetings, and I am not interested in decentralisation. But my child, we are all involved in politics. When my son goes to Cote d'Ivoire to work and sends money back to me, that is politics. When my husband talks to me about a new idea he has, that is politics. When I served you lunch just now, that was also politics. 28 Caitlin Blaser 1 September 2007 ?IF YOUR NEIGHBOR?S HOUSE IS ON FIRE, YOU BETTER HELP IN PUTTING IT OUT. IF NOT, WHEN HIS HOUSE HAS BEEN RAZED, BE SURE THAT YOURS WILL BE THE NEXT TO GO UP IN FLAMES.? GNASSINGBE EYADEMA 2. Setting the Stage As the title suggests, this chapter provides the background information necessary to apply the theoretical information of the previous chapter to the case of Loulouni. Because this is a case study, and as such delves deeply into one particular place, a substantial amount of background information about Loulouni is necessary to apply the above theory in the most appropriate way given the context. The first three sections situate the refugee influx in Loulouni, while the last three sections discuss the decentralisation process in Mali more generally, for understanding of the context in which the decentralisation process in Loulouni is taking place. The chapter begins with a brief history of the migration between Mali and Cote d'Ivoire. The relevance of this might not be immediately apparent. However, the nature of these strong historical ties proves very formative of the relationship Ivorian refugees developed with the community of Loulouni. A policy promoting Malian migration to work on plantations in Cote d'Ivoire was promoted under French colonial rule. This widespread labour migration was central to the economic development of Cote d'Ivoire, both before and after independence. It also forged close ties with southern Mali, since many migrant labourers settled in northern Cote d'Ivoire for generations. Understanding these dynamics is crucial to understanding the relationship displaced Ivorians have with their host community. The next section gives a brief overview of Loulouni, the host community. In the previous chapter, various demographic, geographic, and economic factors are pointed out as shaping the decentralisation process; this section looks at which of these factors are present in Loulouni to get an idea of what challenges to decentralisation the literature predicts. It also compares Loulouni to the rest of Mali to get an idea of how this commune fits into the national decentralisation agenda. The third section looks at the formation of the refugee camp in Loulouni. Building on the overview of Loulouni, this section is a very practical summary of the demography and other main features of the camp, and provides a discussion of the role players involved in the relief effort and their primary interests. The forth section changes gears to give an historical overview of the decentralisation process in Mali. Like a history of migration, if the relevance of colonial policies is not obvious, it will soon 29 Caitlin Blaser 1 September 2007 become clear that these structures have undeniably had a lasting impact. The current decentralisation process has been heavily shaped by former political configurations, whether it is by adopting their ideas, or opposing their failures. Understanding the evolution of public offices also helps explain how certain power dynamics now either hold back or move forward ongoing political transformation. The fifth section of this chapter will discuss the current decentralisation process in Mali. It explicitly ties this concrete situation back to the decentralisation literature discussed in the first chapter in order to provide a theoretical context for the case of Mali. Because of the implications the decentralisation literature has for Mali, contextualising the national process in this way will allow for a comparison when evaluating the state of decentralisation in Loulouni. It is particularly important to get a nation-wide view of the decentralisation process when discussing Loulouni for two reasons. First, it allows for a clear idea of the main shaping factors of the decentralisation process in other areas. Second, Loulouni was a pioneer commune in the decentralisation process. Because of this, it was more heavily influenced by national rhetoric, approaches, and political climates than communes that decentralised later. The final section ties together all of these snapshots by focussing specifically on the actors at play in the contemporary decentralisation process in Loulouni. Initially this may seem redundant, as actors have been a part of every discussion up to this point. However, looking at them explicitly is important. As mentioned in the first chapter, this study adopts Ribot's framework for understanding decentralisation as a method of analysis, and this focuses on actors, power, and accountability. All discussions of the structural elements of the decentralisation process address changes in power and accountability, but in order to recognise the importance of this, it is essential to be clear about exactly who all the actors are, and what their primary interests are. This chapter covers a lot of territory, and may seem like a bit of a pastiche. However, the integration of these elements will be solidified in following chapters, when an analysis of the decentralisation process in the context of a refugee influx draws heavily from all of this background information at once. 30 Caitlin Blaser 1 September 2007 A brief history of Malian ? Ivorian migration One of the most notable trends in the data gathered in this study was the link between economic class and perceptions of Ivorian refugees. Specifically, the wealthier the respondent, the more likely they were to feel negatively about the Ivorian refugees and their impact on Loulouni (see the table in the last section of Chapter 4). This is crucial for understanding the impact the refugee influx had on the decentralisation process, since, among other reasons, wealthier residents were more likely to interact with local government officials. In depth interviews provided two primary explanations for this. This first is that wealthy residents of Loulouni were more likely to have spent time in Cote d'Ivoire, and had negative experiences of xenophobia and racism. The second explanation is that because Cote d'Ivoire has a higher level of development by most indicators than Mali, the Ivorian refugees were more likely to be ?competing? with the wealthier residents of Loulouni for skilled employment, luxury goods, medical care, places in school, etc. Poorer residents were in that respect less affected by the influx. For a more nuanced understanding of the class-based dynamics between refugees and residents of Loulouni, it is necessary to draw a picture of the history of migration and economic ties between Cote d'Ivoire and Southern Mali. Cote d'Ivoire has been a centre of trade and migration in West Africa for centuries (Toure et. al 1992). When much of West Africa was colonised by the French, an explicit policy of promoting (and at various times forcing) migrant labour to Cote d'Ivoire was pursued to develop the plantation economy at the expense of areas seen as having less potential for economic exploitation. As Cooper (1996) states, French colonial policy was not designed ?for the ?total development? of French West Africa but would concentrate on ?islands of prosperity,? well situated for exporting produce.? This has had numerous lasting effects that will be expounded upon throughout this thesis, including defining many of the contemporary social dynamics that exist between Cote d'Ivoire and Mali. The little interest the French showed towards Mali was nearly entirely geared towards exploiting its labour force. This found expression through policies such as relocating entire villages to main transportation routes so labour would be more easily accessible, and imposing head taxes to be paid in cash, demanding an increase in the pursuit of wage labour. These policies failed to promote large- scale migration in the early 20th century because of numerous factors, including the comparative economic equality of the two countries at that time, a neglect of food crops in Cote d'Ivoire, depressed wages partly stemming from forced labour regimes, and resistance as a form of political protest (Cooper, 1996; Amin, 1973). Changes in policy in the mid 20th century, however, brought significant changes in migration patterns. The abolishment of forced labour in 1946 increased wages 31 Caitlin Blaser 1 September 2007 tenfold. This, combined with creation of a workers syndicate, Syndicat Interprofessional d?Acheminement de la Main d?Ouevre (SIAMO) which provided transportation subsidies and health care to migrant workers created a flood of migrants from Mali to Cote d'Ivoire, and largely sparked the boom in Cote d'Ivoire's economy in the years to come (Cordell, et all 1996). After independence, there were significant changes in policy, but ironically their outcomes were remarkably similar to the colonial era, as they continued to marginalise and impoverish rural areas. Open trade and migration combined with a large wealth disparity maintained migration to Cote d'Ivoire. This was reinforced initially by political tensions with Senegal leaving Abidjan as the only accessible port, and later by poor internal economic policies that both encouraged elicit trade over the border, and further impoverished the countryside, making emigration an important livelihood strategy. Attempts were made by Houphouet-Boigny (the president of Cote d'Ivoire), Keita and Traore (both former presidents of Mali), to utilise migration to promote growth of the public sector, but in practice they simply drove migration underground, and undocumented migration remains the primary form of migration between Mali and Cote d'Ivoire today (Domba, 2003). This migration was a huge driving factor in the economic growth in Cote d'Ivoire, widely claimed to be a West African miracle. This growth was largely built by the hands of Malian labourers, however, and the Malian economy ?paid for this in almost total stagnation? (Amin, 1973: 140). The miracle was turning to mirage in the 1980s, when much of West Africa was hit with economic crisis. Structural adjustment programs on the heals of a devastating drought caused a collapse of the Ivorian labour market, which in turn led to a rise in xenophobia. This was not an unforeseen process; as early as 1970, Efrem Sigel noted: Much of the actual or potential political tension stems from that fact that Ivorians do not know if the wealth of their country belongs to them or to foreigners. Not that the actual fact of ownership is so important; in politics it is appearances which count. And as long as there are...foreign Africans who seem to be profiting from Ivorian prosperity more than the Ivorians themselves, the potential for trouble exists (1970: 18). Despite the onset of economic hardship in Cote d'Ivoire, significant disparity remained in the level of development between Mali and Cote d'Ivoire, with the GNP in pre-conflict Cote d'Ivoire over twice that of Mali (Skeldon, 1997 p212). As long as this imbalance was also reflected in the infrastructure, availability of services, and levels of human development, migration continued. As the tide of xenophobia rose, policies began to reflect it, such as the institution of identity cards, seen by many Malians as a precursor of the expulsion of immigrants, many of whom had been in Cote d'Ivoire for generations (Toure et al., 1992). 32 Caitlin Blaser 1 September 2007 This xenophobic movement gained momentum when Houphouet-Boigny, a long-time supporter of Malian migration, died in 1993, and a heated struggle for leadership dominated the next six years of Ivorian politics. The culminated in a coup d'etat in December of 1999, and military clashes soon after. While the political, economic, and social roots of the conflict in Cote d'Ivoire are complex,5 they eventually found articulation in anti-immigrant sentiments, expressed through various forms of identity, including religion (southern Cote d'Ivoire is predominantly Christian, whereas much of the north shares its Muslim faith with Mali), ethnicity (because of arbitrarily drawn colonial lines, northern Cote d'Ivoirians share ethnic ties with southern Malians and Burkinabes), language, place of residence, and national origin. The expulsion of immigrants came to be seen as the only way to restore Cote d'Ivoire to its former economic glory (BBC News, 10/07/02). It is this conflict, which is fundamentally about citizenship rights, and the relationship between Mali and Cote d'Ivoire, that caused the refugee influx. When the dynamics between residents of Loulouni and refugees are discussed, a reflection on this migration context often explains the source of various perceptions. For example, Malians who were themselves migrant workers in Cote d'Ivoire were very likely to have a negative view of the refugees, because they had experienced xenophobia and discrimintation while in Cote d'Ivoire. Others who had a negative opinion of the refugees cited the fact that they were just taking advantage of aid, because they were in fact Malians. This is understandable given the history of migration, and politics of identity in Cote d'Ivoire. Still other aspects of the relationship between Ivorian refugees and their Malian hosts stem from the relative wealth of Cote d'Ivoire, and the fact that Malians participated in creating this (without often sharing in the fruits). How refugees and hosts frame the rights and responsibilities of the other is very much embedded in how they interpret these historical migration ties. 5 For a more in depth account, see Akindes, 2004. 33 Caitlin Blaser 1 September 2007 Introduction to Loulouni Loulouni lies at the heart of these migration ties in a number of ways. It is the capital of a rural commune in the south east corner of Mali, in the region of Sikasso (see appendix 3). Located on the main road between Sikasso and Cote d'Ivoire, it has always played a very important role in regional trade and migration. Even under French colonial rule, it was the seat of the chef du cannon. As mentioned in the first chapter, numerous geographic, economic, and demographic factors have implications for the decentralisation process. Furthermore, the refugee influx did not happen in a political void. This section will begin with a simple introduction to Loulouni and the surrounding area before the discussion on the humanitarian intervention or the decentralisation process can begin. Information was selected with Manor's laundry list of helpful conditions for decentralisation in mind, to see how decentralisation theory suggests the process will unroll. This will be drawn on in the following chapter, where the evidence gathered in this study will point to similarities to and differences from the theoretical path, and suggest reasons for any divergences. The Sikasso region has many economic and institutional ties to Bamako, in a large part due to its reliance on cotton production, a nationalised industry only now taking steps towards privatisation. While subsistence agriculture is still central to the economy of the Sikasso region, cash crops (particularly cotton) have a stronger hold on the economy than in other regions in Mali, due to its higher rainfall. As a result, the region has more net wealth than other parts of the country, but is also marked by greater inequality. Per capita income is higher than in any other region of the country, but it lags behind in human development indicators such as maternal and infant mortality, literacy rates, and life expectancy. While still weak, Sikasso has better transportation infrastructure than the rest of the country. With Abidjan being Bamako's most accessible port, most of the country's imports and exports pass through Sikasso. Its relative accessibility combined with low human development indicators help explain the high presence of NGOs and humanitarian organisations; there are more working in the Sikasso region than any other part of Mali outside of Bamako. There is only one paved road between Sikasso and Abidjan, and Loulouni, with 3000 residents, is the largest town between Sikasso and the border.6 Equidistant from Sikasso and the border town of Zegoua, Loulouni is a commune capital and significant trade centre in its own right. In addition to a large weekly market, it hosts a number of businesses and several services for the region including a second cycle school, health centre, and bank. While no NGOs actually have offices in 6 See appendix 3 34 Caitlin Blaser 1 September 2007 Loulouni, there are numerous extension agents working in both the town and commune, including representatives of agricultural development organisations and literacy trainers. There is a large wealth gap between urban and rural communes in Mali, but Loulouni is one of the wealthier rural communes not adjacent to an urban commune. Because the road through Loulouni goes only to the border and does not connect two major Malian cities, it has lagged behind comparable communes in services like telecommunication and electrification. Subsistence agriculture forms the foundation of Loulouni's economy, but there is a high degree of economic diversity. Maise is the staple crop, but since a World Bank funded project planed a substantial area of farmland adjacent to a river on the outskirts of town in 2005, rice cultivation is increasing. More tropical crops like cassava and oil palms were central only a decade ago, but a drastic drop in rainfall affecting the region over the past two decades means that agricultural production has seen substantial changes in recent years.7 Cotton is the most important cash crop cultivated, but as prices have dropped, it is being replaced by garden produce. Loulouni's advantageous location along the main road makes it possible to grow more profitable, quick spoiling produce, like mangoes and tomatoes for sale in Bamako and Mopti. In addition to trade, migration is an important feature on Loulouni's economic landscape. Much like the rest of Mali and the region of Sikasso particularly, the population of Loulouni is very mobile. Given the lack of economic opportunities in Loulouni, many people migrate as a livelihood strategy. Mali is rapidly urbanising, and with the closest high school being in Sikasso, promising students often go on to study and look for work in either Sikasso or Bamako. The substantially wealthier Cote d'Ivoire is the primary destination of migrant labourers. Southern Mali has a long history of migration to Cote d'Ivoire, and many Malians permanently settled on cocoa and coffee plantations in northern Cote d'Ivoire generations ago.8 In 2000, Malian migrant workers were bearing the brunt of political tensions in Cote d'Ivoire. Many migrants from Loulouni and the surrounding areas returned during this period. When these tensions escalated to a full-scale civil war in 2002, the impact was felt in Loulouni. Not only did many Malian workers and refugees come to Loulouni, the transport route between Sikasso and Abidjan was closed. When traffic was diverted east of Sikasso, bound for Lome, the economic lifeline of Loulouni was strangled. According to Manor's Promises and Limits of Decentralisation, Loulouni is not a textbook example of the best place to implement decentralisation. However, it also has some things working 7 See appendix 2 8 For more reading on migration patterns in West Africa, see Cooper, 1996; Cordell, et. Al 1996 and Diarrah, 1991. 35 Caitlin Blaser 1 September 2007 in its favor. There is very little history of democracy in Loulouni, which make successfully decentralising very difficult, as the accountability structures that are so crucial to democratic decentralisation must be created. However, Loulouni does have actors (to be discussed in more depth in the final section of this chapter) that act in some democratic ways, which could potentially mitigate this drawback. Similarly, Loulouni is quite poor by most standards, and a lack of resources and infrastructure at the local level is another obstacle to decentralisation. However, Loulouni is developed in comparison to other rural communes in Mali, and as such is well placed to enjoy many advantages in terms of the national decentralisation process. There are low levels of inequality in Loulouni, which is an asset to the decentralisation process. From demographic information alone, it is impossible to judge whether or not decentralisation is likely to succeed in Mali. Keeping this information in mind, however, a much more educated guess can be formulated by placing the decentralisation process in an historical and national context. 36 Caitlin Blaser 1 September 2007 Establishment of the refugee camp in Loulouni As the previous section did with migration, this section aims to give a brief overview of the refugee camp in Loulouni. Upcoming chapters will go into a great deal of detail about the nature of the relationships between refugees and residents of Loulouni, and how the refugee camp affected the political, economic, and social landscape of Loulouni. Before this analysis begins, however, it is useful to keep in mind a broad picture of the refugee camp. To draw this picture, this section will outline the establishment of the camp, and its main demographic features and administrative structures. Sporadic fighting broke out in Cote d'Ivoire beginning in 2000, and escalated to full scale civil war by September of 2002. Tens of thousands of the estimated three million Malian citizens living and working in Cote d'Ivoire returned to Mali, both through government assisted emergency repatriation efforts, and of their own accord. Numerous Cote d'Ivoirian citizens also came to Mali as refugees. There are two significant roads between Cote d'Ivoire and Mali; the paved road to Sikasso through Loulouni, and an unpaved yet major transportation route to Bougouni through Faragouaran in the west. When fighting first broke out, many people fled to the border towns of Zegoua and Yanfolila (on the eastern and western roads respectively), but they were both unequipped to host so many people, and not safe themselves due to the possibility of fighting spilling over the border. A transit centre was opened in Faragouaran, and a camp was opened in Loulouni (see map, appendix 3). From the beginning, the government of Mali saw the crisis in Cote d'Ivoire as having the potential of a tremendous impact on the Malian economy. Almost immediately after the escalation of violence, ATT went to Guinea to discuss transportation and access options to the port in Conakry. As a landlocked country, Mali is dependent on its neighbours for all imports and exports, and the majority of them were routed through Cote d'Ivoire before the war. With a closure of the border, the prices of consumer goods shot up. Technical line ministries were asked to do evaluations of the expected impact the crisis would have on their work. It was widely acknowledged that as many of the Malian nationals living in Cote d'Ivoire have remitted significant amounts of money to their families remaining in Mali, many Malian families that were dependent on these remittances would be negatively affected....The possible return of all or a significant amount of Malians living and working in Cote d'Ivoire will put tremendous pressure on the economic and social fabric in Mali (OCHA, 2002). With an official evacuation of its nationals begun by the Malian government, the question in Bamako 37 Caitlin Blaser 1 September 2007 was no longer whether this pressure would be applied, but rather where it would be applied, and how best to respond to it. On the international front, an InterAgency appeal by UN organisations was issued, and by the end of November, a full scale relief effort was in the works. The African Union deployed a mission to Mali to evaluate the situation of refugees, UNDP and IOM sent fact finding missions to Sikasso, and many planning meetings and donor briefings were held. While the UNHCR office in Bamako closed in 2001, a staff member was deployed from Cote d'Ivoire, and implementation partners including the National Commission in Charge of Refugees (CNCR) and the Association of former United Nations Volunteers (AAVNU) were called on to assist in refugee management. Nation-wide, the biggest concern was related to food security. Food production decreased in Mali in the first years of the second millennium, and in 2002, Mali was already going to be dependent on imports to meet its food needs. The food security stock was at an all time low, and a drought had set it that was affecting November harvests, and causing concern about the prospects for food security in 2003. On a local level, however, concerns varied. Loulouni, while not a border town, was generally the first point in Mali refugees and returned migrants came to fleeing the conflict. The initial population of the refugee camp was a mix of people from numerous nationalities (including Burkinabe, Togolese, Nigerian, Guinean, Senegalese, Ghanaian, and Iraqi), although predominantly Cote d'Ivoirian. Rapid migration through the camp meant a substantial fluctuation in the number of people housed, but it remained between 500 and 800 from 2002-2004.9 The Malian Red Cross was the first humanitarian organisation to provide direct assistance to the refugees, meeting people as soon as they crossed the border in Zegoua, and providing various food and non-food items, such as rice, cooking oil, sugar, soap, and sleeping mats. When it was agreed that the refugees would be moved to Loulouni, the Red Cross worked with the former mayor to distribute initial supplies. They were eventually followed by an impressive host of acronyms. AAVNU, UNICEF, WFP, CARE Mali, and Islamic Relief partnered to run the camp itself, although many other organisations (discussed below) provided assistance in other capacities. When the camp was opened, a wide variety of people were staying in Loulouni, both inside and outside the camp. While official statistics are unavailable or unreliable, there is widespread consensus that the biggest population of new arrivals were returned Malian migrants and their friends or relatives, who had friends or relatives in Loulouni to stay with. A smaller number were Cote d'Ivoirians or returned migrants with no previous connections to Loulouni, some of whom arrived on 9 During the years this research was conducted, 2005-2006, the population was significantly lower, around 1/2 that figure, as a closure of the camp began in mid 2006, and refugees were encouraged to relocate to Faragouaran. 38 Caitlin Blaser 1 September 2007 their own while others moved up from Zegoua with the opening of the camp. OCHA estimates that fewer than 20% of non-Malians coming through Loulouni due to the conflict in Cote d'Ivoire attempted to register for asylum. Estimates by local government officials and business people confirm that for the 800 registered refugees, there were approximately 3,000 Malians and Ivorians staying on a temporary basis. Most people used the camp as a stopping-over point to meet their basic needs until they could arrange to go to Sikasso, Bamako, or anywhere else offering better economic prospects, rather than a place of long term or potentially permanent settlement. The majority of the camp population, and the people who stayed at the camp longest were youths (about 65% were under 18), with males outnumbering females in both the child and adult populations (interview, camp manager 8 April 2006; interview, former mayor 29 March 2005). The initial situation in Loulouni left many critical needs unmet. The Government of Mali established coordination services on both a territorial level (with local, communal, regional, and national components), and a sectoral level (with groupings for food security, protection, education, shelter, water and sanitation, health, HIV/AIDS, and Information Management). The planning and coordination needs were substantial to say the least, with partners including CARE, DWHH, ICRC, IOM, AAVNU, CNCR, Handicap International, UNESCO, OXFAM-UK, Islamic Relief, SCF (US), MSF, SDC, UNAIDS, ADAC, UNFPA, UNDP, UNHCR, WHO, UNICEF in addition to local NGOs, CBOs, and government agencies and services (OCHA, 2002). These role players met fortnightly to discuss all issues that required multi-sectoral coordination, including access, security, conflict management, advocacy, and emergency response. The crisis committee at the regional level in Sikasso met on a daily basis (where temporary sites were also set up to assist refugees and returnees), and provided briefings from each meeting for consideration at national level stakeholder meetings. The crisis committee set up in Loulouni included local government officials and relevant community leaders such as the gens d'armerie, school director and doctor. It was led by the sous prefet. Under its guidance, tents were set up,10 latrines were constructed, and potable water was regularly bussed in. A distribution regular of essential food and other items was started through the office of the sous prefet. Soon after the arrival of refugees, many of the Sphere Humanitarian Standards were being adhered to, but there were notable exceptions, primarily with water/ sanitation and health care. The biggest difficulty at the camp and for the town of Loulouni as a whole was that of potable water; in 2002, there was only one borehole pump in Loulouni, located at the health centre. When funds were located, organisations supporting the refugees (led primarily by UNICEF) had two more 10 More permanent housing was eventually built as well 39 Caitlin Blaser 1 September 2007 boreholes drilled and installed pumps near the refugee camp. The camp itself was located in a newly developing part of town, behind the weekly market. It was separated from the main residential area by a cluster of businesses and office buildings, including the offices of the mayor and sous prefet, the health centre, the second cycle school, and a row of artisans. There were, however, a few houses on the same side of town as the camp, primarily belonging to extension workers, or representatives of various services (the gens d'arms camp, school director, etc.). Not only had there been no source of potable water within 500m of the market, there were no wells on this newly developing side of town either. One of the new boreholes was installed near the school with a tap, while the other one was covered with a hand pump behind the market, at the camp itself. This new pump at the camp was the market's only source of water potable water. The challenge of water is inseparable from that of sanitation, and latrines were in demand as soon as the camp was constructed. This is an example of the residents of Loulouni offering a hand during a time of crisis by donating their time, labour, and supplies. The women's association dug eleven provisional latrines that were used by initial residents of the camp. UNICEF then supported their efforts through the provision of construction materials to make them permanent and more sanitary. On the health front, there were many concerns. The health centre in Loulouni, while sizeable for a rural commune-level health clinic in Mali, is only equipped to deal with primary health needs. There is a small pharmacy that is adequate for first aid and primary care, but not sufficient for comprehensive care. Given that many refugees and returnees had travelled long distances under difficult circumstances, demand for health care went beyond the capacity of the clinic. It does have a gas-powered refrigerator for vaccines and other medicines requiring refrigeration, and is staffed by only a midwife, doctor, and secretary to serve the entire commune of more than 35,000 people. At the time of the influx, there was a two way radio with which an ambulance could be called from Sikasso if necessary. This was replaced in August of 2006 when a telephone line was run to Loulouni, greatly improving communications. The most immediate health concern articulated by refugees was for anti-venoms, as there was a problem with vipers in the initial days of the camp. The second priority area identified, however, was HIV/AIDS. The spread of HIV/AIDS was identified at the national level as a potential consequence of the conflict in Cote d'Ivoire. Mali has a relatively low infection rate, around 2% (UNAIDS, 2004), and it is mostly restricted to high-risk populations such as prostitutes, truck drivers, and mine workers. The spread of HIV outside these populations has been very limited, but there is a great concern among the Malian authorities as should this happen, Mali has many features that 40 Caitlin Blaser 1 September 2007 would encourage the spread of the virus, including widely practiced polygamy, high levels of migration, and non-use of condoms. Cote d'Ivoire, on the other hand, has the highest prevalence rate in the region (approximately 12% [UNAIDS, 2004]). The urgency of an intervention targeting the spread of HIV/AIDS was made clear when a Red Cross fact finding mission to Loulouni reported ?the presence of a large number of young, idle, sexually active teenagers,? with ?cases of prostitution becoming rampant (International Federation of Red Cross and Red Crescent Societies, 2002).? This report came out almost simultaneously to an interview with the camp manager who confirmed that there was a refugee dying of AIDS (International Federation of Red Cross and Red Crescent Societies, 2002). The Malian Red Cross began weekly awareness sessions, and distributed condoms. A number of other diseases prevalent in the dry season in Loulouni were exacerbated by displacement, and there were problems of malaria and respiratory diseases, including meningitis. A packet of pharmaceutical supplies was integrated with the food distributions, and this was accompanied by various health related trainings. Because the high demand for minor, first-aid type treatment, between two and four Red Cross volunteers remained in Loulouni at all times to help the health centre cope with consultations, treating minor illnesses, and handling referrals. Schooling was another service provision challenge, given the large number of refugees that were in school, and the already overcrowded classrooms in Loulouni. Some children, generally those who had no immediate plans to leave Loulouni, chose to integrate into the school system. The majority, however, due to uncertain time horizons or other considerations attended a temporary school established by UNICEF that offered flexibility from the rigid Malian curriculum. Given the relative strength of the Ivorian school system, as well as the fact that French, the medium of instruction, is more widely used in Cote d'Ivoire than in Mali, the top student in nearly every grade in Loulouni has been Ivorian since 2002 (Interview, school director, 2 April, 2006). Because the focus of this study is on the impact of the refugee presence on decentralisation, which is by definition a formal, institutionalised transformation, discussions of the relief effort and refugee presence in general will focus on the 'official' sphere, and largely leave out discussions of, for example, the large number of Malian migrant workers in Cote d'Ivoire who came to Loulouni at the same time, and generally under the same conditions as the refugees, but who did not stay in the camp. However, note should be made of the extent to which the ?relief effort? existed outside of the traditional food and supplies rations from the office of the sous prefet, and international organisations working in the community. During the period of greatest influx, in late 2002 and early 2003, 90% of survey respondents 41 Caitlin Blaser 1 September 2007 reported some kind of personal involvement in the informal relief effort, generally through hosting displaced people for periods ranging from a night to a year, but also by helping to raise funds for return or onward passage for migrants, providing meals, etc. While most of this took place on the individual level as acts of charity, there was also institutional engagement. One example of this is when the women's association convinced two chauffeurs to donate their time and vehicles, and in partnership with businesses in the community, pooled funds to transport stranded refugees and returnees to Sikasso where assistance efforts were generally better equipped, and many people expected to meet family or acquaintances. Many other individual acts are widely praised in the community, such as one retired businessman who hosted three Ivorian families, amounting to more than twenty people, in his concession for a span of months, providing them with all they needed during this time, and even constructing an extra house. While stereotypes of African hospitality may or may not apply, cultural beliefs and values were often brought up in discussions about the refugees. The story of Sundiata Keita, one of the rulers of the Empire of Mali who began life with many hardships, was often evoked to illustrate that people in hardship one day may be powerful the next, and vice versa. In considering both the refugee influx and decentralisation in Loulouni, one of the most important things to keep in mind is that neither happened in a void. The history of migration dynamics affected the reception Ivorian refugees received in Loulouni, and national and international economic trends have impacted the process of political transformation on the local level. It is only through understanding how the current social and political context evolved that it will be possible to discuss how the current transformation is unrolling. 42 Caitlin Blaser 1 September 2007 Historical Overview of Decentralisation Movements in Mali Decentralisation has a long history in Mali. Looking at the post-independence waves of decentralisation will first define the trajectory of political transformation in Mali, but will also allow the case of decentralisation in Mali to be tied to the general literature on decentralisation discussed in the introduction. Understanding the different articulations of decentralisation that are present in Mali has a bearing on the places a refugee influx will interact most with the process. It also provides a stronger justification for the measures of success chosen in this study, and helps in explaining the perceptions of local government in Loulouni. Furthermore, understanding the evolution of the current decentralisation process will tie the political transformation in Mali in several clear ways to other decentralisation movements particularly in Africa, but also throughout the world. Mali gained its independence from French colonial rule in 1960. In the first half of 1960, it chose to unite with the wealthier, coastal nation of Senegal to form the Mali Federation. In only a matter of months, however, differences between Senegal and Mali seemed irreconcilable. While Mali held radical stances on the Africanisation of the administration, relations with France, and economic control, Senegal had a more accomodationist outlook. These differences threatened armed struggle, but were peacefully resolved in August of 1960, when each nation declared unilateral independence. While outright fighting was averted, the tension between Mali and Senegal was not only detrimental to the Malian economy, but was central to shaping the Malian political landscape. Modibo Keita had dominated independence struggle politics, and led the first Republic of Mali as a one party state. He sought close ties with the soviet bloc, in part to distinguish himself from Leopold Senghor, the French educated President of Senegal. He was heavily influenced by the socialist, Africanist political ideology that were then prevelant among development philosophies on the continent. The first goal of this nascent state was economic decolonisation, which meant severing ties from French commercial monopolies (Clark in Bingen, 1999). To do this, rural socialism through village collectives were established to encourage the mobilisation of rural resources. While new party committees established throughout the countryside were supposed to breathe life back into rural political leadership, in practice the few structures that offered the potential of a civil society were either pushed away from the government or demolished by this push for organisational conformity (Rawson in Bingen, 1999). For example, commandants de cercle had always been centrally appointed officials, but under French rule they remained for years in the same area, growing increasingly familiar with the local context. These were replaced by Party officials with no knowledge of the local situation. Additionally, the position of canton chief, which formerly provided a 43 Caitlin Blaser 1 September 2007 counterbalance to the representative of the central administration, was abolished. A certain degree of administrative decentralisation was implemented. Territorial collectives were to ?administer themselves freely by elected councils and under conditions foreseen in the law.? However, ?the delegate of the Government has charge of the interests of the Republic, administrative control and respect of laws.? (Republique du Mali quoted in Rawson in Bingen, 1999) This theoretical commitment to local government while balanced against the centralised ?interests of the Republic? clearly reflects the French colonial institution of tutelle in which local administrators are given decision making power, but only with the close oversight and tutelage of the colonial authorities. This spectre of tutelle will continue to appear in later decentralisation efforts. The key to Keita's decentralisation vision was the mobilisation of the rural economy for the state's benefit. The consequences of this were multi-fold. First, this created a tension within the administration, where the Ministry for Plan and Rural Development was disproportionately powerful, and secondly, there were splits within the party between the conservative commercial class and the progressive socialists. To move beyond these divisions, instruments of supervision and control were simply multiplied, magnifying the problems of inefficiency without solving problems of marketing (Diarrah, 1986). All the while, the state-run marketing systems were inefficient, and devastating to rural farmers. Official produce prices were held well below market levels, creating a thriving black market and illicit trade with Senegal and Cote d'Ivoire. Production stagnated, to say nothing of economic growth. While economic deconolonisation was to a degree achieved, it was at a great cost. The economic policies of the Keita regime were increasingly harmful, with ineffective rural development programs being funded by printing more money. When faced with popular resistance, Keita's regime turned to increasingly authoritarian pressures of extracting resources from already impoverished rural areas. In response to the increasing oppression of President Keita, a military coup in 1968 installed Moussa Traore as president. Traore initially pushed towards regional integration, improving relations with neighbouring states and joining the CFA franc zone. In the countryside, the Traore regime instituted a new form of cooperation, creating Tons villageois, political and economic entities modelled after pre-colonial social security organisations. While the rhetoric for collaboration with farmers was positive, the actions of the regime never moved beyond talk of engagement with the long marginalised rural areas. Despite a rhetoric of liberation, these new development initiatives amounted to little more than further attempts of economic control by the state. While there were superficial economic reforms, the economy of Mali did not see improvements during the initial years of Traore's leadership, and the longer he stayed in power, the more the poverty of the nation 44 Caitlin Blaser 1 September 2007 contrasted to the decadent opulence of the clearly corrupt regime. Popular demonstrations began to occur regularly, but were crushed with increasingly repressive measures. There was much talk of return to civilian rule and reform, but there was no meaningful implementation. In the early 1970s, a devasting drought combined with internal and external pressures necessitated some action on the talk of reform. A new constitution was ratified by popular vote in 1974, but it didn't take effect until 1979. When this centrist form of democracy was finally instituted in theory, Traore received 99% of the votes as the head of the unopposed UDPM (Union Democratique des Peuple Malien) party. The cosmetic changes made did include a widespread implementation of administrative decentralisation. Illustrative of the non-democratic nature of this initiative, the reform came not from parliament, but from a military decree (Ordonnance 77-44). This decree cancelled all local governance institutions of the Keita regime, and set up yet another system of territorial administration. Village chiefs were once more elected according to French colonial law, and had an official role in the local administration. However, they were subject to the tutelle of the arrondissement councils, which were upwardly accountable. Much of the motivation for this new wave of decentralisation stemmed from internal and international pressures for the central government to cut spending drastically through Structural Adjustment Programmes (SAPs). Emphasising decentralised structures, privatisation and market liberalisation drew popular attention away from the growing dissatisfaction with one party rule. There was also the recognition that the central government would no longer be able to deliver the same services with a smaller budget, and would need to mobilise local institutions and resources to maintain what little legitimacy it still had (Ribot, 1999). In this way, decentralisation was ?intended to increase efficiency of central government administration rather than promote local autonomy or popular participation? (Conyers, 2000:28).11 As such, the central government created elected local councils, but did not give them any meaningful power (Diallo, 1994 p1). At this point, decentralisation was largely reactionary in nature. For most of the latter half of the twentieth century, the central state in Mali had been continually growing, first spurred by the political imperative of creating a national identity after shedding colonialism, then by attempts at using the state as a strong tool to guide economic development. This centralisation of policy formulation and economic production led to corruption and clientelism. Decentralisation was employed as a way to fight the ills that came with centralisation, but its own merits or disadvantages were rarely examined explicitly. As Fageut (1997 p3) points out, ?Before boarding the decentralisation train, it would seem prudent to establish stronger reasons for embracing it than the fact that it is the opposite of a 11 Conyers is writing about decentralisation in Kenya's 1983 ?district focus? plan. Decentralisation was used as a tool by many central governments in Africa at this time. 45 Caitlin Blaser 1 September 2007 previous, failed strategy.? Some scholars argued in favour of this application of decentralisation on the basis that in large, centralised states that lacked capacity, many services had been locally managed by default in areas far from the capital (Alfonso, 1999). Shaping policies to reflect this reality could offer institutional support to these local management institutions. It is important to point out, however, that this is the cause of many power struggles on the local level. In Mali, services that the central government had neglected or been unable to provide were under the default jurisdiction of ?traditional? authorities. These authorities constitute part of the complex and dynamic political scene that constitutes contemporary Mali. Associations based on age, gender, or profession, village chiefs, religious leaders, workers unions, private businesses, family members who emigrated to find cash labour, and civil society organisations and NGOs all play a role in service provision, and vary greatly in their efficiency, power, and the democraticness of their makeup. By ?decentralising,? powers that were previously de facto privatised or controlled by traditional authorities were given to newly created local government authorities. This led to the governmentalisation of these services, since jurisdiction for them was effectively removed from other organisations. Non-state authorities often rightly perceive this shift as a government takeover of their powers (Ribot, 2002). As Ribot simultaneously points out, traditional authorities are not inherently democratic, so nothing in this move inherently undermines the democratic process. However, there is a predictable power struggle that is overlooked by the implementors of decentralisation who may not be in a position to recognise the de facto transfer of powers that took place due to limited government capacity. Others argued that SAPs strengthened one party governments, and through decentralisation local institutions that had formerly operated autonomously due to lack of central government capacity were penetrated by malicious central governments (Conyers, 2001). Furthermore, there is an underlying assumption that corruption and lack of accountability are natural by-products of centralism, which is not necessarily backed by evidence (Manor, 1997). Regardless of the practical advantages or disadvantages, this articulation of decentralisation ties it theoretically to the broader process of privatisation and market liberalisation (Platteau, 1995). With the emergence of empirical evidence demonstrating the strengths and weaknesses of decentralisation, this framing is often criticised by proponents of decentralisation. Decentralisation has proved unsuccessful at reducing overall government expenditure, and advocates argue that by tying it to SAPs, the true benefits decentralisation may offer are being overlooked (Manor, 1997). However, it does introduce accountability and transparency, efficiency, and the mobilisation of rural resources as key goals of decentralisation. It also speaks to those who argue against decentralisation on the basis that it has been poorly implemented. The United Nations Capital 46 Caitlin Blaser 1 September 2007 Development Fund (2004 p2) condemningly says that ?by the end of the 1970s, it was clear that the state institutional apparatus for decentralised development had neither promoted participation, nor promoted any meaningful economic and social advancement.? However, this is not necessarily a reflection of anything inherent to decentralisation. Rather, one West African survey ?could not find any local government with control over its budget or any with autonomous policy making powers? (quoted in Oyugi, 2000 p17). By the end of the 1980s, there was increasing criticism of President Traore from students and even civil servants in Bamako as well as in the international community, because he was clearly not practicing the economic and political reform he had long promised. Funds set aside for rural development were either caught up in struggles between different levels of administrative bureaucracy that inhibited their distribution, or went to fund party offices at local levels rather than contribute meaningfully to development. Eventually, even nods in the direction of participation were abandoned when Traore abolished the elected community authorities, and replaced them with appointed military officials. Decentralisation was once again used not as a tool for popular participation, but as a way of consolidating power in the centre and undermining individual rights. Furthermore, motions towards economic liberalisation did not correspond to better financial management on the part of the government. The CFA was overvalued, damaging exports, and eventually even public employees were unpaid, both due to general economic deterioration and outright corruption by the ruling elite. This popular dissatisfaction with Traore's government came to a head in 1991, with widespread protests and rioting led by students but often supported by government and military officials. Under intense pressure, a national conference was held to draft a new constitution. The inclusiveness of the national conference, its commitment to democratic principles, and most of all its success in bringing all stakeholders on board in a commitment to move the country forward was a remarkable success, and speaks volumes of the potential in Mali for good governance. Decentralisation is enshrined in the constitution, declaring that ?all territorial authorities in the Republic of Mali shall be freely administered by elected bodies (Republique du Mali, translated by SNV 2004).? This section has primarily used the historical overview of decentralisation in Mali to situate the current process in a theoretically relevant way. However, it is important for a number of other reasons as well. As the next section will emphasise, the political history of Mali has a very real bearing on many aspects of the current decentralisation process. Many of the challenges this study found the decentralisation process in Loulouni was facing have roots as deep as the colonial political systems. Many of the institutions developed for systematic tax evasion, for example, were clearly a response to historical economic policies that empoverished the rural areas (Diarrah, 1986). Similarly, 47 Caitlin Blaser 1 September 2007 much of the lack of legitimacy current local government officials face stems from the frequent rebirth of decentralisation without real changes in the distribution of powers or resources (Bingen, 1999). After looking at the contemporary status of the decentralisation process in Mali at the national level, and keeping in mind the historical context, the next chapter will be in a position to discuss in much more depth the local governance structures in Loulouni, and how the process of transformation has taken place. 48 Caitlin Blaser 1 September 2007 FOR MALI, DEMOCRATIC DECENTRALISATION IS A MATTER OF POLITICAL SURVIVAL. USAID, 2004 Contemporary Decentralisation in Mali This section looks at the status of Mali's decentralisation process at a national level. The measurements of success of decentralisation (appendix 4) and its theoretical implications have been pulled out in the last two chapters. This chapter will end by looking at how implementation has taken place at the national level. It will first provide general information about the powers and responsibilities of many actors in Loulouni. However, it will also demonstrate how some of the conclusions drawn from the case of Loulouni may be applied either to the rest of the country, or different contexts more broadly. Furthermore, an understanding of decentralisation throughout the country is important for measuring the impact of the humanitarian influx. When Loulouni's decentralisation process diverges greatly from the national agenda, and such a divergence cannot be explained by demography, geography, etc., the humanitarian influx is then a possible explanatory factor to be explored when discussing Loulouni. The current decentralisation framework in Mali was established with National Conference in 1991, which readopted the 1960 Constitution's formula for territorial organisation, with local structures being collectives freely administered through elected councils. However, while there are structural similarities, there are also fundamental differences in the logic behind these structures. Earlier constitutions begin with a commitment to socialism and establishing the Party was the first institution of the state, while the 1991 Constitution uses a human rights based vocabulary, talking of individual rights, private property ownership, and freedom of enterprise. This time, decentralisation is a tool for the promotion of the development of rural populations, rather than a more ambiguous national good. In fact, the documentation coming out of the National Conference repeatedly underscores concern for the development of rural areas, and explicitly links this development to decentralisation. The interim President during the Conference, Abdoulaye Toumani Toure (ATT) gave a ?State of the Nation? report claiming that ?the economic and social crisis which Mali has known for the last thirty years has been felt most heavily by the rural populations.? Furthermore, he advocates a ?policy of decentralisation? to ameliorate the situation (Diarrah, 1991, translated by Rawson in Bingen). Decentralisation reforms swept across West Africa in the 1990s, but the reform in Mali has been unique. It has demonstrated a strong political decentralisation, rather than approaching it as a technical exercise in administrative reform to please donors (SNV, 2004). Emphasis has been placed 49 Caitlin Blaser 1 September 2007 on involving the community in decision making (and to a lesser extent resource mobilisation), and enforcing accountability to the population. Community empowerment has been a key goal of this framing of decentralisation, and in practice has been coupled with emphasising government responsiveness to the community, downward accountability, and transparency (Balogun, 2000; Sharma, 2000). Recent critics of this decentralisation literature have called into question whether or not locally accountable representation necessarily means equitable representation in heterogeneous communities (Nejhuis, 2003). Basically, in an effort to apply democracy to rural areas lacking a history of democratic participation, there is a risk that power will be given to the majority (desirable in a place where the entire rural majority has a history of disempowerment), but without adequate considerations for minority protection. Even more dangerously, power could be given to the few speaking for, but not necessarily acting in the best interests of the majority. Somebody may be empowered by democratic decentralisation, but it is likely not the heterogeneous community as a whole. While decentralisation was central to Mali's development agenda, it was simultaneously brought to the country to respond to a different need. It was seen by many African countries as a political tool capable of enhancing political stability and strengthening democratic institutions at the grassroots level in countries facing instability from separatist movements or ethnic tensions. Colonialism left many countries with a legacy of politicised ethnicity, and while there was an initial optimism that socialism could promote unity, the hopes quickly faded. Years later, decentralisation was seen as a way to promote unity in the face of occasionally hostile diversity. Giving local populations more autonomy was an attractive way of avoiding any political tension ethnic diversity could cause (Ribot, 2002). Since independence, the majority of Mali has enjoyed a remarkably low level of ethnic tension. The one exception to this rule is the Tuareg separatist movement in the North of the country. Sharing a different history from the sub-Saharan citizens of Mali, and lacking the access to services sedentary citizens in the South enjoyed, Tuaregs had long opposed the central government, and in 1990s, fighting broke out in the north of the country. In 1992 a peace deal was brokered that was widely heralded as brilliant diplomacy from both the government and Tuareg separatists. The peace deal was based around allowing greater autonomy to the north in a large part through the implementation of decentralisation that had been provided for in theory before, but never implemented. This peace agreement began concretely defining elements of decentralisation. While the actual devolution of authority was slowed by those with vested interests in the south, the agreement makes the ?regional autonomy? articulation of decentralisation crucial to Mali, as it was one of the driving forces behind 50 Caitlin Blaser 1 September 2007 implementing decentralisation in practice, and is the basis for much of the Mali-specific decentralisation literature and policy. This articulation of decentralisation is somewhat less reactionary than those previously discussed. In Mali, it has been applied to solve a problem of regional tensions, while previous articulations simply applied decentralisation to solve problems caused by a highly centralised system of government. This articulation introduces more democratically oriented goals of decentralisation, stemming from allowing local authorities powers of resource distribution. This connects decentralisation to the government's responsiveness to the community, collective grassroots action, and the nature of the relationship between citizens and government officials. It introduces the idea of political participation and ownership of political power as a right; this rights based approach was conspicuously absent from discussions of governance in Mali before the 1990s. The current decentralisation process in Mali, while stemming from this regional autonomy, rights based root, is now responding to a different political context. Providing a higher level of autonomy for the troubled northern region through decentralisation was seen as an overall win-win situation. Implementing this strategy was stalled by the fact that it implied the implementation of decentralisation in the rest of the country as well. In the more densely populated, resource-rich south, there were far more vested interests in holding onto power by upwardly accountable authorities. When decentralisation's focus shifted from peace-making in the North of the country to economic and social development in the South, the architects were faced with a far more complex task than they initially imagined. The five years from 1994-1999 were spent eliciting public commitment to the process, drafting legislation, collecting data, conducting outreach programs, and culminating finally in local elections. The first steps of implementation, in the context of the Pacte National with the north, were taken as defining the initial parameters of decentralised governance. It was problematic, however, in that it came with precise timelines that were not matched by clear legislation. The result was that concrete definition and implementation of the guarantees had to await yet another round of debate. The National Council was clear about its commitment to genuine democratic decentralisation, but that is as far as the clarity went - seven years of legislating began to offer substance to the Constitution's stated decentralisation. Slowly, laws have been filling out a practical framework of decentralisation. The first National Assembly bill established that ?the Region, the District of Bamako, the Cercle and the urban and rural commune are endowed with a corporate personality and financial autonomy (quoted in Rawson, in Bingen 1999).? Later legislation determined that collectives have a 51 Caitlin Blaser 1 September 2007 set of responsibilities for the conception, programming, and implementation of social, economic and cultural development in a manner that will be established by subsequent law. However, State oversight when necessary and appropriate was also provided for, meaning that tutelle, while envisioned as more limited in scope than before, has not been eliminated. Most importantly, it was legislated that the state is to pass down the resources, means, and assets necessary for the exercise of those responsibilities. The first step in this olympian process of implementation was to define the decentralised territories, now called communes. Beginning in 1993, the government began an ambitious plan to wipe the cercles and arrondissements from the map, since they were linked to the predatory regimes of the past, and entirely redraw the administrative lines in the country. Village councils including traditional leaders, women?s groups, heads of associations and other stakeholders were called on to form their own political associations with surrounding villages based on various criteria presented by the decentralisation mission, such as collective will, shared resources, geographic cohesion, and most importantly, economic viability. The eventual result of this process was the creation of 703 communes. These communes were to be governed by elected mayors and municipal councils, which were voted on and therefore effectively established between 1998 and 2000. However, much work remains in consolidating these communes as effective political entities, and they lack capacity in almost every sphere of functionality. One donor report says ?these new local government structures face an uphill task in learning how to carry out their duties, clarify their mandates, and find resources,? (Toulmin, 2000). Subsequent legislation has gone on to make substantial progress in defining how decentralised structures will operate in Mali. While a comprehensive review of this legislation is beyond the scope of this paper,12 an outline of the most central structures will be given to provide contextualisation for understanding the functioning of the commune of Loulouni. The Haut Conseil des Collectivites Territoriales is the highest authority of decentralisation, and is responsible both for representing the interests of regions and communes in the decentralisation process at a national level, and offering advice on the decentralisation process to regional and communal authorities. A commune is primarily run by a mayor and communal council, all of whom are directly elected by universal suffrage. The size of the council is dependent on the population of the commune, but ranges from 10-40 people. The mayor can appoint a certain number of aides (2-5, again depending on the population), subject to the approval of the council. The mayor and council are jointly responsible for most political and administrative activities in the commune, included developing the 12 But available through the Mission de Decentralisation, 1997, updated 2005. 52 Caitlin Blaser 1 September 2007 budget, maintaining the registry, collecting taxes, and providing services. The sous prefet is the representative of the central government at the commune level. He is appointed and upwardly accountable. Before 2000, he performed virtually all the duties now belonging to the mayor and council, and the extent of his involvement now varies substantially by commune. With the responsibility for most activities resting now with the council, the primary role of sous prefets now is that of tutelle, which usually manifests itself either as an advisory role, offering assistance with various procedures, or supervisory, ensuring that the mayor and council are operating within the law. Separate bodies of tutelle are specific to certain activities; for example, a regional treasury is responsible for oversight of communal budgets. The decentralisation of public services is one of the most concrete and crucial changes, as service provision in rural areas goes to the heart of decentralisation's development goals. Unsurprisingly, different sectors have achieved very different levels of success implementing decentralisation due to the different nature of the service being provided, as well as the different ministerial and budgetary structures of the service sector before decentralisation. Generally, the water sector has made a great deal of progress in transferring powers, while education and health have both come up against several obstacles decentralising. Service provision will be discussed at more length in a later section as it relates to Loulouni specifically. Finally, it is often argued that ?decentralisation of territorial authority is fundamentally a question of land (Rawson in Bingen, 1999).? At the heart of the issues of decentralising land is natural resource management policy. All land belonged to the state both under colonial law and the socialist regimes that followed, and it is clear that with decentralisation, this demands redefinition. The debate about the form new natural resource management structures will take is dominating the current debate on decentralisation, and is far from legislative clarity. The extent to which decentralisation has not yet tackled the critical domain of natural resource management is clear in an SNV report on its activities to support the decentralisation process, particularly in the domain of natural resource management. According to the law, the central State will transfer some of its domains to the local government authorities. No precise plan has yet been worked out, as questions relating to exactly what is to be transferred, at what level, at what pace and on what conditions have still not been settled. There is no clear vision or shared position amongst the main stakeholders, whether technical departments, local government authorities or village structures about the process of transferring and managing this 'natural heritage.' Many different interests, forces and powers coexist in the rural environment and controlling access to land and natural resources is vitally important (SNV, 2004). 53 Caitlin Blaser 1 September 2007 While the National Conference was provoked by deeply rooted civil society uprisings, external and international factors in decentralisation still remain very important to the process, and this may be most clear in the debate on natural resource management. Since the National Conference, a substantial amount of decentralisation related activities have been led by the ?international development industry? featuring bilateral and multilateral aid agencies and NGOs. Natural resource management has been a focus of international donor involvement, in part because even before decentralisation, rural resource mobilisation has been seen as a key to development in Mali, and as such, many international donor activities had already been interacting with rural extension or other aspects relating to natural resource management. It is also an area that is particularly contentious in the context of political decentralisation, since lack of capacity or interest on the part of past regimes has meant that most land has been de facto under local customary control, subject to various combinations of traditional and religious law. Because of this, any authority local governments are granted on these matters are seen as an appropriation of authority by the government rather than a transfer of authority from the central state to citizens via elected local authorities. At the same time, it is clear that commune authorities must have a clearly defined role in natural resource management, both because it is central to the economic life of the commune and because it defines so many other relationships and resources over which the local government does have authority. Finally, control over natural resources in rural Mali amounts to economic control, making a focus on the decentralisation of natural resource management central to political transformation. As these changes are underpinned by a human rights discourse that emphasises individual economic rights, natural resource management is seen as the key to rural empowerment, and inexorably linked to other democratic political and social changes. Many questions about the implementation of decentralisation have been answered by legislation, and thus far, legislation has underscored the commitment of the government fully to implement true political decentralisation. However, it has not yet filled the gaps necessary for effective decentralised governance to be practiced on the ground. Remaining questions are primarily in two categories. First is the extent to which systems of tutelle could undermine political decentralisation. Real powers have been vested in local elected bodies, but at the same time legislation provides for oversight by State authorities in several important ways. While instituting a system of checks and balances to ensure that local governments cannot act with impunity is desirable, it is still unclear whether the institutions being set up in Mali will serve that function, or will be abused by central state authorities to reclaim decision making powers theoretically belonging to 54 Caitlin Blaser 1 September 2007 local governments. Secondly, exactly which ?resources, means and assets? will be passed down to whom is still very much a matter of discussion, and has tremendous implications for the shape and capacity of decentralised governance. MATCL (Ministrere de l'administration territoriale et des collectivites locales) set up a working group in 2000 to answer questions about what resources should be devolved to whom, and how. Some decisions taken by the working group were aggressively pro-decentralisation, such as the proposition that the State not only transfer financial resources to communes, but also staff and equipment. When it came to specifics, however, it is clear that the door has been left open for abuse by the State. The working group concluded that because it was difficult to assess precisely what amounts needed to be transferred to which levels, each department should be asked to indicate the share of its budget it was ready to put towards appropriation by local authorities. The mayor of Loulouni pointed out that that was ?akin to asking a cow to indicate which one of its legs should go into the farmer's stew (Interview, 11 June, 2006),? and makes it unsurprising that local governments are still waiting for promised resources. As a final note, lest these laborious details of the Malian political system seem irrelevant to broader issues of governance and political transformation, a passage from Whitaker (2002, p340) describes the political and economic situation in which Tanzania was a refugee hosting community. The refugees arrived in Tanzania during a period of transition. After nearly thirty years of one-party state socialism, the government was moving hesitantly along a path of liberalisation that included a shift toward capitalist development and the adoption of a multiparty system. The implementation of an economic reform package since 1986 was finally producing results, with price incentives shifting in favour of agriculture and per capita income once again increasing. Still, adjustment carried with it inevitable difficulties, including a decline in government investment in social services, growth that favoured some Tanzanians more than others, and increasing dependence on assistance from foreign donors (Barkan 1994). In western Tanzania, as it will appear, these patterns were to become further complicated by the massive influx of refugees and relief resources. Replacing 'Tanzania' with 'Mali' leaves us with a surprisingly similar account of the political context in Mali in 2001, when the Ivorian refugees arrived. The lesson here is two fold. First, that refugees are not coming from a complex political conflict and walking into a void, but rather entering a community that also has a history, with all of the role players and power relations that that provides. Secondly, the global economic system is characterised by certain political and economic dynamics that cannot be divorced from refugee creating environments, and this is often experienced similarly by different countries. Much of sub Saharan Africa has a similar history within this economic system, and as a result, the interactions between forced migration and state transformation share many 55 Caitlin Blaser 1 September 2007 features. This allows for a case study that initially seems limited in scope to find resonance beyond its specific geographical area. 56 Caitlin Blaser 1 September 2007 Actors This historical decentralisation context has converged with a history of migration and the recent refugee influx to create the situation we now see in Loulouni. Well before the arrival of refugees, the political landscape in Loulouni was already a complex one, with many actors exercising often overlapping powers based on different histories of access to resources, legitimacy, etc. The mayor and sous prefet and their respective roles have been discussed in the previous section. However, there are also traditional leaders, women's associations, village associations, technical line ministry officials, NGOs, and the private sector, among others. While Loulouni as a whole does not have a democratic history, some of these institutions function in democratic ways, which affects the creation of accountability structures. Understanding the mandates of all of these actors, as well as how they relate to each other, and to local government is necessary both to see why the decentralisation process is taking place the way it is, and to see how interaction with the refugee community occurs. All of these actors have some form of political power, but they differ substantially in terms of legitimacy, degree of democratic decision making, mandate, etc. Traditional leaders are a particularly interesting case in the Malian political landscape, as they have been moved back and forth from the formal to the informal by different governing powers. In many ways, they epitomise the ambiguity of roles that characterises the decentralised political landscape in Mali today. Under colonialism, village chiefs were formally recognised, very much at the exclusion of other authority figures, such as religious leaders, griots, hunters, etc. However, this was mostly done out of the hope that they would be seen by the population as 'legitimate,' therefore contributing to the legitimacy of the colonial regime, all while providing important tax collection and other services. In terms of real decision making power, they were at various times given a wide reign, to prevent the colonial authorities from needing to deal with ?local issues,? and at other times undermined, such as when forced labour became a central economic policy. The affect of colonialism on traditional authority is a well documented phenomenon, and fundamentally changed power dynamics in rural Mali. The result, as Ole Therkildsen (1993, p87) says, is that ?traditional political authorities have often been viewed as the extended arm of the state in the locality, and usually regarded as inefficient, corrupt, undemocratic and excluding of women.? Another result is that it very much limited any 'staticness' that existed in authority roles in the past. First of all, the French often recognised as ?chiefs? individuals with no local legitimacy. Cooks in the colonial regimes, or even escaped slaves could be made chiefs over villages they had never 57 Caitlin Blaser 1 September 2007 visited. A sign of the extent to which power roles had been muddled was visible in the 1950s, when a group of village chiefs declared themselves to be a trade union (Cooper, 1996). The Keita and subsequent Traore regimes excluded chiefs from their political structures, creating their own parallel village level political institutions. However, village chiefs continued to be the principle authority in most villages, and their relationship with these new institutions varied tremendously. Under decentralisation, village chiefs continue to have a rather ambiguous relationship to formal government structures, and the village level political organisations created under Keita and Traore also continue to operate with some unclear level of formalness. Throughout the decentralisation initiative, as with other political transitions in Mali, there has been a tendency to create new political structures rather than transform previously existing ones. For example, under previous socialist regimes, education initiatives at the village level were delegated to the Parent Teacher's Association (PTA). Widespread dissatisfaction with this has led to the creation of another organisation ? the School Management Committee (SMC). This was created by the Pedagogical Advisory Commission, a cercle level duplicate of another education organisation created by Traore. None of the previous organisations were disbanded. Instead, yet another organisation was created to build partnerships and cooperation between the PTA and the SMC. Grauze (2004) points out that while their mandates may be clearly separated on paper, in reality this is not the case. This has led to immobilising power struggles between the two organisations in the case of Loulouni, where neither function. In the case of Sieouba, a nearby village, membership in both organisations was the same, causing frustration, as the amount of time spent in meetings with extension agents has effectively doubled with no change in capacity. Similar situations characterise other technical line ministries. Occasionally these administrative changes are well thought through on paper, but their implementation is a completely different story. There is now a formalised structure for the selection of village chiefs ? they are selected from a village council, which is elected by universal suffrage for a period of five years. The list of candidates, however, is selected by an appointed state administrator at the cercle level (Republique du Mali, 1995). Ribot (2002) argues that this leaves village chiefs largely as appointees of the central government. This legislation seems to have confused more than it clarified. Not a single village in the commune of Loulouni had actually carried out such elections, and every chief was expected to hold the role for his (and it always was a man) entire life. However, due to extension agents and the commune council having explained the election process, many villages were left not without a chief, but with more than one. For example, one chief would be the man nominated by the village association, another would be the customary chief (son or nephew of 58 Caitlin Blaser 1 September 2007 the previous chief), and yet a third would be the oldest man in the village. As a result, it is largely left to the discretion of the mayor whether or not he chooses to recognise the village chief as a legitimate, formal political authority or not; even if he chooses to, it is largely only for advisory purposes. However, international development agencies doing work ?in the field? nearly always work through village chiefs to establish ties at the village level. Essentially, leaders at the village level in many spheres in Mali today largely have legitimacy based only on the merit of the individual. They are not, in any meaningful political sense, representative of the local population, although they may act in their interests. There has been no meaningful bureaucratisation of the position they hold, although there are various flexible norms about different leadership positions. They are largely seen by central government workers as a hindrance to decentralisation, because many of the powers vested in the mayor, particularly regarding natural resource management, were formerly the domain of these village level authorities. However, because of their continued influence and power at the village level, no actors are willing to dismiss working with them. Ribot (2002, p25) sums up the ambiguity of the role of village chiefs by pointing out that vesting powers in Malian village chiefs could be seen simultaneously as centralisation, if they are to viewed as an auxiliary of the state based on the electoral process, decentralisation, if they are viewed as acting in the interests of the local population as an elected official, or privatisation, if they are seen as an independent local actor. Civil society organisations, however, are seen as a pillar of decentralisation, having the capacity and interest to hold elected authorities to accountability, as well as to speak for large segments of the village population. John Davis (in Bingen, 1999) took various indicators of democratic procedures and applied them to civil society organisations in Mali, including village associations, women's groups, and organisations of hunters. These organisations are seen by the international donor community as the building blocks for democracy. For example, USAID, the leading funder in the democratic governance branch of the decentralisation process, had the promotion of capacity building among these organisations as the sole focus of its program in the year Davis wrote the article, and it has remained a central focus of the program since. For each organisation, Davis looked at factors such as whether membership was ascriptive or mandatory, whether leadership selection was inclusive and systematic, whether decisions were autonomous from outside control, etc., and then looked not only at whether these democratic structures were in place, but whether democratic values were fostered among members and leaders by these structures. What he found is that civil society organisations are not necessarily, as Tocqueville proposed, ?classrooms of democracy.? In fact, he did not find a high level of inclusiveness in decision making or 59 Caitlin Blaser 1 September 2007 leadership selection in a single one of the associations. While this will come as no surprise to those who have long said NGOs do not necessarily work in democratic or accountable ways,13 it does demand questioning the validity of including these organisations as central role players in newly democratic institutions. Davis argues that civil society is growing rapidly in the direction of accountability and inclusiveness, meaning the continued support for civil society institutions is not unwarranted. He continues to argue, however, that because the rapid evolution of civil society institutions towards democratic practices is not yet complete, they are currently unable to play the role of a counter-weight to the state. Another factor worth mentioning for an understanding of the political actors at the local level is that they are not mutually exclusive. While different organisations tend to attract individuals with different profiles, it is not unusual for one individual to fulfil many roles in different organisations simultaneously. Particularly due to widespread illiteracy, it would not be unusual to find a well educated member of the community be the secretary of the village association, the president of the school board, a member of the commune council, and the vice-president of the youth association. Elected officials also tend to maintain their functions in other organisations, be they with associations, private businesses, etc. It was admitted in an interview (anonymous, 8 April, 2006) that: sometimes, it causes problems having people in the commune council who also have allegiances in other organisations. Now it is not so serious, because we rarely have funding for tenders. But when we do, it is difficult. First, we all say that we can separate these things from what is best from the commune. But can we all the time? And the second problem is what people will say. Even if we act in the best way possible, not everyone will see it that way. There are always people who are looking for the worst. But we can't ask people to drop everything because they have been elected. The commune would suffer. First of all, it is not always possible to replace these people. They are leaders, and not just anyone can do the work they do. And secondly, we don't have such a big budget. Sometimes councillors have to go for months without being paid. We cannot tell them not to cultivate with the village, or to quit finding small ways to support their families. It is to this complex and ambiguous landscape that people fled from the fighting in Cote d'Ivoire. In relief efforts, there are numerous actors working with significantly different mandates, goals, capacities, etc., and Loulouni was no exception. While causing damage to the host population is unlikely to be a goal of any of these organisations, ensuring that this does not happen while carrying out their primary functions is also unlikely to be within their capacity and scope. There are countless stories of problems caused within host populations because resources are provided to 13 For a discussion on the accountability challenges faced by NGOs, visit the website of the Accountability charter at www.iangocharter.org 60 Caitlin Blaser 1 September 2007 refugee communities while hosts are neglected due to resource constraints (Kibreab, 1996; Chambers, 1993). Relief related actors, like those already in Loulouni, are involved in complex relationships and do not always have clearly defined mandates. The crisis committees at the various territorial levels, for example, were coordination mechanisms to bring all relevant actors together. At the local level, however, the crisis committee acted as a decision making body with the sous prefet having exclusive decision making powers. With all other actors simply playing advisory roles, the crisis committee can be seen as an upwardly accountable extension of the central government. Relief organisations such as the Red Cross and UNICEF played an advisory role on the crisis committee and served as independent service providers, but were constrained by mandates that would not let them act without approval from both their central offices and the central government (or at least the regional crisis committee). The camp manager was a member of the crisis committee, but also acted with a great deal of autonomy. While he did not have the mandate to partner directly with service providers without the approval of government officials, this was largely a formality on the local level, as the sous prefet's mandate regarding the relief effort was essentially to support the camp manager in all ways. This kind of circular system of mandates and power is one that decentralisation theorists warn against as an easy way to short-cut democratic processes (Ribot, 1999, Prud'homme, 2000). While this was not necessarily caused by the influx of refugees, the increase of resources and diversity of new powers that accompanied them certainly created the environment that facilitated the opportuntiy for such a system to develop. This chapter has provided many contextual factors shaping the decentralisation process in Loulouni, and introduced the ways in which migration could be shaping both the context and the process of decentralisation itself. According to Manor's criteria of the contextual factors helpful to the decentralisation process, Loulouni's demographic, geographic, and economic landscape did not make it a particularly promising candidate. The limited communication and service development infrastructure is certainly a limiting factor by Manor's international criteria, as is the lack of democratic experience. However, putting Loulouni in the context of Mali makes the picture significantly more promising. While Mali has not uniformly reaped the benefits promised by decentralisation, on the whole the transformation in governance has demonstrated many successes, and Loulouni has more contextual indicators of success than many other parts of the country. While discussing the potential for decentralisation, special mention should be made of accountability structures, for two reasons. First, many decentralisation theorists (Manor, 1998; 61 Caitlin Blaser 1 September 2007 Fageut, 1999, Alfonso, 1999) have emphasised clear and effective accountability mechanisms as the single most important factor in determining the success of decentralisation. Additionally, this study has found that this is the area migration, and the refugee influx in particular, has had the most impact. As mentioned at the beginning of this chapter, the history of labour migration between southern Mali and Cote d'Ivoire played a central role in shaping the relationship between Malian citizens, Ivorian citizens, and local government authorities in Mali. The system of taxation and forced labor created a high level of vertical integration in the governance structures, particularly for economic matters in the form of the unite de caisse system that decentralisation has struggled to change. While political decentralisation has attempted to redistribute powers, economic centralisation has pulled against meaningful change. In certain circumstances as discussed in this chapter, it has simply meant the creation of parallel structures at a local level. Through a delicate period of negotiation over the past five years, local officials have worked out a balance of powers in this complicated and occasionaly contradictory environment. A further consequence of Mali's history of state oppression is the precedent of an exploitative relationship between citizens and local government authorities. While decentralisation aims to change this relationship, many of civil society's responses to the oppressive system have remained in place, further complicating accountability structures. For example, systematic tax evasion as a form of protest led to the creation of a parallel, privitised and highly formalised village level system of money sharing for public service work. Because a great deal of service provision in Mali has been in the hands of civil society, decentralisation has struggled with the role of civil society. On one hand, there is a push to rebuild the traditional accountability structures of civil society holding local government accountable for public service provision. However, this same push to strengthen civil society has undermined the very accountability structures it aims to reinforce. This is because efforts that support civil society simultaneously support the service provision for which civil society claimed responsibility. This undermines support for the form of decentralisation that would have civil society holding local government to account for service provision. This study has found that the introduction of the refugee camp in Loulouni changed the scene in which decentralisation unrolled in a number of ways. The argument will be outlined here in some detail, and then expanded on and substantiated in the subsequent chapters. On one hand, the arrival of refugees helped to rebuild the relationship of trust between civil society and local government authorities by reestablishing, to some degree, accountability structures. This happened because the requirements of civil society changed with the introduction of resources from humanitarian agencies, from resources for service provision to conflict management. Before the 62 Caitlin Blaser 1 September 2007 influx, citizens were demanding the provision of infrastructure and services from local government officals, who did not have the resources to deliver. Citizens were not paying taxes, and much service delivery was being done by the voters directly. This was seriously undermining the legitimacy of the decentralisation process. When refugees arrived, and humanitarian organisations supplied certain resources and services, the demands of the citizens of Loulouni suddenly changed. Rather than demanding local government supply goods and services, residences were demanding that the government negotiate their access to the goods and services brought by the humanitarian agencies. Local government was better prepared to provide conflict management services than resources, due to the resource limitations described earlier in the chapter. While the arrival of the refugees helped build the relationship between the citizens and local government authorities, it aggrevated conflicts between different local government officials. Just as the redistribution of powers between the mayor and the sous prefet had been painstakingly hammered out, the arrival of refugees introduced new resources, and with them came new powers and responsibilities. The poorly systematised allocation of these new powers undermined the shift from upwardly to downwardly accountable officials. This argument will be developed and substantiated in the upcoming chapters which will discuss the process of decentralisation in Loulouni in detail. 63 Caitlin Blaser 1 September 2007 DECENTRALISATION IS OFTEN TALKED OF, DESIGNED AND EVEN IMPLEMENTED AS A KIND OF THEATRICAL PERFORMANCE FOR DONORS, PRODUCING AN ELABORATE IMAGE BUT NOT A CHANGE IN POWER DISTRIBUTION. SUCH PERFORMANCES CAN INVOLVE, FOR EXAMPLE, DOWNWARD TRANSFERS THROUGH ONE NEW LAW BUT RECENTRALISATION THROUGH ANOTHER, AS IN THE CASE OF MALI, WHERE THE TRANSFER IS WELL CRAFTED BUT NEW STRUCTURES OF TUTELLE ARE BEING BUILT THAT COULD RETURN TO CENTRAL AUTHORITIES CONTROL OVER APPROVAL OF EVERY DECISION. RIBOT, 2002 IMPLEMENTING DECENTRALISED LOCAL GOVERNANCE: A TREACHEROUS ROAD WITH POTHOLES, DETOURS AND ROAD CLOSURES SHAH AND THOMPSON, 2004 3. Decentralisation in Loulouni: Presenting the Data The decentralisation process in Loulouni shares many of the successes and challenges of the decentralisation process nationally. For example, the lack of clear legislation regarding the decentralisation of resources remains a roadblock to effective local governance in Loulouni. However, there are certain trends in Loulouni that are clearly not in line with the national decentralisation process. This chapter will particularly pick up on these unique aspects. Loulouni does have specific economic and geographic characteristics, personal dynamics between local officials, etc. that are important in shaping the decentralisation process. These will be accounted for in the following chapter. In situations where neither these characteristics nor national trends explain how decentralisation is unfolding, the refugee influx is an additional feature that could be influencing the process. This will be explored in more detail in coming chapters. The first section of this chapter will describe local governance structures in Loulouni, and the relationship the community has to them. The existence of various public offices was explained generally in the last section, and this information will now be concretised, by looking specifically at which powers belong to which offices, how they are exercised in practice, and the implications this has for the residents of Loulouni. Residents of Loulouni are overwhelmingly positive about the relationship they have with the mayor and commune council, a change heralded by decentralisation. However, they are less positive about the capacity of this governing body to act as an agent of development. They also remain sceptical about the ability or willingness of the sous prefet to act in their interests. The next section looks specifically at the accountability and transparency of local government. It is clear that these concepts still lack saliency in Loulouni. Respondents clearly feel the mayor's office acts with a degree of transparency that is lacking from the sous prefet, but perhaps more to the 64 Caitlin Blaser 1 September 2007 point, many people problematise the idea of accountability in a place without democratic traditions or a well educated population. The fiscal elements of decentralisation are then taken up in Loulouni. Most notably, important fiscal considerations, like the payment of taxes by citizens, are an indication of the successes and failures of decentralisation in Loulouni. Some people cite the government's lack of capacity to deliver services as the reason for their nonpayment of taxes; the government, of course, cites the nonpayment of taxes as the reason for their lack of capacity to deliver services. It is here that both the national and historical decentralisation processes are felt most clearly in Loulouni. The final section of this chapter deals with service provision. Drawing from both empirical evidence and interviews with employees of the relevant line ministries, this section looks at the extent to which decentralisation has been achieved in each service sector, and why. Sectors have achieved widely differing levels of decentralisation, largely due to the functioning of national-level line ministries. Understanding this process, however, will be important for the discussion in the following chapter about access to services and allocation of public resources in a refugee context. 65 Caitlin Blaser 1 September 2007 Local government and governance structures Decentralisation has featured prominently in Loulouni's development plan since 1999. Loulouni was chosen as a pilot commune for decentralisation in Sikasso, and as such, held mayoral and council elections in 1999 (whereas other communes observed the elections in Loulouni and other pilot communes, and held their own in 2000). Many concrete steps have been taken to implement the national decentralisation strategy, particularly on an administrative level. Since the mayor and communal council have been installed, many powers have been devolved to them. The responsibility for service provision and planning (for example, formulating the budget and plan de developpement) as well as tax collection have been passed from the sous prefet's office to that of the mayor. Registry functions are shared between the two offices, with the mayor issueing ID cards and birth certificates, while the sous prefet issueing marriage licences and various permits, such as motorcycle registrations, and also holding the commune's archives. This localisation of registry functions was widely applauded by those surveyed. It was mentioned by sixty-eight respondents as a positive element of decentralisation, both by supporters of decentralisation and those who had an overall negative view of the process. Before decentralisation, most routine registry functions had to be done in Sikasso. Of the survey respondents, only sixty percent had ever been to Sikasso before, indicating that it is quite a journey for routine activities such as renewing an expired ID card, or registering a marriage. As one chauffeur explained in an informal discussion, Before if I wanted to register my car, it had to be done in Bamako. And the registration would expire every year. I can't go to Bamako every year, and if my registration is outdated I have to pay double at every checkpoint. Now, I can register it in Sikasso. It is still expensive to keep the registration up to date, but at least it is possible. Before it was impossible. Since 1999, there have been two new sous prefets and one mayoral election, held in 2005. While initially there was a great deal of tension between the mayor and the sous prefet, largely attributed to a lack of clarity of roles, with each individual turnover, this has decreased. Now both the mayor and sous prefet report a generally cooperative relationship between them. While they both indicate a current clarity of roles in carrying out their day-to-day tasks, they also both expressed an uncertainty in how future legislative developments in Bamako will change this. Just a small amount of reading between the lines, however, indicates that this lack of articulation of tension does not mean that the tension is entirely absent. Numerous trainings conducted by NGOs and the government for mayors and sous prefets have addressed a naturally contentious relationship, and 66 Caitlin Blaser 1 September 2007 there has been a great deal of emphasis on reframing the relationship as a cooperative one. The sous prefet largely articulates his role in the community as assisting and advising the mayor and council in their work, creating linkages between them and ministerial authorities, supporting their human resource development, etc. He has articulated this practically, however, by refusing to endorse the commune's plan de developpement for years due to various technical flaws, not allowing a transfer of the archives to the office of the mayor citing inadequate staff capacity to properly maintain the documents, and other such actions that suggest the reported relationship of cooperation and support, while not entirely untrue, is an incomplete description of the dynamics at hand. Measuring the extent to which the political aspects of decentralisation's goals have been achieved in Loulouni is a difficult task. There is both too much and too little information available about the current status of decentralisation in Mali. With a dizzying myriad of NGOs and other organisations involved in support for decentralisation, there is no lack of status reports. However, these are often arbitrarily limited in scope or focus depending on the goals of the organisation in question, and use very different frameworks for analysis and even different vocabulary, making meaningful comparisons virtually impossible. Many of them also use dubious methodologies themselves. In one Management Systems International report (2004), which was contracted to review all USAID country activities, 64% of communes with USAID sponsored governance activities are reported to have included community based organisations in their budgeting processes, and yet only 20% of the same communes are reported to have sought any kind of community participation in the very same processes. The only information offered by methodological explanation of this seeming paradox is that ?one possible explanation for the 44 point spread is criteria and measurement (p83).? As discussed in the discussion on methodology, this study has not had the luxury of a comparative case, or a long enough span of time to measure changes over time. However, through mixed methods including archival information, a survey with a relatively large, random sample, interviews with a wide range of community leaders, some of whom were able to comment on changes in Loulouni over the years, it has been possible both to differentiate between administrative and political implementation of decentralisation, and examine the extent to which both are genuinely taking place and accomplishing their goals. Most indications show that decentralisation is achieving success in increasing local government's responsiveness to the needs of the community. Lest this sound congratulatory, it must be noted that the starting point was quite dismal. As mentioned above, before decentralisation, Mali was not a democracy, and the only government representatives at a local level were sous prefets brought in 67 Caitlin Blaser 1 September 2007 from far corners of the country with no incentive to listen to the needs of the community, let along act on them. The introduction of the position of a mayor, who is a member of the community elected by the commune and is downwardly accountable, has changed the situation entirely. One of the most important factors noted by citizens surveyed about decentralisation is their change in relationship with local government authorities. Before decentralisation, nearly all (86) respondents characterised this relationship as one of intimidation and fear. Now, they overwhelmingly cite open communication. When asked whether they felt able to talk to the mayor about issues facing the community, ninety- nine out of 100 respondents answered yes. The 100th respondent was the mayor's sister, who said that the mayor did not like to talk about work at home. As one respondent says, Now anyone can talk to government officials. I can explain to them my problems, even though I'm not a powerful man, and they will listen. They don't help me each time, but at least we can talk to each other. Maybe they will know things that I don't. This exchange of ideas is important. When asked about the situation ten years ago, only two respondents claimed communication between government officials and the community was open. The problem now, it seems, comes with implementation. When asked if the government can respond to the needs of the community, or help people with their problems, the response was less positive. Fifty-one percent said yes, forty-nine said no. Many people talked about the meetings and lack of action. As one respondent said, Of course the mayor listens to us; he lives here and we elected him. But he spends all his time in meetings. He calls us for meetings to learn about the problems of the people in Loulouni, he goes to Sikasso to explain our problems there. That's all. Is that helping us? His work may be important, but I can't eat his meetings yet. One former mayor (interview, 11 April, 2006) explained, People here are poor. They wanted me to give them cattle and fertiliser. I couldn't do that, I didn't have the resources. What I could do was help people to use the resources that they had better, in a way they would get more from. I could bring people together, and by listening to one another, sometimes they could find real benefits. Whether or not that's helping people, I don't know, but it was all I could do. If I did nothing else as mayor, perhaps I built peoples confidence in the development they can achieve themselves. Interaction between the community and local government officials in itself is an explicit goal of decentralisation, and there is little doubt that this has been a success. Again, this is starting from a baseline near zero. Before democratisation, meetings with local government representatives were rarely held. In fact, the representative was rarely there. Interactions with the mayor, on the other hand, are overwhelmingly reported by all groups of respondents to be strong. A wide variety of people take advantage of the ability to interact with local government officials. 68 Caitlin Blaser 1 September 2007 Interaction, while linked to economic status, does not seem to be exclusively a tool of the powerful. While most ?key role players? were men, and exclusively wealthy ones at that, women were more likely than men to be ?very engaged? in local politics, and engagement of women was less linked to economic status. Registers show women attend meetings at the mayor's office slightly more often than men. There was slight but significant correlation to age as well, with the extremely old and young slightly less likely to be involved in politics than the middle aged. Ethnicity showed no correlation at all to interaction with local government. Not a single respondent who did not interact with the local government claimed either a lack of opportunity or the threat of a negative response from the government officials; the only reasons cited were either a lack of utility in the interactions, or a lack of interest. It is telling, however, that many respondents who did not interact with local government cited the lack of resources or ability to contribute to development. Many expressed very clearly how the lack of capacity for development in the mayor's office was undermining their buy-in to the decentralisation process. One respondent (14) illustrated this very well, saying I have a house by the river, and for the last three years, the rains have flooded it. There are about fifteen of us in this position. Some of them went to the mayor for help. After months of pleading their case, the government officials in Sikasso sent some money to help them rebuild. They did not see one cent of this.14 I went to the hunters. I knew the community would listen to them, and I know that the people in Loulouni are good; if I ask them for help, they listen. The first week of the dry season, I had a new house on the hill above the river. One man had given up some of his land so my house could be built, a group of youth had agreed to make the bricks, the women's association carried the water; that is how development happens here. When I can go to the mayors office and see him ask the community for help, and get the results the hunters got, I can say decentralisation has succeeded. Until then, you won't see me wasting my time at meetings at the mayor's office. He talks to us, yes, but what does he do? When asked about interaction with local government officials before decentralisation, responses were overwhelmingly negative. Ninety respondents never approached local government authorities before decentralisation, and two respondents were jailed for doing so. Most of the rest of the respondents were too young for the question to be relevant. Nearly half of the respondents ? too many to list here ? had specific stories about the predatory nature of the government before decentralisation, and how it affected their lives. The negative views held by the community about local government before decentralisation were so obvious that it is unnecessary to dwell on them. Back in the post-decentralisation era, reviews on interactions with the sous prefet were much 14 This incidence of corruption was cited by 6 different respondents for different reasons; it was a scandal tarnishing the first mayor's time in office. 69 Caitlin Blaser 1 September 2007 less positive than interactions with the mayor. The sous prefet's office does not keep attendance records, but only held two open meetings during the two year period of this research. He was rarely in his office, and often criticised for shirking his responsibilities. As one respondent said, There are a lot of hardworking people here. ACOD,15 they do a tour of Loulouni commune each week. We know the rotation of Red Cross staff coming from Bamako, they do a tour, and the Canadian and American volunteer agencies. Every village in Loulouni commune knows these vehicles. We understand what they do. Loulouni's own sous prefet, if you don't live in the town of Loulouni, if you don't come to the market here, I promise you have never seen him. It's in his contract to travel throughout the commune twice a month. He'll say, who's going to pay for my petrol? His colleagues in Sikasso have money, but in his position, if the money isn't passed down, he does nothing (respondent 2). While many sous prefets are enthusiastic individuals with a great deal of commitment to their post and the development of the country, the nature of their position requires that they be imposed from outside, upwardly accountable, and distant from local decision making institutions. Furthermore, because decision making processes are less participatory, resource allocation runs the risk of being less efficient. As one respondent (16) pointed out The sous prefet means well, but he doesn't understand us. He is like another NGO; many of his projects are a waste. He built a big youth center for the town a few years ago. Anyone from Loulouni knows the youth club hasn't met in years; nobody will use it. What we need are pumps, and a larger health centre. He has good ideas, but they never help us eat. The introduction of the mayor established a position within the local government with both significant executive powers and strong ties to the community. While this has offered a solution to many of the governance challenges caused by the top-down system of government Mali has had the past, political institutions don't change overnight, and the sous prefet and mayor of Loulouni face huge challenges separating their powers and responsibilities. While on most daily issues a balance has been achieved, even slight shifts in available resources or processes to acquire them reignites the debate. The real shift of powers and responsibilities has fundamentally changed the way residents of Loulouni interact with local government officials. While they are overwhelmingly positive about the relationship they have with the mayor and the commune council, the lack of resources in the mayor's office has undermined their trust in his capacity to contribute to development in the commune. While the sous prefet has more administrative support and access to certain resources, they are aware that he is not well positioned to act in their best interests. As a result, there is still no system at the 15 An NGO based in Sikasso that does governance related extension work 70 Caitlin Blaser 1 September 2007 community level to call for accountability of local government officials. Until powers and resources have all been devolved in a transparent, comprehensive way, there is little incentive for such a system to develop. Furthermore, each change to the balance of powers currently established, be it through the introduction of new resources, or new legislation changing the delineation of powers, stalls the process further. 71 Caitlin Blaser 1 September 2007 Accountability and transparency In decentralisation theory, there is a clear link between communication, participation, and accountability and transparency. When citizens are involved in the processes of local government, the logic goes, they are better informed about how resources are spent and also more likely to hold local officials accountable. Conversely, local officials, being accountable through democratic elections, have an incentive to share their challenges with the community and act transparently. In Loulouni, however, there is a marked difference between open interaction with local government, which respondents widely acknowledge exists, and transparency, which is much less widely acknowledged. For the many reasons discussed, pinning down accountability and transparency at the local level is challenging. At the beginning of this study, accountability and transparency were selected from the literature to become two of the prime indicators of the success of decentralisation. However, pilot studies of the survey proved these particularly difficult concepts to operationalise. As stated in the introduction, this study adopted Ribot's (2002) framework of resources, power, and accountability. Establishing the resources and powers in Loulouni was a relatively straightforward process. However, there was a gap found between the rhetoric and practice of accountability. This gap was particularly apparent when soundly designed survey questions designed to pinpoint accountability missed the target completely. Instead, it has been most useful to look at the 'powers' element of decentralisation using a two-pronged approach. The first consideration when looking at power is that of legitimacy. It must be established who has the legitimacy to hold various powers, and where this legitimacy came from ? whether it was through legislation, self-appointment, traditional rights, etc. Then, exactly who exercises these powers must be looked at carefully. Examining how this differs from the powers laid out in decentralisation legislation seems to be the most accurate way of getting at the accountability structures or gaps in place in Loulouni. A substantial amount of this was done in the previous section on local government structures. This section looks more at the impression the community has of these structures, which is also important for the building of accountability. Only eight survey respondents report having tried to find out information from local government officials. All eight were also very engaged in local politics, and reported that this information was easily acquired. Of the ninety-two who had not tried to find information, seventy-eight thought it would be difficult to find. As one respondent put it, It's not that the mayor is trying to hide things, we just don't understand each other. I want to know how my taxes are spent. If I 72 Caitlin Blaser 1 September 2007 asked, he would give me flip charts with lots of numbers, and information about who is responsible for what. That's not what I want. If my neighbourhood doesn't have a pump, I want to know why not. If you don't speak his language, you can't know anything. There are certain transparency requirements put forward by law, and not all of them were adhered to in Loulouni. For example, the commune's planned budget must be posted publicly in the mayor's office, as must the five year development plan for the commune. The budget was not posted, although was available on request (this study was apparently the first time it had been requested), and the five year development plan was 'under negotiation' for the entire two year period of this study. As for accountability, there was a huge gap between the perceptions of the downwardly- accountable mayor, and that of the sous prefet. Only six respondents claimed that the sous prefet was adequately supervised by his superiors. One man said, He is not controlled. He says that he is, because once a month, he goes to Sikasso to meet with his boss. In these meetings, all the powerful men just sit in nice air conditioned rooms, earning big per-diems, talking about how difficult it is to stay in rural areas. They call that control. What can we do? We aren't educated enough to control him. It is the role of the sous prefet to explain the laws and represent the central government in Loulouni. If we understood more, maybe we could hold him accountable, but he is here precisely because we don't understand. It is not a system that works. The situation with the mayor was much less clear. Nobody seemed to know to whom the mayor was supposed to be accountable. Not a single respondent said the mayor was responsible to the electorate, or the people of the commune of Loulouni. However, only thirty-five percent of respondents thought that there was inadequate accountability in the mayor's office. Each of these people had specific stories of corruption. Many people (sixty-two of those surveyed), however, replied that they didn't understand decentralisation well enough to know what accountability meant. To some extent, the confusion inherent in such a fundamental change in government structure could itself be seen as an inhibitor to transparency, and it is strongly compounded by a lack of clear mandates and inadequate education about the process. Interviews shed further light on the matter, when seven people interviewed linked decentralisation to a lack of capacity with implications for accountability. The president of the women's association said: The decentralisation mission came and told us all about decentralisation. Many people liked it. They said, 'it is a chance for us each to have control over our own lives. Nobody can say that is a bad thing.' Well, I say it is a bad thing. There are many things the 73 Caitlin Blaser 1 September 2007 government is supposed to do. It needs to have leadership, and an understanding of the way to solve problems. That is why we have governments, and law, so we are not just running around each man on his own. If decentralisation tells us to solve our own problems, it is not doing the job a government should. It is unfair to the mayor, what is expected of him. He is just a member of the community, who is now told he is a leader, without knowing anything. It is unfair to us, we are also told that we are all our own leaders. Without anyone to follow, how can we know we are moving ahead? So many people think they know, and they can help the mayor move in the right direction. I have been to meetings at the mayor's office, and I can tell you, nobody knows the right direction. Everyone wants to say ?this is right? or ?this is wrong,? but what he really means is ?this is good for me? or ?this is bad for me.? The beginning of the above quotation brings up a point that many extension workers and donor reports focus on. When asked about decentralisation, many citizens start by saying ?the decentralisation mission has taught us that....(Rawson in Bingen, 1999).? While decentralisation is supposed to create a sense of political ownership by civil society, it is still the case that this political ownership is being created by someone other than civil society, which says a good deal about the political environment in which decentralisation is being implemented. As one donor report (USAID, 2002) explains, The PGP16 implementing partners and field agents have encountered difficulties in obtaining the support and sense of ownership that are needed, which are essential to PGP?s success. To develop this ownership, the agents have often conducted an intense campaign on ownership, facilitated the creation of local mechanisms, such as steering and monitoring committees (comit?s de pilotage and comit?s de suivi), in each commune and trained the members in the discharge of their duties. Many of the efforts of decentralisation are being frustrated by the paradox inherent in bringing decentralisation from the central government, and creating grassroots empowerment from above. Accountability is no exception. Decentralisation will continue to fail to provide strong accountability structures if the process is not led by those who must hold the government to account. This is in addition to the challenges of allocating resources and power that were discussed in the previous section. Some of the specifics of these challenges can be seen by looking at the status of fiscal decentralisation. 16 Participatory Governance Program, which is the largest governance element in the larger decentralisation project 74 Caitlin Blaser 1 September 2007 Fiscal Elements In the discussion of resources, powers, and accountability that has taken place so far, the issue of local tax base and budget creation has been briefly touched on, but not explicitly addressed. This section will spell out the financial situation of the commune of Loulouni. This is a particularly important discussion, because it represents, very concrete ways, the convergence of a number of issues at the heart of the decentralisation literature. The ability of the local government to raise taxes can reflect both the capacity for collective action and the participatory nature of the development process at the community level. The creation of the commune budget illustrates the conflict of interest between the sous prefet and the mayor. And finally, the national restrictions on fiscal decentralisation are a manifestation of challenges Mali has had mustering legislative support for a real decentralisation of powers. Budgetary limitations are frequently brought into the discussion by both the office of the sous prefet and that of the mayor, leading us to look at the commune's finances. There is no question that the mayor takes the community's decisions into account when drawing up the annual budget; it is an inclusive process that is almost continuously ongoing. The problem is, the budget assumes 100% collection of taxes, and several times more resources coming in from Bamako than turns out to be the reality. In practice, the budget rarely covers more than the salaries of those working at the mayor's office, with line item loans from different funds maintaining a minimum of activity. In fact, to the extent that payment of taxes is an indication of resource mobilisation, decentralisation has achieved little success in this area. According to the records of the sous prefet, payment of taxes in the commune of Loulouni has actually decreased since the implementation of decentralisation, down from about 40% in the early 1990s to the low 30%s in recent years. Most of what is collected comes from the town of Loulouni. The mayor's secretary (interview, 7 July, 2006) blames a lack of resources. In order to collect taxes from the rural areas, the commune is responsible for paying a tax collector to come from Kadiolo. Tell me, what money are we supposed to use when taxes have not yet been collected? Before, they would send someone to collect the money and go. Either way, there has been no change. We don't see the money. The former mayor (interview, 7 July, 2006), however, blames a lack of understanding of the decentralisation process. Right now, the population is not aware of the situation. There has been an unwillingness to pay taxes, especially since 1992. Just before the 1992 elections, candidates said there would be no more 75 Caitlin Blaser 1 September 2007 taxes. This mentality has entered people's heads. You can say they don't understand democracy, or perhaps you can say they have understood too much. Extensions agents come, and explain that the decision making power rests with the population now. That nobody comes from above and controls their money and resources. This is supposed to be empowering. But now so many people tell the commune, 'this is a democracy, I can keep my money to myself.' There are even sons who are disobeying their fathers, and saying, 'now Mali is a democracy, each person makes his own decisions.' Democracy here means that people think what they want, say what they want, and do what they want.... This is worsened because we don't have any power of enforcement. Even people who do understand that they need to participate in decentralisation for it to succeed, if they don't see a benefit in it for them, they won't adhere do it. Many citizens, however (32% of those surveyed), and farmers in particular, blame a lack of rain. citizens claim that it is not a lack of civic engagement preventing them from paying their taxes, but poverty that, due to climactic changes, has reached new heights. This does seem to be largely supported by statistics; percentage of tax payment closely follows rainfall patterns each year (Association des paysans, Loulouni archives, 2004). Only one respondent put a different explanation on the table. This explanation was also brought up in many informal discussions. In his words, Look, I only have a certain amount of money, and I am going to do what is best for my family. My brothers always pay their taxes. They criticise me a lot, and say I am doing the wrong thing. But I have used my eyes. These people who pay taxes every year, what do they get from it? Nothing. Do we have pumps because we pay our taxes? No. And before an NGO comes in to install a pump, do they look at what village pays taxes to decide where to go? Never. Everyone here pays their taxes, but we have no development to show for it. I'm tired of it; I will keep my money. Every year the mayor comes and says he will put me in jail, and every year I say 'go ahead,' but he never does. When I see the advantage of it, I will happily pay taxes. While this could possibly be explained as a classic case of free-riding, the case of tax paying in rural Mali is in fact much more complicated. When looking at taxes as a form of resource mobilisation, the history of taxes in Mali is an important consideration. Taxing the rural areas was a strong tool employed by the French colonial system to mould Mali under colonial rule. By impoverishing the countryside in certain specific ways, colonial administrators controlled important areas of economic, and by extension social life. Tax evasion, through hiding taxable resources, altering migration patterns, collusion with traditional authorities to deceive colonial employees, and other acts that have become institutionalised, was a form of protest against colonial oppression. 76 Caitlin Blaser 1 September 2007 When the 'development tax' was instituted under the new socialist regimes, promising to return financial resources to the local communities but never delivering, these methods of tax evasion became even more deeply rooted in the rural lifestyle. Non-payment of taxes is often used by communities as a tool if a constituency is unhappy with the actions of the political elite. For example, Sieouba, a village in the commune of Loulouni, had not paid its taxes for two years. When queried, the story emerged that a representative of the council had passed through the village two years ago, and killed a goat with his motorcycle. It was demanded that he replace the goat, but never did, claiming it was not his motorcycle that hit the goat. When, two years later, another representative of the council replaced the goat, the community came together to pay over three million CFA (about USD 6,000) of taxes that were in arrears. As this anecdote demonstrates, the evasion of taxes by no means indicates a lack of collective action. On the contrary, many resource are being mobilised outside the public sphere to be used as strategically as possible. With the large number of NGO extension projects requiring a monetary community contribution to partake in a project, village and association caisses are, in some ways, more important than government resources for community development. The women's association, two different farming associations, a livestock association, and an association of businessmen had more resources available for activities than the mayor's office. Additionally, the two most active NGO extension activities operating in the commune, one providing grain grinders and the other providing grants and loans to start up small businesses, only worked through community based organisations, not local government. Complicated partnerships are formed to match up the mobilisation of resources by some actors to the availability of resources to others, with associations often paying the mayor's contribution of a service available through the government, or one association using funds to systematically buy the goods or services from another association so the first would have access to credit or other services later. Decentralisation has substantially increased the number of NGOs operating in the Sikasso region (CEID, 1997), making a larger array of services available if resources are mobilised in certain ways. This all indicates that this influx of NGOs has more affect on the patterns of resource mobilisation than any political changes brought by decentralisation. While the phenomena are interconnected, it simply underscores the fact that local government is just another actor in a complex arena with many interests vying for resources and power. When asked about impediments to resource mobilisation, residents overwhelmingly cited two reasons. The first was a lack of adequate banking institutions. This was at least in part due to the timing of the research. There was a widely respected savings association with a branch in Loulouni, 77 Caitlin Blaser 1 September 2007 but just before the study began, the president and vice president stole the savings and left town, leaving many quite wealthy associations penniless. The second impediment cited was an uncertainty when dealing with NGOs and government projects, whose agendas change rapidly. Forty-six people mentioned the large number of organisations that came through and made promises of services if contributions were made, but then never delivered. This was also linked to arbitrariness in all aspects of NGO activities, from the topic of the intervention to the selection of participants. There was essentially a call on the part of the community for NGOs to develop a Weberian rationalised bureaucracy, because without it, it was impossible to mobilise resources effectively. A similar sentiment has been expressed by those working with the fiscal elements of decentralisation, but rather than being articulated as a resource mobilisation problem, it is expressed in terms of planning. There is no question that fiscal decentralisation has lagged in Mali, and this can be attributed greatly to the way in which the decentralisation process has been structured. There has been a constant tension between planning and implementation. Local governments bemoan a lack of planning on the national level. A former secretary of the sous prefet (interview, 7 July, 2006) pointed out that one of the biggest sources of trouble we have with decentralisation is that it was implemented without adequate planning. We are here at a commune level, and so is the mayor and his council, before we have been told who is supposed to do what. We wait while the decentralisation mission explains things to us one way, and then as we are doing what they say, they tell us it must be done another way.17 First they say our funding comes in this way for this, then that way for that. Our budgets are never done properly because they themselves haven't decided what a proper budget looks like. And then the community blames us for a lack of action, but our hands are tied. NGOs working in Bamako, however, complain that just the opposite situation is paralysing a functioning decentralised system. There is a real risk in putting too much effort into exhaustive, detailed planning if the link with implementation is tenuous. In Mali, there are already innumerable national and regional schemes and plans, both sector-based and integrated, but....they are not always put into practice. Too often these essentially administrative exercises are unconnected with any real demand. (SNV, 2004) A structured decentralising of the financial system, however, is particularly challenging not only because of the number of individuals and organisations who have a vested interest in maintaining the 17 Here, he was referring specifically to the changes made in the electoral code, which involved a 'recentralising' of many electoral functions that had been placed in the hands of the mayor, without adequate consideration for the fact that he was an interested party. 78 Caitlin Blaser 1 September 2007 status quo, but also because of the French colonial institution of unit? de caisse. A system implemented in theory to promote equitable development of all regions of the country, unit? de caisse brings all of the regional resources to a centralised fund in the capital, to be disbursed to regions and technical ministries by the central government. There are endless features of the entrenched accounting and allocation system that present obstacles to decentralisation. Just to take one example, the education ministry before 1999 did not maintain separate records of primary and secondary school expenditure. Under decentralisation, primary schools are funded on a village level, while secondary schools are funded at the commune level. It is very difficult to determine an appropriate allocation of funds in such circumstances. A representative of the Haut Conseil des Collectivites Territoriales (interview, 8 May, 2005) noted that a more modern version of unite de caisse was also an impediment to implementing the fiscal side of decentralisation. An ideal decentralised political system would be easier in a country whose government is financially self sufficient. A local government would collect taxes, and keep most of them at the local level for its needs, only passing a small percentage to the national level to finance projects of concern to the whole country. Mali is very dependent on aid, however; something like 85% of our public sector is financed by foreign aid. Even though we are working with partners on best practices of coordination for support, that does not change the fact that only the central government can engage in foreign relations. It would be impossible for the local government of the commune of Loulouni to have a bilateral agreement with the United States government. So, all foreign aid must be channelled through Bamako. Yes, we can gave a system to pass this money on to the communes, but no matter what it requires some planning and distribution to take place at the central level, which is not ideal. While there are many reasons for it, the hesitation of the central government to produce coherent legislation on the fiscal elements of decentralisation clearly remains one of the most significant impediments to its implementation at the local level. This will be illustrated in the upcoming sections by looking at service provision, and the decentralisation efforts of technical line ministries. This section has simply underscored the number of factors that affect every aspect of the decentralisation process. Orchestrating them all effectively is necessary for the successful implementation of decentralisation. The nation wide system that had been in place in Loulouni was altered in several ways by the arrival of refugees. The system in place first emphasised the devolution of powers to downwardly accountable authorities, and is only now going back to the devolution of resources. In some ways, the arrival of humanitarian aid turned this system on its head. 79 Caitlin Blaser 1 September 2007 Humanitarian organisations brought significant resources to Loulouni in a way that the lengthy decentralisation process had never managed to do, though it still aims to. However, the power over these resources were never brought into the downwardly accountable governance structures. They were controlled primarily in partnership between the upwardly accountable sous prefet, and humanitarian agencies that are outside the government's system of accountability entirely. 80 Caitlin Blaser 1 September 2007 Service Provision Similar to the fiscal decentralisation process discussed in the previous section, service provision is not so much an isolated phenomenon within the broader decentralisation process that can be analysed alone. It is one of the goals of decentralisation, and as such, is a concrete manifestation of many converging factors. How services are provided is often a concrete indication of how powers are exercised regardless of who is theoretically holding them. This section will also look at which services are provided to whom, and the processes for deciding this. This is essential to determine the extent to which the allocative efficiency goal of decentralisation has been achieved. These processes are also indicative of how accountability structures work in Loulouni, and where they are failing. The record of decentralisation on service provision in Loulouni offers a mixed picture. Perhaps the most notable thing about service provision is that it is certainly not foremost in the minds of Loulouni residence when discussing local government. Like other areas in the focus of the international humanitarian industry, citizens were more familiar with international NGOs and UN agencies than government agencies or departments.18 This is particularly striking given that service provision is the goal around which much of the extension work related to decentralisation has centred. Any observer will note that ?seemingly everywhere in Mali, colourful posters in public buildings, businesses, and stores advertise the government program to shift authority and responsibility to the local level (CEID, 1997),? and almost all of these involve a flow chart or list explaining which level is responsible for which service (see appendix 1). Despite this, when asked to name five responsibilities of government, only two of 100 respondents mentioned a single service. Both of these made reference to security. Education, health, electrification, judicial services, etc. were not mentioned a single time, despite the fact that one of the schools and the health centre are under the charge of Loulouni's mayor and council, even if indirectly through an appointed board. The most commonly cited responsibility was that of dispute resolution, mentioned by seventy-three respondents, followed by holding meetings in Loulouni, mentioned by 48%, and going to meetings in Sikasso, 41%. The fourth most common response was finding and coordinating with NGOs that come to develop the community, mentioned by thirty-six respondents. This was followed by ?nothing? or ?drink tea,? with twenty-four responses, and ?paperwork? with twenty-three responses. While many respondents were pleased with the openness of communication with the mayor's 18 See residents of Kasulu in Landau, 2000 81 Caitlin Blaser 1 September 2007 office, many people did not think this meant an improvement in government behaviour. One woman thought that It was better before. The government was not as friendly, but it was more effective. Now, the mayor holds meetings all the time, but he does nothing. There were no meetings before, people came from above, and told us what to do. But at least something got done. There were good teachers in our schools then. Students could get scholarships to study in Sikasso if they were bright. If you had a high school degree, it used to mean something. Now, you have people who have passed their BAC but they can hardly read and write (respondent 13). The deterioration of higher level education is a widely acknowledged phenomenon, with many professional training institutes throughout the country creating an equivalence by adding three years of study to the qualifications of anyone who completed their studies before 1995. Primary education is often cited as a feather in the hat of decentralisation (SNV, 2004; UNCDF, 2000), with primary school enrolment having increased from 20% to 38% in just the decade 1990-2000 (UNDP, 2003). Attributing this to decentralisation, however, might not be accurate. The number of primary schools in the commune of Loulouni increased from five to eight in the last decade, with all three additions being privately funded (as opposed to government supported). Although rarely mentioned, the quality of these new private schools is also questionable. Not one of the new community schools has classrooms and teachers for each grade offered, and many of the teachers are virtually illiterate themselves. Far more problematic, however, is the fact that education in Mali, modelled after the French system, is highly centralised, and decentralisation has demanded a complete reworking of the country's approach to education. This is combined with other substantial changes, including a shift from primary education instruction in French to local languages. Because of these and other (primarily budgetary) challenges, the department of education did not even develop a framework for decentralising until 2002, and the operational programme it outlines is still years away from being operational. It seems clear that changes in the education sector are more closely linked to donor funding priorities than the decentralisation policy, and once more highlights a need for synthesis among actors throughout the decentralisation process. Changes in service provision seem more closely linked to changes in donor funding priorities than the decentralisation policy. The profusion of community schools is more likely to be explained by the prioritisation of education by a number of large funders than decentralisation. The decentralisation campaign certainly may have played a role, through its involvement in awareness raising throughout the country; many community mobilisation efforts, including those on education issues, are related to the interactions sparked by this exercise, but attributing the service provision itself to decentralisation is inaccurate. 82 Caitlin Blaser 1 September 2007 Administratively, the water sector has made the most progress decentralising its powers and resources, and is held up as a model for other technical line ministries. By 2000, the ministry had developed contracts relating to the transfer of powers and resources between levels of government, and the delegation of management, with particular attention being paid to fiscal structures not only once the transfer had taken place, but throughout the implementation phase. Additionally, it drew up a contract for the responsibility for technical and financial monitoring, which remains the most clear definition of tutelle in practice, effectively limiting any leeway the central government has to recentralise through its supervisory powers. From 2003 on, instruments for transferring water supply infrastructure were tested, trainings were conducted, and by now local management of water related infrastructure is entirely functional. Its success may just serve to underline the importance of donor funding, as despite the success in administrative decentralisation, the number of boreholes drilled in Sikasso has remained unchanged despite a large increase in demand related to changes in rainfall that have left many seasonal rivers dry.19 That indicates that a gap remains in the political control of service provision related resources. The experience of the health sector supports this as well. While demand throughout the commune is for village-level health structures, such as maternities, extension workers trained in first aid, and other initiatives that would make at least a basic level of primary health care available on the village level, spending tends to focus on the CESCOM (Centre de sante commmunautaire), because, as the current mayor points out (interview, 11 June, 2006) and the health committee confirms (interview, 26 July, 2006), ?that is where the fiscal structures direct the resources at the moment.? This is a clear case of a lack of fiscal decentralisation inhibiting true political decentralisation. The health sector was rarely singled out by residents of Loulouni at any stage in the interview of survey process, with the exception of those individuals who were directly involved in it. Only thirty survey respondents said they or someone in their family had accessed health services in the past five years. Because this is a service that is so rarely accessed, many citizens have not built up a track record of expectations from the health center. As there has been little community involvement, issues of health are better understood at a higher-than commune level for the sake of the decentralisation process. The first mayor of Loulouni (interview, 8 June, 2006) explains the situation he found himself in regarding service provision. Education, health, the judiciary, all of these are now the domain of the commune, meaning they are under the charge of the mayor. But what has happened is a transfer of responsibility. It has not been 19 Department d'hydraulique demandes des pompes, 2004. 83 Caitlin Blaser 1 September 2007 accompanied by a transfer of resources. They acknowledge this on a national level, but it makes things very difficult for we mayors. When the decentralisation mission came to explain things to people, they said that the taxes raised by the commune will stay in the commune. But now, I am faced with a situation where even if people pay their taxes, resources for services are still in Bamako. The authorities say they mayor must now provide education, but the money to construct classrooms is in the hands of the ministry of education in Bamako. What am I supposed to use to build these classrooms? According to the legislation, now it is the mayor who must supply water to the population, it is the mayor who must construct health centres, build schools, equip the teachers and so on. But with what? Without support, without resources. The law says that responsibility is not to be transferred without the required resources. You can say now that the central state is acting outside the law. But who will hold them accountable? The population of Loulouni can't; they can only vote for a new mayor. The state and perception of service provision in Loulouni says a lot about the status of decentralisation as a whole. In fact, service provision might well be the key indicator of other aspects of decentralisation. First, it reflects many of the challenges of a lack of resources faced by local governments. However, it also demonstrates how a lack of autonomy in funding allocation can inhibit true political decentralisation. Fageut found in Bolivia that service provision patterns significantly changed under decentralisation, and this has been used to demonstrate the success of decentralisation, because resources are then used according to each community's need, increasing effectiveness. In Loulouni, this change does not seem to have taken place. This can be attributed to two things ? the fiscal structures of decentralisation that still require resources be transferred from Bamako streamline funds in certain directions, undermining any local decision making power. Additionally, the amount of control donors have in setting funding priorities also undermines local control of resources. There is a strong reason to believe that decentralisation has moved many aspects of governance in the direction of participatory democracy. Survey respondents were very supportive of the openness of the mayor's office, and agreed across the spectrum of respondents that the mayor was more accountable and responsive to their needs than previous governing bodies. Additionally, the level of interaction between all members of the community, regardless of class and gender, and local government officials, increased substantially. However, structural aspects of the administrative decentralisation process including the role of the sous prefet in oversight and regulation have the possibility to threaten these nascent supports for democracy. Additionally, the limits the state of fiscal decentralisation places on local control of resources could undermine the relationship built between the mayor's office and the local community, if local government continues to be unable to 84 Caitlin Blaser 1 September 2007 deliver on the concrete development needs of the community. The next chapter will explore the impact the refugee influx has had on Loulouni to determine how this is shaping the decentralisation process. 85 Caitlin Blaser 1 September 2007 4. The refugee experience ?HELP NEVER COMES AS SWIFTLY AS NEED, AND IT RARELY STAYS TO DRINK THE THIRD ROUND OF TEA, WHICH IS WEAK AND BITTER FROM A LACK OF SUGAR.? BOUBACAR SANOGO ?THESE REFUGEES ARE FROM WEST AFRICA, BUT THE MAJORITY OF THEM ARE NOT USED TO THE HARSH CONDITIONS IN LOULOUNI? CAMP MANAGER Trends in the data The experience of the Ivorian refugees in Loulouni, and therefore the influence they have had on the local political situation, has been defined primarily by two factors. First, the history of migration between Mali and Cote d'Ivoire, which played a large role in shaping the integration of refugees in Loulouni. Secondly, the decentralisation process discussed in the previous chapter defines the political stage onto which the refugees entered. To explain the impact these factors have had, this chapter is divided into sections on refugee-community relations, access to common resources and public services, economic patterns, and identity and social perceptions. In very general terms, the relationship between the refugees and the community of Loulouni was good, with many hosts citing the positive contributions refugees have made to the community and many refugees pointing out the warm hospitality the community of Loulouni has offered them. However, certain tensions were underlying this positive relationship. Looking at how the relationship was defined by both refugees and hosts, and the patterns that tensions followed, uncovers many aspects of how the refugee influx has impacted citizen-state interactions. The most common source of tension between refugees and hosts in Loulouni was regarding access to common pool resources and public services. Conflicts around refugee access to these became the leading source of interaction between citizens and local government officials. The refugee influx had very clear, tangible impacts on water, health, and education in Loulouni. These impacts were both positive and negative, and were instrumental in changing the trajectory of the decentralisation process in Loulouni. Economic differences between refugees and hosts also had numerous implications for decentralisation. Because of different economic access, refugees integrated into specific social and economic areas. This naturally affected which issues were brought to the attention of local government, and by whom. It also altered the needs of the local population, in terms of both public services and governance. 86 Caitlin Blaser 1 September 2007 The final section on perceptions and identity brings together the social and economic factors in the previous sections, to look at how those measurable elements of refugee integration are actually experienced by the individuals living the process. The way various trends are perceived is as important for the decentralisation process as the trends themselves, since it is the perceptions that will prompt citizens to act and interact with local government. This section also offers an opportunity to reconnect with the literature on refugee integration, which deals heavily with not only identity, but social and economic integration. It will provide a foundation both in the literature and in the data gathered in Loulouni on which to build the following chapter's argument about the transformative role the refugee presence has had on political transformation. 87 Caitlin Blaser 1 September 2007 Community - refugee relations The relationship between the community of Loulouni and refugees was generally good. This was observed by many of the role-players involved in the relief effort. The Office for the Coordination of Humanitarian Affairs (2002), for example, claims that ?the relation between host communities and refugees/evacuees is good. Host communities have been very forthcoming in their support to refugees in the Sikasso region.? This is supported by the findings of the survey, in which 81% of respondents say the refugees have had an ?overall positive? impact on the community, and 87% claiming the relationship between refugees and the community could be described as having ?no conflict whatsoever.? However, the tension that was found follows certain patterns, and exploring these patterns will help draw conclusions about changes in governance or citizen-state relationships that could be connected to the refugee presence. Because the refugee influx was so large, virtually everyone in the community reported a great deal of interaction with Ivorians fleeing the conflict. Understandably, few people distinguished between Ivorians with official refugee status and those without, or even Malians who had been living in Cote d'Ivoire. In terms of quantity of interaction, 97% of respondents reported having interacted with refugees in the marketplace, and businessmen reported more interactions with Ivorians than other professions. This is most understandable because the refugee camp was located at the doorstep of the marketplace. Upon initial inspection of the data, quantity of interaction with refugees seems strongly correlated to economic status. However, closer inspection reveals that this is explained completely by neighborhood of residence, since the closest neighborhood to the refugee camp is also the wealthiest, while the farthest away is the poorest. The following sections will break down the issues on which residents and refugees interacted, disaggregating both populations by profession, age, economic status, gender, etc. as appropriate given the topic of analysis. Among the most important trends are that conflicts between hosts and refugees tended to be about public services. Wealthier hosts were both most likely to access these services and most likely to describe the relationship with the refugee community as conflictual. Furthermore, they were most likely to interact with local government officials regarding the conflict. Evidence of this, as well as interpretations and consequences, follow. 88 Caitlin Blaser 1 September 2007 Access to common pool resources and services Much of the refugee studies literature dealing with the impact refugees are having on their hosts concentrates on competition for resources or services. A modest amount of research indicates that refugees use a disproportionate amount of common pool resources (Whitaker, 2002; Stonjanov, 2004; UNHCR 2004). Numerous scholars have asserted that when strain is being put on common pool resources by refugees, it is the poorest of the poor hosts, who depend on common pool resources for their livelihood, who are further marginalised (Kibreab, 2001; Whitaker, 2000; Chambers, 1993). This does not seem to describe fully the situation in Loulouni. One explanatory factor is that Cote d'Ivoire generally enjoys a higher level of economic development than Mali. While the assumption that refugees are comparable to 'poor' in host communities is being challenged, Loulouni is a clear case of the inverse scenario ? refugees accustomed to a significantly higher level of development and standard of living than the host community. It is clear that economic patters are different for refugees than for the rest of the population, although this study did not venture into comparing the wealth of refugees to that of the local population. The supposition that refugees use more common pool resources than the host community is to a small extent true in Loulouni. This statement, however, does not come without qualifications. Refugees use only slightly more firewood than their local counterparts. Other studies assert that this is due to the fact that the food rations given to refugees are based on dry food that requires longer cooking times (UNHCR, 2004). The fact that this has proven to be less the case in Loulouni than has been observed in other refugee situations could be explained by the fact that food rations in Loulouni were rice-based, whereas the staple diet in Loulouni is dried maize, which, as it is normally prepared, requires slightly more fuel to cook than rice. Most other common pool resources, such as grass for thatching roofs, resins and gums, fibres for basket weaving, etc. were scarcely used by the refugees at all. While firewood is arguably more important than the others in aggregate, in that more people spend more hours per day devoted to its collection than other materials, it is also true that extra strain will be felt by people nearly regardless of their income category, as nearly all residents of Loulouni across the economic spectrum cook with wood. Those whose livelihoods depend on these other common pool resources, on the other hand, often have specific trades or areas of expertise, and the refugees are clearly not in a position to compete with them and thus displace them in their livelihoods. One refugee specifically addressed this issue, saying: While I was in the bush once collecting wood, I met an herbalist who was looking for medicine. He was upset to see me there, and said 89 Caitlin Blaser 1 September 2007 that I should watch out collecting these barks and roots. He said they were for his work. I told him that I was just collecting firewood, and explained that I was from far away, and didn't know the plants and trees that grew here anyway. He was satisfied. (refugee interview 14). While many respondents (thirty-six, mostly women, and disproportionately poorer respondents) noted an increased scarcity in the past several years of common pool resources, particularly firewood and thatch, it is significant that not a single one cited the arrival of refugees as a source of the scarcity. The most common reason cited was a sustained lack of adequate rainfall affecting the growth of vegetation around Loulouni, followed by general population growth, and the commercialisation of the firewood sector (often attributed to various causes, such as increased transportation to Sikasso, and a growing need for access to cash), which meant commercial exploiters would chop firewood to sell in cities or make charcoal. Access to services was the source of far more tension between refugees and residents of Loulouni than access to common pool resources. The impact of refugee youths on the education system was widely cited, although interpreted differently by different people. To begin with, over half of those at the refugee camp were under twenty-five, meaning the school systems were particularly affected by the influx. Refugees and returnees had benefited from Cote d'Ivoire's superior education system, and as such, had on overall advantage over Malian students. This was compounded by the fact that French, the medium of instruction, is more widely spoken in Cote d'Ivoire than in Mali, since Bambara, the local trade language that most Malians speak in addition to their mother tongue, does not have an equivalent in Cote d'Ivoire. Some parents saw it as a blessing. In one discussion, the mother of a secondary school boy was encouraged by the arrival of the new students. My son is a smart boy, but before the Ivorians came, he was bored at school. The level of education is not very high; the teachers don't expect much from the students. If the work has been completed, they are satisfied. Now he is not doing that much better in school, but I can see the change. He has something pushing him now. You know how boys are, he is competitive, and now he has someone to compete with. I have much more hope for his education than I had before. He is seeing students continuing on to Sikasso, benefiting from education; this wasn't the case before. Another parent, however, did not see the developments so positively. My daughter is in second cycle now, and she has always been very good at school. If she did not come in at the top of her class, she was always in the top 3, ever since her first year. Before the Ivorians came, she was first in her class for two years running. When the conflict started in Cote d'Ivoire, the school in Loulouni has taken on many different kinds of students. The system that we have isn't in 90 Caitlin Blaser 1 September 2007 place any more; we don't know who is studying what. There are students who did 9th standard in Cote d'Ivoire, now in standard 7. Some even lie about their ages. Now the top student in every grade is an Ivorian, and my daughter is discouraged. It's a matter of discipline in the system. If we don't regain it, I'm afraid my daughter will lose interest in school (respondent 16). An interview with the school director (19 March, 2006) revealed a mixed opinion of the new students. There is no doubt that the Ivorian students had a better foundation than students here who come in from remote villages to go to second cycle. Many of them are smart, although some of them have trouble adjusting, since there are some differences in our curriculum. Overall, though, I think the standard of education has been pushed slightly higher since they have come. They have brought with them a problem of discipline, though, that has really been challenging. I can't only blame it on the Ivorians; sometimes it is caused by local students too, but there isn't the same respect for authority that there used to be, that is clear. The culture at school has definitely changed, but I can't say where it is a positive or a negative change. While there were temporary alternative educational facilities provided by UNICEF, it was never suggested by anyone that the refugee youth should not be permitted to attend the mainstream school at Loulouni. While the right to access this service was not a source of conflict, as later analysis will explain, related conflicts found articulation in the school setting. Health services were slightly more contentious than education, though the conflict did not stem from the host community as a whole, but only the local health centre. It is widely recognised that refugees sell a large portion of their aid disbursements (Landau, 2002). In the months after medical kits were included in the disbursements, however, the pharmacy almost went out of business due to the availability of pharmaceuticals being sold at cheap prices by the refugees. An interview with the chef de post (18 January, 2006) does a good job of explaining the dynamics in the health sector. Of course, the biggest problem we as health workers experienced with the refugees was the problem with the pharmacy. That was not a conflict per se, but a logistical problem. It was poor planning by those helping the refugees. They didn't consider that we are also here selling medicines, and that the budget of many of our activities is connected to the pharmacy. We were not trying to make a profit, but running effectively is not free. It's necessary to have a refrigerator for vaccinations, but how can it run without money for petrol? We suffered from the medical supplies from Bamako, and we tried to reach an agreement with the decision makers, but everything was too far above our heads. There was no consultation.....[Other interactions] were generally positive. When people first started to arrive, we had too much work. The new arrivals were tired, and many of them weren't healthy. We were already busy here, serving people from many villages around. The 91 Caitlin Blaser 1 September 2007 Red Cross came with volunteers, though, and that helped us a lot. There were good dynamics here, we worked together well. Nobody ever mentioned that refugees must go here for service, and people from Loulouni must go here. In this health centre, we are care givers, and anyone who is sick is our patient. This is not a police service, and we are not interested in seeing identity documents. There were some challenges; the Red Cross workers had their way of doing things and we had our way, and sometimes it was a challenge to synthesise these differences. It was good to exchange these ideas. I was on the crisis committee, and I gathered from my experience there that some people in Loulouni suffered when the refugees came. But speaking for myself, I can't say that was the case with the health centre. The situation wasn't ideal, but we came out stronger because of any hardships. I think now we are better organised and coordinated than we were before. Sometimes it is good when you have to do things differently, you learn. As mentioned in the previous chapter, residents of Loulouni were generally so far removed from health services that they had little stake in this service at a local level. Refugees also had little to say about the health sector, other than pointing out that the facilities available were sub-standard to what was available in Cote d'Ivoire; refugees with serious health problems were immediately sent to Sikasso. These two things meant that issues of health are better understood at a higher-than commune level, as there has been little community involvement. The most conflictual service provision issue was that of water. As previously noted, water was one of the biggest problems in Loulouni before the arrival of the refugees, and even after the two borehole hand pumps were installed at the site of the camp, Loulouni remains far below the Sphere Project Humanitarian Standards minimum criteria of 250 people per tap, 500 per hand pump, or 400 people per well to allow everyone 15 litres of water daily. Given its population of 3,000 without counting refugees, evacuees, and returnees, there should be at least six hand pumps (Sphere Project, 2004). As it is, the vast majority of residents drink primarily untreated water from uncovered wells, and for activities that require large amounts of water, such as washing clothes, they go to a seasonal river that skirts the town (Site log, 2000). The installation of the two water sources in Loulouni serving 750 people (a hand pump serving 500 people and a tap serving 250 people), approximately covered the population of the camp as per the standards mentioned above, but without consideration for the host population. When UNICEF installed the two new water sources, there was a lack of clarity about usage rights of the pump and tap. Nearly all members of the community assumed that they had free use of both water sources, while nearly all refugees assumed that they had exclusive usage rights. The tap at the school was not terribly contentious. It was further from the camp and therefore used less 92 Caitlin Blaser 1 September 2007 frequently by the refugees not attending school, and few members of the community besides students went to take water there. It was locked outside school hours, and conflict was kept to a minimum. The tap near the market, however, was a constant source of conflict. It was reported by both members of the host community and refugees that for a period of several weeks in 2003, fights broke out at the pump daily. The crisis committee took up the issue of the pump, and concluded that while the tap at the school would be shared equally by the school and the refugee camp, refugees would have exclusive use of the pump near the market. It still took several weeks to hammer down specifics, such as who could be responsible for locking the pump at certain times, and whether fees could be charged for pump usage and who would be responsible for managing them, but agreeing on this system did not immediately reduce conflict with the community. This can be attributed to three things. Firstly, a lack of communication of the decision of the crisis committee to the community meant that many people were unaware of the decision, and continued to think the refugees were being unfairly selfish. Secondly, there was a lack of acceptance of the decision on the part of many women, who perceived it as unfair, and imposed by the sous prefet. Thirdly, because the market attracted vendors and merchants from as far afield as Bamako, Mopti, and Cote d'Ivoire, and conflicts pertaining to pump usage were most frequent on Wednesdays and Thursdays (Thursday being the day of the market, but people travelling long distances tended to arrive on Wednesday to set up), it was difficult to communicate to non-members of the host community the details of the pump usage regime, and even more difficult to induce them to respect it, given the lack of incentive they have to maintain community relationships. When asked about the impact refugees had had on the economic situation in Loulouni, eighty- three respondents made reference to the new water sources as being a positive contribution to the community. The former mayor (interview, 11 June, 2006), however, refused to attribute this service to the arrival of the refugees. The Ivorians have had a very negative impact on services in the town. UNICEF has been working in the commune for many years. Some people say they put in pumps because of the Ivorians, but that is not the case. They have been working to support education, and they put a pump at the school. They have done this in other communes as well. The second pump was installed by the department of waterworks. UNICEF helped with the commune percentage of the funding, and encouraged the department in Sikasso to contribute the rest. That is normally the work of the department. Are we supposed to congratulate the Ivorians because our government department is doing its job? Then the camp manager came and said that that pump must be for the Ivorians, and 93 Caitlin Blaser 1 September 2007 the sous prefet agreed. This is taking away the service development from Loulouni. They did the same damage to our health service, overloading the doctor, and then coming in with their own medicines to put our pharmacy out of business. Other people with a negative view of the refugee influx, however, generally allowed that the refugee relief effort was responsible for the water points, even if they created problems of access, or conflict relating to them. As one respondent demonstrates, Some organisations brought the pumps for the refugees, so that was a big benefit for the community, since there is a problem of water here. The refugees created many problems with the water, though; for months, we were fighting about using the pumps. Every day some would go to the mayor, some would go to the village chief, some would go to the manager of the camp, saying how can we reach agreement about the water? Eventually, we were just quiet, because we knew some day the Ivorians would leave, and we would be left with the pumps. That is worth any trouble we had, it has been a great benefit for the town. Sanitation is an issue closely connected to water, and while it was somewhat less contentious than the question of pump access, there were all the same many sanitation problems raised in the community and at the refugee camp, and the task force dealing with water was also responsible for sanitation. Sanitation issues came to the forefront early on, while the camp was still being installed; however, they were more reflecting existing power struggles than a problem of sanitation per se. At a crisis committee meeting, the sous prefet called on the youth association to require that its members sweep and tidy the camp area once weekly, as a gesture of good will towards the refugees. The mayor was opposed to the idea, reflecting many of the historical dynamics between the sous prefet and the mayor. In his words (interview, 19 March, 2006) At the crisis committee meeting, the government said that the youth association would be required to volunteer their labour to clean that refugee camp. I said that was something I could never agree to. This should not be required, unpaid work. Mali is no longer at a time of forced labour; the government cannot exploit the population at will. If we find that the community is willing to do this work freely, of their own accord, that is fine and they have my blessing, but nobody can say that it is work they are required to do. Perhaps equally significantly, the youth association in Loulouni had not been functioning for nearly a decade due to certain long standing disputes in the community, making the issue moot. The second sanitation issue that came up was related to the sweeping of the market area. Beginning in 2002 before the arrival of the refugees, the mayor's office contracted out the sweeping of the market area every Friday to the women's association. They were paid through a nominal stall fee paid by each vendor at the market. This was very much a case of public service on the part of 94 Caitlin Blaser 1 September 2007 the association, as they did a great deal of work at a reduced rate. In 2003, however, the arrangement was cancelled. The president of the women's association (interview, 14 January 2006) cited mismanagement of funds on the part of the mayor's office. Other women in informal discussions, however, suggested that it was motivated by a dissatisfaction with the way the water conflict with the refugee camp was being handled. The final sanitation issue of debate at the camp was that of latrines. This again was closely tied to the proximity of the camp to the market place. In construction of the camp, eleven latrines were dug. The Sphere Project Humanitarian Standards (2004) recommend a maximum of twenty people per latrine, and given that the camp population was an average of around 800, this is only a quarter of the recommended number of latrines. The hygienic level of the latrines was also a concern. The Sphere Project asserts that latrines are ?more likely to be kept clean if users have a sense of ownership (2004 p75).? This was difficult to promote given the lack of latrines in the market area. Residents around the market area had called for the construction of latrines at the market in the past, citing hygienic concerns. With the number of market-goers using the latrines, the refugees refused responsibility for their upkeep. The eventual result is that they were rarely used. While repairs were occasionally made by NGOs on an ad hoc basis, the sanitary concerns of the refugees were never adequately resolved. Residents of Loulouni and Ivorian refugees generally had positive relationships. However, looking at service provision points to some tensions. First, the experience of the pharmacy is a clear case of unintended consequences on the part of humanitarian agencies. The water sector demonstrates one of the clearest cases of refugees bringing resources into the community and changing the demands citizens have of local government officials, which is a key argument in this thesis. The dynamics around education and sanitation both underline the historical relationship between Malians and Ivorians, and the class element that is a part of both the refugee experience, and the impact the refugees have on the decentralisation process. 95 Caitlin Blaser 1 September 2007 Economic patterns The economic patterns of refugees in Loulouni are in many ways accurate indicators for their integration into the host community, due to the fact that they reflect a wide range of concerns for refugee communities, such as social interactions and economic independence. They also reflect how the historical dynamics and class differences between Malians and Ivorians express themselves in the present situation. There were various differences in the economic habits of refugees and citizens, and exploring these will shed light on the relationship between refugees and citizens. Refugees were less likely to engage in farming as citizens of Loulouni, even though land was made available to refugees should they have chosen to cultivate it. This is unsurprising, given that their time horizons were uncertain, the land set aside was often marginal, and cultivation would have required substantial capital. Additionally, the largest influx of refugees was between the months of September and December ? harvest time. It would be approximately another six months before refugees could sow their own plots. Even so, eleven respondents critically mentioned that refugees did not cultivate their land, often as support for assertions that they were lazy, used to the city life, or not interested in contributing to the development of Loulouni. While refugees were more likely to engage in commerce than citizens, they were less likely to engage in the least profitable kind of commerce - buying a product in bulk and selling it in smaller quantities. Instead, they were more likely to be selling products that they either made or processed, which is generally more profitable, but also requires more skills and capital. The most typical of these activities is making attieke, a fermented cassava dish that is difficult to prepare and a luxury in Mali, but a staple in Cote d'Ivoire. One resident of Loulouni knits hats for a living, and she said the refugee influx had a substantial impact on her work. Ten years ago, I was the only one in Loulouni who knew how to do this. My aunt went to Abidjan with her husband, and she learned there. When she came back to visit, she taught me how to do it, and I had a successful business. It's the kind of thing only people in big cities do. In Cote d'Ivoire, more people learn these things. When so many Cote d'Ivoirians came, they all knew how to do this knitting. It's okay, I still have my clients and now they have theirs, but business was better before they came (respondent 64). The cooking of specialised dishes was particularly common among refugee women. Due to higher rainfall in Cote d'Ivoire, many food products common there are luxuries in Mali, and are generally prepared by Cote d'Ivoirians who have more experience preparing these dishes. This, in combination with cultural norms formed in the relatively wealthier Cote d'Ivoire, could explain the charcoal consumption among refugees being nearly double that of residents. As one woman 96 Caitlin Blaser 1 September 2007 explains, Some of my [Malian] friends laugh at me, because I use charcoal so much. I cook nearly everything on a charcoal furnace. At home, I didn't have a mud stove. I'm not used to cooking with wood. Some things I know how to make using a wood fire; rice and corn dishes are easy. When I'm cooking plantains, though, I don't know how to do it correctly using wood. My friends say that I must be rich; they must think I'm showing off money cooking everything on charcoal. That's not the case; I came here with nothing, and I was not rich, even before. I just know a different way of doing things. I could not manage cooking only with wood; I need to sell plantains, and if they were cooked over wood, I can't say how they would turn out, or if people would buy them (refugee interview, 14 September, 2005). Another occupation for which Ivorians were renowned is brewing dollo, a term which applies to various alcoholic drinks including millet beer and palm wine. As the vast majority of residents of Loulouni are Muslim, alcohol is taboo, and no local residents would openly brew dollo. Twelve of the fourteen respondents who thought that the refugee presence was having a generally negative impact on the community cited dollo as a factor, causing moral degeneration in the community. This was inevitably linked to other debaucherous, disrespectful activities, such as laziness, prostitution and stealing. All of the activities mentioned above are primarily ?women's jobs,? and it is true that women at the refugee camp were, on average, more economically active than men. Some men reported doing wage labour on farms, but most of the refugees came at a particularly bad time of year for temporary agricultural work. The influx took place during harvest, and for the months following, work is almost completely unavailable. The following two years both had very poor rainfall, substantially decreasing the demand for hired workers. A few individuals worked in the construction sector (the primary dry season occupation); however, most refugee men cited difficulty doing this work because mud bricks in Loulouni were a different style than those in Cote d'Ivoire, to suit the drier climate, and thatching was also done differently. Others reported attempting to work as artisans, but facing difficulty primarily due to a lack of access to capital, but also because of either a saturated market, or difficulty breaking into the existing market (mentioned in eight interviews). Petty employment available to men in Loulouni is generally less prestigious than for women. Many youths were reported to work as porters or other small jobs on market day, but beyond that and agricultural labour, job prospects were limited. In terms of material possessions, refugees were less likely to have access to various material indications of wealth than residents. For example, relatively few refugee households had bikes or motos, solar panels, radios, etc. This is unsurprising again given the context of their arrival 97 Caitlin Blaser 1 September 2007 to Loulouni, and that most do not consider it their permanent place of residence. These material indicators do not make a good measurement of economic integration, however, as numerous non- material indicators suggest a lifestyle considered prestigious by residents of Loulouni. The use of charcoal burners mentioned above is once example. They were also more likely to eat cassava than local residents, and less likely to eat corn. A difference in dietary habits was exacerbated by food distributions, which were based on rice. Rice is a relatively prestigious food in Loulouni, and not even the wealthiest residents have a diet based primarily on rice. In addition to material possessions, refugees tended to have cultural capital to associate them with the elite in Loulouni. They were on average more educated than residents, from more urban areas, and had a far better command of French. Finally, through no choice of their own, refugees report more leisure time than their Malian counterparts. As one young man reports (refugee interview 34, 1 October, 2005), People could look at me and say I'm lazy, but that's not it. It's not that I don't want to work, but what can I do here? In Cote d'Ivoire, there were more choices; here, there is nothing. I don't want to spend all my days sitting around drinking tea, but what else can I do? I'm sad every time I go to the office of the sous prefet; I am young and fit, I should be working for the food I eat. I wish the aid workers would distribute jobs instead of food. He goes on to confess Some of my friends pass their days drinking dollo and chasing women. I don't agree with what they're doing, but I also understand why they do it. They are bored, their hands are idle, and they have no power to change their circumstances. I can't blame them for it. The former mayor (interview, 11 June, 2006) also noted his concern about the lack of employment at the camp, citing that it was creating tensions with the community. Right now, there are many refugees staying here that have been given the opportunity to move to Faragouaran, where there is an international mandate for their management. They have been given the opportunity to return to Cote d'Ivoire since the security situation is better. They chose to stay here. This is worrisome for us, because we don't know what work they are doing, we don't understand what they are doing here. When someone is working, we understand there role in the community. But a stranger who comes without work, who stays without work, and who doesn't want to leave for a place that may have opportunities, how can the community trust such a person? The economic aspects of the refugee camp are both a cause and a consequence of the interaction between the camp and the decentralisation process. While certain economic indicators (staple food, for example) are less an indicator of economic class than geography and planning by 98 Caitlin Blaser 1 September 2007 relief agencies. However, the history of classism and discrimination between Cote d'Ivoire and Mali means that all these factors influence the way refugees interact with the community, and how the community perceives the rights and powers of refugees. These interactions then shape the discussions that take place between citizens and local government officials, particularly in the arena of access to resources. This, as we saw in the introduction, and will be emphasised in the following chapter, is a key element in shaping the decentralisation process. 99 Caitlin Blaser 1 September 2007 Perceptions, Identity, and trends It is often the case in refugee contexts that refugees are seen as outsiders, and identities differentiating refugees and the host community serve as polarising factors (see Chapter 2; Bakewell, 2000; Whitaker, 1999). This has not been the case in Loulouni for a number of reasons. An extensive history of migration as well as a shared language, culture, religion, and ethnicity all contribute to the inclusive nature of refugee-host identities. This does not necessarily imply a lack of tension, however, and some conflicts were actually framed around this sameness of identity. It was mentioned in four interviews and by two respondents that there were Malians passing themselves off as refugees. The former mayor (interview, 11 June, 2006), who had very negative opinions of the refugee presence, posited the following view: People are using the camp as a hiding place. These are Malians ? not second, third generation Malian immigrants, but Malians born in Mali, who once had Malian ID cards ? at the camp, receiving aid, claiming to be from Cote d'Ivoire. They've looked at other returned migrants, who are staying with their extended family in Loulouni, and they've seen that life is difficult for them. So they're trying to take a different path. They are just taking what they can get. Are more commonly expressed view, however, was to take pride in the identity shared with the refugee community. One man said You can't come here and say there are the refugees, and these are people from Loulouni. We're all the same. You can see the truth in what I'm saying by watching the young people interact. Whenever there are functions, playing football, and even if there is work to be done, the young people are all together. They are even marrying each other, and we adults are happy with it, because once, we made a distinction, and we have seen that it is wrong. Still others shared the above respondent's observation about young people interacting, but took a very different view on the matter. Previous quotes have alluded to stereotypes in the community of the refugee camp as a den of sin ? a source of alcohol, prostitution, etc. Some of these accusations have truth to them. Even respondents who had generally positive views of the refugee presence tacitly acknowledged the presence of alcohol and prostitution. As one respondent (16) said, Many of the refugees are from cities, where they are used to things we are not used to, and we don't accept all of the views they bring with them. A lot of them drink, they aren't Muslim. It is something we don't like, and they are used to doing things with girls that people here don't accept. Despite this, they come with a lot of new ideas 100 Caitlin Blaser 1 September 2007 that we are interested in hearing; many of them have travelled, and seen things we haven't seen. That is an important asset. Like everything, we have to learn to take the good, and leave the bad. If they have some bad ways of doing things, we must move on and forget them. Their new ideas are giving us many opportunities; when we use them, we can't forget that. Another negative stereotype of refugees that was expressed in Loulouni and other refugee hosting situations (Landau, 2002) is that they are responsible for increased crime in host communities. This seems to have little basis in evidence in Loulouni. While many people who had a negative view of refugees accused them of being criminals, there was no substantial increase in reported crime upon the arrival of the refugees and returned migrants (interview, Kadiolo chief of police, 15 August, 2006). While there was one case of rioting that resulted in the police being called, it was due to perceived corruption of the sous prefet, and did not involve the general population of the host community at all. An additional accusation often levelled against refugees is that they erode social structures, for example, undermining traditional authorities or family structures. This, like crime, is something those with an already negative opinion of refugees evoked to strengthen their case, and the majority of respondents acknowledged, and yet attribute to different sources. Seventy-six respondents brought up a change in traditional structures of authority as a change Loulouni has seen in the past five years; every single older man mentioned it. However, only four linked it to the presence of refugees. That is the same number of respondents who linked the same phenomenon to decentralisation. The majority of respondents linked it to changes with roots in economic patterns, such as more young people migrating, working for cash, or no longer depending on their families for paying bride prices. This is very much in line with what academia suggests, as various studies have found that where agricultural decline and urbanisation lead to high mobility among youth and a concentration of elderly people in rural communities....older people often lose control over important productive resources and traditional support structures become less effective (Whitaker, 2002 p 374, citing Baker, 1995; Sommers, 1995 and Parkin, 1972). While these negative stereotypes that are frequently dealt with in the refugee studies literature could be found in Loulouni, they were not descriptive of the average host perception of the refugee community. Perceptions both by refugees and hosts overwhelmingly centred on the economic differences between Mali and Cote d'Ivoire. The refugees, while not seen as rich during their sojourn in Loulouni, were seen as coming from a place of economic privilege, not accustomed to the difficult living conditions in Loulouni. This defined many levels of engagement between hosts and the community, and was certainly the centre of how other perceptions were articulated. For example, many of the 101 Caitlin Blaser 1 September 2007 hosts who cited the refugees as a source of social degradation linked this to the ?urban way of life? in Cote d'Ivoire. These articulations were widely shared by refugees as well. One refugee woman demonstrated this view discussing the water conflict. Women from Loulouni say it's not fair that we have a pump to ourselves. It's true that they have a problem of water too, and there are more of them than us. But they have been here for a long time, living like this. They are used to it. Staying in Cote d'Ivoire, we are not used to such things. We can barely survive with this pump as it is. While this difference of economic background was a source of tension, it was not always seen negatively. The most commonly cited affect refugees had on the community was that they brought new ideas, mentioned by seventy-four survey respondents. In Mali, travelling is considered an important form of education, and as such, anyone coming from far away is respected as having knowledge of different places, customs, etc. The vast majority of respondents' initial framing of their interactions with the refugee community was based around this exchange of ideas. This was particularly emphasised by poorer respondents, and those who had never travelled to Cote d'Ivoire. One man explained, Many people in Loulouni have been to Cote d'Ivoire, and come back with stories about what it is like there. I have never had the chance to go, so I must keep quiet. Now when I talk to the people coming from Cote d'Ivoire, it is like I have gone myself. I can come to understand their way of doing things. I can also help teach them, since they are strangers here. It turns all of us into wiser people. This has really benefited the town. Now, the boundaries of Loulouni don't stop at the river, they stretch all the way to Cote d'Ivoire. This idea was also expressed through responses about refugees' economic activities. Sixteen women surveyed cited the Ivorian's superior attieke-making abilities. Three even reported working with them to learn the skill. The only two respondents who reported hiring Ivorian agricultural workers also had positive things to say about the exchange of ideas. One mentioned farming in Cote d'Ivoire is not exactly the same as it is here. There, the plots are usually bigger, and some of the things we plant are not the same. Sometimes it is a disadvantage, because the workers don't know how to do everything, and we must teach them. But sometimes I gain a lot, because they might know ways of doing things that are better, or faster. The strongest trend evident in the data, a dynamic which was also evident in other interactions, including the interviews and informal interactions, was that opinions about the impact 102 Caitlin Blaser 1 September 2007 refugees had on the community were strongly correlated to the economic status of the respondent.20 Specifically, the wealthier the respondent, the more likely they were to feel like the refugees had a negative impact on the community. See the following crosstabulation: Refugee Impact * Economic status Impact of refugees Poorest quarter 2nd quarter 3rd quarter Wealthiest quarter Total Positive impact 24 96% 22 88% 21 84% 14 56% 81 Negative impact 1 4% 1 4% 4 16% 8 32% 14 Unsure of impact 0 2 8% 0 3 12% 5 Total 25 25 25 25 100 Reasons for this were articulated in a number of different ways, and its implications are central to thinking about transformations in governance in general and decentralisation in particular. The most striking articulation stems from the fact that wealthy men were more likely to have travelled to Cote d'Ivoire themselves to work than any other sector of the population. Every single male respondent claiming the refugees were having a negative impact on the community had spent time working in Cote d'Ivoire, and they all made a link to the discrimination they themselves faced during their time in Cote d'Ivoire. One man said We who have worked in Cote d'Ivoire, we will turn our backs to the refugees. We went there and suffered. I can't describe what it is to be a Malian working in Cote d'Ivoire; the humiliation when you go to pray. People take advantage of you, stealing because they know you are a stranger, and you have nowhere to turn. Now they are trying to turn to me? No, my eyes are closed to their problems (association leader 20 Economic status was determined by combining responses to a number of questions. First, respondents were asked whether they, or someone in their household, owned each of a list of ten items chosen as indicators of wealth in Loulouni (examples: bicycle, solar panel, radio, etc.). Based on this, the respondent earned a score from one to ten. They received an additional two points if they attended French school, and an additional one point if they had any other formal schooling. They then received between zero and four points based on whether or not they hired workers (and how many). Finally, additional points were assigned based on the respondent's work, the head of household's work, and other sources of income (for example, laborers that worked on others' farms earned no points, subsistence farmers who owned their own land received one point, those who hired workers earned two points, and those who owned land that was farmed in a tenement systems received three points). 103 Caitlin Blaser 1 September 2007 interview, 18 March, 2006) Conversely, a wealthy woman who had also spent time in Cote d'Ivoire experienced similar difficulties, but drew a different conclusion about the impact the refugees were having. I am glad the refugees came. When I went to their doorstep asking for help, they turned me away. I slept outside in the cold, hungry. Now that they are at my doorstep, they will have a warm bed and a full stomach. I used to be angry at the Ivorians, and the way they treated me, but now that they are here, I have the chance to make peace. Being a good host has been a chance for me to regain the respect that I lost when I was in Cote d'Ivoire (merchant interview, 18 July, 2006). In the case of women, negativity in seven of the nine cases was expressed primarily because of the problem of water in the community. As one woman said, When we learned that the Ivorians were getting pumps installed, we were very happy. We had also been obliged to walk far to the river to get water, and then it is not always clean. But then the Ivorians said we couldn't use their pumps, that they were only for them. Even if they were not using them, they would lock them so we couldn't take water. How can we accept them as guests if they behave like this (respondent 78)? The fact that there is a correlation between economic status and a negative opinion of refugees is not on its own important to this study. It becomes important, however, upon learning that economic status was correlated to interactions with local government officials. While the wealthy were not substantially more politically engaged,21 they were more likely than their less affluent counterparts to go to the mayor or commune council to help them resolve any disputes they might be party to in the community. As wealth decreased, respondents were more likely to consult extended family members, religious leaders, or the village chief. The consequence of these two patterns converging is that local government will be more engaged on issues relating to refugees by those who feel they have had a negative impact on the community. The impact of this will be explored in the following chapter. What some of the quotes from this chapter illustrate is that a wide range of members of the host community in Loulouni are making the same observations about the impact of the refugee presence in Loulouni, which seem accurate in that they are supported by other kinds of data as well. The conclusions they are drawing, however, differ wildly. The current refugee studies literature looking at the impact refugees are having on their hosts explains this by disaggregating the local population, and pointing out that different people are experiencing different aspects of the same 21 Political engagement was measured by receiving one point (with a maximum of ten) based on whether or not the respondent was registered to vote, whether they had voted in the last election, whether they reported themselves as involved in politics, whether they were members of associations, whether they were on the board of associations, whether they reported they were involved in, or felt they could influence, decision making processes in Loulouni, and whether they were able to list any organisations active in Loulouni. 104 Caitlin Blaser 1 September 2007 phenomenon. While this is one element, simply disaggregating the population does not go far enough. This study proposes that these differences in articulation can be more accurately explained as a reflection of the pre-existing power dynamics in the community, and provide important insights about stakeholders views and involvement in the local political process. While disaggregation of all populations is important for a robust understanding of social and political dynamics, this chapter has provided an argument that it is inadequate. Rather, the historical political, economic, and social dynamics that make such a process of disaggregation relevant must be brought into the picture as well. These played a large role in setting the trajectory for the process of integration. In the case of Loulouni, the close historical ties between southern Mali and Cote d'Ivoire meant that refugees were not seen as outsiders that the local community could build an identity in opposition to. However, because Cote d'Ivoirians were historically wealthier, at times at the expense of migrant workers from Cote d'Ivoire, discourse over rights and access to resources were often framed in terms of differences of class, level of education, etc. This has then been instrumental in shaping the decentralisation process. By the beginning of this chapter, it was already evident that the implementation of decentralisation in Loulouni was strengthening many aspects of participatory democracy. Citizens were much more likely to interact with local government officials, channels of communication had opened, and there was a widespread perception that, however serious the problems of capacity may be, local government valued consultation, and had become more responsive to their needs. However, from an administrative point of view, the division of powers between the mayor and the sous prefet was delicately balancing many issues of legitimacy, capacity, and service provision. This balance was threatened by limits of fiscal decentralisation, since a lack of resources at the local level was both undermining the relationship between the mayor and the community, and between the mayor and the sous prefet. The arrival of refugees to this scene transformed the situation in several important ways. On one hand, the refugee influx was accompanied by significant resources from humanitarian organisations. This had clear benefits for the community, particularly since many of these resources came in the form of infrastructure. However, rights and access to these resources had to be negotiated between the refugees and the host community, and these negotiations took place in a context of historical inequalities. Additionally, the power to control rights and access had to be negotiated between the mayor and the sous prefet. The process through which this took place, and the way this changed the decentralisation process will be developed in the next chapter. 105 Caitlin Blaser 1 September 2007 WHO REPRESENTS THE LOCAL PEOPLE AND WHO RECEIVES POWERS IN DECENTRALISATION? VARIOUS ACTORS, INCLUDING ELECTED BODIES, CUSTOMARY AUTHORITIES, ADMINISTRATIVE APPOINTEES, LOCAL REPRESENTATIVES OF TECHNICAL SERVICES AND MINISTERS, COMMUNITY GROUPS, ?DEVELOPMENT? COMMITTEES AND NGOS, ARE RECEIVING POWERS IN RURAL AFRICA IN THE NAME OF DECENTRALISATION. THE KINDS OF ACTORS WHO ARE EMPOWERED IN DECENTRALISATION SHAPE THE OUTCOMES THAT CAN BE EXPECTED. DIFFERENT ACTORS ARE EMBEDDED IN DIFFERENT KINDS OF ACCOUNTABILITY RELATIONS, AND THESE RELATIONS SHAPE THE WAYS IN WHICH THEY EXERCISE THEIR POWERS. RIBOT, 2002 5. Impact of Refugee Arrival on Decentralisation The previous two chapters have looked at the impact of refugee arrival on decentralisation first from the point of view of the decentralisation process, and then from the point of view of the refugee influx. This chapter will focus specifically on those areas of interaction already indicated in previous chapters. It will develop and analyse the most important impacts the refugee influx has on the decentralisation process in Loulouni. This chapter breaks down the impacts into three categories: political, administrative, and developmental. While all three are intertwined, distinguishing between them is useful not only to structure the discussion, but also in bringing the analysis to other decentralisation processes, which may be structurally different from either a political or administrative point of view. The impact the refugee influx has had on the administrative aspects of the decentralisation process have been largely negative. As previously discussed, lack of clear mandates on all levels has been a huge challenge to the effective implementation of decentralisation in Mali. Just as local government officials were developing a way of working in the face of this challenge, the introduction of new powers and resources by the refugee influx upset the fragile administrative balance. Power struggles were exacerbated, occasionally to the point of limiting the capacity of the local government to act. Additionally, because of the nature of the relief operation, unprecedented demands for coordination were made on local government. This likely had the positive affect of improving coordination across the board. However, it is also possible that such demands will result in unintended transformations of local government structures. The final outcome remains to be seen as the transformation is very much still work in progress, but evidence from Loulouni suggests that the relief effort moved local government away from participatory, democratic practices by vesting more power in upwardly accountable authorities. The political impact of the refugee influx, however, is largely positive. Before the refugee influx, residents of Loulouni were overwhelmingly frustrated with the inability of local government to meet their needs with regards to basic public services. Local government was frustrated with the demands being made on them from the population while they were unable to access resources to 106 Caitlin Blaser 1 September 2007 respond. Mirroring the administrative conflicts caused by the humanitarian influx, the introduction of resources that accompanied the refugees created conflict in the community. This changed the demands of local government from service provision to conflict resolution. Local government was unable to provide services because of a lack of resource. However, a high degree of success was possible on the conflict resolution front. This built a relationship of trust between local governmnet authorities and the population, bringing the residents of Loulouni on board in the decentralisation process. This confidence is crucial to the success of decentralisation, and without the change in local government role prompted by the refugee influx, decentralisation would have risked being brushed aside as more talk from the central government. The only danger is that ultimately it is local government that is responsible for service provision under decentralisation, and if the local population is no longer going to hold the government accountable for services, it will equally be a failure. Finally, the developmental aspects are briefly considered. While the political and administrative elements fit neatly into the discourse on decentralisation, considering development on its own is useful as well both because development is an explicit goal of decentralisation, and because a large part of the discussion on refugee influxes surrounds the humanitarian aid involved. In this case, more questions were found than answers. There is a discord at present between the importance of nongovernmental and international players in development on one hand, and the developmental role of decentralised government on the other. 107 Caitlin Blaser 1 September 2007 Administrative impacts The arrival of refugees can be seen as impacting the administrative decentralisation process in three places: coordination, capacity, and clarity. In terms of coordination, the multi-sectoral relief effort required a wider variety of stakeholders work together in a way that had ever been demanded in decentralised Mali. It would be difficult to find another situation in which government from the national to the local level, both upwardly and downwardly accountable, is required to coordinate efforts with technical ministries at all levels, and international organisations and NGOs are required to coordinate their work with government at multiple levels, all to one common end. This is linked to clarity, as during the coordinated work, lack of clarity of roles came to the forefront as an obstacle to effective implementation of the relief effort. While some roles may have gained clarity as a result, it is equally true that in some places, ambiguities with which the system had been able to work were no longer considered acceptable, resulting in conflict. Finally, issues of capacity were highlighted time and time again, as new powers and limitations were being tested. The refugee influx can seen as trying, perhaps for the first time since they were implemented, the coordination abilities of decentralised governance structures in Mali, since coordination was at the heart of the relief effort. In fact, far more man-hours were spent in coordination than implementation (interview, Sikasso crisis committee member, 18 February, 2006). It is clear by the way the relief effort was coordinated that the sous prefet, rather than the mayor, was at the forefront of the local government relief effort. Because of the political history of Mali, sous prefets are very responsive to responsibilities passed down to them from the hierarchy of government. One NGO extension worker in a neighbouring commune cited a history of conflict with the sous prefet, as he was seen as working with the mayor on issues surrounding decentralisation. When this was discussed in Sikasso at a planning meeting, he returned to a complete reversal of attitude. Upon inquiry, he found it was really quite simple. All it took was a letter from the office of the governor saying 'decentralisation is our project too, give it your full support.' After that statement was received, the sous prefet was nothing but helpful (ACOD extension agent interview, June 18). From the formation of the crisis committee, it was clear that managing the refugee situation was being placed in the hands of the sous prefet. This explains his strong support of the refugees on so many issues, such as according exclusive use of one pump to the camp, and the encouragement of the community to clean the camp. The mayor, however, felt marginalised throughout the process of the arrival of refugees, and this undoubtedly made already existing tensions with the sous prefet 108 Caitlin Blaser 1 September 2007 worse, and possibly mobilised negative sentiments about the refugees. Allegiances with the mayor by certain segments of the population then became articulated as anti-refugee. In an interview, (6 March, 2006) the mayor at the time of the relief effort explains his experience. The manager of the camp didn't include the commune in the management of things.....Even though someone else represents the state in the commune level, the mayor represents the population. If we, the host community, are going to support the refugees, the manager also must engage the community. I was not at all included in the process. Okay, if a delegation was going to the camp, I was invited along just like that, but as soon as something had to be decided, be it related to protection, services, or resource allocation, the community is completely overlooked. We were ready to engage with the refugees as partners in developing the commune, but as soon as we saw that the manager was only looking to the government, and brushing the population aside, we also stood to the side. There has always been some rivalry between the commune population and the state. That's normal. You hear all over Mali that with decentralisation, this is a problem. The legislation says that mayors and sous prefets are not rivals. But in practice, it's difficult. When I'm given power over something, and it is suddenly taken away and given to someone else, it's only normal not to be happy about it. We all know we must live and work together, but on a psychological level, things are rocky. Of course, it depends on the individuals involved. There are some communes where the mayor doesn't understand a process, and the sous prefet will refuse to explain it to him. In other communes, the sous prefet takes pride in the work the mayor does, and they are partners. Before the refugees came, the sous prefet and I worked together; not without problems, but the work got done. Now, it's much more difficult. Throughout the lengthy interview, he mentioned in over ten places the poor relationship between the central government and the elected local government, and in six cases it was connected to the arrival of refugees. The current mayor (interview, 9 June, 2006) did not note substantive changes in the relationship between he and the sous prefet, although the element of the refugee presence as an exacerbating factor was absent. You could say that the collaboration is better now, because the initial jealousy is over; all government representatives now accept decentralisation, even if they don't like it. The sous prefet has changed, and I am a new mayor, so the initial personalities, who had been hurt by these changes, are not involved. We have all been sensitised about the importance of working together, so now it would be hard for us to say that we disagree. But substantively, nothing has changed. In fact, now you could say it is worse. Before, when there were arguments, you could say it was just the people who didn't get along. But that is not the problem any more; it is not just 109 Caitlin Blaser 1 September 2007 because things are changing, because decentralisation is new. We must admit that the problems now are institutional. The commandant22 and I are friends. I drink tea at his house and we chat, but we have no power over the difficulties at work. The separation of roles is clearer now, but nobody is satisfied with it yet. As long as resources are not distributed according to legislation, it is impossible for anyone to say that there is meaningful cooperation. Here, the mayor linked the poor relationship to resources and decision making powers. While he did not once cite the relief effort (largely over by the time he took office) explicitly as a source of tension, he cites the institutional nature of the lack of cooperation. While the relief effort is certainly not the only cause of this lack of cooperation, it is clearly one factor. Like many survey respondents (see chapter 2), the former mayor cites a lack of consistency as accompanying a relief effort, and he claims that this is due to a lack of clarity in roles. This lack of clarity was at the heart of the power struggle between he and the sous prefet. In his words, Before the refugees came, I knew my work, and the sous prefet knew his. But when the refugees came, everything was shaken up. What it comes down to, is that nobody knows exactly who is responsible for the refugees. And when we started discussing that in detail, it became clear that nobody knew who was responsible for many other things as well. On one hand, you can say the refugees aren't the commune's problem. It is a national, and even international issue. But as long as the community is here, and they are affected, it is the problem of the commune. The [central] State put them here. International partners are responsible for funding the refugee effort, so that must be controlled at the State level. But these funders don't do work themselves, they leave the resources and run. People were all over the place, trying to manage things. For the first 5 months, I was responsible for everything; sleeping mats, food, and other disbursements. Then, there was cooperation with the Red Cross, and State representatives discussed the plan with me. But it didn't last. The State said we must find a place in the village to host the refugees. I told them from the beginning that I was against the refugees coming, because they would cause a big problem in the community. But I could not refuse to find land for them, because who does the land belong to? Yes, I am responsible for managing it in the commune. But at the same time, the land belongs to the State. And at the same time, the land is the property of God.....After I showed them a plot of land for the camp, they went to the sous prefet, and closed the door on me. They weren't interested in the community any more, they got what they wanted. The current sous prefet (interview, 2 August, 2006), while not serving at the time of the influx, had a different explanation of the same confusion, however, saying The funders were in a difficult situation. They want to work with the 22 sous prefet 110 Caitlin Blaser 1 September 2007 mayor, but in a time of crisis, coordination is necessary. The sous prefet is responsible to the government, it only makes sense that when a government decision has been taken, the sous prefet will implement it. And this was an international problem; it can't be spearheaded locally. The mayor was included when the population was affected, but we also must think of the refugees. They need protection, and this cannot be the domain of the mayor. He is concerned with the community, but so are the funding partners. So everyone went back and forth, trying to figure out if decisions were supposed to go up or down. The overlapping mandates and conflicting definitions are apparent in the case of the pumps installed by UNICEF. While the planners of the pump installation were no doubt familiar with common 'best practices' in relief efforts that encourage resources benefit the community as a whole (crisis committee meetings notes, 2002). The placement of one pump at the school is evidence of this, and the second being near the market also follows this reasoning. However, the sous prefet's leadership in the crisis committee has resulted in a kind of tutelle for international organisations. While UNICEF was in a position to determine the placement of the water sources, the sous prefet could determine usage rights. While the crisis committee in some ways played a recentralising role, it also had a coordinating and capacity building role that deserves mention. In three different interviews, members of the crisis committee mention the positive coordination affect the experience had on their sector. The chef de post (interview, 6 June 2006) says These meetings were important. There has never been a problem with communication in Loulouni, the population here is not that big, but sometimes we all get caught up in our work. The crisis committee brought people together from all the different services and sectors, and we talked about concrete issues. Real partnerships came out of this. For example, members of the health committee talked with members of the APE (education association), and we started giving classes on health at the school. Here we were, for years, the school and the health centre only a few hundred meters apart, but it took something to bring us together, to spark the idea. The school director highlighted the capacity building element of the relief effort, saying 111 Caitlin Blaser 1 September 2007 There are so many changes in the education system since decentralisation. When UNICEF workers came down, my staff and I learned a lot. First, people from Bamako and Sikasso work more with policy makers, so there are many decisions that we think are not a good idea, but they understand better. Additionally, because they hadn't only been teaching the national curriculum like we were, they came with different ideas of doing things in the classroom. It was a very good chance for our school. I don't think it ended when they went back to the city either, because now that they have spent time with us, understanding our challenges, they can't forget it. In the case of Loulouni, tensions between the sous prefet and the mayor were often expressed in terms of refugee protection. If the relief effort had adequately considered the existing dynamics, it could have been used as a powerful tool for capacity building and conflict resolution. Instead, refugee protection was framed as separate from and in conflict with the interests of the host community. While this may be necessary in refugee situations characterised by a high level of refugee-host conflict, this was not the case in Loulouni, and this framing of the situation only served to polarise the mayor and sous prefet. Furthermore, by placing refugee protection exclusively in the hands of the sous prefet, a distrust of local officials and lack of responsiveness to the community that decentralisation was going to great lengths to overcome was exacerbated. Sixty-three respondents thought the sous prefet had too much power in general, and this number increased to seventy-two when asked about powers relating to the refugee influx. The former sous prefet's secretary discussed the transformatory role the refugee influx had briefly during his interview (5 July, 2006). When the refugees first came, it was difficult for us. There was more work than we could handle, overseeing everything; it was chaos! In the long run, there have been benefits. We worked closely with NGOs that didn't know about us before, and to this day the relationship remains. Before, they showed no interest in us, but now, they come to the sous prefet to talk about projects. It's a benefit for the community. It has made some people in the community jealous, but these are only people who don't know that we are working for them. The lack of clarity in mandate and at times ad hoc implementation that may be inherent in any emergency relief situations reignited power struggles between local actors. In certain contexts, there may be nothing transformatory about such struggles. In the case of Loulouni, however, and arguably in many refugee hosting countries in a state of political transition, the outcomes of such power struggles will have a defining impact on emerging political structures. This is particularly true in the early years of decentralisation in Mali, since legislation on the decentralisation of powers is not sufficiently developed to provide clarity on roles. As a result, the local reality in Loulouni is in a 112 Caitlin Blaser 1 September 2007 position to shape policy as much as the opposite is the case. As stated in the introduction, the administrative goals of decentralisation in Loulouni are multidimensional. The primary goal is, through the creation of the elected position of mayor, to establish clear downward accountability to the electorate of Loulouni. Similarly, administrative decentralisation also aims to enhance transparency in decision making, and allocative efficiency in service delivery. This, combined with greater transparency, also aims to enhance resource mobilisation, as the advantages of investing in local government become clear. According to the evidence presented in the previous two chapters, the refugee influx has created challenges for the administrative decentralisation process. Much of this comes back again to the difficulty in creating strong accountability structures. The reasons for this are multifold, and by no means all due to the refugee influx. The third chapter discussed problems in this that were independent of the refugee influx. These included a lack of history of democratic political systems, the complexity of the actors and powers divided among them, and the lack of resources. The lack of democratic history, particularly when combined with the complexity of actors and powers, has inhibited the administrative aspects of the decentralisation process simply because it is often too difficult to discern who is to be held accountable for what, and who must do the holding to account. A lack of resources, while at times clarifying roles simply by limiting them in scope, also undermined the creation of clear accountability structures by eroding confidence in local government, making concrete demands for service provision or development seem futile. The refugee influx has threatened the administrative aspects of the decentralisation process. While the influx helped somewhat the problem of resources, it did not necessarily help the adminsitrative aspects of decentralisation along. As the discussion of service delivery clearly showed, with the resources that accompanied the refugees came conflict, both between the host community and refugees about access to the resources, but also between different local government actors about control of the resources. The relative clarity of roles that had been facilitated by the limited powers and resources to control at a local level was obscured by the refugee presence. Humanitarian aid is inherently centralised, both because it is arguably used primarily during crisis situations, where such centralisation is necessary, and also because it is largely funded by multiple international donors, meaning that it much be channeled through the capital. As a result, the balance between regulation being upwardly accountable, but distribution of resources being downwardly accountable, was disturbed. By giving powers for the allocation of resources at the local leve to the sous prefet, the degree to which decentralisation was truly implemented was reduced. This effective recentralisation of powers has had a negative impact on the other adminsitrative 113 Caitlin Blaser 1 September 2007 goals of decentralisation. As mentioned earlier in the section, as well as in previous chapters, the sous prefet works in less transparent ways than the mayor. Furthermore, because he is not accountable to a local electorate, nor does he have consultative decision making processes, he has less allocative efficiency than the mayor. While the refugee influx did not necessarily change the way the office of the sous prefet worked, it did change the trajectory of decentralisation from an administrative point of view, in that new powers were devolved to an upwardly accountable, untransparent governing body. There would be a valid argument to be made that such a system was necessary for the protection of the refugees; an official accountably only to the local population would not have an incentive to protect the refugee population, skewing what precisely constitutes efficient allocation of resources. While this argument is credible, it is also important to note the transformatory impact any shift in powers has on the decentralisation process while legislation is still being solidified. The challenge to the administrative aspect of decentralisation would be less pressing if it were contained to the influx itself. However, as evidence earlier in this section has shown, there was an element of norm creation when the sous prefet took on these new powers. So far, this chapter has looked exclusively at the impact the refugee influx has had on the adminsitrative decentralisation process. The articulation of decentralisation in Mali is such that the political decentralisation process is equally important. Additionally, the goals of both processes are inherently connected, and as such, discussing one without the other is incomplete. While challenges to administrative decentralisation can stall complete political transformation, the political decentralisation process in Loulouni has been moving ahead. The next section will see how it has been shaped by the refugee influx. 114 Caitlin Blaser 1 September 2007 From creating to managing ? political transformation One of the most significant findings of the newly emerging empirical research on decentralisation is that central governments are better than local governments at creating infrastructure ? initially installing the service - whereas local governments are better at managing and maintaining what is already in place (Fageut, 1999). Since the refugee influx created a shift in focus from infrastructure creation to management, it actually created a situation that was more conducive to decentralised governance. Furthermore, since the demands of management and maintenance are clear, and full involvement on the part of the community is inherent to a conflict resolution process, this has created a situation whereby the local government has a clear mandate that is visible to the community, facilitating accountability and transparency. Chapter 2 mentioned a lack of ownership of the decentralisation process as an inhibition to the creation of strong accountability mechanisms. A lack of ownership of many local government functions is understandable, as there has not been a history of public involvement in health, water, education and other services, and much of the process of service provision and financing, even under the new decentralisation regime, is relatively technocratic and complex. The management of access, however, and conflicts related to it, inherently include all involved parties, making a lack of civic engagement effectively impossible. This makes management of services and conflict resolution two areas where decentralisation in Mali is based placed to succeed. Decentralised control of management issues, such as access to services, is relatively uncontroversial. It is clearly within the mandate of locally elected officials to manage services that they are theoretically responsible for providing, and negotiate citizens' access to the same. The arrival of refugees raises many questions about access to services for host communities. Complex negotiations inevitably take place to determine the rights of refugees to make use of public services, as well as the rights of hosts to access relief related services. While it is possible to follow certain universal guidelines, in practice, each local situation is individual and specific enough that no one formula can be applied to every circumstance. Local governing authorities are particularly well placed to deal with these issues, and at the same time they are institutionally strengthened by being given the mandate to make decisions on issues that were formerly nonexistent. The conflict resolution element may need to be treated with a bit more caution, as it could imply a potentially undesirable or undemocratic institutional change. While the need for leadership on issues of services access can only strengthen local governments in a desirable way, the merging of negotiating access to services with the broader role of resolving conflicts has different implications 115 Caitlin Blaser 1 September 2007 for the political decentralisation process. In Loulouni, more than half of those surveyed cited conflict resolution as an important role of the mayor's office, and of those respondents citing specific experiences they had interacting with the mayor, more than half were parties to a conflict. Outside of conflicts related to the relief effort, these conflicts were rarely connected to access to services. They dealt, in order of frequency, with land use, tenure, management of the weekly market, crime, and management of civil society associations.23 In the cases of land use and tenure, as well as the management of associations, the mayor was often being asked to create legislation without a legal framework from which to work. In other land use and tenure disputes, as well as those related to crime, the mayor was being asked to perform essentially judicial functions. The lack of judicial and legislative support at a local level is a challenge to decentralised institutions. From a short term point of view, it may have positive benefits. Many of the interactions citizens in Loulouni have had with the mayor surrounded conflict resolution, and as such, communication and interaction was fostered between the population and local government, which is crucial to the success of decentralisation, particularly around issues of transparency and accountability. Much of the power of traditional authorities comes largely from a legitimacy vested in them by the population, even if they do not technically have the capacity for enforcement. Having local government behave similarly could be seen as a positive, capacity building step. The long term danger of this, however, is two-fold. First, it is likely to create a confusion among executive, legislative, and judicial functions. The complete lack of judicial functioning on a local level in Mali, while perhaps in the short term is strengthening the commune council, will certainly be a hindrance in the long term. At best, in the medium term it will cause a power struggle between the sous prefet and the mayor, as the sous prefet remains responsible for policing and security; in other words, enforcement. While this could turn into a desirable system of checks and balances, it is firstly not a sign of strong governance to trust in the goodwill of all involved to create this system, and secondly another opportunity for the central government to add another layer of recentralisation. The second long term danger of having local communes responsible for some unspecified collection of executive, legislative, and judicial functions is that it risks, as some critics of decentralisation have suggested (Benjamiasen, 2004), local government simply being another layer, or another role player, in a situation where all parties already lack various elements of legitimacy, capacity, etc. The PGP, and therefore governance related aspects of decentralisation, are based on the principles of shared governance. As one donor report explains, Shared governance is supposed to bring together all of the 23 This clearly demonstrates that these civil society organisations are not yet fulfilling expectations as democratic counter-balances to local government. 116 Caitlin Blaser 1 September 2007 structures and actors with legitimacy in the eyes of the people so as to yield well-respected decisions and higher levels of citizens? participation. Up to now, the PGP has not sufficiently incorporated the legitimate structures and actors that exist outside the realm of modern governance. If the PGP is to gain the full participation and ownership of all sectors and groups of the population, it will have to intensify and extend its work in this regard, by reaching out formally to the other structures and actors -- or rather, persuading the communes to do so. As the section on actors above maintains, there seems to be a disconnect between aiming for participation, and aiming for democracy; the two are not necessarily equal as long as those targeted for participation are not themselves democratic. The influx of refugees and the relief effort that accompanies them raises unique questions about participation. The principle behind democratised decentralisation in Mali is that elected officials are accountable to their politically engaged electorate. This way, resources are allocated according to the wishes of the people. This makes two assumptions. First, that the resources to be allocated are in the hands of the elected leaders. Second, that either the electorate or the law will provide adequate protection for minorities and non-voters (for example, refugees, immigrants, children, etc.). Proponents of decentralisation argue that it is better positioned than other governing systems to account for the second assumption, because ideally under decentralisation what will happen is an internalisation of externalities (Oates, 1999). Much of the Sahelian decentralisation literature, grounded primarily in natural resources management, maintains this as an underlying assumption. Critiques of this are now emerging (Prud'homme, 2000; Nejhuis, 2004; Breton, 2001) that argue this ideal is not achieved. One of the implications of this is that there is not adequate minority protection. In fact, authorities may be required to act in a discriminatory fashion if that is the will of the local majority. During the refugee influx, Loulouni was also hosting many returned migrants and others 'non native' to the commune. This made determining who the local electorate was challenging, essentially raising questions about whose participation in the democratic process was trying to be elicited. Perhaps more significantly, however, what this participation is being elicited for is called into question when the relief effort brings in numerous resources outside the allocatory powers of the local government. The relief effort brought in resources that were worth many times the annual budget of the commune. Landau noted in a refugee affected area in Tanzania that some residents thought international organisations related to the refugee influx had done more to help residents in terms of development in the last five years than the government has in the past twenty (Landau, 2004). While 117 Caitlin Blaser 1 September 2007 in Tanzania district authorities denied this, both residents and government officials in Mali tended to blend actors' roles in development. A close cooperative relationship was always expressed, although like any relationship, there were still frustrations on both sides. As the mayor said (interview, 11 June 2006), you can say international organisations do more for development than we do, but it's just a matter of perspective. Our work overlaps. They cannot work without us, we cannot work without them. We must be realistic ? other countries have money, and Mali is poor. We cannot develop without their resources. But at the same time, just money is not development, we know the population here, and we understand how to turn their resources into something that can be used to benefit the people. It is a partnership.....Of course there are factors that effect community perceptions. The commune council can't drive around in a big new car, and we don't put signs and stamps on all the work we do. Some of our work is quiet. Not everybody realises how many teachers' salaries we pay, or broken equipment for the school we repair. It is not as visible as a big new pump, but it is also important, and it adds up. We are not trying to catch the eye of the community, we are trying to get work done. When it came to talking about international organisations or NGOs working in Loulouni, many survey respondents echoed the sentiments of the mayor. As one respondent said, so many organisations come to Loulouni with their big cars, they have money, but it's no use to us. Only the mayor and leaders in the city understand how these resources work. They understand how to work together with these organisations for our development. They have their work for our development and the NGOs also have their own ideas, but neither can accomplish anything without the other. An interesting feature of democratic decentralisation is that because local populations distinguish between powers and responsibilities, authorities are only as responsible for service provision as their constituents think they should be, regardless of what powers have been devolved administratively. This makes understanding the relationship between the electorate and the local government particularly important. One frequently levelled criticism of decentralisation is that it creates an environment in which local elites are able to hijack the resource allocation process, exacerbating inequalities. While this study does not suggest that this accurately describes the process in Loulouni, it is true that the wealthy and educated are disproportionately involved in the decentralisation process. While this was true before the arrival of refugees, the establishment of the camp exacerbated the situation (mayoral meeting records, 1999-2005, and survey responses). The most commonly cited reason by poor respondents on their lack of engagement in politics is the most intuitive as well ? a lack of time. According to a time use survey done by a Peace Corps volunteer in Loulouni (Site log, 2001), household respondents that this study identified as in the 118 Caitlin Blaser 1 September 2007 wealthiest quarter had on average six times more leisure time than those in the bottom quarter. As one respondent explained, When decentralisation was new, in 1998, 1999 and 2000, I went to many meetings at the mayor's office, and the sous prefet's office. Many people came out to explain the changes, and I went because it is important to learn. Many of the things they discussed in the beginning will be important to me some day, such as where to go to get an ID card, or register a marriage. Decentralisation is not so new any more. The mayor still has many meetings, and they talk about important things, but I don't have time for all of that. Now that I understand how the new government works, I have my fields to work in. People who are involved in managing the issues in Loulouni, they can sit in the meetings and exchange ideas. If something affects my household some day, maybe I will go too, but so far it has not happened. Another possibility is that the wealthiest 25% of residents were far more likely to claim that the refugee influx was having a 'strong' impact on the community. Of the interactions they reported having with the mayor's office, over half of them were refugee-related issues. Only one of the respondents outside of the wealthiest 25% reported interacting with the mayor or local government officials on issues relating to the refugees. This should not be divorced from the dynamics that underpin wealth and political allegiance in Loulouni. There is a strong differentiation made between 'original' inhabitants of Loulouni, and new immigrants. The village chief, mayor, and a majority of the wealthiest quarter of respondents form a part of the former category. The other strata of respondents in the wealthiest quarter are either employees of organisations generally based in Sikasso who are on short term contracts in Loulouni, or successful businessmen from other areas who chose to settle in Loulouni. These migrants were no more likely to have a negative view of the refugee influx than the rest of the population. Additionally, they were far more likely to talk about either national or regional economic implications of the Ivorian crisis than the original wealthy inhabitants, who framed the situation as primarily political. On a national level, it is likely true that the primary impact of the conflict in Cote d'Ivoire is economic rather than political, since Mali is landlocked, and very dependent on Senegal and Cote d'Ivoire for the transportation of most of the country's manufactured goods. Locally, while there has been an economic impact, it has likely been a temporary one. The political impact, on the other hand, has been transformatory. In some ways, the Ivorian refugees came to the rescue of the decentralisation process in Loulouni. Support for decentralisation was fading because local government authorities did not have the resources to provide infrastructure and services in the 119 Caitlin Blaser 1 September 2007 commune, reducing the level of community involvement and interest in the process. This was revived when humanitarian agencies injected the commune with resources to support the refugees and reduce the burden their arrival would have on the community. At the same time, the arrival of refugees may be hindering the establishment of the kind of accountability mechanisms decentralisation needs to succeed. On one hand, when the demands the community made of local government officials changed from the provision of infrastructure to the negotiation of access and rights to use existing infrastructure, the local government was much better able to respond to the needs of the community. However, this is enforcing a system that has been in place in Mali for a long time, in which the public sector is not providing services. One of the political aims of decentralisation is to change this pattern. As long as the citizens of Loulouni are not going to expect service delivery on the part of local government, however, the system of tax payment and accountability envisioned by Mali's leaders at the national level, and partners at the international level, will not be established. This is closely related to the developmental implications of the refugee arrival on decentralisation, which will be discussed in the following section. 120 Caitlin Blaser 1 September 2007 Development Aspects A central consideration in evaluating the impact of refugee related activities is the resources of the host community. The lack of resources on the public sector to undertake development activities in Loulouni and other rural communes means that many of such activities had long since been privatised by default, largely left in the hands of NGOs or individuals. There are numerous economic models for analysing the effectiveness and efficiency of decentralisation, but if there is not financial support for decentralised governing structures, these models can all be thrown out the window. In Loulouni, any role player who invests resources in services or infrastructure can be seen as having a positive contribution towards the development goals of decentralisation simply because the decentralised governing structure does not have the resources to implement their own development activities. However, it is important to point out that while this may work towards the developmental aims of decentralisation, it is coincidental, and does not represent any institutional change. The significance of relief services delivered in a decentralised context is threefold. First, they ease some of the pressures on local government who is unable to provide these services with the resources available. This is particularly true of resources which the host community and refugees both use without distinction, such as school supplies at the secondary school attended by both refugees and hosts. As Whitaker suggested, this seemed to foster a higher degree of satisfaction with the local government than would otherwise be the case. More respondents cited that local government should be responsible for coordinating NGO interactions to develop the commune than to actually provide services (sixty-four vs. two), and on the whole, respondents were satisfied with the government's work in this department, some of them specifically mentioning refugee-related aspects. For example, one respondent noted the mayor does a good job cooperating with outside projects. When the Ivorians came, he saw that there was a chance to develop the commune. Without his leadership, we would not have the pumps we have now. He also helped bring improvements to the school, and to the road. These are things not just anyone in the community understands how to facilitate. The second is that these services are instituted outside of the accountability structures of local government. While the relief effort may add legitimacy to local government by creating powers and responsibilities that previously did not exist, it is simultaneously undermining them through the back door, by taking control over services that would normally reside with the local government. This could inhibit the creation of accountability structures and a solid tax base as the population does not link tax payment to services provision, and service provision to local government. It essentially ensures that 121 Caitlin Blaser 1 September 2007 service provision and infrastructure development remains outside of the institutional and political transformation decentralisation is trying to promote. Finally, they create yet another ambiguity in mandate that must be clarified for decentralisation to function effectively. One of the cornerstones of decentralisation as formulated by the central government in Mali is that local communities will be politically and financially responsible for the provision of many of their own services. The challenge to decentralisation here is multi-fold. First, it is unclear whether the central or local governments will end up deciding how decentralisation will unfold on this issue. Secondly, it raises questions about what kind of relationship decentralised authorities will have with international donors, who will inevitable be providing the resources for many different aspects of the process. Finally, it demands questioning the extent to which communities can be autonomous in defining the roles of local government officials. It would certainly be an unprecedented (and potentially dangerous) experiment in democracy were there to be no parameters legislated instituting checks on local government officials. 122 Caitlin Blaser 1 September 2007 6. Conclusion This study has examined the impact a refugee influx has had on the process of democratic decentralisation in one commune in southern Mali. It has found that the influx of refugees has had a profound impact on the transformation of governance in the commune. It has encouraged the process of political decentralisation to become more participatory, through creating incentives for interaction between residents of Loulouni and local government officials. At the same time, some degree of recentralising took place, since many new powers and resources that were brought be the humanitarian influx were placed in the hands of the upwardly accountable sous prefet. From an administrative point of view, the impact of the humanitarian influx in Loulouni has threatened the decentralisation process. The way decentralisation has been implemented in Mali is such that powers were devolved often while legislative, financial, and other forms of support were still evolving. This created a delicate time period (in which this research was conducted) during which local government authorities had to designate their roles and responsibilities with very little support or enforcement. Just as the mayor and sous prefet had reached certain working norms in line with the little guidance they had, the humanitarian influx came, introducing a large number of resources and powers that had not existed before in the commune. The result was a significant degree of recentralisation. Had this been explicitly temporary, the impact of the humanitarian influx could not accurately be described as transformatory. Because the decentralisation process in Mali is still so nascent, however, and the process of political transformation was still very much in progress, and also very much defined by practice, a large amount of norm creation in the offices of the sous prefet and mayor went on with the arrival of the refugees. From a political point of view, however, the humanitarian influx granted decentralisation several victories. Before the influx, a lack of resources at the local level was holding back the political goals of decentralisation. Citizens saw little utility in political involvement, since local government was not in a position to provide services. When resources and infrastructure accompanied the refugee influx, the demands citizens made of local government authorities changed from service provision to negotiating rights and access to services. This is something the local government had the capacity to do, and because of this, a good deal of trust was restored between residents of Loulouni and their elected leaders. A cautionary note must be added, however. While it is too early to know if this will be the case, there is a possibility that the shift from demanding services to conflict resolution will inhibit the longer term creatio nof accountability structures at the local level. 123 Caitlin Blaser 1 September 2007 The conclusions that can be drawn from this study speak first to the literature in forced migration studies that is connected to state transformation. It supports the findings of Landau and Whitaker who claim that refugees are having a fundamental, transformatory impact on the state. Both Landau and Whitaker were studying the case of Tanzania, but neither focussed on sepcific bureaucratic changes the state was undergoing. Rather, they defined state transformation in a broader social sense, looking primarily at affiliations citizens expressed towards the state. This study has taken the premise of their findings, and applied it to a specific case of state transformation: decentralisation. Its findings have been in line with those of Landau and Whitaker; essentially, that refugees are transforming the way citizens interact with governments in important and fundamental ways. This study particularly bolsters their claims because they both used Tanzania as a case study, while this study was done in Mali, a significantly different political context. While using refugee studies as an analytical framework for studying decentralisation may initially seem like a poor choice, a strong justification for this approach can be found in 'mainstream' political science literature. In Cages of Reason, Silberman (2000) argues that the rise of Weberian rationalised bureaucracy in State function is not the conquest of one specific institutional structure, but rather that rationalised bureaucracies have taken two different shapes. He divides these into a position-oriented structures, and a professionally-oriented structures. Silberman argues that which structure evolved in these states is a function of leadership disputes at the turn of the century. It is not only the leadership disputes that shape states, but rather a lack of sufficiently rationalised State institutions that led to leadership disputes and transformation of State institutions. As such, leadership disputes can be seen as flags for times of transformation of State institutions, and Silberman has shown that the nature of the leadership disputes will play a crucial role in determining the shape of emerging institutions. The current decentralisation process in Mali is undeniable a time of tremendous institutional change in the State's bureaucratic system. This paper has argued that the refugee presence is playing an important role in shaping local political processes, particularly redefining power dynamics and leadership disputes. A refugee presence can then be seen as having a substantial transformatory impact on the shape of the institutions currently emerging from this period of political evolution. The short-term political consequences of the refugee impact may have a positive affect on governance. This study found that the refugee influx in Loulouni has promoted interaction with local government officials, and shifted demands away from the provision of services that officials lack the resources to provide to negotiation of management and conflict resolution, which local governments 124 Caitlin Blaser 1 September 2007 are better equipped to handle. This has played an important confidence building role that is crucial to the decentralisation process. The long-term consequences, however, could be more problematic, as power struggles between upwardly and downwardly accountable officials regarding their responsibilities to the refugees could result in an institutional shift away from downward accountability. While this study has exclusively focused on the commune of Loulouni, there are implications for other parts of Mali and other countries currently undergoing a process of decentralisation. As mentioned in the introduction and throughout the paper, many countries, particularly in Africa, share certain historical and economic links, and are experiencing decentralisation in similar ways. While each case will have important differences, there are lessons from Loulouni that can be generalised. This generalisation must be done with caution; much of this thesis was devoted to background information about the history and context of decentralisation in Loulouni, to completely understand the process and the dynamics surrounding it. In generalising, the same process of contextualising must take palce. Decentralisation, like any process of political transformation, is cross cutting, requiring coordination in administrative, political, and fiscal policies that can be applied across a range of social, economic, and political realities. In Loulouni, the administrative fragility that was aggrevated by the arrival of refugees highlights the need for coordination of each step along the way. Otherwise, any change in context, such as the refugee influx in Loulouni, can create norms and become transformatory. Having political transformation come before there is legislative guidance and fiscal support risks undermining the decentralisation process as a whole. A second lesson is for the humanitarian organisations that operate in areas of political transformation. As noted previously, many refugee crises happen in weak states, with challenges of institutional capacity, and ongoing transformation of the roles and responsibilities of state institutions. It is then particularly necessary to be sensitive to the political realities on the ground. In Loulouni, humanitarian organisations had such capacity when compared to local government structures, that it was nearly inevitable that they would shape the political realities in Loulouni. Given the potential to impact the decentralisation process tremendously, they could have been a strong tool for promoting participatory governance without compromising the protection of refugees that was their mandate. By working through upwardly rather than downwardly accountable officials at the local level, they took the decentralisation process in Loulouni several steps back. This element of planning should be introduced into all relief operations. While this study did answer many of the questions it set out to ask, it created far more 125 Caitlin Blaser 1 September 2007 questions in the process. There is a clear need for more research linking the influx of refugees to specific kinds of political transformation. The causal linkages of political transformation as a cause of forced migration, and even migration more generally, are well explored in the literature. However, comparably little work has been done looking at the impact forced migration has on political transformation. It is clear in both bodies of literature that there is a demand for this to be done, and this study confirms the need for further exploration of the issue. While this study has broadly treated the impact the influx of one group of refugees has had on one decentralisation process, there is a need for comparative work to be done on the subject. Much of the comparative work that has been done on decentralisation has been at the international level; very few of the conclusions are applicable to local contexts. There are many challenges in ensuring such large studies are methodologically rigorous, or sufficiently in-depth to be relevant for comparison given varying national contexts. Additionally, there is a need to link humanitarian influxes to state transformation in many more concrete ways. Decentralisation is only one example of state transformation. There is an equal need for links to be created between a refugee influx and the creation of legislation, electoral processes, judicial precedents, and other political processes that have the potential to fundamentally transform the way states function, and citizens interact with them. Some decentralisation research is now being done that looks at the decentralisation process of certain ministries in various countries in a comprehensive way, to increase the technical understanding of legislative and other processes, and the most effective and efficient ways of supporting decentralisation in each arena. However, most of these studies are exceptionally case specific, and there is a need for methodological synthesis for solid conclusions to be drawn. There is a need for a standardisation of language and indicators of success in decentralisation, so that effective and conceptually sound comparative studies can be done. With the current literature, such a study is virtually impossible at anything but a broad, international level. Finally, there is a need for the research question to be turned on its head, to explore the implications of decentralisation on refugee protection and camp management. As the autonomy of local governments becomes better understood, their autonomy related to a refugee influx will provide an important missing link between theory and practice in forced migration studies literature. This gap is already widely acknowledged in the forced migration studies literature (Misago, 2005), but as norms develop for interaction between humanitarian agencies and local governing bodies, there will certainly be a call better to understand these dynamics. 126 Caitlin Blaser 1 September 2007 Bibliography Agrawal, Arun, and Jesse Ribot. 1999. Analysing decentralisation: A Framework with South Asian and West African Environmental Cases. Journal of Developing Areas 33. Akeratane, Mohamed Ag. 2005. ?Cooperatives, Development and decentralisation in Mali.: Experiences in building successful farmer group businesses.? Sustainable Development Department, Food and Agriculture organisation of the United Nations. Akind?s, Francis. 2004. ?The Roots of the Military-Political Crises in Cote d'Ivoire.? Research Report 128, Uppsala: Nordiska Afrikainstitutet. Alchian, Armen and Harold Demsetz. 1973. ?The Property Rights Paradigm.?Journal of Economic History. 33(1) The Tasks of Economic History, p16-27. Alfonso, Haroldo Dilla. 1999. ?Political decentralisation and Popular Alternatives: A View from the South.? Ontario: International Development Research Center. Amin, Samir. 1973. Neo-Colonialism in West Africa. Trans. Francis McDonagh, Penguin Books: Middlesex, England. Anthony, Constance G. 1991. ?Africa's Refugee Crisis: State Building in Historical Perspective. International Migration Review 25(3) p 574-591. Aredo, Dejene. 2004. ?Fuzzy Access Rights in Pastoral Economies: Case Studies from Ethiopia.? Paper prepared for the Tenth Biennial Conference of the International Association for the Study of Common Property. Ba, Lassana. 1998. ?Community Based Management of Natural Resources in the villages of the ?Siwaa? Test Area of the CMDT's Southern Mali Zone.? CMDT. Ba, Sadou Oumar. 1999. ?D?centralisation et Pratiques Locales. Mise en Relation du Capital Social des Institutions Locales Traditionnelles et des Processus de D?centralisation en Vue d?informer les Politiques de la D?centralisation. Etude de Cas au Mali: Syst?mes Fonciers et Gestion des Ressources Naturelles de l?Espace For?t du Samori.? Consultancy report. Bamako: FAO/AVES. Baker, Johnathan and Tade Akin Aida. 1995. The Migration Experience in Africa. Uppsala: Nordiska Afrikainstitutet. Bakewell, Oliver. 2000. ?Uncovering Local Perspectives on Humanitarian Assistance and Its Outcomes.? Disasters 24(2): 101-116. Baland, Jean-Marie and Jean-Philippe Platteau. 2002. ?Collective action on the commons: 127 Caitlin Blaser 1 September 2007 the role of inequality.? CRED. Balogun, M.J. 2000. ?The scope for popular participation in decentralisation, community governance and development: Towards a new paradigm of centre-periphery relations.? Regional Development Dialogue, Vol. 21, No. 1, Spring, pp. 153?173. Banzhaf, M., Drabo, B., and Grell, H., 2000, ?From Conflict to Consensus: Towards Joint Management of Natural Resources by Pastoralists and Agro-Pastoralists in the Zone of Kishi Beiga, Burkina Faso,? Securing the Commons: 3. SOS Sahel/IIED, London. Bascom, Johnathan. 1993. ?The Peasant Economy of Refugee Resettlement in Eastern Sudan.? Annals of the Association of American Geographers 83(2) 320-346. BBC On-line News, ?Timeline: Ivory Coast? 10/07/2002. www.bbc.co.uk. Becker, L.C. 1990. ?The collapse of the family farm in West Africa? Evidence from Mali.? Geographical Journal 156(3): 313-321. Benjaminsen, Tor. 1997. ?Natural Resource Management, Paradigm Shifts, and the decentralisation Reform in Mali,? Human Ecology 25 (1): 131. Benjaminsen, Tor.1998. decentralisation and Management of Village Land in Rural Mali. World Bank/ EDI's CBNRM Initiative. Benjaminsen, Tor. 2001. Power, Property, and Politics in the West African Sahel. Uppsala: Nordiska Afrikainstitutet. Bennet, R.J. 1990. Decentralisation: Local Governments and Markets, London: Clarendon Press. Beramendi, Pablo. 2003. ?Political Institutions and Income Inequality: The case of decentralisation.? Berlin: Wissenschaftszentrum Berlin fur sozialforschung. Bingen, James. 1998. ?Cotton, Democracy and Development in Mali.? The Journal of Modern African Studies. 36(2) p265-85. Bonfiglioli, Angelo. 2003. Empowering the Poor: Local Governance and Poverty Reduction New York: United Nations Capital Development Fund. Bratton, Michael, Massa Coulibaly, and Fabiana Machado. 2000. ?Popular Perceptions of Good Governance in Mali.? Agrobarometer Paper 9. Breton, Albert. 2000. ?An Introduction to decentralisation Failure.? Unpublished Paper. International Center for Economic Research, Turin. Bruijn, Mirjam de and Han van Dijk. 2003. ?Changing Population Mobility in West Africa: 128 Caitlin Blaser 1 September 2007 Fulbe Pastoralists in Central and South Mali.? African Affairs 102 p285-307. Chambers, Robert. 1993. ?Hidden Losers: The Impact of Rural Refugees and Refugee Programs on Poorer Hosts.? Pp. 29-46 in Refugee Aid and Development Theory and Practice, edited by R. F. Gorman. Westport: Greenwood Press. Cheema, G.S. and D.A. Rondinelli (eds.). 1983. decentralisation and Development: Policy Implementation in Developing Countries. Beverly Hills, CA: Sage. Clark in Bingen. 2001. Democracy and Development in Mali. Madison: University of Wisconsin Press. Cohen, John and Stephen Peterson. 1997. ?Methodological Issues in the Analysis of decentralisation.? Administrative decentralisation in Late Developing Countries. Kumarian Press. Conyers, Diana. 2000. ?Decentralisation: A conceptual analysis (Part 1).? Local Government Perspectives: News and Views on Local Government in Sub-Saharan Africa, Vol. 7(3): 7?13. Conyers, Diana. 2001. ?Whose Elephants Are They? decentralisation of Control over Wildlife Management through the CAMPFIRE Programme in Binga District, Zimbabwe,? Working Paper, January 31, 2001. Washington, D.C.: World Resources Institute. Cooper, Frederick. 1996. Decolonisation and African Society: The Labor Question in French and British Africa. Cambridge Univeristy Press: Cambridge. Cordell, Dennis et al. 1996. Hoe and Wage : A Social History of a Circular Migration System in West Africa. Westview Press: Boulder, CO. Crisis Committee Meeting Notes. April 14 2002, Loulouni. Crisp, Jeff. 2000 ?Africa?s Refugees: Patterns Problems and Policy Challenges? Journal of Contemporary African Studies, Vol. 18 (2): 157-178. Crook, Richard. 1991. "State, Society and Political Institutions in Cote d"Ivoire and Ghana." In James Manor (ed.) Rethinking Third World Politics. London: Longmans. Crook, Richard and James Manor. 1994. Enhancing Participation and Institutional Performance: Democratic Decentralisation in South Asia and West Africa. A Report to ESCOR, the Overseas Development Administration. Crook, Richard C., and James Manor. 1998. Democracy and decentralisation in South-East Asia and West Africa: Participation, Accountability, and Performance. Cambridge: Cambridge University Press. 129 Caitlin Blaser 1 September 2007 Crook, Richard C., and Alan Sturia Sverrisson. 2001. ?Decentralisation and Poverty- Alleviation in Developing Countries: A Comparative Analysis or, Is West Bengal Unique?? Institute for Development Studies (IDS) Working Paper 130. Brighton: IDS. Dante I., J.F. Gautier, M.A. Marouani & M. Raffinot (2002), Institutionalising the PRSP approach in Mali, PRSP Institutionalisation Study, Final Report (ODI, DFID), SPA. De Mello, Luis R. 2004. ?Can Fiscal Decentralisation Strengthen Social Capital?? Public Finance Review 32(1) p4-35. Diallo, C.S. & M. Raffinot .1999. Evaluation du programme National de Lutte contre la pauvret? au Mali, PNUD. Diarrah, Cheick Oumar. 1986. Le Mali de Modibo Keita Paris: Editions l'Harmattan. Diarrah, Cheick Oumar. 1991. Vers la Troisieme Republique du Mali Paris: Editions l'Harmattan Dick, Shelly. 2002. ?Responding to protracted refugee situations: A case study of Liberian refugees in Ghana?, Evaluation and Policy Analysis Unit:UNHCR. Domba, Jean-Marc. 2003. "L?Immigration Burkinab? et les Enjeux Identitaires en C?te d?Ivoire" INSS/CNRST: Ouagadougou. Ebel, Robert and Serdar Yilmaz. 2001. ?Concept of Fiscal decentralisation and Worldwide Overview.? International Symposium, Qu?bec Commission on Fiscal Imbalance. Eyadema, Gnassingbe in Doudou, Baba. 2003. ?Art of Setting a Region on Fire.? World Review. Faguet, Jean-Paul. 1997. ?Decentralisation and Local Government Performance.? Technical Consultation on decentralisation Rome: FAO. Fiszbein, Ariel. 1997. ?Decentralisation and Local Capacity: Some Thoughts on a Controversial Relationship,? Paper presented athe the FAO/UNCDF/ World Bank Technical Consultation on Decentralisation, Rome, 15-8 December 1997. Freund, Paul and Katele Kalumba. 1986. ?Spontaneously Settled Refugees in Northwestern Provence, Zambia.? International Migration Review 20(2) Refugees: Issues and Directions p299-312. Gibson, Clark C. 2001. ?The Politics of Decentralised Environmental Policy: Theoretical Considerations and Preliminary Findings.? Paper prepared for the Open Meeting of the Human Dimensions of Global Environmental Change Research Community. 130 Caitlin Blaser 1 September 2007 Grauze, Anton de. 2004. ?Decentralisation and School Improvement: A Comment on research in West Africa? International Working Group on Education. Hilhorst, Th?a and Adama Coulibaly. 1998. Une Convention Locale pour la Gestion Participative de la Brousse au Mali. Dossier no. 78. London: IIED. International Federation of Red Cross and Red Crescent Societies. 2002. ?Mali Red Cross Aids Thousands Fleeing Cote d'Ivoire Conflict.? Press Release. Jacobsen, Karen. 2001. ?The Forgotten Solution: Local Integration for Refugees in Developing Countries? New Issues in Refugee Research, Working Paper No 45. Geneva: UNHCR. (Available online at: http://www.jha.ac/articles/u045.pdf. ) Jacobsen, Karen. 2002. ?Can Refugees Benefits The State? Refugee Resources and African Statebuilding. Journal of Modern African Studies. Vol. 40(4): 577-596. Jennings, Michael. 2003. ?We Must Run While Others Walk: Popular participation and development crisis in Tanzania, 1961-9.? Journal of Modern African Studies 41(2) p163-87. Kassibo, Brehima. 2004. Historical and Political Foundation for Participatory Management and Democratic decentralisation in Mali: A Synthesis of two Case Studies.Working Paper 16. World Resources Insitute: Institutions and Governance Program. Keita, Sikoro. 2000. ?Overview of the Privatisation Process in Mali.? USAID Mali. Keita, Sikoro. 2002. ?Assessing the Economic Impact of USAID Assistance in Mali Over the Period 1998-2002.? Economic Report Series. USAID. King, Gary, Robert Keohane, and Sidney Verba. 1994. ?Causality and Causal Inference? Designing Social Inquiry: Scientific Inference in Qualitative Research. Princeton: Princeton University Press: 115-149. Koser, Khalid and Van Hear, Nicholas. 2002. ?Asylum migration: implications for countries of origin?, paper presented at the WIDER Conference on Poverty, International Migration and Asylum, Helsinki, 27?28 September 2002. Landau, Loren. 2003. ?Beyond the Losers: Transforming Governmental Practice in Refugee- Affected Tanzania.? Journal of Refugee Studies. Vol 16(1): 19-43. Landau, Loren B. 2004. ?Challenge without Transformation: Refugees, Aid, and Trade in Western Tanzania.? Journal of Modern African Studies. Vol.42 (1): 31-59. Litvack, Jennie, Junaid Ahmad and Richard Bird. 1998. ?Rethinking Decentralisation in Developing Countries.? Sector Studies Series. Washington, D.C.: World Bank. 131 Caitlin Blaser 1 September 2007 Mamdani, Mahmood. 2001. ?Defining the Crisis of Postcolonial Citizenship: Settler and Native as Political Identities.? Chapter One in When Victims Become Killers. Princeton: Princeton University Press. Mamdani, Mahmood. 1996. Citizen and Subject: Contemporary Africa and the Legacy of Late Colonialism. Princeton: Princeton University Press. Management Systems International. 2004. Democratic Decentralisation Review: Mali. Unpublished report. Manor, James. 1997. ?The Promise and Limitations of Decentralisation? Technical Consultation on Decentralisation Rome: Sustainable Development Department, Food and Agricultural Organisation. Manor, James. 1999. The Political Economy of Democratic Decentralisation. Washington, D.C.: The World Bank. Martin, Susan F, Schoenholtz, Andrew I, and Fisher, David. 2002. ?Impact of asylum on receiving countries?, draft version (30 July 2002) of paper presented at WIDER Conference on Poverty, International Migration and Asylum, Helsinki, 27?28 September 2002. MATCL/DNCT, Mise en OEuvre de la D?centralisation ? Recueil des textes l?gislatifs et r?glementaires, March 1999?April 2000. Messer, Norman. 2001. ?Mapping Traditional Structures in Decentralisation Policies: illustrations from three countries in Sub-Saharan Africa and the Near-East.? Household Livelihood Strategies and Local Institutions: A Working Paper Series. Rural Development Division: Food and Agriculture Organisation. Misago, Jean Pierre. 2005. ?Responses to Displacement in Africa: The Irrelevance of Best Practice.? Conflict Trends Durban: ACCORD. Mission de D?centralisation, ?La Commune en Question ??, R?publique du Mali, avec Appui du Programme des Nations Unies pour le D?veloppement (PNUD), no date. Mission de D?centralisation, ?Lois et D?crets de la D?centralisation?, R?publique du Mali, February 1997. Nejhuis, Karen. 2003. ?Does Decentralisation Serve Everyone? The Struggle for Power in a Malian Village.? European Journal of Development Research 15:2. p67-92. Nkrumah, Stephen A. 2000. ?Decentralisation for good governance and development: The Ghanaian experience.? Regional Development Dialogue 21:1. p 53?67. Oates, W. 1999. An essay on fiscal federalism. Journal of Economic Literature 37:1120-49. 132 Caitlin Blaser 1 September 2007 Olowu, Dele. 2001. ?Local Political and Institutional Structures and Processes? A summary report prepared for the UNCDF symposium Decentralisation and Local Governance in Africa. Ostrom, Elinor, Larry Schroeder, and Susan Wynne. 1993. Institutional Incentives and Sustainable Development: Infrastructure Policies in Perspective. Boulder: Westview. Ostrom, E. 1999. ?Coping with Tragedies of the Commons.?Annual Review of Political Science, 2:493-535. Ostrom, E. 2000. Private and Common Property Rights. p332-352. Oyugi, Walter O. 2000. ?Decentralisation for Good Governance and Development,? Regional Development Dialogue. 21(1). PADLOS. 1998. Decentralisation and Local Capacity Building in West Africa: results of the PADLOS-Education Study, Club du Sahel/OECD/CILSS, Paris. Parkin, D.J. 1972. Palms, Wine and Witness: Public Spirit and Private Gain in an African Farming Community. San Fransisco: Chandler Publishing Company. Phillips, Melissa. 2003. ?The Role and Impact of Humanitarian Assets in Refugee-Hosting Countries? New Issues in Refugee Research: Working Paper No. 84. (March 2003): Geneva: UNHCR. Platteau, Jean-Philippe. 2003. ?Droits de propri?t? et gestion efficace des ressources naturelles.? Les s?minaires de l'Iddri, 10. Platteau, Jean-Philippe. 1995. Reforming Land Rights in Sub-Saharan Africa - Issues of Efficiency and Equity. Gen?ve, UNRISD (Discussion Paper no. 60). Prud?homme, R. 1994. ?On the Dangers of Decentralisation,? World Bank Policy Research Working Paper No. 1252. Washington, DC: World Bank. Prud?homme, Remy. 2001. ?Fiscal Decentralisation and Intergovernmental Fiscal Relations,? Draft Paper for the Participatory Symposium on decentralisation and Local Governance in Africa, United Nations Capital Development Fund, Cape Town, March 26-30, 2001. Quinn, Mick, ?More than implementers: civil society in complex emergencies ? a discussion paper?, International Alert, August 2002, p4. Rawson in Bingen. 2001. Democracy and Development in Mali. Lansing: University of Michigan Press. Republique du Mali. 2001. ?Joint World Bank-IMF Staff Assessment of the Progress of the Poverty Reduction Strategy Paper.? Report no 23287-MLI. 133 Caitlin Blaser 1 September 2007 Republique du Mali. 2002. Poverty Reduction Strategy Paper. Unpublished document. Ribot, Jesse. 2002. ?African Decentralisation: Local Actors, Powers, and Accountability.? Democracy, Governance, and Human Rights Working Paper: 8. UNRISD. Ribot, Jesse C. 1999. ?Decentralisation, Participation, and Accountability in Sahelian Forestry: Legal Instruments of Political-Administrative Control,? Africa 69(1): 23-65. Ribot, Jesse. 1998. ?Decentralisation and Participation in Sahelian Forestry: Legal Instruments of Central Political-Administrative Control.? Harvard Center for Population and Development Studies Working Paper Series: 98.06. Robson, Mark. 2004. ?Complementarity Between Formal State Policy and Informal Norms, An Institutional Determinant of Success in Environmental Policy Interventions? The Case of Deforestation in Developing Countries.? Paper prepared for EGDI-WIDER Conference, Helsinki, Sept 17-18. Rondinelli, Dennis A. 1990. ?Decentralisation, Territorial Power, and the State: A Critical Response,? Development and Change 21: 491-500. Rondinelli, D., G. Shabhir Cheema, J. Nellis. 1984. Decentralisation in Developing Countries: A Review of Recent Experience, World Bank Staff Working Papers No.581. Washington, D.C.: World Bank. Runge, C. R. 1983. "Common Property Externalities: Isolation, Assurance and Resource Depletion in Traditional Grazing." American Journal of Agricultural Economics 63 (4): 595- 606. Ruttan, Lore and Monique Borgerhoff. 1999. ?Are East African Pastoralists Truly Conservationists?? Current Anthropology 40 (5). Sall, A. 1989. L'organisation du monde rural du Mali (1910-1988). Bamako: Editions Imprimerie. Santiso, Carlos and Augustin Loada. 2003. ?Explaining the Unexpected: Electoral reform and democratic governance in Burkina Faso.? Journal of Modern African Studies 41(3) p395-419. Schroeder, Larry and Matt Andrews. 2001. ?Key Issues in Sectoral Decentralisation.? A summary report prepared for the UNCDF symposium on Decentralisation and Local Governance in Africa. Sewell, David O. 1996. ??The Dangers of Decentralisation? According to Prud?homme: some Further Aspects,? The World Bank Research Observer 11(1). Shah, Anwer and Theresa Thomson. 2004. ?Implementing Decentralised Local Governance: A Treacherous Road with Potholes, Detours and Road Closuers.? World Bank Policy 134 Caitlin Blaser 1 September 2007 Research Working Paper 3353. Sharma, Keshav C. 2000. ?Popular participation for good governance and development at the local level: The case of Botswana.? Regional Development Dialogue. Vol. 21(1): 177? 191. Sigel, Efrem. 1970. "Ivory Coast: booming economy, political calm? Africa Report April p8-21. Site log. 1998-2000, Loulouni, Sikasso: Peace Corps, unpublished document. Skeldon, Ronald. 1997. Migration and Development: A Global Perspective, Longman Press: Essex, England. Smoke, Paul. 2000. ?Fiscal Decentralisation in Developing Countries: A Review of Current Concepts and Practice,? Paper prepared for the United Nations Research Institute for Social Development (UNRISD), Geneva, October 2000: UNRISD. Smoke, Paul. 2003. ?Decentralisation in Africa: Goals, Myths, and Challenges.? Public Administration Development 23(7): 7-16. SNV. 2004. ?La decentralisation au Mali: du discours a la practique? Instutut Royal des Tropiques: Amsterdam. Sommers, M. 1995. ?A New African World: Rwandan and Other African Youth.? Global Justice 1(3): 69-74. Sphere Project. 2004. Humanitarian Charter and Minimum Standards in Disaster Response. Geneva: Oxfam Publishing. Sperl, Stefan. 2002. ?International Refugee aid and Social Change in Northern Mali.? Refugee Survey Quarterly. 21(1). Skinner, J. 1988. ?Towards Better Woodland Management in the Sahelian Mali.? Pastoral Development Network. 25d. Sugden, R. 1984. "Reciprocity: The Supply of Public Goods Through Voluntary Contributions." The Economic Journal (94)376: 772-787. Tache, Boku and Ben Irwin. 2003. ?Traditional institutions, multiple stakeholders and modern perspectives in common property: Accompanying change within Borana pastoral systems.? Securing the Commons: 4 SOS Sahel. Tessougu?, Martin (dit Moussa), Pippa Chenevix-Trench and Philip Woodhouse. 1998. ?Terre, Eau et Administration Locale au Mali: Riziculture et Gestion des Ressources Naturelles dans la Vall?e du Sourou Cercle de Bankass,? Rural Livelihoods Working Paper series. Manchester: Institute for Development Policy and Management. 135 Caitlin Blaser 1 September 2007 Therkildsen, Ole. 2001. ?Efficiency, Accountability, and Implementation: Public Sector Reform in East and Southern Africa,? Draft Copy of Democracy, Governance and Human Rights Programme Paper No. 3, United Nations Institute for Social Development: Mimeo. Therkildsen, Ole. 1993. ?Successful African Local Government: Some Methodological and Conceptual Issues,? Public Administration and Development, 13(5): 51-5. Therkildsen, Ole.1993. ?Legitimacy, local governments and natural resource management in Sub-Saharan Africa.? In Henrik Secher Marcussen (ed. Institutional Issues in Natural Resources Management: With Special Reference to Arid and Semi-Arid Areas in Africa. Occasional Paper No. 9, Roskilde University : International Development Studies. Toulmin, Camilla, Rebecca Leonard, and Thea Hilhorst. 2000. ?Mali Poverty Profile.? Drylands Project, International Institute for Environment and Development (IIED), London. May. (Available online at http://www.iied.org/pdf/poverty.pdf) Tour?, Moriba et al. 1992. ?Population Dynamics and Development Strategies in C?te d'Ivoire? in eds. Moriba Tour? and T.O. Fadayomi. (1992) Migrations, Development and Urbanisation Policies in Sub-Saharan Africa. CODRESIA: Dakar, Senegal. Turrittin, Jane. 1988. ?Men, Women, and Market Trade in Rural Mali, West Africa.? Canadian Journal of African Studies 22(3) p583-604. Turton, David. 2003. ?Refugees and Other Forced Migrants? Refugee Studies Centre Working Paper. No. 13. Oxford: Queen Elisabeth House. (Available online at: http://www.rsc.ox.ac.uk/wp13.pdf) UNAIDS. 2004. ?Country Report: Mali.? Geneva: United Nations. (Available online at: www.unaids.org/Mali) United Nations Capital Development Fund. 2004. ?Companion Report: Mali.? Independent Programme Impact Assessment (PIA) of the UNCDF Local Development Programme. ECI: Woodsmead. UNCDF. 2000. ?Africa: Decentralisation and Local Governance Conference Concept Paper,? Draft Paper for the Decentralisation and Local Governance Conference, Capetown, March 2001, United Nations Capital Development Fund. United Nations High Commissioner for Refugees. 1993. ?Mali.? 2003 UNHCR Statistical Yearbook. Geneva: United Nations. United Nations Office for the Coordination of Humanitarian Affairs (OCHA). 2002. ?Humanitarian Situation Report No. 01I.? Humanitarian Highlights. Vedeld, Trond. 1998. ?State Law Versus Village Law: Law as exclusion principle under 136 Caitlin Blaser 1 September 2007 customary tenure regimes.? Erling Berge and Nils. Chr. Stenseth (eds.) 1998: Law and the Management of Renewable Resources, Institute of Contemporary Studies, ICS Press: San Francisco. Vengroff, Richard. 1993. ?Governance and the Transition to Democracy: Political Parties and the Party System in Mali.? Journal of Modern African Studies 31(4) p541-562. Waters, T. 1999. ?Assessing the Impact of the Rwandan Refugee Crisis on Development Planning in Rural Tanzania, 1994-1996? Human organisation 58: 142-152. Weber, Max. 1946. ?Politics as a Vocation? in From Max Weber: Essays in Sociology. Edited by H. H. Gerth and C. Wright Mills. New York: Oxford University Press. Whitaker, B. E. 1999. ?Disjunctured Boundaries,? Ph. D. Thesis, Political Science, University of North Carolina, Chapel Hill. Whitaker, Beth Elise. 2002. ?Refugees in Western Tanzania: The Distribution of Burdens and Benefits Among Local Hosts. Journal of Refugee Studies. Vp; 15(4): 339-358. World Bank. 2000. ?Decentralisation: Rethinking Government.? World Development Report 1999-2000. p107-124. Wunsch, James S. and Dele Olowu. 1995. The Failure of the Centralised State: Institutions and Self-Governance in Africa. Boulder, CO: Westview Press. p334. 137 Caitlin Blaser 1 September 2007 Appendixes Appendix 1: Illustrative tables of how powers are being decentralised in Mali 138 Caitlin Blaser 1 September 2007 Appendix 2: Change in rainfall in Mali over the past century (Source: Meteological Association, Bamako) 139 Caitlin Blaser 1 September 2007 Appendix 3: Map of southern Mali and surrounding regions 140 Caitlin Blaser 1 September 2007 Appendix 4: Table of dependent and independent variables in measuring the success of the decentralisation process. Dependent Variables Indicators Accountability and transparency ? Current laws regarding information provision ? My own ability to access information ? Perceived availability of information by local officials and citizens ? Local government's oversight and monitoring structures ? Accountability perceived by local officials ? Accountability perceived by voters Responsiveness to community ? Local officials' reported priorities in taking budgetary and planning decisions ? Responsiveness perceived by the community ? Specific examples Resource mobilization ? Tax records ?History of village caisse spending ? Resources available to local government ?Resources available to community associations ? Resources available to refugee camp Efficiency in service delivery ? Survey of services provided by local government, refugee camp related organizations, and other parties ? Perceived responsibility for service provision by community members and service providers ? Reported satisfaction about services provided by government, refugee related organizations, and other parties Interaction between government and populace ? Minutes of local government meetings ? Attendance of local government meetings ? Local officials' schedules ? Community perceptions of interactions ? Local officials perceptions of interactions 141 Caitlin Blaser 1 September 2007 Dependent Variables Indicators Collective grassroots action ? Presence of associations and other CBOs ? Resources available to associations present ? Perception of collection action by the community and local government ? Specific examples Independent Variables Indicators Human capital ? Level of formal education achieved ? Other trainings and/or informal education ? Number of languages spoken Social cohesion ? Degree to which gender impacts responses ? Degree to which ethnicity impacts responses ? Perceptions of the extent to which the community works together ? Specific examples Involvement in political process ? Reported interest ? If registered as a voter ? If reported voting in the last election ? Voter turnout in last election ? Ability to name elected officials Economic equality ? Number of responses in each economic bracket as determined by primary profession, other sources of income, head of household's profession, if and how many employees are hired, and whether they are paid in cash. ? National Department of Statistics information on poverty levels and equality in Loulouni Refugee impact ? To be determined by controlling for other independent variables ? Specific examples 142